Here is one intriguing tweet:
Wllierslev: 24,000-yr-old Siberian Mal'ta person geneticall similar to native amer and west eurasians. No east asian #paleoamericanodysseyand another:
Willerslev: Native Americans formed by an admixture of east Asian ancestors and the ancestors of western Eurasians #paleoamericanodysseyand another:
Willerslev: Based on genomes, "the Mal'ta is much darker if you want than the iceman (Otzi)" #paleoamericanodysseyI'll be occasionally looking at the #paleoamericanodyssey tag, but feel free to point to any other interesting tweets from the conference in the comments.
UPDATE:
From the Met:
The Mal'ta tradition is known from a vast area spanning west of Lake Baikal and the Yenisey River. The site of Mal'ta, for which the culture is named, is composed of a series of subterranean houses made of large animal bones and reindeer antler which had likely been covered with animal skins and sod to protect inhabitants from the severe, prevailing northerly winds. Among the artistic accomplishments evident at Mal'ta are remains of expertly carved bone, ivory, and antler objects. Figurines of birds and human females are the most commonly found items.From a review article:
Debetz (1946) identified the remains of “nothern Asian Mongoloids” at the site ofUPDATE (Oct 26):
Afontova Gora 2; they included a fragment of the frontal bone. Mongoloid features had
been originally acknowledged in the skeletal remains of a child found at the site of
Malta. Alexeev (1998, 323) in his later publication was more cautious, stating that this
area was “inhabited by a population of Mongoloid appearance.”
Michael Balter has an article on this topic in Science:
Yet the child's Y chromosome belongs to a genetic group called Y haplogroup R, and its mitochondrial DNA to a haplogroup U. Today, those haplogroups are found almost exclusively in people living in Europe and regions of Asia west of the Altai Mountains, which are near the borders of Russia, China, and Mongolia.This suggests that the Mal'ta boy was not ancestral to Native Americans (since Native Americans don't possess Y-haplogroup R and mt-haplogroup U), although obviously is in some way related to them based on the autosomal evidence. It's hard to read between the lines, but I guess a paper in Nature will come out soon enough as it is currently "in press".
131 comments:
See more on Willerlev's talk at http://anthropogenesis.kinshipstudies.org/2013/10/paleoamerican-odyssey-conference-ancient-malta-dna-back-migrations-to-the-old-world-and-hallway-discussions/
"Native Americans formed by an admixture of east Asian ancestors and the ancestors of western Eurasians"
Boy. Have I got into trouble every time I've suggested exactly that. To me it was the only concusion possible when this paper came out in 1999:
http://www.ncbi.nlm.nih.gov/pubmed/10053017
"And there is almost no evidence besides mtDNA X2 that native Americans have any west Eurasian blood".
That 1999 paper shows that the Native Americans Y-DNA (Q) is 'West Eurasian' while almost all their mt-DNA is 'East Eurasian'. To me that has always made total sense while no other interpretation of the data has done so. It means that men with Y-DNA Q moved east through uninhabited land along the northern fringe of already inhabited land, and picked up local women to the south as they moved east.
"But be careful to note it says 'ancestors.' Meaning, proto-Caucasians and proto-Mongoloids".
Can you explain what the terms proto-Caucasians and proto-Mongoloids actually mean. I understand that a 'proto-hereford' could mean a cattle beast from a region that later produced hereford cattle, but the fully-hereford breed only came about as the result of selective breeding of the cattle in the region. But I fail to see how that could be in any real way similar to what you're proposing as 'proto-Caucasian' or 'proto-Mongoloid'.
From the 2003 Japanese Review article: 'The next stage in the settlement of northern Eurasia by early modern humans occurred during the time-span of 32-18 thousand years ago. This was the coldest period of the Last Ice Age that included the Last Glacial Maximum (LGM).
Remarkably this episode coincided with the quasi-total depopulation of western and central Europe with all Palaeolithic sites virtually disappearing in that area [...] Significantly, during the same period the AMH population density rose on [the] East European Plain, the frequency of Palaeolithic sites markedly increasing at 29-26 ka BP and forming a clear maximum at 24-18 ka BP.
[...]
This suggests a large-scale eastbound migration of groups of modern humans during the coldest episode of the Last Ice'
This scenario of heavy population movements seems to have had a profound influence on the contemporaneous racial composition af far as the Zhoukoudian Upper Cave near Peking, whose Ainu characteristics "may be extended
to the entire Ice Age population of Northern Eurasia". This may have mitigated the Mongolic morphology for a while in this northern part of the world, and probably in the Americas as well.
Apparently some admixed groups of descendents also returned to the west again to make their 'Amerind' contribution in northern Europe, although I don't think the European plains were ever completely abandoned. However, that much of the European Plain may have been repopulated rather from the east could make an elegant alternative to the moribund Iberian glacial refuge hypothesis.
It seems the data has already been analyzed by a genome blogger. Autosomal breakdown for Mal’ta, courtesy of Verenich.
[2,] “33.7% Brahui + 66.3% Udmurd” “21.9804″
[3,] “34.5% Makrani + 65.5% Udmurd” “22.357″
[4,] “34.3% Balochi + 65.7% Udmurd” “22.413″
[5,] “33.3% Sindhi + 66.7% Udmurd” “24.1198″
[6,] “36.5% Burusho + 63.5% Udmurd” “24.211″
[7,] “39.7% Pashtun + 60.3% Udmurd” “24.3389″
[8,] “34.3% Pathan + 65.7% Udmurd” “24.716″
[9,] “32.2% Pakistani + 67.8% Udmurd” “24.753″
[10,] “41.4% Tadjik + 58.6% Udmurd” “24.852
What does the affinity to Northwest South Asian populations imply? Very unexpected, and rather cool. My knowledge of the Byzantine world of haplogroups is pathetic, but could this have anything to do with the R haplogroup? I thought this haplogroup originated in South Asia (I might be wrong, never really kept track of the literature, and when I did read some material, I never read closely). Perhaps Dienekes could analyze the data, apparently it's available via the Estonian Biocentre.
terryt,
Can you explain what the terms proto-Caucasians and proto-Mongoloids actually mean.
At this point, we're pretty much going to have to sit back and let the aDNA results do the talking for us. Each new revelation concerning this matter supports what I've been saying since August 2012 and you've been saying for much longer. But we keep making the mistake that, because we argue on the basis of fact, other people also do this.
Onur,
Nothing to do with the ancestors of Native Americans, who probably then lived in locations that lied more to the east than Mal'ta. If anything, the Mal'ta individual has some ancestry from the ancestors of Native Americans (his paleo-northern Mongoloid admixture seems to be from them), not the other way around.
Still? While I admire your dedication and tenacity in misrepresenting facts and drawing illogical conclusions, you really seem to be much more of the broken record you accused me to be than I am.
Rokus,
Apparently some admixed groups of descendents also returned to the west again to make their 'Amerind' contribution in northern Europe
The signal needs to be revised in light of the mixed origin of Amerindians. Also, the ancient North Eurasian signal is essentially equal in North and South Europeans (including Basques and Sardinians) and West Asians, based on the Adygei figure, when accounting for their recent East Eurasian admixture (Lipson et al. (2012)). This needs to be taken into account in any consideration of what the signal signifies.
That 1999 paper shows that the Native Americans Y-DNA (Q) is 'West Eurasian' while almost all their mt-DNA is 'East Eurasian'.
How do you know that Q is West Eurasian? It does not seem that Native American Q has ever been in West Asia, and the North European ancestry component of Selkups and Kets is hardly the result of yDNA Q. I think that Q is neither West Asian nor East Asian, but instead it is ONE possible proto-mongoloid component.
At the moment, I think that mtDNA A was with Q from the beginning, hgs C and D (the second and the third possible proto-mongoloid components) are from North China and Siberia and picked up there and B (the fourth possible proto-mongoloid component) has arisen somewhere in South China and arrived to the North with yDNA C. Hg B could in fact explain the Oceanian component in the North.
Don't forget that those original NA mtDNA haplogroups (especially X2) have been reclassified (renamed) at least a few times since first discovery. Also keep in mind that research on North American NA DNA has been severely limited.
@TerryT
"Boy. Have I got into trouble every time I've suggested exactly that. To me it was the only concusion possible when this paper came out in 1999."
@Onur
"The Mal'ta individual cannot be ancestral to Native Americans..."
Onur is right, Terry is wrong. Amerindians likely contributed genes to a Mal'ta-kind of West Eurasian population. Mal'ta's lack of association with East Asians but the existence of an association of Amerindians suggests that there were no "East Asians" at 25,000 YBP, while there were "Amerindians."
What does the affinity to Northwest South Asian populations imply? Very unexpected, and rather cool.
Seinundzeit,
It looks like both Aurtignacian and Gravettian Europeans entered via a northerly route - but given today's haplogroups, the people must have originated in the NW part of the subcontinent. The way I see it, "Gedrosia" is the local continuation of that source population, "W Asian" the continuation of that population from E Anatolia through Iran and Afghanistan, and the two European components (Mediterranean and N European) are the continuation of that source population in Europe. The Gravettian seems to me the connection to S Siberia - but given the origin of haplogroups P, Q, and R, those people originally likely came from a similar NW subcontinent source like all F-descendants (otherwise they also wouldn't be Caucasian-like 24,000 ya). Also, it seems to me that the new Gravettian arrivals had a stronger impact in N Europe, and it appears very reasonable that their closest descendants today are the Uralic people(s).
There is no good reference population left for the original source population, but "Gedrosia" seems like a decent approximation.
Of course, in today's European and W Asian populations, we have a lot of additional admixture, since (W Asian, SW Asian, S Asian, Turkic, Mongolian, etc.). So, for example, the visible C Asian admixture in Uralic people is likely rather recent, in comparison (or ongoing "diffusion" since LGM).
Terry,
what I mean by "proto-" is rather simple: when modern humans expanded to W Asia, Europe, NE Asia, and Siberia, you cannot yet talk about the categories that are useful today. The people living in S Siberia 35,000 - 24,000 ya cannot possibly derive from Caucasians or Mongoloids, because people with those characteristics where just then in the process of forming. It is pretty clear, anatomically, that the first people moving into the Americas had no or almost no Mongoloid features (but later Na-Dene did, and Inuits the most). This is likely a space-time effect: Siberians and Beringians initially appear not to have had Mongoloid features and likely originate from what I mentioned above (which would make them proto-Caucasian); when people with somewhat Mongoloid features finally moved northward and spread W and E, they were still "proto-Mongoloid" and came "too late" to much affect the Beringians and first wave of Americans. Only later, probably just after LGM, did they spread over time in Siberia and and Beringia.
A couple of contributors (whose opinions I respect very much) have used the expressions 'paleo-northern Mongoloid' and 'proto-Mongoloids'.
I'm sure everyone here agrees that each gene mutation must happen in a single individual. The mutation can then spread through a population by drift or selection. Selection is much more rapid than is drift, but both happen most effectively in small, isolated populations. The possibility that selection on a set of genetic mutations operated across the huge expanse of territory now occupied by people exhibiting what we might call 'the Mongoloid phenotype' is so unlikely as to be not worth considering.
A far more effective and rapid method of spreading a genetic mutation widely, as any animal breeder knows, is the introduction of a population containing that mutation. I agree there may have been several population expansions containing the several individual genes involved in producing the Mongoloid phenotype and therefore, collectively, we may talk of several 'proto-Mongoloid' populations. But it is equally possible that there was a single region in which the phenotype largely developed, and proto-Mongoloid populations are simply the product of various degrees hybridization with that population.
And we have research that tells us exactly where one of the genetic mutations involved in the formation of the Mongoloid phenotype developed, and expanded from: the EDAR370A mutation. The mutation occurred some time before anyone had reached America, and is certainly prominent there. The relevant maps are most easily accessed here:
http://blogs.discovermagazine.com/gnxp/2013/02/is-girls-generation-the-outcome-of-the-pleistocene-mind/
It is also very much possible that the EDAR mutation was not the only Mongoloid gene that reached fixation in that region. Perhaps not surprisingly the mutation became fixed in exactly the region Dienekes' previous two posts showed the Chinese Neolithic (and probably Y-DNA O) developed. What's more it is found through southern Eurasia in exactly those regions Y-DNA O is found.
"each gene mutation must happen in a single individual."
I don't believe that is necessarily always the case. Environmental stressors might potentially cause the same identical mutations en masse in many individuals (or at least, in *more than a single person*) who happen to carry the same genetic code (such as families or tribes). It's generally a chemical reaction, after all.
Also, if the mutations are triggered by hybridization (such as that which produces Mules) -- it's just a matter of breeding.
Of course, mutations are inherited once they've been established... but that doesn't mean they must necessarily always be linked to single or even small numbers of individuals. (Perhaps they only seem to be so random...)
"The possibility that selection on a set of genetic mutations operated across the huge expanse of territory now occupied by people exhibiting what we might call 'the Mongoloid phenotype' is so unlikely as to be not worth considering..."
It would be worth considering, if you allow for the possibility of certain kinds of mutations being a lot more common, or more likely to occur, than others. And if you allow for the possibility of certain environmental factors which could effect the Human population on a more global (or at least, wide-spread) level.
Just because Science has discovered from genetic research the importance of heredity through sexual selection, doesn't mean that environmental factors have been entirely ruled out or discredited... imho.
If I am wrong, please enlighten me.
"How do you know that Q is West Eurasian? It does not seem that Native American Q has ever been in West Asia"
Q seems very likely to have originally been west Eurasian. The haplogroup splits into two at an early stage: Q1a-L472 and Q1b-L275. The second is entirely west Eurasian from South Asia, SW Asia and Europe. Q1a then splits into three: Q1a1-F1096, Q1a2-L56 and a mysterious Q-MEH2 which may be the eastern Q found in Koryaks and Eskimos. Both of the others have representatives in the west and it is not until Q1a2a-L53 appears that we get definite Siberian and American representatives. And its two closest relations (Q1a2b-L940 and Q1a2c-M323) are also west Eurasian.
"the existence of an association of Amerindians suggests that there were no 'East Asians' at 25,000 YBP"
Unlikely to be the case. The EDAR mutation appears to have become well established by that time.
"The people living in S Siberia 35,000 - 24,000 ya cannot possibly derive from Caucasians or Mongoloids, because people with those characteristics where just then in the process of forming".
Any people living in any region who were not Caucasians or Mongoloids cannot be called 'proto-Caucasian' or 'proto-Mongoloid' because they cannot have led directly to what we know as Caucasoid or Mongoloid today. Those categories are the result of expansion of the populations who, as you say, were 'just then in the process of forming'. I agree genes from ancient populations in the region survive in modern 'Caucasians' and 'Mongoloids'. That's why neither population is homogeneous.
"It is pretty clear, anatomically, that the first people moving into the Americas had no or almost no Mongoloid features (but later Na-Dene did, and Inuits the most)".
True. The major introgression of the Mongoloid phenotype into Siberia post-dates the eastward expansion of the Americans' ancestors. But those first people still appear to have some level of a Mongoloid element which gets progressively greater with continued migration.
terryt,
And we have research that tells us exactly where one of the genetic mutations involved in the formation of the Mongoloid phenotype developed, and expanded from: the EDAR370A mutation. The mutation occurred some time before anyone had reached America, and is certainly prominent there.
A more detailed breakdown of the same data can be found in figure 1 here: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0002209
As expected, within 'Europe' it's present only in North Russians and the Adygei, as consistent with recent Mongoloid admixture in those populations. It is completely absent in Western Europeans, as is consistent with a lack of Mongoloid admixture and features in those populations.
Its minor presence in the Druze is the only thing unexpected, but the Middle East is a genetic crossroads, and if we pool the Druze with the other Middle Eastern populations included, the net occurrence of EDAR is low enough to be explicable in terms of tiny recent long-distance geneflow.
The theory of 'proto-Mongoloid' admixture is thoroughly unconvincing to me also. Pontification (especially from Onur) seems to be the preferred modus operandi of its advocates, maybe with the occusional spurious example thrown in for bad measure.
It's obvious to me that a quite physically distinct population (i.e. Mongoloids) has largely displaced and/or absorbed ancient Siberians, who may have been West Eurasian, or, as Kristiina said, not really East or West Eurasian in the modern sense. Both possibilities are consistent with the fact that the 'Amerindian' signal in ADMIXTURE analyses doesn't translate well (if at all, in some cases) to 'East Asian' when an Amerindian component is lacking.
If they were a distinct population, then their features no doubt contribute to what we now consider West Eurasian, but certainly not in the way many people seem to think; i.e. not in terms of tending individuals toward what we intuitively see as 'Asiatic', which is obviously based on populations that have since expanded into Siberia and which have affected West Eurasia only to the extent of Y-haplogroup N (mostly). Considering N-carriers moving into Europe likely carried both East Asian (Mongoloid, EDAR-carrying) and ancient North Asian (non-Mongoloid, non-EDAR carrying) genes, this likely explains why ADMIXTURE may confuse the 'Amerindian' and 'East Asian' proportions in some populations (such as Finns, whose 'Amerindian' proportions often convert to 'East Asian' at different K-levels) and not in others (such as Basques, whose 'Amerindian' proportions don't convert to 'East Asian' at different K-levels).
I suppose it's natural that people should equate ancient and modern Siberians, but I called pretty much everything that people now are just beginning to realise, and physical anthropology was a major part of that.
"Each new revelation concerning this matter supports what I've been saying since August 2012 and you've been saying for much longer. But we keep making the mistake that, because we argue on the basis of fact, other people also do this".
Thanks for the acknowledgement. I mentioned your comment to a friend and he said, 'Ahh, he's on the bus'. Welcome.
@hamarfox, in what sense would Lipson's additional elements for European ancestry influence the apparent back migration of glacial emigrants having Amerindian admixtures to the west? La Brane mesolithic inhabitants appear to be closer to those original emigrants for having less to none Amerindian admixture, and still they share mtDNA with ancient Amerindian relatives in Siberia.
The origin of hg P is in Central Asia or South Asia. I have never heard that hg P would have arisen in the West. Q* is found with low frequency in India and Pakistan. Q1a is found both in the East and the West, and also Q1b is found both in the East (China) and the West (Near East), and, of course, in India and Central Asia. The deepest split of Q1a is between Koryaks and the rest. Q1a has spread to 1) China, Korea, Tibet, and Indo-China (M120), 2) West Asia (M25), 3) North Eurasia (M346), including Central Asia, Mongolia, Siberia, Scandinavia, Europe, and America and India. On the basis of this, I still insist that the ancestors of Native Americans came from Central Asia and did not go first to the west and from there again to the East.
@TerryT
""the existence of an association of Amerindians suggests that there were no 'East Asians' at 25,000 YBP"
Unlikely to be the case. The EDAR mutation appears to have become well established by that time."
Through a back migration from the Americas. So at 25,000 YBP East Asians didn't exist but Amerindians as more extreme "East Asians" did exist. Such phenotypical features as shove-shaped incisors have an east-to-west gradient between Asia and Africa/Europe and their frequencies are the highest in America pointing to the place of origin for EDAR.
We know that the same effect can be produced by very different mutations.
Stuff like fair hair or eyes for example. That may happen without linkage, but in that case with different mutations (who all cause the same effect.
Like...."fair hair" = the water tank is dry.
Why is the water tank dry? Because there is a bullet hole (mutation) that someone shot into it.
There may exist more than one dry tank in the world but then the bullet holes that damaged them may not be identical because they had been fired at different times by different pistols and different persons.
More than one water tank with an identical bullet hole shoot into an identical place may indicate that they are copies from an original water tank wich had been hit.
"I don't believe that is necessarily always the case [each gene mutation must happen in a single individual]."
I don't think you'd find many geneticists who agree with you on that.
"if the mutations are triggered by hybridization (such as that which produces Mules) -- it's just a matter of breeding".
And that's exactly what Hamarfox and I are suggesting is the case with native Americans and East Asians. And is the sort of thing that would 'cause the same identical mutations en masse in many individuals (or at least, in *more than a single person*) who happen to carry the same genetic code (such as families or tribes)'.
"mutations are inherited once they've been established... but that doesn't mean they must necessarily always be linked to single or even small numbers of individuals".
That is what is known as 'evolution', the spread of sets of genes through populations.
"doesn't mean that environmental factors have been entirely ruled out or discredited"
There is some interesting research that suggests environmental conditions can lead to particular mutations that aid survival in the particular environment. Epigenetic factors seem to be involved. If the idea eventually is shown to be correct we will be back at Lamarkian evolution; the giraffe grew a long neck to help it reach higher into the trees. Obviously the process is not as simple as that though.
Thanks for the link Hamarfox.
I still insist that the ancestors of Native Americans came from Central Asia and did not go first to the west and from there again to the East.
Kristiina,
many people would agree that P originated in the subcontinent, likely in the NW portion of it, like all descendants of F -- and that the ancestors of Native Americans originated from Central Asia/ S Siberia. But since we have the Himalayas, the question is, how did they get there?
I strongly believe there were some migrations straight from NE India into China - but not in this case. In part, because R has a very westerly distribution, but also, the trip +- straight north from Afghanistan can be the simpler one during the right climatic conditions.
If P moved East that way, Q would have simply extended that movement on its later stages - there is no large-scale back-migration of Q required except for the usual diffusion of some subgroups.
Rokus,
In what sense would Lipson's additional elements for European ancestry influence the apparent back migration of glacial emigrants having Amerindian admixtures to the west?
Because the figures don't suggest an introgression from any direction. Couple this with the fact that no large-scale admixture into Europe (or West Asia) could have occurred after the West/or North Eurasian element of Amerindians mixed with the ancestors of East Asians and picked up substantial ammounts of the EDAR variant, then we have a scenario resembling more of a directionless connection than an admixture event.
Even if we consider ADMIXTURE proportions, which do show evidence of a cline, once we account for the confusion in the signal related to N-carriers (which I detail above) it shows more of a Mesolithic/Neolithic-related cline and not really evidence of introgression.
La Brane mesolithic inhabitants appear to be closer to those original emigrants for having less to none Amerindian admixture, and still they share mtDNA with ancient Amerindian relatives in Siberia.
There are a few weird things about the La Brana results, such as the fact they plot much closer to Asians than do modern Europeans despite apparently lacking any of the Amerindian element that modern Europeans seem to possess.
One thing to consider is that, if Lipson et al.'s paper was accurate, ADMIXTURE only captures a fraction of the phenomenon, meaning that La Brana may have the admixture/element, they just have less than modern Europeans. Or, maybe they even have roughly the same levels of the element, but the same factors that disguise the admixture in Sardinians relative to Basques (who Lipson claimed had equal levels) in ADMIXTURE runs also disguise it in the La Brana samples.
Another thing to consider is that it's not the only conundrum the La Brana DNA presents. Taken individually, they seem to be about 10% SSA, and this seems mostly to be Southern African, but when they are pooled as one individual, this admixture almost disappears, so maybe we're just looking at paradoxes created by a smallish number of SNPs.
terryt wrote: "The Mal'ta individual cannot be ancestral to Native Americans as he carries Caucasoid haplogroups (Y-DNA hg R and mtDNA hg U) that are lacking in pure Native Americans and, according to the latest news, autosomally he is mainly Caucasoid with some paleo-northern Mongoloid admixture".
I think the jury is still out on whether or not R1a and/or R1b are founding haplotypes of Native Americans. Even though R1b is primarily associated with western Europe, its origin is in Asia and there are significant populations outside western Europe in which R1b is well represented, such as the Bashkirs and Turkomens. Moreover, the Turkomens are believed to have originated in the Altay area and moved farther east. I think someone should be taking a closer look at the data on North American R1 and compare it to Bashkir, Turkomen, and Siberian subclades to determine if at least some of the NA samples might be affiliated with south Asian or Siberian populations.
For what it's worth, some DNA results have been published which claim that some Cherokee women from eastern North America have tested as mtDNA types T and U5, with very few instances of typical European H in the sample. While the person who published this data has an agenda (i.e., trying to demonstrate that Cherokees are somehow related to ancient peoples described in the Christian Bible), the proportions are similar to modern Finnic populations around the Baltic. A paper mentioned in an earlier blog, "Ancient Mitochondrial DNA From Pre-historic Southeastern Europe: The Presence of East Eurasian Haplogroups Provides Evidence of Interactions with South Siberians Across the Central Asian Steppe Belt" by
Jeremy R. Newton, Grand Valley State University, also discusses generic flow backwards from Europe to southern Siberia as well, including mtDNA types T and U.
Eurologist
Do you mean that Q arose only in Siberia? I agree that it seems a good possibility. If you believe that there were some migrations straight from NE India into China - but not in this case, what haplogroup(s) do you mean? Now it appears that R had not so very westerly distribution if it was around Lake Baikal 25.000 years ago.
the figures don't suggest an introgression from any direction [...] no large-scale admixture into Europe (or West Asia) could have occurred after the West/or North Eurasian element of Amerindians mixed with the ancestors of East Asian
There are a few weird things about the La Brana results, such as the fact they plot much closer to Asians than do modern Europeans despite apparently lacking any of the Amerindian element that modern Europeans seem to possess.
To me this just looks like La Brana-like expansive elements were partly ancestral to "Asians" themselves, even before the Amerindian admixture came into play to introduce the Amerindian-admixed backmigration of their kin in northern Europe. This way it is only logical that La Brana was closer to Asians than modern Europeans. After all, also according to the Lipson study:
[...] either [Mesolithic European hunter-gatherers] or [migrants from northern or Central Asia] could be related to the “ancient
northern Eurasian” branch. Present-day Europeans differ in the amount of drift they have experienced since the admixture and in the proportions of the ancestry components they have inherited, but their overall profiles are similar.
"Q1b is found both in the East (China)"
Is it? I wasn't aware of that. However it could have come from India or Central Asia during Buddhist times, for example.
"Q1a has spread to 1) China, Korea, Tibet, and Indo-China (M120)"
According to ISOGG Q1a-M120 is Q1a1a1, a considerably downstream mutation.
"West Asia (M25)"
Also reasonably downstream: Q1a1b.
"I still insist that the ancestors of Native Americans came from Central Asia and did not go first to the west and from there again to the East".
I agree they probably 'did not go first to the west and from there again to the East'. But the deeper ancestry of Q is western and the haplogroup progressively moved east, diversifying as it went.
"their frequencies are the highest in America pointing to the place of origin for EDAR".
You'd have great difficulty trying to prove the EDAR3790A variation originated in America.
"I strongly believe there were some migrations straight from NE India into China"
So do I. Many mt-DNA M haplogroups for a start. And probably NO, even KMNOPS(perhaps xP).
@TerryT
"You'd have great difficulty trying to prove the EDAR3790A variation originated in America. "
What's so difficult? EDAR3790A is intrusive in the eastern provinces of the Old World, decreases form east to west, and is highly frequent in the New World, which was isolated for tens of thousands of millennia and could naturally have evolved a new mutation. After the retreat of the ice shield, the new mutation migrated to the Old World. "Our estimate of the time since fixation of 370A in a sample of 45 Chinese chromosomes is 10,740 years" (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0002209) which is consistent with the end of the last Ice Age.
What does the Asian origin of EDAR370A have that the New World origin doesn't?
Rokus,
To me this just looks like La Brana-like expansive elements were partly ancestral to "Asians" themselves, even before the Amerindian admixture came into play to introduce the Amerindian-admixed backmigration of their kin in northern Europe.
But this would require an Amerindian-like but not East Asian-admixed, EDAR-lacking population to have survived as late as 7000 years ago or later and to have expanded into an unpopulated Europe (e.g. no cline) or to have committed unparalleled genocide (for their age) in their expansion.
This would also have been around the time that N-carriers also moved into Europe, who were East Asian-admixed and EDAR-carriers, and also did create a very obvious cline. So in what region did the Amerindian-only population survive without contact with East Asians before their one last push into Europe and then complete disappearance? And how were they able to distribute their genes differently from Y hapologroup N carriers?
This way it is only logical that La Brana was closer to Asians than modern Europeans. After all, also according to the Lipson study:
[...] either [Mesolithic European hunter-gatherers] or [migrants from northern or Central Asia] could be related to the “ancient
northern Eurasian” branch. Present-day Europeans differ in the amount of drift they have experienced since the admixture and in the proportions of the ancestry components they have inherited, but their overall profiles are similar.
But if you look at the parameters for the six Western European populations, the upper and lower boundaries, and therefore also the midpoint (sorry, Onur :P) are almost identical. The only European outliers are the two known to be recently admixed groups: Russians and Adygei.
I agree with you, though, that the lack of the Amerindian signal in the La Brana samples is confusing. I just think that placing too much stock in them can be a mistake, since low numbers of SNPs can cause weird results.
I think the jury is still out on whether or not R1a and/or R1b are founding haplotypes of Native Americans. Even though R1b is primarily associated with western Europe, its origin is in Asia and there are significant populations outside western Europe in which R1b is well represented, such as the Bashkirs and Turkomens. Moreover, the Turkomens are believed to have originated in the Altay area and moved farther east. I think someone should be taking a closer look at the data on North American R1 and compare it to Bashkir, Turkomen, and Siberian subclades to determine if at least some of the NA samples might be affiliated with south Asian or Siberian populations.
Surely you do not think that Turkmens brought most of their R1b haplogroups from the Altai, do you (the Altai area and any region closeby probably have never been rich in R1b)? Most of Turkmens' R1b haplogroups must be from the pre-Turkic locals of where they live now (what is now Turkmenistan in their case). The same for Bashkirs.
Hamarfox, I think you lost me. Could you please imagine "Asian" as different from EDAR genes? The influence of East-Asians during the Amerindian gene flow to the west should not be an issue, possibly these were two completely different and entities. What matters is the geographic gravity of northern Eurasian peoples, that during LGM shifted east where they came in contact with Amerindian genes. Only when the gravity of northern Eurasian populations shifted back west again (I would say already much earlier than 7 kya), Amerindian admixed populations could actually become the neighbours of older non-Amerindian admixed "La Brana"-like western populations (eventually the latter may have been "pushed" forth as far as Iberia), while in Asia the Amerindian element eventually diminished or vanished in favor of new East Asian expansions in northern direction (including eg. Y-Hg N).
Mind this hypothesized shift of gravity back to the west comprise just a single chapter of a much more ancient, entangled history of back and forth movements West-East and NE-SE that could include the history of all Hg F+ as far south as Melanesia, and even Denisovan genes. This aside.
[...] for the six Western European populations, the upper and lower boundaries [...] are almost identical. The only European outliers are the two known to be recently admixed groups: Russians and Adygei.
Why shouldn't they be almost identical (in relatedness to Asian populations I guess), if all of these European populations have always been in close geographic contact? They certainly must have been already diverged before related groups moved their genes to Asia. However, apparently their relatedness over time in close contact one to the other was preserved.
"EDAR3790A is intrusive in the eastern provinces of the Old World"
Not according to the original paper that showed the similarity in effect in mice and humans. The authors showed the mutation began in northern China and expanded from there.
"decreases form east to west"
Decreases from northern China to the west and south, almost in direct proportion to the amount of Y-DNAs N, O and C3.
"is highly frequent in the New World, which was isolated for tens of thousands of millennia"
Yes. I doubt Americans have been 'isolated for tens of thousands of millennia'. Perhaps a little between ten and twenty millenia. The mutation's high proportion in America is easily explained if we assume the mutation provided some advantage in the north Eurasian habitat. It would have been selected for quite heavily as the population moved north and east into America.
"Our estimate of the time since fixation of 370A in a sample of 45 Chinese chromosomes is 10,740 years"
That age is less than half the age suggested in the earlier paper. But the paper you refer to contains only Han Chinese, most of whom are presumably Y-DNA O3 and are paert of a Neolithic expansion that began about the 10,000 year mark. To get the real age it is necessary to include all populations containing the mutation.
"What does the Asian origin of EDAR370A have that the New World origin doesn't?"
Have you checked the relevant map in the link I provided above, as well as the one Hamarfox provided? I presume not, so here it is again:
http://blogs.discovermagazine.com/gnxp/2013/02/is-girls-generation-the-outcome-of-the-pleistocene-mind/
http://en.wikipedia.org/wiki/HLA-A68
from eurogenes site.
"Could you please imagine 'Asian' as different from EDAR genes?"
I presume you mean 'East Asian' not
just 'Asian generally. What you ask for is very difficult to do. The EDAR370A looks to be one of the major defining East Asian genes. It is obviously not the only one of course but trying to imagine an East Asian phenotype that excludes the gene is impossible. And hardly a useful excercise.
"The influence of East-Asians during the Amerindian gene flow to the west should not be an issue, possibly these were two completely different and entities".
Probably not two completely different entities. The EDAR mutation is very strongly represented in Amerindians.
"Mind this hypothesized shift of gravity back to the west comprise just a single chapter of a much more ancient, entangled history of back and forth movements West-East and NE-SE that could include the history of all Hg F+ as far south as Melanesia, and even Denisovan genes".
I agree with that entirely although I think the surviving haplogroup representatives from the earliest wave are Y-DNA C/mt-DNA N rather than F. As for back and forth movemment I definitely support the idea it is very ancient. In fact I'd go as far as suggesting that our whole evolution, as well as that of other species, is a product of waves of genetic expansion from various points within the overall geographic range of the species. In fact this Dienekes post suggests such movement goes right back to when H. erectus (or whatever) first emerged from Africa:
http://dienekes.blogspot.co.nz/2013/10/d4500-and-unity-of-early-homo.html
As CleverPrimate posted in the comments:
"This pattern of similarities cannot be easily explained under a model of bifurcating change. It implies some degree of continued dispersal and mixture. I expect that the dispersal and mixture were biased in direction, probably more moving out of Africa than back into Africa. The dispersal and mixture happened episodically, not continuously. But viewed within the time-averaged record of 500,000 years, the overall appearance was one of gradual concerted evolutionary change of a highly diverse metapopulation. It should go without saying that this pattern is multiregional evolution; but it is multiregional evolution with an especially episodic rather than continuous nature".
The same pattern seems to emerge here when we consider the North Eurasian/American relationship. It is what a friend has termed 'the wave theory of evolution'.
Onur wrote:
"Surely you do not think that Turkmens brought most of their R1b haplogroups from the Altai, do you (the Altai area and any region closeby probably have never been rich in R1b)? Most of Turkmens' R1b haplogroups must be from the pre-Turkic locals of where they live now (what is now Turkmenistan in their case). The same for Bashkirs."
Both R1a and R1b are well represented in some east Asian groups such as the Uighurs. Keep in mind that most ethno-historians and genetics researchers believe that the Eurasian steppes - all the way through Mongolia - were once dominated by Indo-European speaking R1 populations until they were swept away by the expansion of the Huns, the Mongols, and the Turks. The high incidence of R1b in the Turkomans and the Bashkirs might, in fact, be the contribution of descendants of groups that were once Indo-European speakers. Indeed, Europe was sort of a refugium for IE-speaking R1 populations until the rise of the Russian Empire reversed the course of conquest. I might add that the discontinuous distribution of R1 in Siberia has parallels with haplotype C3, which disappears in the extreme NE end of Siberia only to reappear in North America.
As for the issue of R1 in North America, one observation that contradicts the theory that the presence of R1 is primarily due to recent Eurpoean admixture is that the blood group profiles of many of the populations with high incidence of R1(M173) exhibit the typically Type O-heavy distribution of Native Americans with comparatively little Type A, B, or AB, which is not what we would expect in a highly admixed population.
Unfortunately, it appears that some genetics researchers have developed a case of tunnel vision, arbitrarily discarding data points because they don't fit with the "Out of Beringia" model.
@TerryT
"The authors showed the mutation began in northern China and expanded from there."
They didn't even consider America, as their America-less heat map indicates. But if you look at the heat map closely, you'll see that they placed the origin of the mutation right where the frequency of EDAR370A is the highest in Asia. It's central China, indeed. But North America shows the same high frequency of EDAR370A (full red circle), and with much greater uniformity (as we know from other studies). The ancestral allele is found in Central and South America suggesting that the ancestral allele had a history in the Americas just as it did in the Old World, but then EDAR370A likely originated in America and migrated back ad expanded in the America, too. This is precisely why we see the ancestral allele almost disappearing in the Americas because the populations there were the closest to the epicenter of the derived allele.
"Decreases from northern China to the west and south, almost in direct proportion to the amount of Y-DNAs N, O and C3."
Agree that decreases from north to south (America is another example) and from east to west. The absence of N and O in the Americas militates against the migration of EDAR to America but it's consistent with the migration of EDAR out of America with subsequent evolution of Y-DNA N and O in Asia as EDAR spread west and south.
"That age is less than half the age suggested in the earlier paper. But the paper you refer to contains only Han Chinese... To get the real age it is necessary to include all populations containing the mutation."
The authors of the mice paper postulate the origin of EDAR among Han Chinese, so adding more populations shouldn't affect the age of the mutation. The age of 30,000 is inflated and is probably serves the purpose to accommodate the ever-growing evidence for human antiquity in the Americas. The authors are aware of the fact that they have to explain EDAR470A in the Americas, hence the date can't be 10,000 years.
"http://blogs.discovermagazine.com/gnxp/2013/02/is-girls-generation-the-outcome-of-the-pleistocene-mind/"
Terry, why don't you stop visiting Razib's MinuteClinic and actually read the paper itself?!
Do you mean that Q arose only in Siberia? I agree that it seems a good possibility. If you believe that there were some migrations straight from NE India into China - but not in this case, what haplogroup(s) do you mean? Now it appears that R had not so very westerly distribution if it was around Lake Baikal 25.000 years ago.
Kristiina,
Yes, I think it's quite likely Q arose in Siberia.
In addition to what Terry mentions, perhaps y-DNA HG D (-> DM55).
As to R, what I meant is: sufficiently far west to make an entry of P west of the Himalyas more likely.
I just wonder if the whole R initially originated in Siberia.
Valikhan,
I think the diversity or R on the subcontinent is two high, and that outside too low for that to be the case. But an origin in the NW portion of the subcontinent, near where I suspect P arose, seems quite likely. R1a and R1b sub-branches could have migrated N and NW just before the Gravettian, and mixed with the earlier (pre-Mongoloid) Siberian population, there.
Both R1a and R1b are well represented in some east Asian groups such as the Uighurs. Keep in mind that most ethno-historians and genetics researchers believe that the Eurasian steppes - all the way through Mongolia - were once dominated by Indo-European speaking R1 populations until they were swept away by the expansion of the Huns, the Mongols, and the Turks. The high incidence of R1b in the Turkomans and the Bashkirs might, in fact, be the contribution of descendants of groups that were once Indo-European speakers.
Modern Uyghurs are a population of the Tarim Basin, and the Tarim Basin has nothing to do with the Altai region and Turkmen origins. The Altai region and environs have very little, if any, R1b, and according to the ancient DNA evidence the situtation was no different as regards to the amount of R1b in the past. R1b haplogroups of Turkmens and Bashkirs must be overwhelmingly from places much further west such as what is now Turkmenistan and Bashkiria respectively as there is no population so high in R1b on the routes taken by the Turkic ancestors of Turkmens and Bashkirs and in the probable Turkic homelands.
As a side note, the Tarim Basin is not in East Asia, as it is a part of Central Asia.
Lastly, I am not talking about R1a but R1b. R1a is much more represented than R1b in the eastern parts of Central Asia (no doubt a legacy of Indo-Europeans).
The EDAR370A looks to be one of the major defining East Asian genes. It is obviously not the only one of course but trying to imagine an East Asian phenotype that excludes the gene is impossible.
Then what again about this tweet? Wllierslev: 24,000-yr-old Siberian Mal'ta person geneticall similar to native amer and west eurasians. No east asia
Maybe I am dreaming, but stylistically the Mal'ta - Buret' venuses in Siberia (http://donsmaps.com/malta.html) aren't that far removed from the European fat-ass figurines found from Hohle Fels up to Willendorf. Except for the hair and breasts, that in Siberia appear heavily influenced by the unique effects of EDAR370A! If so, this would support the (paleo) Amerindian identity of the gene rather than being truly East Asian. The puzzle is complicated, though, since the selective advantage of EDAR370A is often attributed to more abundant sweat glands, rather typical of warmer, more southern climates. This kind of details makes me wonder if East Asians aren't really quadruple hybrids and altogether anachronistic to the north eaurasian (i.e. mixed west earasian/amerindian) issue here.
Interesting piece from Coon's Racial Adaptations book:
Let us return to Professor Thoma's theory about Neanderthal's relationship to the living Mongoloid race. With the help of recent Russian discoveries , he has traced the Near Eastern Neanderthal line from a skull found in the cave of Jebel Amud in northern Israel, through Shandar in Iraqi Kurdistan, to Teshik Tash in Uzbekistan, and then as far north and east as Afonta Gora, near Krasnoyarsk, in Siberia. Afonta Gora is at 56.05 degrees north latitude, 0.20 degrees farther north than Moscow. Its date is 20,000 years ago, according to Russian carbon 14 analysis, placing it at the height of the last Wurm glacial advance in Europe. Siberia was then partly forested and partly covered with tundra, and the access to America over the Bering Strait was bridged and free of ice. Access to the rest of North America was clear as far south as about 63 degrees south latitude.
Archaeological evidence indicates that some of the pioneers moved across the Bering Strait before the land bridge had been covered by the melting of the polar ice, but others went southeastward in Siberia, as indicated by the skulls found at Mal'ta and Buret, in the Irkutsk District, just north of Lake Baikal and close to the home of the Buryats, the most extremely Mongoloid people yet studied.
During this journey, the Near Eastern Neanderthal had become increasingly cold-adapted until they had acquired a Mongoloidal face form. That face form includes reduced brow ridges, which are vulnerable to frostbite; facial flattening, which makes the nasal passages internal rather than beaky; and fat, narrrow-slitted, double eyelids, which protect their flush eyeballs from chilling."
"Terry, why don't you stop visiting Razib's MinuteClinic and actually read the paper itself?!"
As far as I'm aware the paper is still pay per view. If you can provide a link to the paper please do so.
"They didn't even consider America, as their America-less heat map indicates".
The authors didn't consider the possibility of the mutation having arisen in America because it doesn't fit without major manipulation of the facts. If, as you claim, the mutation arose in America you have to concoct some complicated theory in order to explain why it has reached fixation in the Northern Han but is more thinly spread through the rest of East Eurasia.
"The absence of N and O in the Americas militates against the migration of EDAR to America but it's consistent with the migration of EDAR out of America"
No. It is consistent with the mutation having been introduced by the female line (mt-DNAs A, C and D, and possibly through B as well) into the mixed population that entered America. Selection increased its presence in that population.
"Another point, races are not defined based on haplogroups but based on overall genetics".
Yes. Haplogroups can introgress into resident populations. They are an indication of population spread though, and probably provide a good clue as to the genetic combination of the haplogroup's source population.
"Native Americans are racially Mongoloid, not hybrid".
Of the many Native Americans I've met very few of them look to be more than just a little Mongoloid.
"24,000-yr-old Siberian Mal'ta person geneticall similar to native amer and west eurasians. No east asia"
The haplogroups of the two populations are completely different. That is hard to fit with the quote.
Native Americans are racially Mongoloid, not hybrid.
Onur,
I agree in the case of Inuits and Na-Dene speakers, but not with indigenous peoples from Central and South America. While they appear to have some Mongoloid (or strong proto-Mongoloid) admixture, many look more like Pacific Islanders, others most closely resemble Tibetans.
As to the sweat glands, I don't see a contradiction to cold adaptation. We all have sufficient sweat glands to cover our skin with water when it is hot and humid. More sweat glands would only reduce the time before that happens, which in people with rounder shape, reduced limb and neck dimensions, and more subcutaneous fat could prevent heat stroke under sudden, strenuous activity.
Modern Uyghurs are a population of the Tarim Basin, and the Tarim Basin has nothing to do with the Altai region and Turkmen origins.
and has nothing to do also with the Bashkir origins
Onur:
"Modern Uyghurs are a population of the Tarim Basin, and the Tarim Basin has nothing to do with the Altai region and Turkmen origins."
Uyghurs in the Tarim Basin arrived there ca 842 CE, after the Kyrgyz invasion destroyed the Uyghur Khaganate. These refugees were originally from Mongolia.
"As a side note, the Tarim Basin is not in East Asia, as it is a part of Central Asia."
Inner Asia might be more appropriate. Portions of the Tarim were first brought under Chinese control in the second century BCE, but Chinese cultural influence, and quite possibly some Chinese people, had been present since at least 1000 BCE.
Uyghurs in the Tarim Basin arrived there ca 842 CE, after the Kyrgyz invasion destroyed the Uyghur Khaganate. These refugees were originally from Mongolia.
The majority of modern Uyghur ancestry (including the majority of modern Uyghur R1b haplogroups) is from the pre-Turkic populations of the Tarim Basin. Also the Tarim Basin had already some yet-do-be-determined level of Mongoloid admixture before the Turkic (original Uyghur) migration to the Tarim Basin. Finally, what is now Mongolia and environs have never been rich in R1b.
Then what again about this tweet? Wllierslev: 24,000-yr-old Siberian Mal'ta person geneticall similar to native amer and west eurasians. No east asia
I understood this to mean that the Mal'ta boy was genetically related to modern West Eurasians and the West Eurasian portion of modern Native Americans. All the previews of Willerslev's upcoming paper that I've read seem to suggest this.
Only when the gravity of northern Eurasian populations shifted back west again (I would say already much earlier than 7 kya), Amerindian admixed populations could actually become the neighbours of older non-Amerindian admixed "La Brana"-like western populations (eventually the latter may have been "pushed" forth as far as Iberia), while in Asia the Amerindian element eventually diminished or vanished in favor of new East Asian expansions in northern direction (including eg. Y-Hg N).
Why shouldn't they be almost identical (in relatedness to Asian populations I guess), if all of these European populations have always been in close geographic contact? They certainly must have been already diverged before related groups moved their genes to Asia. However, apparently their relatedness over time in close contact one to the other was preserved.
If we take aDNA ADMIXTURE results at face value, then the Amerification of Europe was ongoing even during the Neolithic and after. Oetzi had a bit of the Amerindian signal at k=4 (0.8%), but less than modern Italians, including Southern Italians and even Sardinians. So we'd have to conceive of a movement of one mesolithic group displacing not only another, but also even Europe's highly successful Neolithic and Neolithic/Mesolithic mixed populations. Also note that, considering how late the Amerification process seemingly continued, then if diffusion of these genes within Europe occurred to the extent of masking any original clines, then the contemporaneous or earlier clines forged by the Neolithic migrations should also have been similarly scrambled by the same process. But they weren't.
Perhaps Lipson's signal was an expression of the fact that Amerindians are uniformly related to unadmixed Europeans, and so, when the mistake is made of assuming Amerindians to be an unadmixed population, a uniform signal is instead projected into Europe.
Or it could mean that all Europeans and, as far as I can tell, also West Asians and ancestral North Indians derive, through shared ancestry with one another, roughly equal proportions of their ancestry from the same 'Amerindian-related' population. But I don't see much evidence of an admixture event within Europe after the ancestors of various European populations had already split.
As for the ADMIXTURE signal, which is a little different from Lipson's, I suppose it's possible that a population with stronger Amerindian signals (but not necessarily admixture) moved into Europe after the Neolithic. Logic points to R1 carriers. After all, these possibly arrived later than the Neolithic, and maybe, owing to a closer relationship with Q, showed a stronger amer. signal without having the evidence of real admixture we should see on global plots of modern R-influenced Europeans relative to La Brana if an actual major admixture event occurred.
But even then, the Amerindian signal at k=4 corresponds much more to Mesolithic proportions than to levels of R1a & b. For instance, Scandinavians have more of the Amerindian component than Britons or Basques, and still have no less than Britons even when accounting for more recent N-related stuff in Scandinavians -- and all this despite having much less R1.
None of these inconsistencies can be resolved until we look at the other side of the equation: namely, admixture in Amerindians.
@TerryT
"As far as I'm aware the paper is still pay per view. If you can provide a link to the paper please do so."
You can go to my site and download it from http://anthropogenesis.kinshipstudies.org/2013/10/web-gems-october-30-2013/
" If, as you claim, the mutation arose in America you have to concoct some complicated theory in order to explain why it has reached fixation in the Northern Han but is more thinly spread through the rest of East Eurasia. "
If EDAR370A arose in Asia and migrated to America, how did it raise to fixation all over North America? That's a bigger problem that its fixation in Han. If the Amerindian origin of the mutation is assumed, then it's easy to see that its patchy distribution in the Old World is consistent with an incoming smaller population carrying the mutation admixing with some larger pre-existing populations.
"No. It is consistent with the mutation having been introduced by the female line (mt-DNAs A, C and D, and possibly through B as well) into the mixed population that entered America. Selection increased its presence in that population. "
That's a new theory, right? Remember that blood group B also very frequent in Han Chinese and all over Asia is almost non-existent in the Americas. If EDAR moved to the New World from Asia, we would have seen blood group B there. But we don't, and you can't explain it's lack in the Americas with some uniparental transmission of blood groups.
Part of this what you describe as "ongoing Amerification" in Europe must be related to the apparent Middle Neolithic Northern European expansion I wrote about recently, as eg. Di Gaetano’s team established that even in the Sardinian population the average admixture proportions for Northern European ancestry is nowadays as high as 14.3%. French, Northern Italian, Central Italian and South Italian populations show a NE ancestry of 70%, 56%, 52% and 43,6% respectively. Indeed, Amerindian-related gene flow should have much older roots to explain this vaporization of clines, though your statement that all clines have been erased is too confident. These clines still exist, and were much more pronounced before the Middle Neolithic as testified by La Brana and still even by Ötzi.
Lipson's ancient northern Eurasian component is about equivalent to the common ancestor of 'Americans', apparently still pronounced in Russians and Yakut but dropping steep especially in (other) Asian provinces. The East Asian component just can't be taken responsible for the measured admixtures among Europeans, while apparently the Amerindian element was already West-Eurasian admixed in Mal'ta. However, "common ancestor of Americans" is just not good enough as a distinctive group to tell europeans apart from another common Eurasian element, whose origin could as well have been, at least partly, in Europe itself - as suggested by the combined evidence from La Brana and Mal'ta. This makes Lipson's analyses less useful to really clarify, or mystify, Europe's constituent genetic components. Let us rather focus on the paleogenetic evidence.
Onur,
Compare to the Ainu. Even when less admixed, 100 years ago or so, the women looked Mongoloid, while the men did not. Today, most Ainu look rather Mongoloid.
Sure, after 20,000 or so years, things are hard to tell - but IMO many indigenous Central and South American people look like they only carry a subset of Mongoloid features.
Of course, another way of saying that is that today's NE Asian Mongoloids are highly derived and those features are only reflected in Inuits and Na-Dene, while the remainder of natives only carry proto-Mongoloid features (and even those not necessarily to 100%).
"when the mistake is made of assuming Amerindians to be an unadmixed population, a uniform signal is instead projected into Europe".
I think that is exactly the problem we have here. We are seeing in many comments a huge emotional commitment to the belief Amerindians are a 'pure' population. What is the philosophical position that demands such a belief?
"You think that the more derived branch (modern Asian Mongoloids) is more representative of the original Mongoloid characteristics than the more conservative branch (modern Native Americans) is".
I would assume that to be the default position. What selection process do you propose led to modern Asian Mongoloids being 'more derived' than 'original Mongoloids'? I can certainly think of no possible selction pressure. And what leads you to claim Native North Americans represent 'the more conservative branch'.
"Like many, you have a biased view of the Mongoloid physical characteristics".
Onur, I think it is you who have the unrealistic view of the relationship between Eurasian and American populations:
"I agree in the case of Inuits and Na-Dene speakers [Native Americans are racially Mongoloid], but not with indigenous peoples from Central and South America. While they appear to have some Mongoloid (or strong proto-Mongoloid) admixture, many look more like Pacific Islanders, others most closely resemble Tibetans".
Exactly.
"You can go to my site and download it from http://anthropogenesis.kinshipstudies.org/2013/10/web-gems-october-30-2013/"
Thanks very much for that.
"If EDAR370A arose in Asia and migrated to America, how did it raise to fixation all over North America? That's a bigger problem that its fixation in Han".
Simple. Environmental selection along the route. From the paper:
"Spatially explicit simulation, haplotype, and maximum likelihood analyses suggest that 370A originated once in central China more than 30,000 years BP with
a selective coefficient that is one of the highest measured in human populations. Our results are consistent with previous inferences that 370A must have arisen prior to 15,000 BP (Bryk et al., 2008; Peter et al., 2012) and the first peopling of the Americas (Goebel et al., 2008; O’Rourke and
Raff, 2010) but also suggest that the allele likely emerged in East Asia even earlier".
That is a pretty good summary.
"That's a new theory, right?"
No. Haven't we known for ages that the Amerindian mt-DNA was East Asian?
"If EDAR moved to the New World from Asia, we would have seen blood group B there. But we don't, and you can't explain it's lack in the Americas with some uniparental transmission of blood groups".
Once one is prepared to accept Amerindians derive from a hybrid population the difficulty you see disappears entirely.
Sorry. Me again. I've just found time to follow the Michael Balter link. He doesn't wriggle around the topic of hybrid formation:
"The finding suggests that about a third of the ancestry of today's Native Americans can be traced to 'western Eurasia,' with the other two-thirds coming from eastern Asia, according to a talk at a meeting here by ancient DNA expert Eske Willerslev of the University of Copenhagen".
But wait, there's more:
"The team proposes a relatively simple scenario: Before 24,000 years ago, the ancestors of Native Americans and the ancestors of today's East Asians split into distinct groups. The Mal'ta child represents a population of Native American ancestors who moved into Siberia, probably from Europe or west Asia. Then, sometime after the Mal'ta boy died, this population mixed with East Asians. The new, admixed population eventually made its way to the Americas. Exactly when and where the admixture happened is not clear, Willerslev said. But the deep roots in Europe or west Asia could help explain features of some Paleoamerican skeletons and of Native American DNA today".
Davidski has some interesting maps in his blog on the subject. He suggests megafauna hunters occupied much of the mammoth-steppe of northern Eurasia and account for the commonality between European and American genes:
http://eurogenes.blogspot.co.nz/2013/10/surprising-adna-results-from.html
Onur:
"Also the Tarim Basin had already some yet-do-be-determined level of Mongoloid admixture before the Turkic (original Uyghur) migration to the Tarim Basin."
I am pretty sure this was strongly implied in my comment and you're tilting at windmills.
I am just going to quote from the Science preview article with few comments - everything matches the interpretation I have voiced before.
"The finding suggests that about a third of the ancestry of today's Native Americans can be traced to western Eurasia..."
"They found that a portion of the boy's genome is shared only by today's Native Americans and no other groups."
That is, Europeans, Western Eurasians, and Native Americans share ancestry that at least in part did not come from nor make it into extant NE Asians.
"The boy's genome showed no connection to modern East Asians."
That is, at most proto-Mongoloid attributes - if at all.
"The west Eurasian [genetic] signatures that we very often find
in today's Native Americans don't all come from postcolonial admixture"
Of course not.
Also, carbon dating confirms 24 kya.
Onur wrote: "The majority of modern Uyghur ancestry (including the majority of modern Uyghur R1b haplogroups) is from the pre-Turkic populations of the Tarim Basin. Also the Tarim Basin had already some yet-do-be-determined level of Mongoloid admixture before the Turkic (original Uyghur) migration to the Tarim Basin. Finally, what is now Mongolia and environs have never been rich in R1b."
From what I have read, the pre-Uighur populations were almost exclusively R1a with no mention of R1b at all. Either the Turkic core of the Uighurs originally had a considerable proportion of R1b, or they acquired it by absorbing relic formerly IE-speaking populations in the course of their migration to the Tarim basin.
Central Asia has seen massive amounts of population replacement and displacement in historic times. There is also no doubt that both R1a and R1b originated in Central or Southern Asia and spread to Europe, and not vice versa. At any rate, R1b is found in presently Turkic-speaking populations of both extremes of Central Asia, and was probably common all throughout the region before the Mongol and Turkic expansions. (The expression "Turkoman", BTW, indicates that the group was not regarded as one of the traditional Turkic tribes, and implies that the group was assimilated from non-Turkic groups.) Here is a link to a handy page consisting of pie charts showing Y-haplogroups of various Turkic and other populations: http://s155239215.onlinehome.us/turkic/60_Genetics/TurkicGeneticsGraphs.htm
In Greenberg's Three Migration model of the peopling of the Americas, the second wave consisted of the ancestors of current Na-Dene speakers to North America and brought R1 to the Americas. I would amend this model to include the Iroquoian and Algonquian-speaking groups, which also exhibit high levels of R, or assign them to a separate wave, and suggest that they might have originated from deeper in Siberia than the Lake Baikal area.
hamarfox wrote: "If we take aDNA ADMIXTURE results at face value, then the Amerification of Europe was ongoing even during the Neolithic and after. Oetzi had a bit of the Amerindian signal at k=4 (0.8%), but less than modern Italians, including Southern Italians and even Sardinians. So we'd have to conceive of a movement of one mesolithic group displacing not only another, but also even Europe's highly successful Neolithic and Neolithic/Mesolithic mixed populations. Also note that, considering how late the Amerification process seemingly continued, then if diffusion of these genes within Europe occurred to the extent of masking any original clines, then the contemporaneous or earlier clines forged by the Neolithic migrations should also have been similarly scrambled by the same process. But they weren't."
The Amerification of Europe may be a result of Indo-Europeanization. The Amerindian signal in modern Italians may be related to to the fact that the Latin word for water, 'aqua' resembles the Nottaway (a coastal Native American language in North America) equivalent 'awa', and the word for 'I' in Latin, 'ego', resembles the equivalent in Nottaway and Mohawk, 'i'. Many linguists believe that the IE resulted from a mix of two different languages, one of which was Central Asian. This Central Asian language apparently shared a common origin with the Iroquoian family of North America. It is therefore highly likely that the original Indo-Europeans included a component from a population that also gave rise to the second wave of Native American migration in Greenberg's model of the peopling of the Americas which included Y haplogroup R1.
@TerryT
"Simple. Environmental selection along the route."
The same argument works the other way, too. But there's no specific evidence for either, though.
"a selective coefficient that is one of the highest measured in human populations."
The authors don't show the other populations with highest selective coefficients. They must be in America, they just don't show them.
"Haven't we known for ages that the Amerindian mt-DNA was East Asian?"
Mal'ta DNA suggests that there were no East Asian markers in Northeast Asia at 25,000 BP but the Mal'ta population still showed proximity to Amerindians. It's easy to see that Amerindians is an older population than East Asian and since there's indeed a number of matches between East Asian and Amerindians, it's the former who must have descended from the latter.
"Once one is prepared to accept Amerindians derive from a hybrid population the difficulty you see disappears entirely."
What does it have to do with blood groups? And also Amerindians, according to recent research are the "admixers" in West Eurasia and not "admixees." Same works for East Asians.
Terry,
I think that is exactly the problem we have here.
That's my bet too. We don't even have to wait for Willerslev's paper to see a glaring mistake in Lipson's model that hints at the above, since it considers the bonafide East Asian admixture in North Russians and Agygei to be of the same 'Ancient North Eurasian' signal it detects universally in Europeans. It simply assigns those populations slightly increased proportions consistent with their higher recent East Asian ancestry relative to the other European groups.
This can only be explained if the algorithm wrongly takes the East Asian admixture in Amerindians and designates the signal 'ANE' on the logic that those genes are in Amerindians and that Amerindians are unadmixed. Very likely the same thing occurs with the West Eurasian signals in Amerindians: the algorithm assigns the portion of European ancestry most closely related to that in Amerindians the designation 'ANE' too.
I'm not even entirely convinced any North Eurasian population ever existed. It's quite possibly just the result of a statistical compromise between reality and some deeply faulty assumptions.
Gary Moore,
The Amerification of Europe may be a result of Indo-Europeanization. The Amerindian signal in modern Italians may be related to to the fact that the Latin word for water, 'aqua' resembles the Nottaway (a coastal Native American language in North America) equivalent 'awa', and the word for 'I' in Latin, 'ego', resembles the equivalent in Nottaway and Mohawk, 'i'. Many linguists believe that the IE resulted from a mix of two different languages, one of which was Central Asian. This Central Asian language apparently shared a common origin with the Iroquoian family of North America. It is therefore highly likely that the original Indo-Europeans included a component from a population that also gave rise to the second wave of Native American migration in Greenberg's model of the peopling of the Americas which included Y haplogroup R1.
There seems to be logic in pursuing any possibility for the origin of the signal, but also logic against the same possibilities.
I agree with you that the R1 connection makes sense, but if we look at the k=4 Dodecad calculator, some R1-depleted populations (e.g. Scandinavians) have more of the signal than R1-heavy populations, like the Basques, and the Amerindian proportions seem much more influenced by proportions of Mesolithic and Neolithic ancestry.
Moreover, if we look at the 'deepest' attempt to quantify the signal, which is Lipson's paper, we find something else.
I took a SS for you. The values are of Amerindian-centered ancestral components in Sardinians and Basques:
http://i674.photobucket.com/albums/vv103/camelsloop/hgjh.png
We see an R1-heavy population (Basques) and an R1-light population (Sardinians) with basically equal proportions. We also see a mesolithic-light/absent population (Sardinians) and a moderately mesolithic population (Basques) being undifferentiated in deep ancestry, which hurts the two most reasonable explanations for the West Eurasian end of the signal. This is one of the reasons I think we're chasing ghosts here, as I mentioned to Terry above.
With ADMIXTURE, the Amerindian component likely suffers from a Kalash effect, which creates a similar effect to Lipson's model, where it projects its own heterogeneous make up outwardly as a homogeneous element in other populations. If that's the case, then perhaps mesolitic Europeans are the most strongly related to the West Eurasian-like component in Amerindians, hence a stronger signal in them.
That doesn't discount any linguistic connection, though, since traces of a common ancestral language or family of languages could easily have survived in some groups and not others.
thank you Mr Moore
"The Amerification of Europe may be a result of Indo-Europeanization. The Amerindian signal in modern Italians may be related to to the fact that the Latin word for water, 'aqua' resembles the Nottaway (a coastal Native American language in North America) equivalent 'awa'"
finally some one is sneaking up on it..
and so also Porteugese and Periqui and Por(t) agui's ( as related by the ones who are laughed at for not being real enough natives called "melungeons" ) and what is that the mexican ( tribe on baja) that was just wiped out a few(20 or so ) years ago it is also a corruption Port aqua / these are the Port going sea/water people . these are all parts of the sea tribes of the western world that the religion of science wishes would remain buried in some watery grave and as some kind of underworld myth .
This paper ”Y-chromosome distributions among populations in Northwest China identify significant contribution from Central Asian pastoralists and lesser influence of western Eurasians” is very interesting.
There you can see that Northwest China is very diverse. YDNA C is typical of Mongolians and Kazakh. YDNA D is typical of Yugu. YDNA G is found in Tajike and Ozbek. YDNA H is only found in Uygurs. Tataer have almost exclusively yDNA I and R1a1-M17. J2 is found in some groups, but in particular in Uygurs, Tajike and Ozbek. N-M231 is found in small quantities in almost every group. O*-M175 and O3-M122 are also found in almost every group. O1 is found mainly in Tu and Xibo. O2 is typical of Bao’an. Q* (xM120) is found mostly in Dongxiang. Q-M120 is found in low frequencies in Mongolians, Uygurs and Russ. R* is found in particular in Tajike. R1*-M173 is found in Yugu. R1a1-M17 is found in every group except Yugu, but the highest frequencies are in Tataer, Uygur, Kirghiz, Tajike, Dongxiang and Salar. R1b is rare, and the highest frequency is in Salar.
It seems that Tataer could be linked to Westerners, and it would be interesting to know what their I haplotype is. Uygurs seem to have had Silk Road connections with India. Tajike and Ozbek seem to be more West-Eurasian with yDNA G and J2. R1a1, N and O are found everywhere. The frequency of Q is patchy and quite low. Also R1b is rare in the area. Yugu D is interesting, in particular if it is old in the area.
In my eyes, the most western group is Tataer.
It would be interesting to know if this Mal’ta R is close to Tajike R* or Yugu R1* or to something else.
"I'm not even entirely convinced any North Eurasian population ever existed. It's quite possibly just the result of a statistical compromise between reality and some deeply faulty assumptions".
This whole discussion supports a diagram I placed in one of my essays from 2009. Hopefully some will find it useful to refer to it:
http://humanevolutionontrial.blogspot.co.nz/2009/06/human-evolution-on-trial-human-star.html
In the diagram I show 'America' as a subpoint on the human star. At the time I couldn't think of a term to cover the middle of the star but if I were to do the diagram today I would take note of what Va_Highlander said a while ago: 'Inner Asia might be more appropriate'. I would call the middle of the star 'Inner Asia'. In the essay I made this comment:
"For any species the individuals at the geographical extremities of its distribution will be the most different. This has been recognised for a long time (Mayr and Diamond 2001). From the middle genes or ripples can flow in many directions but at the points gene flow reaches a dead end (if it reaches the point at all). On the other hand while ripples can flow from a point into the middle of the star or to a neighbouring point they are less likely to reach a more distant point".
The non-existent 'North Eurasian population' is probably an 'Inner Asian' population. Probably something to do with Davidskis steppe hunters.
"Europeans, Western Eurasians, and Native Americans share ancestry that at least in part did not come from nor make it into extant NE Asians".
There it is. Inner Asians. East Asia received only minimal population from that region.
"That is, at most proto-Mongoloid attributes - if at all [the boy's genome]".
Specifically: none. The Mongoloid element is what makes Native Americans different from him.
"Terry, as I said countless times, purity is relative. Races are not pure entities, they exchange genetic material with neighboring races, if there are any, through mixing"
And that is exactly what my human star diagram explains.
"Polynesians in general seem to be Mongoloid-Australoid hybrids"
Yes. They make up another of the human star subpoints. Closer to mainland SE Asia the population forms a steep cline with the Australoids/Papuans.
"The same argument works the other way, too. But there's no specific evidence for either, though [environmental selection]".
But there is evidence. The EDAR mutation exhibits no possible centre of expansion within America, just a scatter of varying amounts, whereas it very much does show a centre in East Asia.
"The authors don't show the other populations with highest selective coefficients. They must be in America, they just don't show them".
Their diagram 1b does show the mutation reaches a very high level in parts of America though.
"Mal'ta DNA suggests that there were no East Asian markers in Northeast Asia at 25,000 BP"
No it doesn't. It shows there were none at Mal'ta. Besides which the proto-American population may have picked up those haplogroups more recently than 25,000 years ago, before they came close to America.
"the Mal'ta population still showed proximity to Amerindians".
Yes. They both share that 'Inner Asian' element.
"It's easy to see that Amerindians is an older population than East Asian"
Really?
From what I have read, the pre-Uighur populations were almost exclusively R1a with no mention of R1b at all. Either the Turkic core of the Uighurs originally had a considerable proportion of R1b, or they acquired it by absorbing relic formerly IE-speaking populations in the course of their migration to the Tarim basin.
All of the existing ancient Y-DNA haplogroup results from the Tarim Basin are from about 4000 years ago. So there is about a 3000-year period immediately preceding the Turkicization of the Tarim Basin for which we do not have any info about the Y-DNA haplogroups of the Tarim Basin. 3000 years is a long enough time for the transformation of the Tarim Basin from a R1a-dominant region to a region where R1a is a little above in frequency than R1b. So no need to explain the increase in the frequency of R1b with the Turkic migration, which would in any case be nonsensical as R1b is in very low frequencies in the probable Turkic homelands (including what is now Mongolia, the region where the original Uyghurs came from) and probably has always been so. R1b in the Tarim Basin must already have come to the levels similar to the current levels during the 3000-year period prior to the Turkicization.
There is also no doubt that both R1a and R1b originated in Central or Southern Asia and spread to Europe, and not vice versa.
West Asia is another probable source of origin of those haplogroups (this is especially true for R1b).
The expression "Turkoman", BTW, indicates that the group was not regarded as one of the traditional Turkic tribes, and implies that the group was assimilated from non-Turkic groups.
No, the expression "Turkmen" was originally used for the Islamized Turkic peoples in general but gradually came to be only used for the Islamized Oghuz among all the Islamized Turkic groups, and, no, the Oghuz are among the traditional Turkic tribes. Consequently, the expression "Turkmen" cannot imply being assimilated from non-Turkic groups. Rather, it seems to have a meaning related to Islamization.
Onur:
"Also the Tarim Basin had already some yet-do-be-determined level of Mongoloid admixture before the Turkic (original Uyghur) migration to the Tarim Basin."
I am pretty sure this was strongly implied in my comment and you're tilting at windmills.
Va Highlander, I did not say that you did not imply that. Read my post again, and not just that part but as a whole.
This paper ”Y-chromosome distributions among populations in Northwest China identify significant contribution from Central Asian pastoralists and lesser influence of western Eurasians” is very interesting.
That paper is pay per view. How can I find that paper for free?
Onur, send me an email to the address sanava@pp.inet.fi.
@Kristiina,
Thanks for the sharing. I have just sent you an email.
@TerryT
""Mal'ta DNA suggests that there were no East Asian markers in Northeast Asia at 25,000 BP"
No it doesn't. It shows there were none at Mal'ta. Besides which the proto-American population may have picked up those haplogroups more recently than 25,000 years ago, before they came close to America. "
Admittedly, Ma'ta is our only direct reference point of that antiquity in that region but that's all we got. Ma'ta is smack where we would expect East Asians to be at 25,000 YBP and instead we find a connection to Amerindians.
""It's easy to see that Amerindians is an older population than East Asian"
Really?"
Really. It's simple logic and it fits perfectly with ALL the cultural evidence we have.
"The EDAR mutation exhibits no possible centre of expansion within America, just a scatter of varying amounts, whereas it very much does show a centre in East Asia. "
They derived the center from the highest frequencies and the highest frequencies are found in abundance in the Americas. They just don't show East Asia directly compared to North America.
"Their diagram 1b does show the mutation reaches a very high level in parts of America though."
Duh. That's what I'm saying.
German Dziebel wrote:
"Mal'ta DNA suggests that there were no East Asian markers in Northeast Asia at 25,000 BP but the Mal'ta population still showed proximity to Amerindians. It's easy to see that Amerindians is an older population than East Asian and since there's indeed a number of matches between East Asian and Amerindians, it's the former who must have descended from the latter."
In the Berinigian Hypotheses for the founder populations of the the Americas, the Beringian population was formed by the merger of two groups: an interior Siberian population which entered the region from the west and a coastal population moving north along the coast. The interior Siberian population probably corresponded to the population represented by the Mat'ta and the coastal stream probably was more "East Asian", accounting for the observed pattern seen in Native American populations today.
Hamar Fox wrote:
"We see an R1-heavy population (Basques) and an R1-light population (Sardinians) with basically equal proportions. We also see a mesolithic-light/absent population (Sardinians) and a moderately mesolithic population (Basques) being undifferentiated in deep ancestry, which hurts the two most reasonable explanations for the West Eurasian end of the signal. This is one of the reasons I think we're chasing ghosts here, as I mentioned to Terry above."
The Basques are an interesting case. I've not been able to do as much research on this language, but there may also be links to attested Native American languages. Various linguists have proposed a link between Basque and the Na-Dene languages of North America. However, I've found some interesting points of similarity to the Algonquian languages. For instance, the word for "I" in Basque and Lenape, an eastern North American Algonquian language, is "ni". The Basque words for "knife" (aizto), "ax" (aizkora) and "hoe" (aitzur) could be related to Algonquian words for stone, such as "asin" in Anishinaabemowin, another Algonquian language.
Another possible explanation for the introduction of Native American DNA into neolithic Europe might be via the transatlantic route. Some archeologists have suggested that similarities between Archaic North American cultures and cultures of western Europe in the same time frame may have been as result of transatlantic contact. Normally, this would be the stuff of crackpot science were it not for the following facts:
- The deep water navigation skills of Archaic North Americans is attested in the archeological record. Archaic North Americans pursued Great Auks and swordfish in open water, and traded up and down the East Coast of North America. Native American techniques for building substantial watercraft were still being used by European Americans in the Chesapeake Bay area well into the 20th century.
- The prevailing currents favor a a West to East crossing from North America to Europe.
- A number of documented cases have been described of Japanese castaways surviving the accidental voyage from Japan to the Pacific Northwest of North America. (However, I am not aware of any systematic effort to determine their genetic impact if any on the populations of the North America.) The distance between North America is far shorter and the Gulf Stream is faster than the north Pacific current, decreasing the transit time and increasing the success rate of North America to Europe voyages.
@Gary Moore
"In the Berinigian Hypotheses for the founder populations of the the Americas, the Beringian population was formed by the merger of two groups: an interior Siberian population which entered the region from the west and a coastal population moving north along the coast. The interior Siberian population probably corresponded to the population represented by the Mat'ta and the coastal stream probably was more "East Asian", accounting for the observed pattern seen in Native American populations today. "
I agree that in pre-LGM times the boundary between West Eurasians and East Asians may have just been located in mid-Siberia. But recent genetic findings contradict your model. All New World populations have the "Amerindian" component (see Reich et al. ). This by itself contradicts any admixture theory for the origin of Amerindians. Plus this component is found in Paleoasiatic peoples and in West Eurasians, so Amerindians can't be a mix of West Eurasians and East Asians. (THis is where Eske Willerslev and Reich apparently disagree and I'm with Reich on this.) If we stretch Mal'ta DNA's significance a bit further, we can argue that ancestors of modern East Asians did not exist at a time when ancestors of modern Amerindians did exist. So, with new genetic discoveries, the picture that emerges is that Amerindians contributed their genes to both ancient Asians and this admixture event yielded modern East Asians and to ancient West Eurasians and this admixture event yielded modern Europeans.
@Onur
Yes, Terry is an old-school human origins thinker and as such closer to pre-scientific world-views.
"You are wasting your time with Terry. He is a Garden of Edenist when it comes to Mongoloid origins".
Ahh. Great to see my expression is catching on.
"Ma'ta is smack where we would expect East Asians to be at 25,000 YBP"
It's not actually. From the best information we have at present we should place the East Asian phenotype at that time to be more to the southeast of that region.
"it fits perfectly with ALL the cultural evidence we have [Amerindians is an older population than East Asian]".
But doesn't fit the genetic or archeolgical evidence one little bit.
"In the Berinigian Hypotheses for the founder populations of the the Americas, the Beringian population was formed by the merger of two groups: an interior Siberian population which entered the region from the west and a coastal population moving north along the coast. The interior Siberian population probably corresponded to the population represented by the Mat'ta and the coastal stream probably was more 'East Asian', accounting for the observed pattern seen in Native American populations today".
Absolutely.
@Kristiina,
I remember you posted some information regarding the distribution of various Y-DNA haplogroups at Maju's blog. I didn't copy the information off at the time and now can't find it. Could you be so kind as to do it all again here. We may then be able to work out the pattern of the haplogroup's northward movement. I found the information that a different Q subgroup in the Selkups and the Kets most interesting.
Sorry. One more thing:
Onur wrote:
"Polynesians in general seem to be Mongoloid-Australoid hybrids with more Mongoloid ancestry than Australoid ancestry".
Great. It is not so very long ago most were claiming Polynesians to be 'proto-Mongoloid Southeast Asians'. How things change. I presume it will not be very far in the future that people will be talking of Native Americans as being 'Mongoloid-Central Asian hybrids.
@TerryT
"It's not actually. From the best information we have at present we should place the East Asian phenotype at that time to be more to the southeast of that region. "
What best information? Back to the EDAR map? Mongoloids are known to have moved south into Southeast Asia and admixed with "Australoids," but it's an open question where they came from. Mal'ta is located smack where we find Mongoloids and East Asian genes historically but ancient Mal'ta DNA is not East Asian, just like Denisovans at 40,000 YBP from the same general area are not East Asians.
"But doesn't fit the genetic or archaeolgical evidence one little bit. "
What does archaeology have to do with the Mongoloid phenotype or East Asian genes? But even if we look at archaeology, the increasing frequency of bifaces in Siberia in the Holocene is seen to by some archaeologists as compatible with a back migartion out of America. The genetics of modern human populations is either validated by ancient DNA or dismissed. In the light of Mal'ta DNA, the so-called "east Asian" mtDNA haplogroups (A, C and D) are likely a back-flow from America. The opposite is not possible because hgs B and X are rare to non-existent in Siberia, while blood group and Y-DNA hgs N and O are rare-to-non-existent in America.
"I have always regarded Polynesians as Mongoloid-Australoid hybrids. Polynesians' racial hybridity was apparent to me even before seeing their genetic results".
Just as Amerindian racial hybridity was apparent to me long before seeing the currently available evidence. I don't recall arguing with you specifically when I kept suggesting Polynesian hybridity but certainly Maju (and several others0 became almost apoplectic whenever I suggested it.
"Terry is an old-school human origins thinker and as such closer to pre-scientific world-views".
Obviously each individual gene has its own individual origin and so in that sense each individual gene has its own Garden of Eden. What is 'pre-scientific' about that?
"What best information? Back to the EDAR map?
Yes. The EDAR370A gene is reponsible for a very high proportion of the Mongoloid phenotype and obviously has its own Garden of Eden. The gene is unlikely to have expanded on its own. Presumably it would have been carried by some population containing other genes as well as the EDAR370 mutation.
"Mongoloids are known to have moved south into Southeast Asia and admixed with 'Australoids,' but it's an open question where they came from".
It's not an 'open question' at all. Everybody knows it was introduced by a southward-moving Neolithic population. The Chinese Neolithic started somewhere between the Yangtze and Yellow Rivers.
"In the light of Mal'ta DNA, the so-called 'east Asian' mtDNA haplogroups (A, C and D) are likely a back-flow from America".
See if you can find anyone who agrees with you on that matter.
"All New World populations have the 'Amerindian' component (see Reich et al. ). This by itself contradicts any admixture theory for the origin of Amerindians".
No it doesn't. That 'Amerindian' component is shared with most Europeans but Amerindians have another component that is not shared with Europeans. Surely that suggests 'hybrid' somewhere.
"If we stretch Mal'ta DNA's significance a bit further, we can argue that ancestors of modern East Asians did not exist at a time when ancestors of modern Amerindians did exist".
But that's stretching things a hell of a lot. It indicates you've reached the conclusion that you want the evidence to show, rather than examining the evidence before coming to a conclusion.
"So, with new genetic discoveries, the picture that emerges is that Amerindians contributed their genes to both ancient Asians and this admixture event yielded modern East Asians and to ancient West Eurasians and this admixture event yielded modern Europeans".
By what method of logic do you reach that conclusion?
"What is 'pre-scientific' about that? "
Your reasoning is pre-scientific because you don't distinguish between hard facts and interpretations and because you fall in love with an idea and then defend it against all the incoming facts. And you don't read enough scientific literature. It takes time for people to learn how to think scientifically and you're still closer to a pre-scientific worldview although you use more "scientific" verbiage.
"Yes. The EDAR370A gene is reponsible for a very high proportion of the Mongoloid phenotype and obviously has its own Garden of Eden. The gene is unlikely to have expanded on its own. Presumably it would have been carried by some population containing other genes as well as the EDAR370 mutation. "
We've established that the origin of EDAR370A in America hasn't been tested by scholars, although the strength of expression of some of the phenotypes associated with this gene is demonstrably the strongest in the Americas (e.g., dental shoveling). Let's not use EDAR as proof of East Asians' origin in China.
" Everybody knows it was introduced by a southward-moving Neolithic population. The Chinese Neolithic started somewhere between the Yangtze and Yellow Rivers."
Yes, we agree on the southward movement. But you only see half of it. Right now just extend the southward arrow all the way from Beringia to the Yangtze. You don't believe that Amerindians got their EDAR370A from the Chinese neolithic, I hope...
"See if you can find anyone who agrees with you on that matter."
I don't care if anybody agrees with me or not. When people first constructed phylogenetic trees, they already believed in the origin of Amerindians from Asia, so the bias is baked in all of our thinking. Ancient DNA will hopefully show which interpretation is right.
"That 'Amerindian' component is shared with most Europeans but Amerindians have another component that is not shared with Europeans. Surely that suggests 'hybrid' somewhere."
The "Amerindian" component is ascertained in Karitiana and all South American Indians have the "First American" component, according to Reich et al. Where Reich found "admixture" is in Na-Dene and Eskimos, which is in the extreme north of the New World and it has no bearing on the origin of the "First American" component. But since the "First American" component is found in Paleoasiatic peoples in addition to Na-Dene and Eskimos, it's just as likely that this one or two additional components originated in America and migrated back. EDAR370A fits this interpretation very well.
"By what method of logic do you reach that conclusion?"
See above.
"But that's stretching things a hell of a lot. It indicates you've reached the conclusion that you want the evidence to show, rather than examining the evidence before coming to a conclusion."
Not at all. If the Mal'ta boy is 24,000 years old and is related to modern Amerindians but not to modern East Asians, we can be certain that Amerindians existed for at least that long but we don't have this certainty about East Asians. From all we know, Mongoloids is a recent population that appeared in the eastern provinces of the Old World in post-glacial times. This is precisely the time when Amerindians could walk out of the New World and enter Asia from the North.
Here you have another map on the distribution of the EDAR 1540C allele frequency worlwide
http://mbe.oxfordjournals.org/content/28/2/1013/F4.expansion.html
The highest frequencies are in northwestern South America, and the highest frequency in Asia is east of Baikal in the area traditionally inhabited by Evenki, Oroqen, Manchu, Daur and Buryat people. It seems to be the area of the highest frequency of mtDNA D and C. All these groups have quite much yDNA C3 and some C3c. Yenisei Evenki have 10% of R1a1, Buryats have 4% of R lines. Oroqen, Daur and Manchu have 7%, 6% and 5% of P lines, respectively. Q lines seem to be rare in all of them. Of course, all of them carry N lines, but I think that yDNA N is too young to be relevant here.
Terry, do you mean this?
At long last, I managed to find detailed yDNA data from Khanty, Mansi and Selkup who live near that Baraba Taiga Steppe area. IMO these could be the missing Y lines.
http://ling.tspu.edu.ru/files/PDF/articles/volkov_v_g_79_96_1_1_2013.pdf
Northern Selkups:
Q1a3-(xL330) 66.4%, R1a1 – 19%, N1b - 6.9%, R1b - 6.1%, C - 1.5%,
Mansi:
N1c - 16%, R1a1 - 6%, N1b 6%, R1b 2%, I - 4% (here something is missing! – according to my papers it should be
N1c - 16%, R1a1 - 8%, N1b - 60%, R1b - 4%, I - 8%, J- 4%)
Khanty
N1b - 48%, N1c1 - 38%, R1a1 - 7%, Q-L330 - 1%, G - 1%
Northern Khanty
N1b - 57%, Q-L330 - 21%, R1a1 - 21%, N1c1 - 7%, R1b1b2 - 0.9%
Kets:
Q-L330 - 84%, N1c - 8%, N1b - 4%, R1a - 4%
According to that Tomsky paper, R1b of southern Selkups is M73, so that goes probably also for Northern Selkups. It is interesting that Selkups and Kets have a different Q haplotype!
In that same paper you find also a pylogenetic tree of Q branches (p. 84). There you can see that Selkups and Chechens share the same Q haplotype which is different from Ket and Khanty haplotype. According to the paper, Scandinavian Q-L804 is very close to Amerind Q-M3, but, in Scandinavia, also Q-L527 is found, and it is probably very close to Selkup and Chechen haplotypes. According to Malyarchuk et al. 2011 paper, Balkars share with Altaians also R1b-M73 haplotype.
@Kristiina
"Here you have another map on the distribution of the EDAR 1540C allele frequency worlwide
http://mbe.oxfordjournals.org/content/28/2/1013/F4.expansion.html
The highest frequencies are in northwestern South America, and the highest frequency in Asia is east of Baikal in the area traditionally inhabited by Evenki, Oroqen, Manchu, Daur and Buryat people. "
Thanks! I remember seeing this map but you pulled it up right on time. Where do you get the frequencies? My color vision doesn't allow me to distinguish between NW South America and North America.
NW South America is interesting because that's where we find Y-DNA C3* and mtDNA D4H3a (rare in North America but found in ancient remains in British Columbia at 10,000 YBP). A coastal migration would make sense, it seems.
"Terry, do you mean this?"
Yes, that seems to be the one.
"It is interesting that Selkups and Kets have a different Q haplotype!"
Really interesting. And surprising. Their Q's are obviously from separate sources, with the Selkup source being more 'western' and Ket being closer to American.
"Your reasoning is pre-scientific because you don't distinguish between hard facts and interpretations"
How about this then:
"Mal'ta is located smack where we find Mongoloids and East Asian genes historically but ancient Mal'ta DNA is not East Asian"
A moment's reflection would have told you exactly why that is. The Mal'ta boy has Y-DNA R1a and mt-DNA U, neither of which are at all common in the region today. Over to Kristiina:
"It seems to be the area of the highest frequency of mtDNA D and C [east of Baikal in the area traditionally inhabited by Evenki, Oroqen, Manchu, Daur and Buryat people]. All these groups have quite much yDNA C3 and some C3c. Yenisei Evenki have 10% of R1a1, Buryats have 4% of R lines".
There has been a degree of population replacement. By haplogroups almost certainly associated with the spread of the Mongoloid phenotype. And, further, we can be fairly sure that Q was a fairly early presence in the region because:
"Q lines seem to be rare in all of them".
It's distribution has been fragmented by the later expansions (probably including that of R1a), just as C2's has been fragmented in the Pacific by the expansion of New Guinea/Papuan haplogroups.
"You don't believe that Amerindians got their EDAR370A from the Chinese neolithic, I hope..."
Of course not. The mutation is much older than the Chinese Neolithic although its spread south seems to be associated with the Neolithic's spread south.
"EDAR370A fits this interpretation very well [one or two additional components originated in America and migrated back]".
The only reason you are at all able to claim that is because you insist the currently widely accepted phylogenetic trees are completely wrong. You have yet to provide any evidence at all that your position is valid.
"From all we know, Mongoloids is a recent population that appeared in the eastern provinces of the Old World in post-glacial times".
Again the only reason you are at all able to claim that is by ignoring generally accepted evidence, especially that recently posted at this blog.
"you fall in love with an idea and then defend it against all the incoming facts".
As a result of this discussion it has become very obvious to me that is exactly what you have done. The only idea I fell in love with is that collections of genes, including haplogroups, have moved around the earth in what we can imagine as a series of waves. The wave theory of evolution. I have consistently interpreted the evidence available from that perspective.
I would be a bit careful with the interpretation of the EDAR frequency plot. There are no data points between the US/Mexican boarder, Northern China, and Greenland. The only thing we can say is that between Mexico and northernmost S America the frequency seems to fluctuate between 70% and 90%. Assuming they didn't distinguish between single and double copies, ~75% is actually not that high - it means only about 1/4 of the population carries a double set.
Please compare the frequency map of hg C (http://www.theapricity.com/forum/showthread.php?5836-What-is-your-mtDNA-Haplogroup/page14) and D (http://theweeklyjo.files.wordpress.com/2013/04/23andme.jpg
) with the Edar distribution map (http://mbe.oxfordjournals.org/content/28/2/1013/F4.expansion.html). Sorry that I could not get a better source for hg C map! It seems that mtDNA C could be a better match to this EDAR variant, as the EDAR frequency seems to be high in Mexico and southwest USA. I would rather link mtDNA C to yDNA Q than C3.
@TerryT
"The only reason you are at all able to claim that is because you insist the currently widely accepted phylogenetic trees are completely wrong. You have yet to provide any evidence at all that your position is valid. "
I think the very assumption that gene phylogenies reflect the historical reality of population movements has been weakened as of late in view of the growing amount of evidence in favor of admixture as the driving force behind observable genetic variation.
"Again the only reason you are at all able to claim that is by ignoring generally accepted evidence, especially that recently posted at this blog."
What do you have in mind? I remember this one (http://dienekes.blogspot.com/2010/12/paleoamerican-morphology-in-context-of.html) arguing that Amerindians differentiated prior to the formation of Mongoloids and Caucasoids.
"A moment's reflection would have told you exactly why that is. The Mal'ta boy has Y-DNA R1a and mt-DNA U, neither of which are at all common in the region today. "
First of all, we don't know the subclade of R that the Mal'ta boy belonged to. But I'm at a loss trying to understand what you mean and why you are disagreeing with me. The ancient pattern of variation in Siberia was more like the one we find in West Eurasia (and autosomally in America) now, with East Asian genes coming to dominate it later.
"Their Q's are obviously from separate sources, with the Selkup source being more 'western' and Ket being closer to American."
Terry, I know this Russian paper from which Kristiina took this data: in fact Amerindian Q is closer to Northern European Qs followed by Selkup and Chechen Qs. Ket Qs cluster with South Siberian and Ob-Ugrian Qs and they all are further removed from the Amerindian ones.
"As a result of this discussion it has become very obvious to me that is exactly what you have done."
OK, no more mutual insults, Waveman.
"Assuming they didn't distinguish between single and double copies, ~75% is actually not that high - it means only about 1/4 of the population carries a double set".
Which does provide rather convincing evidence the mutation did not arise in America. Besides which I remember this rather interesting paper:
http://dienekes.blogspot.co.nz/2013/04/y-chromosomes-of-native-south-americans.html
Perhaps some population rich in the EDAR mutation was involved, as German suggests:
"NW South America is interesting because that's where we find Y-DNA C3* and mtDNA D4H3a (rare in North America but found in ancient remains in British Columbia at 10,000 YBP). A coastal migration would make sense, it seems".
From Asia, of course.
"I think the very assumption that gene phylogenies reflect the historical reality of population movements has been weakened as of late in view of the growing amount of evidence in favor of admixture as the driving force behind observable genetic variation".
Such admixture affects the phylogeny not one little bit. In fact it can often inform us as to the sources of the admixture. What we see in the phylogeny distribution is the individual movement of each haplogroup, regardless of whether it actually moved independently or not.
"I remember this one ... arguing that Amerindians differentiated prior to the formation of Mongoloids and Caucasoids".
2010. From more recent research we can now see exactly why that appeared to be so.
"The ancient pattern of variation in Siberia was more like the one we find in West Eurasia (and autosomally in America) now, with East Asian genes coming to dominate it later".
Yes. Ancient Siberian Y-DNA R and mt-DNA U, carrying a phenotype present also in modern Europeans, was later overlain by an East Asian phenotype consisting of Y-DNAs N and C3 as well as mt-DNAs C and D. Simple.
"Amerindian Q is closer to Northern European Qs followed by Selkup and Chechen Qs. Ket Qs cluster with South Siberian and Ob-Ugrian Qs and they all are further removed from the Amerindian ones".
That doesn't easily fit the phylogeny accepted at ISOGG. Perhaps Ebizur, who seems to know a great deal on the subject, will be able to contribute.
"OK, no more mutual insults, Waveman".
Agreed. Although to be honest you were the one that started the insults.
"A moment's reflection would have told you exactly why that is. The Mal'ta boy has Y-DNA R1a and mt-DNA U, neither of which are at all common in the region today. "
but this is just another assumption based only in and founded in assumptions created by todays assumptions about dna.. and nothing has been proved yet.
Science can't tell WHY mtdna mutates! they seem to like not knowing answers to very real questions and important questions. I think they really like being able to make stuff up as they go.
So to me the most interesting part of the boys mtdna and his adna is that he was both a U and had a fish diet ... and he isn't even by any oceans.
So what happens when he gets land bound and has to start eating buffalo or rats or beans or only rice for that matter . what is it that makes those 270's jump to 217 and thus they become a B ( or whatever else it can jump too ) .. I think the fact he was a sea food/fish eater is a monumental bit of information ! considering where Most U people now live.
throw in some tribes moving west not east..
http://westerndigs.org/thirteen-prehistoric-villages-discovered-in-wyoming-mountains-may-redraw-map-of-tribal-migrations/
@TerryT
"Such admixture affects the phylogeny not one little bit. "
Of course it does. Admixture shifts the root around, creates false terminal branches, etc.
"From more recent research we can now see exactly why that appeared to be so."
Too bad, you can't quote any of it.
"Although to be honest you were the one that started the insults."
No, it started with an accurate description of your approach to human origins. This stays unchanged. I'm just willing not to repeat it.
"Yes. Ancient Siberian Y-DNA R and mt-DNA U, carrying a phenotype present also in modern Europeans, was later overlain by an East Asian phenotype consisting of Y-DNAs N and C3 as well as mt-DNAs C and D. Simple. "
There's nothing simple in it. This East Asians phenotype could come from America - to everybody's surprise and my delight. Y-DNA is not found in the Americas, hence east Asians didn't populate America, but hence there are many genetic similarities between East Asians and Amerindians and Amerindians often have the earliest attested and most strongly expressed "East Asian" features (EDAR and its phenotypcial expressions, craniofacial morphology, etc.), the population movement likely went from New World into Asia.
"Amerindian Q is closer to Northern European Qs followed by Selkup and Chechen Qs. Ket Qs cluster with South Siberian and Ob-Ugrian Qs and they all are further removed from the Amerindian ones".
I've also had an interesting time going back to this paper:
http://dienekes.blogspot.co.nz/2013/08/y-chromosome-haplogroup-q-and-native.html
Quote:
"Two main founding lineages, Q1a3a1a-M3 and Q1a3a1-L54(xM3), were detected"
Under the ISOGG classification these would be Q1a2a1a1 and Q1a2a1a. According to the paper outside America Q1a2a1a1-M3 has been found only in Koryaks. This is not surprising. We surely would expect to see some connection between Eurasia and America. But Q-L54 has a wider distribution, as far as Mongolia. The above paper includes Ket Q-L330 within the clade although ISOGG places it one step further removed. ISOGG has Q1a2a1a-L54 as 'brother' haplogroup to both Q1a2a1c-M330 (the 'Ket' haplogroup) and to Q1a2a1b-Z780. But according to Wikipedia the Z780 mutation actually lies within the L54 mutation:
http://en.wikipedia.org/wiki/Haplogroup_Q-Z780
Quote:
"Haplogroup Q-Z780 is a subclade of Y-DNA Haplogroup Q-L54 ... Q-Z780 has descendants across much of the pre-Columbian Americas. It is the second most common branch of Q-M242 in the Americas".
So we have a little confusion. But if the above statement is correct the Q1a3a1-L54(xM3) mentioned in the above paper is actually Q1a2a1b-Z780. And it becomes straightforward: the Ket Q-L330 and American Q-M3 are 'brothers'. You can't get much closer than that. Selkup Q1a3-(xL330) can therefore only be more distant from American Q than is Ket Q. The closest it can be is some other branch of Q1a2a-L53, or perhaps even further distant, to Scandinavian Q1a2b-L940.
"Perhaps Ebizur, who seems to know a great deal on the subject, will be able to contribute".
I realised I did have some data that I'm fairly certain Ebizur provided. I have been using the information for some time, including the above.
"Admixture shifts the root around, creates false terminal branches, etc."
How does admixture affect any aspect of haplogroup phylogeny? Any admixed population will have the haplogroups of the populations that make up the admixture, originally in direct proportion to the numbers in each population. With time one or more of the original haplogroups may come to dominate but the wider phylogeny will remain unaltered apart from the dissappearance of some links in the chain. It will still be possible to see the chain though.
"No, it started with an accurate description of your approach to human origins".
Your above claim shows you have nothing but a very minimal understanding of population genetics. You may have university degrees but they are not in biology of any kind. Because of that I presume you are completely unable to follow the logic in my comment from yesterday. Get some first year genetics student to explain it.
"Y-DNA is not found in the Americas"
No Y-DNA in the Americas? What are you on about here? What do you call Y-DNA haplogroups Q-M3, Q-Z780 and C3?
"Too bad, you can't quote any of it".
The paper Dienekes linked to in this post, plus several other recent ones he has posted, and the other links various of us have provided here is surely all the evidence required. Unless you've made up your mind in advance you're not going to accept any new evidence. Native Americans are formed from admixture between at least two populations that have undergone separate processes of selection before they mixed. That is so obvious that surely everyone can now see it.
"This East Asians phenotype could come from America - to everybody's surprise and my delight".
That is so unlikely as to be not worth even considering. Unless you're prepared to ignore whole swathes of genetic evidence.
terryt wrote:
"Amerindian Q is closer to Northern European Qs followed by Selkup and Chechen Qs. Ket Qs cluster with South Siberian and Ob-Ugrian Qs and they all are further removed from the Amerindian ones".
With regard to the YH Q presence in Scandinavia, it might be useful to take a look at the following two maps:
http://en.wikipedia.org/wiki/File:Haplogroup_Q_%28Y-DNA%29.PNG
http://en.wikipedia.org/wiki/File:North_Atlantic_Gyre.png
I think there is definitely a possibility that YH Q (M242) may have reached Europe from North America via an oceanic route. Keep in mind that it is widely accepted in the archeological community that Native Americans were able to populate the Americas rapidly because of their use of watercraft, and that Native Americans were able to settle the islands of Caribbean and the Bahamas at an early date. Native Americans of the eastern coast of North America built substantial catamaran vessels, and evidence of deep water sailing is attested in the archeological record for the Archaic period. As you can see from the map of the North Atlantic Gyre, voyagers from North America would be steered into the North Sea and towards the coast of Norway exactly were the frequency of YH Q peaks. Any that veered south would and drifted past the British Isles would likely be caught by the Rennell Current and be directed into the Bay of Biscay towards the Basque areas.
Another possible source of Q in Europe that is frequently put forward is the Hunnic invasion. The Huns are widely identified with the people called by the Chinese "Xiongnu". The language of the Xiongnu is poorly attested, but many linguists think on the basis of grammar that their language was Yeniseian (that is, related to Ket) and not Turkish or Mongolian.
German says that the population movement likely went from New World into Asia. This is a highly curious statement, as America is full of yDNA Q, and what we have of Q in East Asia is not very close to L-53. Instead, a flow of Q clades (under L53) from Northeast Asia to the West is possible.
German’s postulation of genetic similarities between East Asians and Amerindians could be and must be explained with mtDNA, as that is what is shared, to a bigger extent, between Native Americans and East Asians. One possibility to understand these similarities is that yDNA Q went across Siberia with MtDNA C and A, and certain features developed in them, perhaps as an adaptation to the cold climate they were facing. In my opinion, the Australoid look should be linked to YDNA C carrying people, and that must also have contributed to Native American looks. The ”African” contribution of yDNA D and its mtDNA companions (D?) is exciting, but I do not have a clear idea how that contributed to East Asian and Native American looks.
Everybody knows that Han Chinese carry mostly O clades. The biggest three mtDNA types of the ancient people from the Central Plains of China were M10 (28%), M7c (20%) and D5 (12%). The frequency of C in modern south Chinese is only 3.86% and in north Chinese 6.5%. I really cannot see how the Chinese could have given the Mongoloid look to Native Americans. Chinese should be mainly a mixture of yDNA O and, to a smaller extent, of C and their mtDNA counterparts. I think that it is more probable that the Han Chinese got EDAR from the northerners as they settled in more northern areas.
Everybody wants to have yDNA N to have given Mongoloid phenotype to the world. According to the Chinese Super-Grandfathers paper, yDNA N started spreading 16.000 years ago. I do not really think that yDNA N can be the origin of EDAR and Mongoloid looks, and certainly not in the least to have brought EDAR to America. That phenotype must have arisen earlier in the north. According to that same Chinese paper, the split between O and N is deep, at 30.000 years. In this paper they also propose important changes to the structure of N tree but I am not sure what they mean. One of the changes involves N-LLY22g’s position. If N developed in the North, it was certainly part of the Mongoloid looks and its expansion. If it developed in the South, e.g. in Yunnan, that connection is much weaker.
Anyway, I firmly think that the looks of people are an areal feature and not something brought by intruders of all sorts.
Correction: Chinese should be mainly a mixture of yDNA O and, to a smaller extent, of C and N and their mtDNA counterparts.
@ Gary Moore:
"terryt wrote:"
That is actually German's statement. I hope I showed above that his statement is incorrect.
"I think there is definitely a possibility that YH Q (M242) may have reached Europe from North America via an oceanic route".
A possibility. But any American Q arriving that way has not survived. American and European Q belong to separate branches.
@ Kristiina:
"Instead, a flow of Q clades (under L53) from Northeast Asia to the West is possible".
That could only apply to the 'Ket' Q-L330 though. And although Q-L54 and Q-Z780 both have representatives in East Asia their presence there more probably represents remnant populations from the movement through Northeast Asia to America. This is especially likely when we consider that the third ISOGG Q-L53 haplogroup is the 'Ket' haplogroup, not found in America at all.
"One possibility to understand these similarities is that yDNA Q went across Siberia with MtDNA C and A, and certain features developed in them, perhaps as an adaptation to the cold climate they were facing".
I agree with much of what you say from this point on, especially regarding the problems you bring up. However I propose a slightly different solution to the problems than you offer. I believe it provides a better fit to all the evidence:
By 30,000 years ago humans in Eastern Eurasia had become confined to no further north than 40 degrees N. This population, long isolated at a geographic extremity of the human range, had undergone a long period of selection. This produced a population completely homozygous for the EDAR370A mutation (and presumanly other mutations). It contained Y-DNAs C3, D and NO, the latter possibly already separating into O1, O2, O3 and N. The mt-DNAs were A, C, D and (possibly) B.
Further west, at the southern edge of the Central Asian steppe, a different population emerged from south of 40 degrees N containing Y-DNA P and its derivatives R and Q along with various N- and R-derived mt-DNAs. The steppe population invented (or adopted) a method(s) of surviving extremely cold winters, and were able to move north. The earliest immigrants utilized the most easily available resources in the untouched habitat and spread rapidly, but thinly, through the steppe. Y-DNA Q and mt-DNA X perhaps became the vanguard. So much so that Y-DNA Q outpaced mt-DNA X and began mixing with the East Eurasian population as it met them on its eastward journey. The men from that eastern population were a little slow in joining the movement to America but Y-DNA C3b managed to do so eventually. Other men from the East Asian population adopted the introduced Upper Paleolithic technology and began to press southward. Later still N was able to move even further north, but not early enough to reach America before it was already fully occupied.
On the steppe the main population developed from the slower-moving Y-DNA R1a and mt-DNA U who must have been able to set up longer term survival strategies.
"I really cannot see how the Chinese could have given the Mongoloid look to Native Americans".
So it wasn't 'the Chinese' who provided the 'Mongoloid look to Native Americans' but an East Asian population from the higher altitudes south of 40 degrees north.
"the Australoid look should be linked to YDNA C carrying people, and that must also have contributed to Native American looks".
That might (and I stress 'might') be where mt-DNA B comes in. It certainly seems to be more 'southern Chinese' than any of the other mt-DNAs involved. However that exposes another problem in the mt-DNA B is almost certainly responsible for much of the Mongoloid element in the Pacific island population.
@terry
"How does admixture affect any aspect of haplogroup phylogeny?"
Considering that you read Dienekes and Razib here's a post you can access and learn from: http://dienekes.blogspot.com/2011/09/latent-admixture-causes-spurious-serial.html
"No Y-DNA in the Americas? What are you on about here? What do you call Y-DNA haplogroups Q-M3, Q-Z780 and C3? "
I meant hgs N and O that are all over Eurasia but not in America.
"Your above claim shows you have nothing but a very minimal understanding of population genetics. You may have university degrees but they are not in biology of any kind. Because of that I presume you are completely unable to follow the logic in my comment from yesterday. Get some first year genetics student to explain it."
Dude, I worked at the Joanna Mountain Lab at Stanford while you were milking cows.
"The paper Dienekes linked to in this post, plus several other recent ones he has posted, and the other links various of us have provided here is surely all the evidence required. Unless you've made up your mind in advance you're not going to accept any new evidence. Native Americans are formed from admixture between at least two populations that have undergone separate processes of selection before they mixed. That is so obvious that surely everyone can now see it."
Just give me a reference, will you?
@Kristiina
"German says that the population movement likely went from New World into Asia. This is a highly curious statement, as America is full of yDNA Q, and what we have of Q in East Asia is not very close to L-53. Instead, a flow of Q clades (under L53) from Northeast Asia to the West is possible."
Not sure what you meant here. The Volkov paper, Skhema 1, p. 84, has Amerindians above L53, while all of the Qs in east Asia below it. That's the back migration signature I'm talking about.
Just popping back to talk about a rumour I've read. Apparently, the Mal'ta boy breaks down in terms of modern populations as 2/3rds Udmurt and 1/3rd Kalash (or similar pred. Europid South Asian populations).
I've no idea whether that analysis is leaked from the upcoming paper or someone just ran their own test on the boy's leaked data, but assuming it's true, it has some really interesting implications.
A quick Google image search shows Udmurts to be an admixture of a majority Upper Paleolithic (Lithuanian-like) Europid component and a minority Mongoloid component. I can't find any genetic analyses of the Udmurt, but the related Komi's ADMIXTURE data are consistent with my observation, being just under 20% East Asian at k=4:
http://img843.imageshack.us/img843/4837/russiangwaadmixk2k8.png
Since the Mal'ta boy lacks East Asian admixture and the East Asian in the Udmurt was brought later by N-bearers (which clade the Udmurt now overwhelmingly belong to), it must be the UP European element in the Udmurt that draws the boy close to them.
Perhaps this suggests that indigenous pred. West Eurasian Siberians, especially the European-like ones (i.e. Finno-Ugrics) have deep -- really deep -- ancestry in the region, since the boy is most related to the most local (at least ancestrally) Europid populations.
Gary Moore,
I think there is definitely a possibility that YH Q (M242) may have reached Europe from North America via an oceanic route.
But keep in mind that the Science article mentions the boy shares a portion of his genome with Amerindians and no other population. This suggests minimal interaction between the Mal'ta people's descendants and populations to the West (or so far East that they end up being West again).
Apologies to Seinundzeit. I see he already mentioned the analysis. Well, at least if it's from Vadim Verenich, it's credible.
"Considering that you read Dienekes and Razib here's a post you can access and learn from: http://dienekes.blogspot.com/2011/09/latent-admixture-causes-spurious-serial.html"
And that paper concerns haploid genes how exactly? I repeat: 'Your above claim shows you have nothing but a very minimal understanding of ... genetics. You may have university degrees but they are not in biology of any kind'.
"I meant hgs N and O that are all over Eurasia but not in America".
The expansion of those haplogroups in East Eurasia obviously post-dates the arrival of Americans in that continent. And those Americans moved somewhat north of where the early members of N and O lived at the time.
"Dude, I worked at the Joanna Mountain Lab at Stanford while you were milking cows".
I'm surprised you learned nothing there. By the way I have always tried to avoid actually milking cows.
"Just give me a reference, will you?"
Reread Dienekes' original post and al the comments here. People have provided no end of references.
"Perhaps this suggests that indigenous pred. West Eurasian Siberians, especially the European-like ones (i.e. Finno-Ugrics) have deep -- really deep -- ancestry in the region, since the boy is most related to the most local (at least ancestrally) Europid populations".
That seems to be what the evidence German is so unable to see is telling us.
Sorry Onur. I was a bit rushed earlier and didn't explain myself well. I think we can assume the movement east was neither particularly rapid nor continuous. It is possible (in fact probable) that the 'original' movement was made up largely of Y-DNA Q and mt-DNA X. Haplogroups A, B, C, and D would have gradually, and progressively, joined the movement. Offspring of these latter haplogroups would at first have mostly had children with the progeny of the Y-DNA Q/mt-DNA X people. As a result the 'Mongolification' would have been gradual, not sudden. The eastern haplogroups would have gradually replaced the original X, not abruptly.
Terry, your solution is clearly geared to keep Q clean from Mongoloid ancestry and to maintain the R-Q superiority over East Asians through attributing Upper Palaeolithic technology only to P clades.
According to a Chinese study on the Upper Palaeolithic Shuidonggou site “the data suggests there were two peaks of occupation falling around 32,000 to 24,000, and 13,000 to 11,000 years ago. "The earliest human occupation at Shuidonggou is substantially younger than in the initial Upper Palaeolithic from western Eurasia. Thus, it is quite plausible that modern humans migrated into northern China from western Eurasia during this time period", says PEI Shuwen, first author of the study.” (http://www.stonepages.com/news/archives/004846.html)
On the basis of this, it seems that the Upper Palaeolithic technique came to North China with yDNA D and NO.
If Q came to Northeast Asia only after NO, D and C3, I do not understand why only Q and C3b would have made it to America. When I look at the maps of the Chinese Super-Grandfathers paper, I see that only Q is in the Baikal area and only C3 is on the same North Eastern path to America with Q, while O is spreading in the south and NO probably survived in the area ranging from West China to Japan, as well as taking a refuge in Tibet, Yunnan and Sichuan.
I disagree with your statement that ”his produced a population completely homozygous for the EDAR370A mutation (and presumanly other mutations). It contained Y-DNAs C3, D and NO”, in that this could not have produced a situation in which EDAR370A is more common in America than in areas with mostly C3, D and NO lines.
When you say that ”mt-DNA B is almost certainly responsible for much of the Mongoloid element in the Pacific island population”, my question is: can we see this Australoid vs. Mongoloid division in Admixture analyses? Han Chinese are usually marked with only one colour, although they have considerable amounts of yDNA C and mtDNA B. Moreover, in the recent Thai ancestry analysis, Samoan and Tongan are very different from mainland East Asians and share almost only K = 2 with mainland East Asians.
German, ”the Volkov paper, Skhema 1, p. 84, has Amerindians above L53, while all of the Qs in east Asia below it. That's the back migration signature I'm talking about.
I am looking at ISOGG 2013 hg Q tree, and it seems that Native Americans are under L-54, i.e. Q1a2, and that covers also Ket L330 and Scandinavian L804. Selkup and Northern Altaian L940 seems to be the brother clade of L53. East Asian M120 is under NWT01, i.e. Q1a1. Do you mean that Q1a2 backmigrated from Beringia to the West (not to China)? If you want to insist in the backmigration of whole Q, you could say that Q1a-MEH2 arose in Beringia and all the rest migrated to East Asia, West Asia and Russia-Europe, respectively!
I think that I am increasingly of the opinion that Australoids should be linked to yDNA C and mtDNA B, F and N9 and Mongoloids in Eastern Eurasia and America to yDNA MNOPS and D and mtDNA M. Perhaps I will be proven wrong! Negritos probably came through India. Could they be of yDNA K and mtDNA R?
BTW, here you have nice maps and a Japanese perspective to this origin issue: http://www.eladna.com/index.html.
Correction: when I referred to Samoan and Tongan admixture components, I was looking at the wrong column! Tongan and Samoan is SP15, and in this column, 4/9 consists of Pacific-specific ancestry, and, for the rest, Tongans and Samoans resemble in particular Beijing Chinese cluster with a small amount of South Asian ancestry. I propose that yDNA C and mtDNA B were first to reach Beijing area, and O arrived c. 30.000 years ago. Then O1 developed and spread on the East Cost of China and at some point started settling in the southern islands.
"I think that I am increasingly of the opinion that Australoids should be linked to yDNA C and mtDNA B"
There is virtually no B in Australia. And the C there is neither C2 nor C3, but C4. C's diversification is obviously very ancient. B is closely associated with the Austronesian expansion into the Pacific, along with C2 from Southern Wallacea, although B was possibly originally 'Australoid' early in its existence in South China/Vietnam.
"Mongoloids in Eastern Eurasia and America to yDNA MNOPS and D and mtDNA M".
MNOPS as 'Mongoloid' doesn't fit at all. M and S are definitely not Mongoloid, and neither are the closely related K1, K2 and K3. Of the other two haplogroups within the clade the P-derived R cannot be considered Mongolid either. I believe its relation Q is Mongoloid only through later introgression. NO is the only 'Mongoloid' Y-DNA haplogroup within the group.
"On the basis of this, it seems that the Upper Palaeolithic technique came to North China with yDNA D and NO".
I don't think so. By the Upper Paleolithic both these haplogroups were already resident in East Asia.
"it is quite plausible that modern humans migrated into northern China from western Eurasia during this time period"
NO for a start is almost certainly not from 'western Eurasia' at all. It is related to the KMNOPS of Sundaland although perhaps P formed from this haplogroup somewhere further west in South Asia. As for D, I suspect it had been isolated in eastern Tibet/western China for a long time. That leaves just the P-derived haplogroups R and Q as forming any possible movement from western Eurasia.
"I disagree with your statement that 'his produced a population completely homozygous for the EDAR370A mutation (and presumanly other mutations). It contained Y-DNAs C3, D and NO', in that this could not have produced a situation in which EDAR370A is more common in America than in areas with mostly C3, D and NO lines".
How do you conclude that the 'EDAR370A is more common in America than in areas with mostly C3, D and NO lines'? These three haplogroups all contain a very high proportion of the mutation in most regions where they are present. The only regions this is not so is where it is obvious there has been considerable admixture with non-Mongoloid populations.
"If Q came to Northeast Asia only after NO, D and C3, I do not understand why only Q and C3b would have made it to America".
Q passed to the north of the existing East Asia population, picking up only the female lines. C3 managed to join just the later stages. This is demonstrated by your comment, 'I see that only Q is in the Baikal area and only C3 is on the same North Eastern path'.
"while O is spreading in the south and NO probably survived in the area ranging from West China to Japan, as well as taking a refuge in Tibet, Yunnan and Sichuan".
I strongly suspect that NO originated 'in Tibet, Yunnan and Sichuan' and spread respectively north and south from there after adopting the Upper Paleolithic and then developing the East Asian Neolithic.
"can we see this Australoid vs. Mongoloid division in Admixture analyses?"
We can in the Pacific. In fact such admixture is almost universally accepted.
"Tongan and Samoan is SP15, and in this column, 4/9 consists of Pacific-specific ancestry, and, for the rest, Tongans and Samoans resemble in particular Beijing Chinese cluster with a small amount of South Asian ancestry".
yes.
Sorry. A further comment. German wrote some time back:
"the growing amount of evidence in favor of admixture as the driving force behind observable genetic variation".
And such admixture is a fundamental assumption in the wave theory of evolution. Evolution is a product of the movement of various waves of genetic change through populations.
"Terry, the fact that the Native American Y-chromosome haplogroup pool is so limited in diversity while their mtDNA haplogroup pool is much more diverse strongly suggests that they descend from a small group of males who admixed with a lot of Mongoloid females, and almost only Mongoloid females looking at their mtDNA haplogroups, during their migration towards the Americas".
But the presence of X indicates that the admixture was gradual. And that explains the apparent one third West Eurasian genetic element in Native Americans.
"component that peaks in East Asians at low K's before the appearance of the Amerindian component does not equal to the Mongoloid race, as it is more representative of Asian Mongoloids, who are more derived than Native Americans"
Surely Asian Mongoloids simply appear to be 'more derived' because they are less admixed with other populations. Or do you insist that the apparent admixture mentioned in Dienekes' blog here is exactly that: apparent?
"So it does not necessarily indicate non-Mongoloid admixture in Native Americans".
Am I to understand that you absolutely dismiss all these statements:
"24,000-yr-old Siberian Mal'ta person geneticall similar to native amer and west eurasians. No east asian"
And:
"Native Americans formed by an admixture of east Asian ancestors and the ancestors of western Eurasians"
And Michael Balter's comments:
"His DNA shows close ties to those of today's Native Americans. Yet he apparently descended not from East Asians, but from people who had lived in Europe or western Asia. The finding suggests that about a third of the ancestry of today's Native Americans can be traced to 'western Eurasia,' with the other two-thirds coming from eastern Asia"
And:
"One expected relationship was missing from the picture: The boy's genome showed no connection to modern East Asians. DNA studies of living people strongly suggest that East Asians—perhaps Siberians, Chinese, or Japanese—make up the major part of Native American ancestors. So how could the boy be related to living Native Americans, but not to East Asians?"
And again:
"The Mal'ta child represents a population of Native American ancestors who moved into Siberia, probably from Europe or west Asia. Then, sometime after the Mal'ta boy died, this population mixed with East Asians. The new, admixed population eventually made its way to the Americas. Exactly when and where the admixture happened is not clear, Willerslev said. But the deep roots in Europe or west Asia could help explain features of some Paleoamerican skeletons and of Native American DNA today".
????
"...the fact that the Native American Y-chromosome haplogroup pool is so limited in diversity while their mtDNA haplogroup pool is much more diverse strongly suggests that they descend from a small group of males who admixed with a lot of Mongoloid females, and almost only Mongoloid females looking at their mtDNA haplogroups, during their migration towards the Americas. So their mtDNA haplogroups represent their autosomal genetics way more than their Y-chromosome haplogroups do. In fact, it is very likely that that the original Y-DNA hg Q-carrying ancestors of Native Americans could not leave more than 0% of their genome to present-day Native Americans."
Onur,
I agree with everything in this except:
(i) For most early native Americans, it was decidedly proto-Mongoloids who admixed - not Mongoloids in any traditional or modern sense.
(ii) The 0% statement is absolutely ridiculous based both on phenotype and all autosomal DNA analyses carried out, to date.
Terry, so in your opinion we have c. 45.000 years ago (in Super-Grandfathers paper it is 34.000 years ago) KMNOPS in Sundaland, and yDNA C/Australoids in Australia and in continental East Asia. I suppose that KMNOPS people were negritos at that time. At the same time, “African” YDNA D and mtDNA G/D (?) were already in North China. Then, YDNA P goes to India with mtDNA R through the areas inhabited by ancient South Indians, and from there to Central Asia c. 40.000 years ago. There it leaves behind mtDNA R and M and takes along mtDNA X. In Central Asia these people become Caucasoids with bloodtype O.
While P is in India, NO originates in Yunnan, c. 40.000 years ago. We know that according to ancient finds, people in South China were Negrito and Jomon looking (yDNA D or C?). O starts spreading in China and arrives to the North Asia very quickly. Before the Ice Age NO and O go up to the Baikal area with mtDNA C and A and develop Mongoloid traits together with yDNA C3 and mtDNA B and yDNA D and mtDNA D.
Circa 35.000 years ago yDNA Q-M120 brings the Upper Palaeolithic Technique to China and again leaves behind its original women and mixes with women carrying Eastern mtDNA. These Q-M120 people never make it to America.
The mainstream Q is heading to Baikal and meets Mongoloid yDNA C3, D and NO people. YDNA Q passes over the earlier inhabitants. Before the onset of the Ice Age, YDNA Q presses on to Beringia and again leaves behind mtDNA X and takes along women carrying D, C and A lines. (I conclude that Eastern women were prettier as they are always preferred).
These paleo Native Americans become bloodtype O, although their women come from the areas with a high frequency of bloodtype A and B. NO, D and C3 recede back to the South in big numbers to change the looks of all southerners.
If this is not possible, we can change the route and say that yDNA Q passes through North China instead of Baikal area. Q brings the Upper Palaeolithic Technique to China and takes along women carrying D, C and A lines. However, these men do not leave any yDNA trail in China (Q-L53 lines) or any mtDNA X, they just pass through China changing wifes and leaving a new technique to previous inhabitants and press forward to Amur and further to America.
If this theory is not working, we can go back to the theory that Mongoloid traits are not linked with the cold climate but developed in the Chinese heartland, but then again it is not so simple to explain why these traits were passed on to Native Americans and why EDAR mutation is most frequent in Baikal area and America.
You also say that “the presence of X indicates that the admixture was gradual. And that explains the apparent one third West Eurasian genetic element in Native Americans.“ However, according to admixture analyses, http://www.biomedcentral.com/1471-2148/13/127/figure/F6?highres=y , Southern Native Americans do not show any admixture with Western Eurasians or East Asians. If Northern Native American X reflects West Eurasian gene flow, it must be more recent than the ancestry from the first settlers.
"Not all researchers agree with Willerslev's conclusion of partial Caucasoid ancestry in Native Americans".
No, but those researchers are going to be left behind in the dust as new evidence comes to light.
"The 0% statement is absolutely ridiculous based both on phenotype and all autosomal DNA analyses carried out, to date".
Absolutely.
"results of formal admixture tests such as ADMIXTOOLS and MixMapper and haplogroup distributions all support my interpretation".
The tools rely on what the researcher assumes to be the 'basic' populations. If those assumptions are incorrect the results will be.
"all of them seem to possess 100% of the Amerindian component ... seems to indicate that the Native American sub-race is a racially virtually pure branch of the Mongoloid race".
No. What the results indicate is that Native Americans are a racially virtually pure branch of the Native American race.
"For most early native Americans, it was decidedly proto-Mongoloids who admixed - not Mongoloids in any traditional or modern sense".
So we're back to this meaningless 'proto-Mongoloids' expression. What no-one has come up with is an explanation as to what the term actually means. How did the whole population of 'proto-Mongoloids' become 'modern Mongoloids'? Doesn't make sense.
"Mongoloids were already differentiated into northern and southern branches"
What is the 'southern branch of the Mongoloids'? I suspect what you mean here is SE Asians, who we know were not Mongoloid at all until some 5-6000 years ago.
"The Mal'ta individual cannot be ancestral to Native Americans as he carries Caucasoid haplogroups (Y-DNA hg R and mtDNA hg U) that are lacking in pure Native Americans"
We all know that haplogroups are not a close indicator of aDNA. The researchers showed (to my satisfaction at least) that the Mal'ta boy does have a genetic element found in Native Americans but not in East Asians.
Sorry Dienekes. This is three posts. However I think it raises important points.
"so in your opinion we have c. 45.000 years ago (in Super-Grandfathers paper it is 34.000 years ago) KMNOPS in Sundaland, and yDNA C/Australoids in Australia and in continental East Asia".
Yes. But C3 in East Asia was already differentiating autosomally from C4 in Australia, C5 in South Asia, C2 in Wallacea and C* in South China/SE Asia. And possibly C1 was forming in Japan, although some research suggests it branched off C3 some time more recently.
"I suppose that KMNOPS people were negritos at that time".
Possibly, but the Negrito phenotype may be a response to selection in tropical jungle.
At the same time, 'African' YDNA D and mtDNA G/D (?) were already in North China".
And mt-DNA A and N9, as well as Y-DNA C3.
"Then, YDNA P goes to India with mtDNA R through the areas inhabited by ancient South Indians, and from there to Central Asia c. 40.000 years ago".
Yes. That combination of haplogroups must have utilised the South Asian environment in a manner the earlier inhabitants did not. My guess is that the boating technology require to cross Wallace's Line enable a more efficient use of rivers and lakes, and even the coast.
"There it leaves behind mtDNA R and M and takes along mtDNA X. In Central Asia these people become Caucasoids with bloodtype O".
Basically, although R-derived mt-DNA haplogroups are very common west of Altai/Hindu Kush. They must have accompanied Y-DNAs Q and R to start with as they moved north.
"While P is in India, NO originates in Yunnan, c. 40.000 years ago".
Yes.
"O starts spreading in China and arrives to the North Asia very quickly".
I'm sure the major part of O's expansion is Neolithic. But that expansion is from north of the Yangtze River.
"Before the Ice Age NO and O go up to the Baikal area with mtDNA C and A and develop Mongoloid traits together with yDNA C3 and mtDNA B and yDNA D and mtDNA D".
Y-DNA O was yet to differentiate from NO, and perhaps the movement was not as far north as the Baikal region.
"Circa 35.000 years ago yDNA Q-M120 brings the Upper Palaeolithic Technique to China and again leaves behind its original women and mixes with women carrying Eastern mtDNA. These Q-M120 people never make it to America".
I suspect Q-M120 arrived pretty much with Q-L56 although the latter may have been a more northerly group. Just the latter made it to America.
"The mainstream Q is heading to Baikal and meets Mongoloid yDNA C3, D and NO people. YDNA Q passes over the earlier inhabitants. Before the onset of the Ice Age, YDNA Q presses on to Beringia and again leaves behind mtDNA X and takes along women carrying D, C and A lines. (I conclude that Eastern women were prettier as they are always preferred)".
I don't think 'preference' is involved. The region north of the eastern Y- and mt-DNAs was uninhabited until Q and entourage arrived. Men would have moved ahead of the advancing wave and been the first of the advancing group to meet locals along their southward margin. They would have brought them back to join their Y-DNA Q and mt-DNA X tribes.
"These paleo Native Americans become bloodtype O, although their women come from the areas with a high frequency of bloodtype A and B".
It is possible that some sort of selection is involved in the O bloodtype of Native Americans.
"NO, D and C3 recede back to the South in big numbers to change the looks of all southerners".
Actually N goes north to increase the Mongoloid element in the north.
"we can change the route and say that yDNA Q passes through North China instead of Baikal area".
Wherever it passed it would have been in uninhabited regions to the north of any earlier inhabitants.
"we can go back to the theory that Mongoloid traits are not linked with the cold climate but developed in the Chinese heartland, but then again it is not so simple to explain why these traits were passed on to Native Americans and why EDAR mutation is most frequent in Baikal area and America".
That is the central problem Eurologist and Onur are unable to explain.
For clarity, from now on I will say "paleo-northern Mongoloids" and "paleo-southern Mongoloids" rather than "northern Mongoloids" and "southern Mongoloids".
Onur, that is getting closer to my terminology. Except that I think that autosomal DNA indicates that Native-Americans-to-be separated in NE Siberia and Beringia 35,000 - 15,000 ya, and as such, at the beginning of this process, there was only one paleo-Mongoloid group/ culture available for admixture - the split you are indicating happening later, albeit near simultaneously, in part due to this admixture.
Also, in such an old population, tools like Admixture will always make them look pure and unadmixed - that carries little meaning by itself, because we have too few ancient reference populations (ancient DNA).
I do not how you can reconcile the idea that the ancestors of Native Americans passed through uninhabited regions with the idea that they were however gradually and progressively absorbing Eastern mtDNA from East Asian Mongoloids.
Then I do not understand why so many people want to have Mongoloid yDNA NO to have developed in the south in the middle of Australoids and Negritos and yet to be the vector of traits that can in many cases be understood as an adaptation to the cold northern climate.
BTW, according to Wikipedia, the Ordos culture [in North China] is documented from the Upper Palaeolithic. The points and sides of their tools indicate a "Moustero-Levalloisian" element. They seemed to have a masterful knowledge of Upper Palaeolithic technology, producing blades as much as fifteen centimeters long. The human fossil remains of the Ordos Man from Salawusu site dated between 50,000 and 35,000 BCE show strong Mongoloid features, specifically on the fore-tooth and occipital bone.
Ordos man and Ting-ts’un man also had shovel shaped incisors.
"It should be clear by now that you and I won't reach consensus on this issue"
Yes. I'm reminded of Hamarfox's comment above.
"there was only one paleo-Mongoloid group/ culture available for admixture"
I agree absolutely. And I'll get back to that.
"in such an old population, tools like Admixture will always make them look pure and unadmixed - that carries little meaning by itself"
Correct.
"I do not how you can reconcile the idea that the ancestors of Native Americans passed through uninhabited regions with the idea that they were however gradually and progressively absorbing Eastern mtDNA from East Asian Mongoloids".
I can't see where you see the problem. Surely as the ancestors of Native Americans expanded through uninhabited regions they would have filled all the available habitat. At the southern margins of that habitat they would have encountered eastern groups who had already reached the northern limit of their own expansion. In fact the eastern people probably occupied habitat that the incoming people would have regarded as preferable to what they were actually expanding through.
"I do not understand why so many people want to have Mongoloid yDNA NO to have developed in the south in the middle of Australoids and Negritos"
That is certainly not the scenario I envisage, although Onur seems to see things that way. To me it is obvious that NO developed somewhere north of at least the Middle Yellow River. Y-DNA O later moved back south through regions occupied by Australoids and Negritos.
"the vector of traits that can in many cases be understood as an adaptation to the cold northern climate".
I am sure both the Mongoloid phenotype and NO haplogroups expanded from some region to the northwest of the Middle Yellow River. And (surprisingly??) that is exactly the region the authors of the paper on the subject postulated to be geographic origin of the EDAR370A mutation. Even more interesting is this:
"The human fossil remains of the Ordos Man from Salawusu site dated between 50,000 and 35,000 BCE show strong Mongoloid features, specifically on the fore-tooth and occipital bone".
And carried the EDAR variation?
That is exactly where we would expect to find such a fossil human, and fits exactly my comments above. See:
http://archaeology.about.com/od/sterms/g/salawusu.htm
"The Salawusu site is located in Ordos, Nei Menggu (Inner Mongolia), China. Excavated in 1922, the site is an early Homo sapiens site, where bone fragments named Ordos Man have been dated between 50,000 and 37,000 years old".
It is almost in the centre of the EDAR gene's geographic origin. Right in the northern loop of the Yellow River.
Well, well, well. Check this out:
http://blogs.discovermagazine.com/gnxp/2013/11/long-first-age-mankind/#.Uo0eOcSkrIU
"I do not how you can reconcile the idea that the ancestors of Native Americans passed through uninhabited regions with the idea that they were however gradually and progressively absorbing Eastern mtDNA from East Asian Mongoloids".
Here's what Edward Vajda is quoted as saying:
"The gist of his argument is that ca. 15kya, there was a population in south central Siberia that developed a micro-blade technology, which proved to be a 'killer app' that allowed for rapid expansion. This group probably spoke Proto-Dene-Yeniseian (or, at least, the PDY speakers were early adopters). Male-only groups formed the leading edge of this expansion, absorbing existing female lineages in northeastern Siberia; these groups were the first to cross Beringia into the New World, bringing with them the non-Dene-Yeniseian languages spoken by most Indians. Meanwhile, the core micro-blade population, including males and females, expanded more gradually into the whole of Siberia. After thousands of years, one of these Dene-Yeniseian-speaking groups crossed the Pacific and settled what’s now the Pacific coast of Canada; this was the originally Athabaskan-Eyak-Tlingit population".
Seems I'm not alone thinking that way. And, of course, I have to make sure you've all seen this:
http://dienekes.blogspot.co.nz/2013/11/ancient-dna-from-upper-paleolithic-lake.html
Quote:
"Similarly, we find autosomal evidence that MA-1 is basal to modern-day western Eurasians and genetically closely related to modern-day Native Americans, with no close affinity to east Asians ... Furthermore, we estimate that 14 to 38% of Native American ancestry may originate through gene flow from this ancient population".
terryt wrote,
"Yes. But C3 in East Asia was already differentiating autosomally from C4 in Australia, C5 in South Asia, C2 in Wallacea and C* in South China/SE Asia. And possibly C1 was forming in Japan, although some research suggests it branched off C3 some time more recently."
I have neither seen nor heard of any research that suggests that C1-M8 has diverged from C3-M217 more recently than it has diverged from C5-M356 or C6-V20. C1-M8 has diverged from European C6-V20, South/Central/West Asian C5-M356, and perhaps also Wallacean/Oceanic Austronesian C2-M38 more recently than it has diverged from C3-M217.
Furthermore, you should take care to remember that C3-M217 is a very old haplogroup, with not only an ancient split from the ancestor of the other extant branches of haplogroup C, but also an ancient internal split into one branch that has been found almost exclusively in "northern" populations (Mongols, Manchus, etc.) and another branch that has been found widely throughout eastern (and southeastern) Asia, including some cases from the Altay-Sayan-Baykal area of northern Asia. This split between "exclusively northern" C3-M217 and "ubiquitously Eastern Eurasian" C3-M217 has occurred in roughly the same era as the earliest attested split within P-M45 (between Q-M242 and R-M207) and the earliest attested split within O-M175 (between O3-M122 and O1'2-F75).
"I have neither seen nor heard of any research that suggests that C1-M8 has diverged from C3-M217 more recently than it has diverged from C5-M356 or C6-V20".
I was thinking of a paper on the age TMRCA of various Y-DNAs where CF was placed at 56,000 years ago and C1-C3 at 35,000 years go. This is obvioulsy more recent than the C4-C3 split if, as seems very likely, C reached Australia some 45,000 years ago. Unfortunately I can't find the paper at present. The following phylogeny seems to agree with what you've written but makes a simple split between C3 and C1 very unlikely:
http://www.phylotree.org/Y/tree/
Of course this phylogeny was not available when I made my comment above. Perhaps you have contributed to this phylogeny?
I've found Dienekes' post on the subject:
http://dienekes.blogspot.co.nz/2012/07/estimating-age-of-y-chromosome-adam_30.html
And the C1-C3 split date specifically at his update:
http://dienekes.blogspot.co.nz/2012/08/dates-of-major-clades-of-y-chromosome.html
"C3-M217 is a very old haplogroup, with not only an ancient split from the ancestor of the other extant branches of haplogroup C, but also an ancient internal split into one branch that has been found almost exclusively in 'northern' populations (Mongols, Manchus, etc.) and another branch that has been found widely throughout eastern (and southeastern) Asia, including some cases from the Altay-Sayan-Baykal area of northern Asia".
I'm having trouble reconciling the ISOGG and the Y-tree I linked to above. The big split you mention is, in ISOGG, obviously C3b-L1373 and C3e-M546. But that leaves ISOGG's C3a-M93, C3c-P53.1, C3d-P62 and C3f-JST002613-27 not assigned to either main group. Do you think these are genuinely independent or do they actually belong to one of the two main C3 clades?
terryt wrote,
"Of course this phylogeny was not available when I made my comment above. Perhaps you have contributed to this phylogeny?"
No, I have not contributed to the construction or publication of that phylogenetic tree.
terryt wrote,
"I'm having trouble reconciling the ISOGG and the Y-tree I linked to above. The big split you mention is, in ISOGG, obviously C3b-L1373 and C3e-M546. But that leaves ISOGG's C3a-M93, C3c-P53.1, C3d-P62 and C3f-JST002613-27 not assigned to either main group. Do you think these are genuinely independent or do they actually belong to one of the two main C3 clades?"
Please consider my "time slice" (to borrow a term from Eurologist) analysis of the data from Yan et al. 2013:
"Era of the Common Ancestor of Modern Greater Eurasian Y-DNA"
TMRCA of CF-P143 and DE-YAP: 54,100 [95% CI 50,600 to 58,200] YBP
TMRCA of F-M89 and C-M130: 54,000 [50,100 to 58,000] YBP
"Era of the Initial Diversification of F-M89"
TMRCA of IJK-M523 and G-M201: 35,800 [33,300 to 38,300] YBP
TMRCA of NO-M214 and P-M45: 33,000 [30,900 to 35,200] YBP
"Era of the Initial Diversification of NO-M214"
TMRCA of N-M231 and O-M175: 30,000 [27,900 to 32,000] YBP
TMRCA of C3-F1144 and C3-F1396: 25,900 [22,500 to 29,400] YBP
"Era of the Initial Diversification of O-M175 and the Initial Diversification of P-M45"
TMRCA of C3-F1144 and C3-F1396: 25,900 [22,500 to 29,400] YBP
TMRCA of O3-M122 and O1'2-F75: 24,700 [23,000 to 26,500] YBP
TMRCA of Q-M242 and R-M207: 24,100 [21,500 to 26,700] YBP
TMRCA of O1a-M119 and O2-M268: 23,400 [21,600 to 25,300] YBP
TMRCA of O2a-F1462 and O2b-M176: 21,500 [19,500 to 23,600] YBP
"Era of the Initial Diversification of O3a-M324"
TMRCA of O2a-F1462 and O2b-M176: 21,500 [19,500 to 23,600] YBP
TMRCA of O3a1-KL1 and O3a2-P201: 18,900 [17,300 to 20,300] YBP
TMRCA of O3a2b-M188 and O3a2c-P164: 18,100 [16,600 to 19,600] YBP
"Era of the Initial Diversification of O3a2-P201 and the Initial Diversification of N-M231"
TMRCA of O3a1-KL1 and O3a2-P201: 18,900 [17,300 to 20,300] YBP
TMRCA of O3a2b-M188 and O3a2c-P164: 18,100 [16,600 to 19,600] YBP
TMRCA of N-F1206 and N-F2930: 15,800 [13,800 to 18,000] YBP
"Era of the Initial Diversification of O3a2c-P164 and the Initial Diversification of N-M231"
TMRCA of O3a2b-M188 and O3a2c-P164: 18,100 [16,600 to 19,600] YBP
TMRCA of N-F1206 and N-F2930: 15,800 [13,800 to 18,000] YBP
TMRCA of O3a2c1-M134 and O3a2c2-N6: 15,500 [13,900 to 17,100] YBP
"Era of Diversification of N-M231, O3a1-KL1, and O3a2c1-M134"
TMRCA of N-F1206 and N-F2930: 15,800 [13,800 to 18,000] YBP
TMRCA of O3a2c1-M134 and O3a2c2-N6: 15,500 [13,900 to 17,100] YBP
TMRCA of O3a1a-002611 and O3a1b-F964: 13,900 [12,500 to 15,400] YBP
TMRCA of O3a2c1a-M117 and O3a2c1b-F444: 13,300 [11,800 to 14,900] YBP
"Era of Diversification of O3a2c1-M134, O3a1a-002611, and O1a-M119"
TMRCA of O3a2c1a-M117 and O3a2c1b-F444: 13,300 [11,800 to 14,900] YBP
TMRCA of O3a1a1-F11 and O3a1a2-F449: 11,200 [9,800 to 12,500] YBP
TMRCA of O1a1-P203 and O1a2-M110: 10,400 [8,400 to 12,400] YBP
"Era of the Chinese Neolithic Expansion"
TMRCA of O-F11: 6,800 [5,900 to 7,800] YBP
TMRCA of O-F46: 6,500 [5,500 to 7,500] YBP
TMRCA of C3e1a-M407 and C3-F1144(xM407): 6,500 [5,200 to 7,900] YBP
TMRCA of O-M117: 5,400 [4,100 to 6,700] YBP
(Continued...)
Note that the linkage between C-M130 and F-M89 is very weak (or brief); F-M89 and C-M130 are each almost as closely related to DE-YAP as they are related to each other.
Then, for about one third of the time between the split of CF and DE and the present, F-M89 produced only one line of direct patrilineal descent that has persisted to the present day. After this long span of time had passed uneventfully, F-M89 appears to have produced a myriad of descendant lines in quick succession, almost simultaneously (G-M201, H-L901, I-M170, J-M304, NO-M214, P-M45, etc.).
C3-M217 initially splits into C3-F1144 (equivalent to ISOGG C3e-Z1338, which includes Z1300, which includes M407) and C3-F1396 far back in the "Era of the Initial Diversification of NO-M214" or "Era of the Initial Diversification of O-M175 and the Initial Diversification of P-M45." However, each of the two primary subclades of C3-M217 does not internally diversify until much, much more recently (Mesolithic/Neolithic). It seems likely to me that Yan's Northern (F1396) subclade of C3-M217 is roughly equivalent to ISOGG C3b-L1373, which includes North American C3b1-P39 and North Asian (mainly Tunguso-Turkic) C3b2-M48.
The other subclades of C3-M217 that you seem to be troubled about are very minor or singleton-type ("private") clades that have been found in Japan or China IIRC, so they might turn out to belong to the "Southern" branch of C3-M217 (ISOGG C3e-M546/Z1338/F2613). I am unsure whether they already have been tested for any of the SNPs for ISOGG C3e. SNPs that have been found in single individuals (like M93 and P62 IIRC) are not really informative (you can't track the individual's history; his patrilineal ancestor might have immigrated from anywhere in a quite recent era).
By the way, in contrast to the North American distribution of C3b1-P39 and the North Asian distribution of C3b2-M48, the 1000 Genomes Project and HGDP samples include representatives of C3e-F2613/Z1338(xC3e1-Z1300) in a closely related pair of Dai (ethnic Thais) in Xishuangbanna, Yunnan, a Kinh (ethnic Vietnamese) in Ho Chi Minh City, Vietnam, and a Japanese in Tokyo, Japan. C3e1-Z1300(xC3e1a-M407) is found in three Han in Beijing and a Bengali in Bangladesh. C3e1a-M407 is found in a Japanese in Tokyo, Japan. There is also a Han from Hunan or Fujian in southern China who appears to be outside of both C3b-L1373 and C3e-F2613/Z1338, but who I suspect is probably a representative of a sort of "proto-C3e."
Thanks for taking the time to provide all that data.
"F-M89 and C-M130 are each almost as closely related to DE-YAP as they are related to each other".
That makes sense.
"for about one third of the time between the split of CF and DE and the present, F-M89 produced only one line of direct patrilineal descent that has persisted to the present day. After this long span of time had passed uneventfully, F-M89 appears to have produced a myriad of descendant lines in quick succession, almost simultaneously"
Which suggests F did not take part in the C, D, E expansion. In other words it was probably 'trapped' for a considerable time in the region CT had reached soon after its exit from Africa.
"It seems likely to me that Yan's Northern (F1396) subclade of C3-M217 is roughly equivalent to ISOGG C3b-L1373, which includes North American C3b1-P39 and North Asian (mainly Tunguso-Turkic) C3b2-M48".
That is the conclusion I came to.
"The other subclades of C3-M217 that you seem to be troubled about are very minor or singleton-type ('private') clades that have been found in Japan or China IIRC, so they might turn out to belong to the 'Southern' branch of C3-M217"
They might not either. I know Maju disregards 'private lineages' but I believe they are extremely interseting in their own right. In many cases they probably represent lineages once more common than they are today.
"SNPs that have been found in single individuals (like M93 and P62 IIRC) are not really informative (you can't track the individual's history; his patrilineal ancestor might have immigrated from anywhere in a quite recent era)".
That cautionary element would also apply to the single individuals you list within the 'southern' C3.
terryt wrote,
"They might not either. I know Maju disregards 'private lineages' but I believe they are extremely interseting in their own right. In many cases they probably represent lineages once more common than they are today."
They might represent ancient lineages that once have been more common than they are today...or they might represent some descendants of that one sampled individual's father or grandfather. It is not possible to disprove either of these two hypotheses with only a singleton, so, for scientific purposes, the SNP is not informative unless and until it has been detected in another individual.
"That cautionary element would also apply to the single individuals you list within the 'southern' C3."
No, it would not apply (unless you have intended to limit the scope of your comment to M407 or my hypothetical "proto-C3e," and even then only in the context of the limited data from the 1000 Genomes Project/HGDP). Even in that limited data set, C3e-F2613/Z1338(xC3e1-Z1300) has been found in several individuals of unrelated ethnicity (Thai, Vietnamese, and Japanese) in Yunnan, Vietnam, and Japan, and C3e1-Z1300 (which includes M407) has been found in several individuals of unrelated ethnicity (Han, Bengali, and Japanese) in Beijing, Bangladesh, and Japan. This means that "Southern" C3 (C3e-F2613/Z1338) is distributed through a wide and ethnically diverse area that extends to the southwest at least as far as Bangladesh, to the southeast at least as far as Ho Chi Minh City in southern Vietnam, to the northwest at least as far as Beijing in northern China, and to the northeast at least as far as Tokyo in eastern Japan. That is much more informative than any clade that is represented in the literature by only a single individual.
By the way, just in case anyone might have failed to notice: I have produced my "time slice analysis" by separating the TMRCA estimates of various clades published by Yan et al. (2013) into groups in which the 95% confidence interval of each TMRCA estimate overlaps to some extent with the 95% confidence interval of every other member of the group. In other words, they represent groups the order of appearance of whose constituent clades we cannot be certain, but that can be confidently placed in chronological order relative to every other group.
Post a Comment