February 04, 2011

Sailing across the Pacific to settle Polynesia (Soares et al. 2011)

I always have a hard time following these easy archaeological interpretations on the basis of uniparental markers, both because I'm convinced that they are not supported by the wide confidence intervals of age estimates, and because, as I've argued countless times, age of colonization != age of most recent common ancestor of colonists' descendants.

With respect to the latter point, let me just reiterate that:
  • The common founder of a set of lineages may postdate the colonization event, if the number of colonists was small enough so that attrition was high enough, and a founder that lived long after the colonization effect contributed most of the present-day population
  • The common founder of a set of lineages may predate the colonization event, if the number of colonists was high enough, so that multiple related lineages with a founder who lived before the event, survived into the modern population.
We should all take modern genetic-archaeological-prehistorical correlations (including my own!) with a huge grain of salt. Polynesia seems like the last place on earth likely to yield ancient DNA of any antiquity, due to the combination of heat and moisture, but, on the whole, a single well-dated and authenticated ancient DNA sample carries much more weight than all the modern DNA put together.

From the press release:
Surprising new evidence which overturns current theories of how humans colonised the Pacific has been discovered by scientists at the University of Leeds, UK.

The islands of Polynesia were first inhabited around 3,000 years ago, but where these people came from has long been a hot topic of debate amongst scientists. The most commonly accepted view, based on archaeological and linguistic evidence as well as genetic studies, is that Pacific islanders were the latter part of a migration south and eastwards from Taiwan which began around 4,000 years ago.

But the Leeds research – published today in The American Journal of Human Genetics – has found that the link to Taiwan does not stand up to scrutiny. In fact, the DNA of current Polynesians can be traced back to migrants from the Asian mainland who had already settled in islands close to New Guinea some 6-8,000 years ago.

...

Professor Richards and co-researcher Dr Pedro Soares (now at the University of Porto), argue that the linguistic and cultural connections are due to smaller migratory movements from Taiwan that did not leave any substantial genetic impact on the pre-existing population.

The American Journal of Human Genetics, 03 February 2011
doi:10.1016/j.ajhg.2011.01.009

Ancient Voyaging and Polynesian Origins

Pedro Soares et al.

Abstract
The “Polynesian motif” defines a lineage of human mtDNA that is restricted to Austronesian-speaking populations and is almost fixed in Polynesians. It is widely thought to support a rapid dispersal of maternal lineages from Taiwan ∼4000 years ago (4 ka), but the chronological resolution of existing control-region data is poor, and an East Indonesian origin has also been proposed. By analyzing 157 complete mtDNA genomes, we show that the motif itself most likely originated >6 ka in the vicinity of the Bismarck Archipelago, and its immediate ancestor is >8 ka old and virtually restricted to Near Oceania. This indicates that Polynesian maternal lineages from Island Southeast Asia gained a foothold in Near Oceania much earlier than dispersal from either Taiwan or Indonesia 3–4 ka would predict. However, we find evidence in minor lineages for more recent two-way maternal gene flow between Island Southeast Asia and Near Oceania, likely reflecting movements along a “voyaging corridor” between them, as previously proposed on archaeological grounds. Small-scale mid-Holocene movements from Island Southeast Asia likely transmitted Austronesian languages to the long-established Southeast Asian colonies in the Bismarcks carrying the Polynesian motif, perhaps also providing the impetus for the expansion into Polynesia.

Link

44 comments:

German Dziebel said...

"Polynesia seems like the last place on earth likely to yield ancient DNA of any antiquity, due to the combination of heat and moisture, but, on the whole, a single well-dated and authenticated ancient DNA sample carries much more weight than all the modern DNA put together."

MtDNA samples from Lapita remains didn't contain 9 bp deletion, the main differentiator of haplogroup B. See Erika Hagelberg's work from the late 1990s. This undermines the out of Taiwan theory and is consistent with the results of the current paper.

ren said...

"We should all take modern genetic-archaeological-prehistorical correlations (including my own!) with a huge grain of salt."

You have a very good point there. These genetic papers and their rewriting of prehistory, overturning all the linguistic, archaeological, physical anthropological evidence, is the biggest scam of the genetics era. There is no archaeological evidence for migration from East Asia 9,000 years ago. Just because the mtDNA is (faultily) dated to that time does not mean it happened contrary to the findings of other disciplines. The arrogance of these geneticists, who usually don't have any grasp of other disciplines to have a good feel and interpretation, is just jaw dropping.

terryt said...

"Surprising new evidence which overturns current theories of how humans colonised the Pacific"

More than a little bit of journalistic overstatement there. It hardly overturns anything. I'm sure no-one actually believes that the Polynesians left Taiwan and headed straight out to New Zealand. For as long as I've been studying the pattern it has been assumed that the migrants mixed with people they met along the way. The presence of Y-haplogroup C2 in the Polynesians of New Zealand is alone enough to show that. The only 'new' bit is:

"This indicates that Polynesian maternal lineages from Island Southeast Asia gained a foothold in Near Oceania much earlier than dispersal from either Taiwan or Indonesia 3–4 ka would predict".

But, as yoy say, the truth of that statement depends on how many women were involved in the original movemt to Near Oceania.

lkdjf said...
This comment has been removed by the author.
Jack said...

I don't fully understand what TruthPlease is writing.
"any "find" that seems to politically benefit western europeans is false, obviously."
"the only thing you can find here is steps in the direction of benefit for europeans one after the other"
It's not important, but if you could Please clarify.
Do you claim that "Europeans" (I suppose you mean white people) are altering data?
Just curious.

lkdjf said...
This comment has been removed by the author.
|\/| said...

Διηνέκη, δες εδώ. ;)

terryt said...

"allot of asians made it here from the south pacific".

There's no evidence for that at all. The only thing that can be said regarding the South Pacific and America is that some evidence suggests that some Polynesians may have made it that far. The sweet potato was spread through the South Pacific, presumably from South America.

"from what i gather there was indians in the america's from the pacific also".

Yes, but they entered North America from Northeast Asia. And many probably entered by walking although it is at least very likely that some entered by boat along the coastline.

"the only thing you can find here is steps in the direction of benefit for europeans one after the other".

That's possibly why New Zealand has an advantage in indigenous genetic studies. Many Maori are involved in the study and most New Zealanders of both European and Polynesian ancestry agree totally on the broad outline of indigenous origin and settlement.

terryt said...

Good point. I used to say, during the genocide in Rwanda and Burundi, that the people there were just trying to make room for everybody when we all went back to where we came from.

German Dziebel said...

"These genetic papers and their rewriting of prehistory, overturning all the linguistic, archaeological, physical anthropological evidence, is the biggest scam of the genetics era. There is no archaeological evidence for migration from East Asia 9,000 years ago."

There's no archaeological evidence for a founding migration out of Africa, there's no archaeological evidence for a founding migration to the New World.... Archaeology is not always helpful with providing evidence for a population process. Everything that it finds is compatible with diffusion or independent invention.

I think the article (judging by the abstract, as I haven't had a chance to read it yet) articulates the best existing alternative to the out of Taiwan theory pretty well. 1) The presence of a "Paleolithic" lineage C2 in Polynesians at high frequencies; 2) the fact that at all touchpoints between Austronesians and non-Austronesians in the Pacific C2 admixes from the former to the latter and not the other way around; 3) the sheer amount of Pleistocene lineages, especially Y-DNA, in Austronesians; 4) the recently suggested linguistic link between Austronesian and Andamanese languages are, IMO, the strongest indicators that the recent out of Taiwan theory for the origin of Austronesians is likely wrong. See my discussion with Terry well before this paper came out: http://dienekes.blogspot.com/2010/08/social-selection-in-y-chromosome.html

Archaeologists just need to dig deeper...

ren said...

TruthPlease claims to be of European descent. I don't know if this helps anyone in framing a preconception.

pconroy said...

German,
Are you aware of the paper from a few years ago that found mtDNA M in ancient samples from British Columbia?
http://mathildasanthropologyblog.wordpress.com/2008/11/03/mitochondrial-haplogroup-m-discovered-in-prehistoric-north-americans/

What do you make of that?

pconroy said...

Ren,
Whether TruthPlease is European, African or Asian is of no matter to me.

The fact is that any group/tribe or society will commit genocide - either intentional or unintentional (disease) - if/when they run low on resources in their own territory and face famine or are being marginalized by another group and so move to another territory. Just as most ethnic groups have been slave owners/owned as slaves at one point in their history, just like all societies/ethnicities reach a Golden Age at some time or other, then later a not so Golden Age(s). It's just human nature, that people under duress or facing death will do whatever they can to survive, even if it means the non-survival of other groups - that's all.

To be super angry about that is IMO irrational!

lkdjf said...
This comment has been removed by the author.
terryt said...

"1) The presence of a 'Paleolithic' lineage C2 in Polynesians at high frequencies"

Some time back Ebizur provided a whole heap of evidence that showed C2 originated in Nusa Tengarra. It moved north about the time the convention claims the Austronesian-speaking people would have arrived there.

"2) the fact that at all touchpoints between Austronesians and non-Austronesians in the Pacific C2 admixes from the former to the latter and not the other way around"

That's correct. Members of it dropped off along the way.

"3) the sheer amount of Pleistocene lineages, especially Y-DNA, in Austronesians"

Which Pleistocene lineages, apart from C2?

"4) the recently suggested linguistic link between Austronesian and Andamanese languages are, IMO, the strongest indicators that the recent out of Taiwan theory for the origin of Austronesians is likely wrong".

A far as I know that link is not yet widely accepted. However we do know that just to the south of the Andamans Austronesian languages are spoken on the Nicobars.

A 1984 book I glanced at yesterday ("Where the Waves Fall", K. R. Howe) started off thus:

"The story of the first human settlement of the Pacific islands is now known at least in broad outline"

As I said, that is from 1984. A few questions remain unanswered from slightly before the entry to the Pacific though.

Are the Hoabinhian, the Austo-Asiatic language and the expansion of Y-hap O2a all the same phenomenon or are all three separate? The Hoabinhian is the most restricted of the three geographically but the Y-hap and the language could have spread beyond the technology. And the Andaman Islands seem to have been first settled by a people whose technology sprang from the same source as did the Hoabinhian.

Another question is: were the Philippines, Sulawesi and Halmahera no more than sparsely inhabited until the Austronesian-speaking people arrived on them? And possibly Borneo? Mitochondrial DNA haplogroup M7 is certainly a likely candidate as is some sort of Y-hap C(xC2,C3) and F2.

A third question: Where did mtDNA B coalesce? It has become spread around the pacific from New Zealand to South America. It certainly liked to be beside the sea.

German Dziebel said...

@pconroy

"Are you aware of the paper from a few years ago that found mtDNA M in ancient samples from British Columbia?"

Absolutely. They found a new basal M lineage without any derived mutations found in the Old World. So it's Mx. This is very consistent with 1) the fact that better sampling and ancient DNA in the New World furnish more and more distinct Amerindian lineages; 2) Amerindian lineages tend to share only basal mutations with their Old World counterparts. It's noteworthy that studies that argue for a South Asian origin for macrohaplogroup M continue to overlook the discovery of Mx in North America and, correspondingly, the fact that basal M lineages span all of East Asia, Oceania, Australia and North America, which only suggests that, just like macrohaplogroup N, mhg M originated east of India and then migrated west all the way into East Africa (M1). This confirms the huge geographical gap between the putative ancestors of M and N in Africa (L3) and the areas of likely origin of M and N clusters. It's a huge biogeographic problem for the out-of-Africa theory.

From the out-of-America perspective, Amerindians are a refugium of a bunch of small, residual mtDNA lineages from all the non-African clusters such as D (D1), M (Mx, C), N (X and A) and R (B). These are the stay-behind lineages that weren't affected by the massive expansion of the rest of D, M, N and R clusters around the globe, with East Asia and South Asia being the main centers of lineage proliferation.

German Dziebel said...

@Terry

"Some time back Ebizur provided a whole heap of evidence that showed C2 originated in Nusa Tengarra."

I don't care where, geographically, C2 "originated." What's important is that it's an Austronesian-specific lineage that 1) is of "Pleistocene" origin; 2) wasn't picked up from any known Papuan/pre-Austronesian groups but in fact admixed into them from incoming Austronesians; 3) not found in Taiwan.

"Which Pleistocene lineages, apart from C2?"

All "downstream" lineages from C are found in Austronesians outside of Taiwan, especially C (Cx, C2), K, O. Only O is found in Taiwan. (There may be C* there, too, but in minuscule proportions. In Melanesia and Polynesia, only 10-20% of lineages have East Asian origin. It takes some time to absorb them.

"A far as I know that link is not yet widely accepted. However we do know that just to the south of the Andamans Austronesian languages are spoken on the Nicobars."

It just came out of the blue and people don't know how to react to it. Nobody has shown that Blevins's comparison is flawed. Linguists have good ways of distinguishing a recent Nicobarese influence from an ancient kinship between Ongan-Jarawan and the common stock of Austronesian (including Formosan). There's a good match between linguistic and genetic evidence, though: Andamanese are y-DNA D and mtDNA M31'32. Austronesians, outside of Taiwan, have Y-DNA C (same phylogenetic level as D) and mtDNA M27 (Solomon islands) (see massey.genomicus.com/.../Ricaut_2010_JArchaeolSci_v37_p1161.pdf). So, overall, Austronesians look like Andamanese on Holocene steroids, and not like an agricultural population that replaced an indigenous substratum.

German Dziebel said...

"Where did mtDNA B coalesce? It has become spread around the pacific from New Zealand to South America. It certainly liked to be beside the sea."

You know I'm biased towards America as the place where B coalesced. It's most frequent and pervasive there (found even in Eskimos, while otherwise rare in NE Asia and northern parts of North America). But, whatever it is, it's geographical footprint in the Old World suggests antiquity. It's found in the Ainu and in South Siberia but not in mainland NE Asia. Ainu is also one of the rare populations that harbors Y-DNA hg D. mtDNA B is a sister clade of P, which is frequent in Papua New Guinea. At the same time, it's not directly attested in any human remains prior to 2500 BC. Not in Korea, not in Japan, not in Lapita. This could also mean antiquity, as in all those places it had a long time to get lost through drift.

lkdjf said...
This comment has been removed by the author.
terryt said...

"I don't care where, geographically, C2 "originated." What's important is that it's an Austronesian-specific lineage"

Because the Austronesian language has taken over the region where C2 originated. It's expansion is post the arrival of Autronesians in Nusa Tengarra, so it picked up the language at that time.

"1) is of 'Pleistocene' origin"

But the Austronesian language is hardly of ''Pleistocene' origin'.

"2) wasn't picked up from any known Papuan/pre-Austronesian groups"

We don't know what was happening in Nusa Tengarra before the Austronesian speaking people arrived there. My bet is that the people were a sort of Papuan-type people. Just as the people of Timor today are hardly typical Malay. Many exhibit Papuan features yet all are basically Austronesian speakers.

"3) not found in Taiwan".

Completely irrelevant. As I've said elsewhere, many people seem to expect the Austronesian expansion to have been like some biblical Abrahamic expansion. A single small group who, against all odds, maintained their genetic 'purity'.

"It's found in the Ainu and in South Siberia but not in mainland NE Asia".

Confirms its association with coastal migration.

"mtDNA B is a sister clade of P, which is frequent in Papua New Guinea".

And I've long argued with Maju that the two haplogroups originated somewhere around there. He has now conceded an SE Asian origin for mtDNA R, but not as far into SE Asia as Wallacea.

pconroy said...

German,
Are you on 23AndMe?

There was a fascinating discussion there yesterday by an Italian from Southern Italy, whose Y-DNA was C (just C, nothing downstream), and was wondering how this could be. Then another Northern Italian reported also being just C.

This is very strange, what do you think?

Gioiello said...

I don't know if German has something to say. I wrote to this Italian-American (Caldararo)privately and on 23andme hypothesizing either an ancient Italian origin or a recent Gypsy one (given also his surname which could be linked with Calderas etc.) but without a STRs test I think we cannot say anything deeper. Unfortunately this test isn't for now in his budget.

terryt said...

"the most striking thing about phylogenetics is racism weather geopolitical or within the science itself as we see here".

And it goes way back to beyond the Upper Paleolithic. Haplogroups have continuously replaced others. In fact modern haplogroups replaced those of earlier humans. I agree that we should be trying to discourage it these days.

German Dziebel said...

"We don't know what was happening in Nusa Tengarra before the Austronesian speaking people arrived there. My bet is that the people were a sort of Papuan-type people. Just as the people of Timor today are hardly typical Malay. Many exhibit Papuan features yet all are basically Austronesian speakers."

We know, at least from the Philippines, which is a good case study area, that both "papuan-looking" and 'asian-looking" Austronesians have both Late Pleistocene and late Holocene lineages.

You propose a solution whereby asian-looking agriculturalists brought with them Holocene genes and a Holocene language. They imposed their language and their genes onto the indigenous "papuans" but borrowed only "papuan" genes but not "papuan" languages. They absorbed one of "papuan" lineages, namely C2, and drove it to high frequency as they continued their eastward expansion. But everywhere we have non-Austronesian populations still in existence we see that C2 is admixed from Holocene intruders into indigenous "papuans." So, it's more logical to interpret the incoming Austronesians as another "papuan" population (or better to say an indigenous Southeast Asian population with roots in the Late Pleistocene foraging adaptation) that transitioned to agriculture and absorbed East Asian/Mongoloid inflows over millenia that may have brought cultural items with them but that didn't result in language replacement.

"Confirms its association with coastal migration."

South Siberia is not coastal. Is Y-DNA D (high in Tibet, high in Japan) "coastal", in your opinion?

""3) not found in Taiwan".

Completely irrelevant. As I've said elsewhere, many people seem to expect the Austronesian expansion to have been like some biblical Abrahamic expansion. A single small group who, against all odds, maintained their genetic 'purity'."

I don't imagine anything to be that simple. But Taiwan just doesn't have pre-O Y-DNAs (you can correct me on that, as I'm talking from memory), while SE Asia, Melanesia and Polynesia have them in abundance. You interpret it as substratum influence in the latter areas - fine - but even under this scenario I don't think such intense intermingling and complete language replacement could've taken only 5,000 years. Especially since language and gene adoption apparently went in opposite directions: from immigrants to indigenes in the case of language and from indigenes to immigrants in the case of genes (especially, Y-DNA).

"And I've long argued with Maju that the two haplogroups originated somewhere around there. He has now conceded an SE Asian origin for mtDNA R, but not as far into SE Asia as Wallacea."

Genetics is pretty arbitrary as to where exactly individual haplogroups originated. I'm working off of a macropicture: if the New World has representatives of all globally representative clusters (B from R, X from western N, A from eastern N, C and Mx from M and D1 from D), then it's likely that they all dispersed from there (with linguistics, kinship and other data lurking in the background). Population expansion resulted in the proliferation and re-distribution of other lineages around the Old World. America wasn't affected, hence we don't find, e.g., mtDNA P in the New World. If P and B arose where they are currently found, i.e. Oceania, then we would've found them together in America as well.

German Dziebel said...

Paul (I assume p- in pconroy stands for Paul),

No, I'm not on 23andMe. I should be, I suppose. Would you mind sending me the exact link? I poked around Spitoon but didn't find the interesting discussion you mention. At face value, this personal report fits with the discovery of mtDNA C in Iceland. We don't really know what Europe looked like in pre-U times. Maybe it was all C, D, X, and A....

terryt said...

"We know, at least from the Philippines, which is a good case study area, that both 'papuan-looking' and 'asian-looking' Austronesians have both Late Pleistocene and late Holocene lineages".

The lineages and the genes have become thoroughly mixed. As a result we don't find 'papuan-looking' people with Papuan haplogroups and 'asian-looking' people with Asian haplogroups.

"You propose a solution whereby asian-looking agriculturalists brought with them Holocene genes and a Holocene language. They imposed their language and their genes onto the indigenous "papuans" but borrowed only "papuan" genes but not "papuan" languages. They absorbed one of "papuan" lineages, namely C2, and drove it to high frequency as they continued their eastward expansion".

That seems to be what happened.

"But everywhere we have non-Austronesian populations still in existence we see that C2 is admixed from Holocene intruders into indigenous 'papuans'".

What's so surprising about that? Languages are not intimately associated with haplogroups. C2 has obviously spread into the local populations abandoning the language it arrived with.

"So, it's more logical to interpret the incoming Austronesians as another 'papuan' population (or better to say an indigenous Southeast Asian population with roots in the Late Pleistocene foraging adaptation)"

So why do Southeast Asians look far more 'East Asian' than 'Papuan'? Presumably the population was thinly spread before the East Asian phenotype arrived.

"Is Y-DNA D (high in Tibet, high in Japan) 'coastal', in your opinion?"

What has Y-DNA D have to do with things?

"while SE Asia, Melanesia and Polynesia have them in abundance".

It's my understanding that the SE Asian haplogroups have not been further defined. They are actually unlikely to be basal O. To me it is obvious that O moved south with the expansion of the early Chinese Neolithic, but the Chinese are unwilling to suggest that ancient Chinese were expansionist. Their neighbours are already worried that modern Chinese are expansionist.

German Dziebel said...

"The lineages and the genes have become thoroughly mixed. As a result we don't find 'papuan-looking' people with Papuan haplogroups and 'asian-looking' people with Asian haplogroups."

Then why do find phenotypical differences between them if their genes got so thoroughly mixed?

"What's so surprising about that? Languages are not intimately associated with haplogroups. C2 has obviously spread into the local populations abandoning the language it arrived with."

I think I allowed myself to be misunderstood. The problem for me is that, although the recent out of Taiwan theory requires that Y-DNA C2 in Austronesian-speaking Melanesians and Polynesians is explained as emerging as a result of gene flow from indigenes to immigrants, all the facts we actually have at our disposal (C2 frequencies data from both Austronesians and Papuans) seem to indicate that C2 is admixed from immigrants to indigenes, and not the other way around.

"What has Y-DNA D have to do with things?"

Y-DNA hg D is an example of a haplogroup that's found in "coastal" areas (Ainu, Andamanese) but also in the deep mainland (Tibet). The same seem to be the situation with mtDNA hg B: it is indeed found in coastal areas but also thousands of miles away from the coast (South Siberia). In North America, B is not the most frequent haplogroup in California (mtDNA A is), it's absent on the Northwest Coast, but is prominent in the Southwest (including Southern Athabascans), which is very far from the coast.

My point is that its coastal distribution in SE Asia, Oceania and South America may be a coincidence.

"To me it is obvious that O moved south with the expansion of the early Chinese Neolithic."

I have no problem with that. But I don't identify Austronesians with Chinese Neolithic. Austronesians got established in SE Asia and Oceania before the spread of Chinese Neolithic.

"So why do Southeast Asians look far more 'East Asian' than 'Papuan'? Presumably the population was thinly spread before the East Asian phenotype arrived."

Because gene flow from East Asia over millenia made them look Asian. But the Austronesian language was established in SE Asia and Oceania prior to gene flow from East Asia.

terryt said...

"Because gene flow from East Asia over millenia made them look Asian. But the Austronesian language was established in SE Asia and Oceania prior to gene flow from East Asia".

I agree with the first sentence but the second one is where we disagree. I'm certain the Austronesian language was established in SE Asia BY gene flow from East Asia.

"Then why do find phenotypical differences between them if their genes got so thoroughly mixed?"

The genes got thoroughly mixed only in the hybrid zone through what is now Indonesia and Malaysia. To the east the Papuans remain largely unmixed, representing the original phenotype through SE Asia/New Guinea. The East Asian phenotype is the result of a more recent movement south, mixing with the Papuan phenotype in the hybrid zone. Representatives from this hybrid zone moved into the Pacific and became Polynesians.

"Y-DNA hg D is an example of a haplogroup that's found in 'coastal' areas (Ainu, Andamanese) but also in the deep mainland (Tibet)".

But the haplogroup is certainly not 'coastal'. It's not found in coastal regions between The Ainu and Andamans. It's origin in both those regions is almost certainly from a common region inland, perhaps from Tibet or nearby.

"The same seem to be the situation with mtDNA hg B: it is indeed found in coastal areas but also thousands of miles away from the coast (South Siberia)".

But its expansion is far more 'coastal' than is that of Y-hap D. The South Siberia and American distributions are easily explained as a result of subsequent movements from the coast.

"the recent out of Taiwan theory requires that Y-DNA C2 in Austronesian-speaking Melanesians and Polynesians is explained as emerging as a result of gene flow from indigenes to immigrants"

Only in Nusa Tengarra. It was carried elsewhere after being absorbed from there by the Austronesians.

"all the facts we actually have at our disposal (C2 frequencies data from both Austronesians and Papuans) seem to indicate that C2 is admixed from immigrants to indigenes"

Exactly.

German Dziebel said...

"I agree with the first sentence but the second one is where we disagree. I'm certain the Austronesian language was established in SE Asia BY gene flow from East Asia."

To clarify: I don't think Austronesians came to occupy their present territories from anywhere but East Asia (via SE Asia). But I think it preceded the spread of agriculture, the out of Taiwan gene flow and the southward spread of the (Southern) Mongoloid phenotype. It preceded all of these developments. But I can't say by how many millenia.

"But its expansion is far more 'coastal' than is that of Y-hap D. The South Siberia and American distributions are easily explained as a result of subsequent movements from the coast."

I don't think it's explained that way any "easier" than the opposite way. But it's definitely a strong possibility. There's just so much "coast" in SE Asia and Oceania (as well as parts of the New World) that some aspect of human genetic and cultural variation will easily end up looking coastal, whether or not there was a single coastal migration connecting Circumpacific populations. E.g., the Maya Indians who bridge the Pacific and Atlantic coasts of Mesoamerica are rich in mtDNA hg B. Does it make it a signature of a Circumpacific coastal migration? Hg B also reaches fixation in some Andean populations. Does it make it a coastal or an alpine haplogroup?

It's noteworthy, though, that hg B is found at much higher frequencies in the tropical and subtropical than in the arctic and subarctic areas.

Alvah said...

Professor Terry Jones of Cal Poly San Luis Obispo suggests that the Chumash had contact with the Polynesians http://cla.calpoly.edu/~tljones/ fishhook – plank canoe – and name for it “Tomolo”. Geneticists have found the mtDNA B2 (I presume) from Chumash descendants that trace themselves back historically to the Channel Islands off Santa Barbara (Dr. John Johnson SB Museum Natural History). Several papers on ancient mtDNAs studies from ancestral Asians burials older than 4,000 y.b.p. did not find the mtDNA B. Is this just a coincidence? Can anyone site an article where we have either living or ancient mtDNA B2 in Asia (or the precise ancestor for mtDNA B)?

German Dziebel said...

Good points, Alvah. I especially like Lawler's paper "Beyond Kon-Tiki: Did Polynesians Sail to South America?" available for free from his website. I wonder if Terry Jones and Terry Toohill are one and the same person as the identity of their first names strongly suggests?

It's true that hg B wasn't detected in Lapita remains, which is a puzzle. It's also true that in the Old World hg B is not attested in ancient remains prior to 3-4,000 years. Same for South America. In North America, however, hg B dates back to at least 12,300 years BP in Oregon (Gilbert et al. 2008) and 8,000 years BP in Colorado.

I doubt that it came to Polynesia from the Americas recently across the Pacific, though. As far as cultural similarities between Polynesians and Amerindians go, there are also very strong similarities between Amazonian Indians and Highland New Guinea Papuans (some people even invented the term Melazonia). See, e.g., Gender in Amazonia and Melanesia, by Gregor and Tuzin, 2001. But there's no way one could reasonably invoke trans-Pacific contacts to explain those. A better explanation for both Polynesian-Amerindian and Papuan-Amerindian cultural similarities is common heritage variably retained from the time when all these populations shared a homeland in the New World (according to the out-of-America hypothesis) or in the Old World (according to the out-of-the-Old World orthodoxy).

terryt said...

"I wonder if Terry Jones and Terry Toohill are one and the same person as the identity of their first names strongly suggests?"

No.

"It's noteworthy, though, that hg B is found at much higher frequencies in the tropical and subtropical than in the arctic and subarctic areas".

As would any self-respecting haplogroup. It's only people who've been recently forced into undesirable habitats by population pressure that are found there. But that doesn't mean that other's have not moved through such environments before being able to emerge into more desirable habitats.

"I don't think Austronesians came to occupy their present territories from anywhere but East Asia (via SE Asia). But I think it preceded the spread of agriculture, the out of Taiwan gene flow and the southward spread of the (Southern) Mongoloid phenotype. It preceded all of these developments. But I can't say by how many millenia".

That's what I disagree with. For a start, the southward spread of the Mongoloid phenotype almost certainly preceded the Austronesian spread. It may have started with the spread of Austro-Asiatic though.

terryt said...

"It's true that hg B wasn't detected in Lapita remains, which is a puzzle".

What haplogroups were found in the early lapita remains? It's not really suprising that B wasn't found in just the few Lapita examples tested (how many samples?) because other haplogroups accompanied mtDNA B and Y-hap C2 during the early stages from the Admiralties and Bismarks.

"I doubt that it came to Polynesia from the Americas recently across the Pacific, though".

Extremely unlikely. The Polynesians reached Hawai'i no earlier than 400 AD. And I read recently they may have arrived there just 1000 years ago. But what I wouldn't be surprised about at all is if Polynesians made it to America. They sailed widely across the Pacific so it would be almost surprising if they'd failed to reach America.

German Dziebel said...

"For a start, the southward spread of the Mongoloid phenotype almost certainly preceded the Austronesian spread. It may have started with the spread of Austro-Asiatic though."

Mind you, though, that Austro-Asiatic, Austronesian, Tai-Kadai and Ongan-Jarawan may be all related as part of Narrow Austric. If confirmed, this will make SE Asia home to a more diverse bunch of related languages than Taiwan, with Formosan languages possibly forming their own island clade post-dating the breakup of the Austroasiatic, Tai-Kadai, Ongan-Jarawan and Malayo-Polynesian (extra-Formosan Austronesian) languages. Non-Mongoloid phenotypes are found across Austro-Asiatic (including Munda), Ongan-Jarawan and Malayo-Polynesian speakers, which suggests that Southern Mongoloid gene flow affected an already-diversified and localized bunch of communities of speakers of these languages. You suggest that these languages were brought by Southern Mongoloids and imposed onto indigenous non-Mongoloids. This will require the complete adoption by indigenous non-Mongoloids of incoming (likely related) languages in four different regions (India, mainland SE Asia, Philippines and Oceania) within the last 5,000 years. This is less parsimonious than my idea, don't you think? If, on the contrary, we associate Narrow Austric languages with a) Munda, Onga-Jarawa, SE Asian Negritos, Philippine Negritos and Austronesian-speaking Melanesians and b) with Pleistocene genes (Y-DNA D, C2 and mtDNA M31'32, Q, B), then we are only left with the spread of agriculture, the spread of the Southern Mongoloid phenotypes and some genes (Y-DNA C3, O, etc.), which we can now attribute to the Neolithic times. We end up having two dispersal events with a similar geographic signature: NE India, Andaman islands, SE Asia, Oceania for Narrow Austric speakers and NE India, SE Asia, Taiwan, Oceania (not as far east as the Pleistocene one) for Mongoloid and agricultural expansions. The colonization of Polynesia post-dated both of them but was carried out not by Taiwanese but by SE Asian/Wallacean. Austronesians.

terryt said...

"The colonization of Polynesia post-dated both of them but was carried out not by Taiwanese but by SE Asian/Wallacean. Austronesians".

The Polynesians are neither Mongoloid nor Papuan. They are a mixture of the two.

"Austro-Asiatic, Austronesian, Tai-Kadai and Ongan-Jarawan may be all related as part of Narrow Austric".

I'm reasonably sure that they are.

"with Formosan languages possibly forming their own island clade post-dating the breakup of the Austroasiatic, Tai-Kadai, Ongan-Jarawan and Malayo-Polynesian (extra-Formosan Austronesian) languages".

I agree to a large extent, except that I believe that Malayo-Polynesian languages can be best explained as being a branch of Taiwanese Austronesian.

"Non-Mongoloid phenotypes are found across Austro-Asiatic (including Munda), Ongan-Jarawan and Malayo-Polynesian speakers"

To me all those groups, with the possible exception to some extent of Ongan-Jarawan, exhibit a degree of admixture with Mongoloid phenotype.

"You suggest that these languages were brought by Southern Mongoloids and imposed onto indigenous non-Mongoloids".

I would be very surprised if the original languages spoken in South China and SE Asia were other than related to Papuan or Australian Aborigine languages.

"This will require the complete adoption by indigenous non-Mongoloids of incoming (likely related) languages in four different regions (India, mainland SE Asia, Philippines and Oceania) within the last 5,000 years".

Those four regions are contiguous though. And I'd place the incoming languages as beginning their spread south and west as being some 10,000 years ago. And I wouldn't be too sure in claiming the populations that 'adopted' the languages were anything other than significantly admixed with the Mongoloid phenotype.

"the spread of the Southern Mongoloid phenotypes and some genes (Y-DNA C3, O, etc.)"

Y-DNA O has come to completely dominate the region though. It hardly represents a 'minor' spread.

"We end up having two dispersal events with a similar geographic signature: NE India, Andaman islands, SE Asia, Oceania for Narrow Austric speakers and NE India, SE Asia, Taiwan, Oceania (not as far east as the Pleistocene one) for Mongoloid and agricultural expansions".

I strongly suspect many more migrations than just those two. First to Australia carrying the Aborigine languages. Next to New Guinea/Melanesia with the Papuan languages. Next a coastal movement south from the Yangtze region which brought the Austro-Asiatic languages. Then a movement between Taiwan/Philppines, which carried, and then spread, the Austronesian languages. Then a quite recent one of modern Chinese spreading the Han languages. Except for the first two, each of these movements brought more Mongolian phenotype southwards.

German Dziebel said...

"I strongly suspect many more migrations than just those two. First to Australia carrying the Aborigine languages. Next to New Guinea/Melanesia with the Papuan languages. Next a coastal movement south from the Yangtze region which brought the Austro-Asiatic languages. Then a movement between Taiwan/Philppines, which carried, and then spread, the Austronesian languages. Then a quite recent one of modern Chinese spreading the Han languages. Except for the first two, each of these movements brought more Mongolian phenotype southwards."

I'm fine with all of them. I just think that the spread of Austro-Asiatic and Austronesian languages predated the spread of the Mongoloid phenotype.

"Y-DNA O has come to completely dominate the region though. It hardly represents a 'minor' spread."

No, not minor. And the footprint of hg O does match with the footprint of the (Southern) Mongoloid phenotype. But then it forms a clade with N, which corresponds to the "Uralic" phenotype, and S, which is frequent in Papua New Guinea (comp. mtDNA B among Austronesians and P among Papuans). At the same time, Northern Mongoloids are high in hg C and don't have their own "member" within the MNOPS clade.

"The Polynesians are neither Mongoloid nor Papuan. They are a mixture of the two."

I doubt it. Polynesians may form their own phenotypic taxon (just like Uralics above), which of course exchanged genes with Papuans and absorbed some of the Mongoloid gene flow, too.

"To me all those groups, with the possible exception to some extent of Ongan-Jarawan, exhibit a degree of admixture with Mongoloid phenotype."

Yes. The Narrow Austric unity seems to precede this admixture, just like the Austronesian unity.

terryt said...

"I just think that the spread of Austro-Asiatic and Austronesian languages predated the spread of the Mongoloid phenotype".

That's where we disagree. We'll have to wait until Chinese geneticists are no longer obsessed with denying any ancient move south from China, and so become able to look at the evidence honestly.

"I doubt it. Polynesians may form their own phenotypic taxon"

Why do you doubt it? The Polynesians have both Papuan and Mongoloid haplogroups and give every appearance of having become a stabilised hybrid between the two groups. That has resulted in their forming 'their own phenotypic taxon'.

"the footprint of hg O does match with the footprint of the (Southern) Mongoloid phenotype".

Exactly.

"But then it forms a clade with N, which corresponds to the 'Uralic' phenotype"

N is brother haplogroup to O. And N certainly corresponds with the Mongoloid phenotype in Northeast Asia. It becomes less associated with the Mongoloid phenotype as you move towards the western end of its distribution (Uralic), but in that region presumably becomes mixed with the 'caucasoid' phenotype.

"and S, which is frequent in Papua New Guinea"

I don't see how O has any connection to the New Guinea haplogroup S.

"At the same time, Northern Mongoloids are high in hg C and don't have their own 'member' within the MNOPS clade".

And I'm reasonably sure that C3 was associated with the 'original' Mongoloid phenotype. Haplogroup O 'borrowed' it when the two haplogroups met in the northern half of China, before O began its move back south. And, as far as I'm aware, Mongolians do have a reasonably high proportion of Y-hap O.

German Dziebel said...

"I don't see how O has any connection to the New Guinea haplogroup S."

O is part of MNOPS. What this tells me is that instead of having a clear difference between Pleistocene and Holocene haplogroups corresponding to earlier Papuan and later Mongoloid phenotypes, they subdivide into more granular lineages (M and S are Papuan, O Southern Mongoloid, R is European, N "Uralic", etc.) and these lineages are all coming off a single root.

"N is brother haplogroup to O. And N certainly corresponds with the Mongoloid phenotype in Northeast Asia. It becomes less associated with the Mongoloid phenotype as you move towards the western end of its distribution (Uralic), but in that region presumably becomes mixed with the 'caucasoid' phenotype."

N is found at highest frequencies among Balto-Finnic speakers, Samoyeds and Altaic speakers. "Uralics" (Balto-Finns, Volga Finns and Samoyeds) are phenotypically distinct but related to Mongoloids. They hybridized in the west with Caucasoids and in the east with Mongoloids (see, e.g., here http://dienekes.blogspot.com/2004/12/dissection-of-y-chromosome-haplogroups.html).

"he Polynesians have both Papuan and Mongoloid haplogroups and give every appearance of having become a stabilised hybrid between the two groups. That has resulted in their forming 'their own phenotypic taxon'."

I understand your position, Terry. But I'm inclined to reverse it in the case of Uralics and and in the case of Polynesians. I don't deny the reality of admixture but I think you can't explain one population as the product of admixture between the other two. Populations branch off from each other, evolve in isolation, then admix with neighbors, but their "core" resulted from descent with modification, not from hybridization. Polynesians have high frequencies of hg C. It doesn't come from Papuans or from Mongoloids. Also in blood groups Polynesians are high in blood group O. Mongoloids have low frequencies thereof (high in B). Australians and Amerindians have frequency range similar to Polynesians. This tells me, again, that the spread of Austronesian speakers pre-dated the spread of the Mongoloid phenotype. It's rooted in a Pleistocene SE Asian refugium. Later it expanded out of it and absorbed Papuan and Mongoloid "blood."

terryt said...

"This tells me, again, that the spread of Austronesian speakers pre-dated the spread of the Mongoloid phenotype".

I still completely disagree, even after all the evidence you've offered. I find the alternative explanation much more likely.

"It's rooted in a Pleistocene SE Asian refugium".

Wallacea was presumably very sparsely populated, if populated at all, during the Pleistocene.

"O is part of MNOPS".

Not quite that simple. MNOPS broke up first into M, NO, P and S. NO broke up some time later into N and O, possibly somewhere far to the north of SE Asia. Perhaps in North or Central China. O then moved back south and N carried on north and then west.

"What this tells me is that instead of having a clear difference between Pleistocene and Holocene haplogroups corresponding to earlier Papuan and later Mongoloid phenotypes, they subdivide into more granular lineages (M and S are Papuan, O Southern Mongoloid, R is European, N 'Uralic', etc.) and these lineages are all coming off a single root".

I agree that M and S are New Guinea/Melanesian, and R is West Eurasian (rather than being purely European), but I certainly don't agree that O is Southern Mongoloid. It is found throughout China, north and south, and in my opinion originated in the northern part of its current range. N originated slightly to the north of O. And N is hardly 'Uralic'. It's extrememly common in Northeastern Eurasia, certainly a region of predominantly 'Mongoloid' phenotype.

"N is found at highest frequencies among Balto-Finnic speakers, Samoyeds and Altaic speakers. 'Uralics' (Balto-Finns, Volga Finns and Samoyeds) are phenotypically distinct but related to Mongoloids. They hybridized in the west with Caucasoids and in the east with Mongoloids"

The higher N frequencies in the west is a product of their dilution in the northeast by the arrival of other haplogroups, especially C3.

"e.g., here http://dienekes.blogspot.com/2004/12/dissection-of-y-chromosome-haplogroups.html)".

Even in the the link we can see that the Mongoid phenotype is dominant in all those groups. Q is dominant only in the Selkup and Ket, and is presumably the oldest haplogroup through the region. N has largely replaced it as it moved north from China and then westward towards Europe.

"I don't deny the reality of admixture but I think you can't explain one population as the product of admixture between the other two. Populations branch off from each other, evolve in isolation, then admix with neighbors, but their 'core' resulted from descent with modification, not from hybridization".

But in the case of both Uralics and Polynesians we are not dealing with 'core' groups. Both regions were occupied relatively recently. Almost certainly by populations from the surrounding regions. In the case of Uralics by eastern 'Mongoloids' and western 'Caucasians', predominantly by the eastern population. In the case of Polynesians by northern 'Mongoloids' and southern 'Papuans'. The mix developing in Wallacea before the Polynesians' departure.

"Polynesians have high frequencies of hg C. It doesn't come from Papuans or from Mongoloids".

In a way it does come from 'Papuans'. It comes from southern Wallacea, a region presumably occupied by people similar in phenotype to current Papuans from further east.

"Later it expanded out of it and absorbed Papuan and Mongoloid 'blood'"

I'd say it almost certainly expanded FROM a combination of Papuan and Mongoloid blood.

German Dziebel said...

"N originated slightly to the north of O. And N is hardly 'Uralic'. It's extrememly common in Northeastern Eurasia, certainly a region of predominantly 'Mongoloid' phenotype."

The Uralic phenotype is hardly a mix of Caucasoid and Mongoloid phenotypes. It's distinct and underlies some of the Altaic-speaking groups. See http://www.ncbi.nlm.nih.gov/pubmed/12018120. Hence the elevated frequencies of N is the Yakuts, for instance.

"In a way it does come from 'Papuans'. It comes from southern Wallacea, a region presumably occupied by people similar in phenotype to current Papuans from further east."

This is not based on any evidence. I've been saying it all along: Austronesians gave C2 to Papuans as they moved east into Polynesia, and hence we don't have any evidence for pre-Austronesian Wallacean languages and C is not original in Papuans, the most parsimonious explanation is that Austronesians can't be reduced to a mix of Mongoloids and Papuans. Their core is distinct, both craniologically, and genetically. It's clearly reflected in Y-DNA C2 and mtDNA B4. mtDNA B4 is not Papuan and it's not Mongoloid but is related to B2 in Amerindians, who are not Asian Mongoloid either. C2 is not Papuan or Mongoloid but is related to C3 in Amerindians and C4 in Australians.

"The higher N frequencies in the west is a product of their dilution in the northeast by the arrival of other haplogroups, especially C3."

I agree that C3 is a late arrival to Siberia, likely from America.

German Dziebel said...

Terry,

I just noticed a curious distributional thing: in the same way as B2 (New World) and B4 (Polynesia) are related on the mtDNA side, C3b (New World) and C3c (Siberia) are related on the Y-DNA side. mtDNA B is very rare to non-existent in Siberia, while C3 is very rare in SEA and non-existent in Polynesia. Between the two areas - Siberia and Oceania - the frequencies of Y-DNA C3 and mtDNA B2/B4 are low and patchy. In the New World, we see the same complementarity along the south-north axis: C3b is very common in the North among Na-Dene but very rare-to-non-existent in South America; whereas B2 is very frequent in South America, Mesoamerica and American Southwest but is rare to non-existent in northern North America.

We end up with two fully complementary Circumpacific regions based on male vs. female-biased lineages. C3 in the North and B2/B4 in the South. We don't see the same frequency range and degree of common descent between South America and Polynesia in terms of Y-DNA haplogroups as we see between mtDNA B2 and B4. And we don't see the same frequency range and degree of common descent between North America and Siberia in terms of mtDNA as we see in terms of Y-DNA C3b and C3c.

I wonder if here we may be dealing with the same ancient demographic process connecting the New World and the Old World and expressing itself in Y-DNA terms as C3 (North) and in mtDNA terms as B2/B4 (South). Subsequent events (drift, incoming migrants from other regions, etc.) drove the frequency of Y-DNA C3 chromosomes down in the South and the frequency of mtDNA B2/B4 down in the North on both sides of the Pacific and cut this original single population in half.

terryt said...

"I wonder if here we may be dealing with the same ancient demographic process connecting the New World and the Old World"

I think there is definitly some sort of connection around the Pacific, but it is not as direct as you seem to be claiming. The improvement of boating technology allowed humans to move all around the Pacific. But:

"C3 is very rare in SEA and non-existent in Polynesia"

C3 was never in SEA in ancient times. Its prtesence there is a result of much more recent movement south, poosibly starting with the movement south of influences from the Chinese Neolithic and even as late as the Han southward movement.

"Between the two areas - Siberia and Oceania - the frequencies of Y-DNA C3 and mtDNA B2/B4 are low and patchy".

Various mtDNA Bs are spread right through China though.

"Subsequent events (drift, incoming migrants from other regions, etc.) drove the frequency of Y-DNA C3 chromosomes down in the South and the frequency of mtDNA B2/B4 down in the North on both sides of the Pacific and cut this original single population in half".

That certainly seems to have been the case for mtDNA B on either side of the Berring Strait. But, as I said, C3 was never part of the pre-Austronesian makeup of SE Asia. It's a northern haplogroup, probably spreading originally from somewhere round northeast Mongolia. So American mtDNA B and Y-hap C3 originate in totally different regions.

"C3b is very common in the North among Na-Dene but very rare-to-non-existent in South America"

I don't think C3 was ever part of an 'original' movement into America. It arrived much later, and so any presence in the south is a product of recent founder effect, not a result of drift in a previously more common haplogroup.

"B2 is very frequent in South America, Mesoamerica and American Southwest but is rare to non-existent in northern North America".

I have a map that claims it is fairly common in the western half of North America.

German Dziebel said...

"I don't think C3 was ever part of an 'original' movement into America. It arrived much later, and so any presence in the south is a product of recent founder effect, not a result of drift in a previously more common haplogroup."

This is unlikely, as C3 is found in all the corners of North America, not just among the Na-Dene. All studies confirm that Q and C were part of the same migration. But it's noteworthy that it's precisely C3 and B2 that sometimes look younger than other haplogroups in the New World.

"I have a map that claims it is fairly common in the western half of North America."

In California hg B is not as common as hg A. B's absent in northern Na-Dene and extremely rare in Eskimos. It's attested in a pre-Clovis site in Oregon, but overall it's frequencies are very low in northern North America.

"C3 was never in SEA in ancient times. Its prtesence there is a result of much more recent movement south, poosibly starting with the movement south of influences from the Chinese Neolithic and even as late as the Han southward movement."

yes, agree. We're left with C2, which is a sister clade to C3.

"It's a northern haplogroup, probably spreading originally from somewhere round northeast Mongolia. So American mtDNA B and Y-hap C3 originate in totally different regions."

I don't believe we can pinpoint the geographic origin of a haplogroup with any precision. Note, however, that mtDNA B and Y-DNA C have rather similar distributions: Japan, Northeast India, Oceania, America, rare but consistent in SE Asia, absent west of India. They look like relic haplogroups overrun by mtDNA M and N and Y-DNA MNOPS clusters.

terryt said...

"yes, agree. We're left with C2, which is a sister clade to C3".

And C1 in Japan and the Ryukyus.

"I don't believe we can pinpoint the geographic origin of a haplogroup with any precision".

But we have a string of C haplogroups from C4 in Australia, C2 in Southern wallacea, C1 as above and C3 in Northeast Eurasia and North America.

"C3 is found in all the corners of North America, not just among the Na-Dene".

But it's nowhere near as common in South America as it is in North America. Haplogroups very seldom remain with the particular cultural group with which they enter a region.

"They look like relic haplogroups overrun by mtDNA M and N and Y-DNA MNOPS clusters".

Possible I suppose. But to me it doesn't seem their distribution is closely related.