New blogs

Leherensuge was replaced in October 2010 by two new blogs: For what they were... we are and For what we are... they will be. Check them out.

Thursday, February 11, 2010

Eurasian Y-DNA note


Just a brief working note on my current understanding of Eurasian Y-DNA diffusion and diversification patterns after the out of Africa migration.


I am assuming here that the newly proposed Y-DNA macro-haplogroup MNOPS (downstream of K) is for real and that it includes (for parsimony reasons) haplogroups K2, K3, K4, M, NO, P and S. This is not sure at all but is a quite reasonable possibility.



I used color codes for the main lineages (blue for DE, orange for C and red for F). The pink line represents the main route of the Eurasian migration, the red arrows represent the main flows of F and K derived lineages, which are the most complex ones. The light red dotted ellipse represents the most likely urheimat of F, IJK and K, as well as other derived lineages such as H and P. For simplicity reasons I have not depicted the expansions of C and D, nor of P and NO.

In order to infer the possible homelands of each of the lineages I used common sense, of course, but specially I used the following method: halfway point between the centroid of the lineage and the estimated homeland of upstream clades, when no centroid was available, midpoints between downstream centroids were used. It's notable that Northwest South Asia seems to be the homeland of the most important Eurasian Y-DNA lineages, not just of F but of several major nodes downstream of it (H, IJK, L, K, P, R), however the MNOPS "boomerang" migration, if confirmed, also has its own significance - and not just for Eastern Asia and Oceania. I wonder if this peculiar Y-DNA flow into SE Asia and back into South Asia is related with the flows of mtDNA N and R.

132 comments:

terryt said...

"I am assuming here that the newly proposed Y-DNA macro-haplogroup MNOPS (downstream of K) is for real and that it includes (for parsimony reasons) haplogroups K2, K3, K4, M, NO, P and S".

And why not K1? If Y-hap P could so easily move all that way from Indonesia K1 could easily have accompanied it part of the way, or followed later. We'll leave aside the possibility of L and T also being carried along.

"halfway point between the centroid of the lineage and the estimated homeland of upstream clades"

You've made assumptions 'estimating' KMNOPS's homeland.

"the MNOPS 'boomerang' migration, if confirmed, also has its own significance - and not just for Eastern Asia and Oceania".

I agree, 100%. The only disagreement I have with you here is that you've still placed KMNOPS too far west.

Regardless of that, because for years I've been fairly sure that P originated in SE Asia I've had plenty of time to consider some of the boomerang migration's significance. What was it that enabled just one or two haplogroups to spread back through most of Eurasia? I doubt it was some particular quality of the Y-chromosome itself. Technology and males tend to move together. Perhaps the fact that NO and P's closest relations crossed Wallacea provides a clue.

"I wonder if this peculiar Y-DNA flow into SE Asia and back into South Asia is related with the flows of mtDNA N and R".

Again, I've been convinced for years that mtDNA R was carried along with that Y-hap back-migration. After all we find mtDNA P confined to the eastern side of Wallacea. R31 is found in both Australia and India. F/R9/R21/R22 spreads up into China, but is especially common on Taiwan. B/R11 is spread along all of eastern Eurasia, from America in the northeast to Polynesia in the southeast. R23 is also East Asian. R8 is Indian, but is centred on the east coast, Orissa.

Far to the west we have three lineages that have spread far beyond India: R0/H, R2/JT and U. By my count that leaves six basal R mtDNA lines in India: R1, R5, R6/R7, R12, R14 and R30. Do they represent an origin in India, or a more recent starlike expansion there? My guess is the latter. If that is so then the male and female haplogroups must have entered a previously unexploited environment. Was their success related to the crossing of Wallacea?

Once we take mtDNA R out of India, as I've been trying to persuade you to do for years, we're left with virtually no mtDNA N there. And the few lines that are there could easily have entered relatively recently. Even the presence of some western R-derived haplogroups may be the product of yet another back-migration.

So how did mtDNA N get to Wallacea?

terryt said...

I've just realised a mistake you've probably made with your estimated homeland for KMNOPS. For a start its centroid would be weighted considerably by the expansion north and west of subclades PQR and NO, just as you mentioned elsewhere that O's centroid is shifted by the Polynesian expansion. Secondly I'll bet you've completely ignored the centroids of S, M, K2, K3 and K4 in your calcluations. They're not available in your link but they would be pretty easy to guess. Their distributions are relatively confined but would shift the overall centroid considerably to the south and east. I'm sure that if you get around to adjusting for both these factors you will get a more accurate homeland for KMNOPS.

Maju said...

"And why not K1?"

In your own words less than 24 hrs ago: "You may well be correct in placing K1 tentatively with T and L in India".

It's a mere matter of geographic parsimony: we don't have any reason at this stage to think K1 had ever any relation whatsoever with the MNOPS proposed clade. Instead we have many reasons to think that P has "oriental" connections, at least NO and probably M and S as well.

I'm, like everybody, in wait of confirmation for this macro-lineage and its fine detail. It's probable that we get to know something before housing prices plummet (as they should).

"You've made assumptions 'estimating' KMNOPS's homeland".

Like not placing it in Wallacea? Or estimating its gravity center based as much on P and NO as on M and S? Whatever.

... that you've still placed KMNOPS too far west.

Actually that part of the World is known as the Far East. ;)

Technology and males tend to move together.

Not among other great apes: in Chimpanzees and Bonobos it's females and children who teach and learn cultural aspects, memes, including technology. Adult males are just too "autistic" to bother.

But going to the grain, there's no known technology migrating from SE Asia to South or West Eurasia in the Paleolithic. Also the pattern of spread of Y-DNA P (or for the case NO as well) doesn't seem to hold any peculiar relation with your technological fetish of boats, except maybe of Q in America, but at a much later date.

Instead Andamanese and proto-Ainu, both with loads of Y-DNA D, seem to have used them. If anything boating in East Eurasia seems to be an areal feature, not a lineage-related one. It seems more determined by ecological reasons (living in coastal/insular areas) than by paternally-transmitted memes.

"I wonder if this peculiar Y-DNA flow into SE Asia and back into South Asia is related with the flows of mtDNA N and R".

Again, I've been convinced for years that mtDNA R was carried along with that Y-hap back-migration
.

I have a problem with this matter: mtDNA N may have migrated westward from SE Asia but R migrated eastward from South Asia instead. If N would have migrated westward with (pre-)P, then how did R migrate eastward without any males?

At the moment I'm more inclined to ponder that mtDNA R might have migrated eastward with (pre-)MNOPS, specially because it seems to have spread precisely at about the same locations: New Guinea and Eastern Asia. This still leaves the presence of N in South and West Asia ill-explained, asking for further refining of the model at the earliest stages.

"By my count that leaves six basal R mtDNA lines in India: R1, R5, R6/R7, R12, R14 and R30. Do they represent an origin in India, or a more recent starlike expansion there?"

You are going a bit too far by daring to exclude R2, R8, R31 and U. F and B have also been found in South Asia. Overall, I think there's no doubt that mtDNA R is South Asian by origin. There are in fact more doubts re. the mtDNA N urheimat.

"I'll bet you've completely ignored the centroids of S, M, K2, K3 and K4 in your calcluations".

I only used S and M. I really suspect that K2, K3 and K4 are likely to belong to some higher level subclade, not yet described because nobody seems to pay too much attention to such minor and exotic clades. You also have to consider research bias in these matters, right?

"Their distributions are relatively confined but would shift the overall centroid considerably to the south and east".

Depends on how you weight them. Centroids are after all just the center of gravity based on distribution, not diversity. Careful!

Whatever the case, I'm pretty sure that whatever entered Wallacea, did never really get out. It's a cul-de-sac.

terryt said...

I remember you saying you were not too interested in SE Asia because it was a cul-de-sac, and not really relevant to Eurasia as a whole. It's been exhausting enough getting you to accept P's origin as being as far east as modern Thailand, so I won't labour the issue of Wallacea, let alone mtDNA N and Y-hap C for now. But I'm sure you'll come round eventually. But a few commnents in passing:

"In your own words less than 24 hrs ago"

I changed my mind when I'd thought about it a bit more.

"we don't have any reason at this stage to think K1 had ever any relation whatsoever with the MNOPS proposed clade".

We equally have no reason to think its status has changed.

"Not among other great apes"

But we're not talking about cracking nuts with stones here. I'd assume that spear making was passed from father to son in organisationally complex societies (certainly mostly from male relations), rather than from mother to offspring.

"Also the pattern of spread of Y-DNA P (or for the case NO as well) doesn't seem to hold any peculiar relation with your technological fetish of boats"

Y-hap R1a is fairly concentrated along the Ganges Plain and O is quite possibly coastal originally, and then spread inland. As for S, M, K2, K3 and K4: they're pretty obviously associated with boats, in quite a fundamental manner.

"Instead Andamanese and proto-Ainu, both with loads of Y-DNA D, seem to have used them".

Yes. But the obvious question arises: when? There's actually no evidence for human presence anywhere in the Andamans until more recently than about 15,000 years ago. And D is quite common in Burma, so the conclusion is obvious to me.

"but R migrated eastward from South Asia instead".

What makes you so certain of that? Last time I checked (some months ago now) Ian Logan still had mtDNA R derived from Y/N9, a Northeast Asian haplotype.

"then how did R migrate eastward without any males?"

The only reason you're forced to consider that problem is that you're assuming it originated in India. I do agree that something must have accompanied KMNOPS eatward but there are plenty of mtDNA M haplogroups that fit the bill.

"specially because it seems to have spread precisely at about the same locations: New Guinea and Eastern Asia".

But mtDNA M lines fit even closer.

"You are going a bit too far by daring to exclude R2, R8, R31 and U. F and B have also been found in South Asia".

I didn't exclude them. Read again. I pointed out that R31 is found in both India and Australia so we can't draw definite conclusions about its origin, R8 is centred on Orissa so could have come from an easterly source, and that R2, U, F and B could easily be immigrants into India rather than indigenous, the first two back-migrations from the west and the other two from the east.

"You also have to consider research bias in these matters, right?"

Correct.

terryt said...

Just going back to your comment:

"I'm more inclined to ponder that mtDNA R might have migrated eastward with (pre-)MNOPS"

Mitochondrial DNA R's distribution certainly seems to accompany Y-hap KMNOPS from Thailand once the Y-hap had broken into its subclades. Hence mtDNA R's presence in 'New Guinea and Eastern Asia', not to mention South and West Eurasia. But various mtDNA M lines, such as D, E and Q, must have entered SE Asia from India accompanied by some Y chromosome or other. Indifferentiated KMNOPS and F2 seem obvious candidates.

Maju said...

Man, you are flippant: I have never ever said that Indochina, mainland SE Asia, was any cul-de-sac. Don't insult my intelligence, ok? Indochina: central, Wallacea: dead end - that's what I've held all the time since I can recall (i.e. since before I began this blog).

Long before that I once briefly flirted with the idea of Melanesia (not Wallacea) being origin of Y-DNA K (because of diversity) but soon I realized that such extremist idea was not just unnecessary but most likely very wrong. I never had a boating fantasy myself and I also realized that SE Asia (or beyond) is not central for lithic technology (blades) and that therefore, even if played any role re. West Asia, must have been after cultural filter in South Asia and not on its own.

This is another reason to think that the boomerang K>MNOPS>P must have happened before the main colonization of West Eurasia, i.e. before 50-40 Kya, when blade technologies became popular in the West. I also think that the R1b and Q spread should have happened before 38 Kya when microlithic bladelets began expanding in South Asia (not permeating westward until much later dates).

In contrast to your hypothesis on boats, I base my reasoning on hard material evidence that can be tracked properly. However, I'm widely open to further archaeological findings.

I also suspect that P did not bring any particular boat technology back to South and West Eurasia, because its spread happened largely though inland regions such as Central and North Asia and not the coasts. But this does not mean that they did not know the concept of boat, as this must have existed west of Bengal before all the process, just that they don't seem specialized in such ecological niche, which was surely fully exploited before their return from Indochina.

"O is quite possibly coastal originally, and then spread inland".

Maybe. But what about (Y-DNA) N and what about C? Ainu are not characterized by any F or K derived lineage, for example, nor is what describes best Australian Aboriginals fundamentally. Coastal exploitation was surely an areal feature and does not exclude the exploitation of other areas, like the edges of the Tibetan plateau or whatever other inland regions.

I think that macro-haplogroup C relates better with coastal exploitation in the eastern range of Greater Eurasia. However, as they were not doing any paternity tests back then, other lineages also participated in the process, including D and K. This may imply that pre-C journeyed along the coastal route already in South Asia, while its "brother" F (and also K) was more generalist, taking maybe some more time to reach the Far East but doing so in larger numbers (as K=>MNOPS transition probably) or with improved technologies.

Maju said...

(continued from above)

"As for S, M, K2, K3 and K4: they're pretty obviously associated with boats, in quite a fundamental manner".

Not more than C1, C2, C3, C4 and C5, for instance.

"And D is quite common in Burma"...

Is it? AFAIK Burma remains unsampled (for political reasons). So please provide a source, which would be most interesting if real (but I fear it's just another figment of your fantasy).

"Last time I checked (some months ago now) Ian Logan still had mtDNA R derived from Y/N9, a Northeast Asian haplotype".

The graphs in that site don't seem to be actualized since long ago. It's not reliable in that aspect.

"But mtDNA M lines fit even closer".

Not to my eyes: Y-DNA NO seems to have spread over D and C in East Asia, pushing them to the edges, roughly what we see with mtDNA B and F (though not as impressively but totally within the apportions I'd expect).

Similarly Y-DNA M (and other K) spread to New Guinea and neighboring areas, where mtDNA P also gained a very solid foothold. Again the spread of Y-DNA seems wider and more dense than that of mtDNA but that's only logical and expected.

So I think we have a very good case to associate mtDNA R with K/MNOPS in Eastern Eurasia.

"I didn't exclude them. Read again. I pointed out that R31 is found in both India and Australia"...

Ignoring that is believed to have migrated to Australia only recently.

"R8 is centred on Orissa so could have come from an easterly source"...

But is in South Asia and unknown to exist further East. It's a wild speculation that I'm not willing to accept, very specially when oriental Y-DNA (O) in inland Orissa and neighbor Chattisgarh is very much restricted to tribal groups (Munda).

"...and that R2, U, F and B could easily be immigrants into India rather than indigenous".

This is not totally impossible but, considering the rest, rather not. Even if you count them as West and East Eurasian lineages, South Asia still has by far greater basal diversity for mtDNA R.

The South Asian origin of mtDNA R is pretty much unquestionable.

(Plus splitting the origin of R into West and East Eurasia just makes no sense whatsoever. Spare me that nonsense, please).

Ebizur said...

How many times will I have to remind you two that no Australian aborigine has ever been found to belong to Y-DNA haplogroup S-M230?

K-M9(xM1-M4, M2-M353, NO-M214, P-M74, S-M230)
18/60 = 30.0% Arnhem Land, Australia (Kayser et al. 2006)
6/35 = 17.1% Sandy Desert, Australia (Kayser et al. 2006)

K-M9(xL-M20, M1-M5, NO-M214, P-P27, S-M230)
14/33 = 42.4% Australian Aboriginal People (Hammer et al. 2006)

The most common Y-DNA haplogroup among Australian aborigines is C4-M347/P309. The aforementioned K-M9(xL-M20, M1-M4, M2-M353, NO-M214, P-M74, S-M230) type is the only other Y-DNA haplogroup that is suspected of having a pre-colonial presence in Australia. The rest of the Australian aboriginal individuals for whom Y-DNA test results have been published are almost certainly descended from recent European and Asian colonists, as they have been placed in subclades of R1-M173 (and especially R1b), I1-M253, and O3a3c-M134.

Maju said...

Where am I saying the opposite here, Ebizur?

Ebizur said...

Maju said,

"Where am I saying the opposite here, Ebizur?"

I am referring to your recent discussion in the comments section of the "African R1b is a single haplogroup" thread. I do not want to dig up that old thread again!

In any case, the present thread seems like a more appropriate place for a discussion of the distributions of Oceanian Y-DNA haplogroups. Australian aboriginal K-M9 Y-DNA may turn out to belong to M3-P117/P118, M*-P256(xM1, M2, M3), T-M70/M184/M193/M272, or K-M9(xL, M, NO, P, S, T), but it certainly does not belong to haplogroup S-M230 or any other Melanesian haplogroup with which you two are apt to confound it.

Ebizur said...

Addendum:

Australian K-M9 also might belong to one of those K# clades, i.e. K1-M147, K2-P60/P304/P308, K3-P79/P299/P307, or K4-P261/P263. K3-P79 is pretty common in some parts of Melanesia and western Polynesia (data are from Kayser et al. 2008 unless otherwise noted):

K3-P79
10/19 = 52.6% Mussau (Scheinfeldt et al. 2006)
3/6 = 50% Tokelau
28/100 = 28.0% Tuvalu
92/395 = 23.3% New Britain (Scheinfeldt et al. 2006)
4/19 = 21.1% Tolai New Britain
1/7 = 14% Manus (Scheinfeldt et al. 2006)
12/105 = 11.4% Fiji
16/152 = 10.5% New Ireland (Scheinfeldt et al. 2006)
5/50 = 10.0% East Futuna
2/37 = 5.4% New Guinea (Scheinfeldt et al. 2006)
6/147 = 4.1% Admiralty Islands
3/75 = 4.0% Bougainville (Scheinfeldt et al. 2006)
3/77 = 3.9% Cook Islands
1/29 = 3.4% Tonga
1/31 = 3.2% PNG Highlands
1/33 = 3.0% PNG Coast
1/61 = 1.6% Samoa
0/9 Niue
0/17 Malaysia
0/31 Nusa Tenggaras
0/33 Moluccas
0/35 Bereina PNG
0/37 Philippines
0/40 Southern Borneo
0/46 Kapuna PNG
0/52 Trobriand Islands
0/53 Java
0/56 Sumatra
0/89 WNG Lowlands
0/94 WNG Highlands

K1-M147 is a rare clade that has been found only in a couple South Asians as far as I know. I am not aware of any data regarding the distribution of K2-P60/P304/P308 or K4-P261/P263.

terryt said...

"I also suspect that P did not bring any particular boat technology back to South and West Eurasia"

It must owe its expansion to something.

"Not more than C1, C2, C3, C4 and C5, for instance".

C3 got to Mongolia by boat?

"How many times will I have to remind you two that no Australian aborigine has ever been found to belong to Y-DNA haplogroup S-M230?"

I was pretty sure they hadn't, but I was giving Maju the benefit of the doubt. The separation between New Guinea and Australian Aborigine haplogroups is fairly complete, especially the Y-haps. To me that argues eloquently for at least two movements across Wallacea.

Maju said...

I believe we already discussed S at that other thread. You do well in not digging up in it but I'm pretty sure you already intervened there to warn us of the peculiarities re. S and Australian K*.

"I am not aware of any data regarding the distribution of K2-P60/P304/P308 or K4-P261/P263".

Karafet 2008 briefly mentions them as Melanesian. That's all I know.

Maju said...

Actually it reads:

"Haplogroup K1 was found at low frequencies in India/Pakistan (Underhill et al. 2001). Lineages K2, K3, and K4
are found in Oceania, Indonesia, and/or Australia".

Quite vague. They may not be Melanesian but found somewhere in the SE fringes of Greater Eurasia in any case.

Maju said...

"C3 got to Mongolia by boat?"

Why not? Specially considering that is one of the two only Y-DNA lineages found among Native Americans and that all its direct relatives all have insular or coastal distributions. Y-DNA C seems very much the "boater" type.

Maju said...

"It must owe its expansion to something".

Cultural hybridization maybe? Aggressiveness? Exploitation of marginal niches like Central Asia/Siberia and Europe? Whatever else we have not yet thought of?

I'm really not sure if to liken the expansive dynamic of P to that of NO or to that of other western lineages like IJ. The first would be more appropriate for Q maybe (somewhat parallel to N in its specialization for the Far North), the latter for R1(b).

Btw, this reflection on N and Q tells me something in relation with your previous objection to C3 having arrived to NE Asia by boat. Unlike N and Q, C3 did not penetrate too far inland (not in Asia at least), what really makes total sense if they were "boaters".

So, I reaffirm myself: C were originally boaters, K/MNOPS maybe not so strictly so (though obviously they must have used boats to reach to Australasia, of course). It just makes total sense.

terryt said...

"I am not aware of any data regarding the distribution of K2-P60/P304/P308 or K4-P261/P263".

I have found somewhere (sorry I can't remember where, but I have it in my notes) that K2 and K4 are amoung the few, or only, Y-haps common to both Australia and New Guinea.

"So I think we have a very good case to associate mtDNA R with K/MNOPS in Eastern Eurasia".

Yes. But at the beginning of KMNOPS's expansion, which you have placed as far east as Thailand and I would place in Wallacea. So how can you persist in placing mtDNA R's origin in India? It must originate somewhere within or between Thailand and Wallacea.

"Plus splitting the origin of R into West and East Eurasia just makes no sense whatsoever. Spare me that nonsense, please"

But hang on. Isn't that exactly what the scenario you propose requires? To make mtDNA R fit with both an original eastward-moving Y-hap KMNOPS and the subsequent westward Y-hap P back-migration you actually do need to postulate a two-stage movement for mtDNA R. First, a few of the available mtDNA R lines must move east from India into Thailand with Y-hap KMNOPS. Then, several of these mtDNA R lines move back west again with Y-hap P, right through India and out onto the Eurasian Steppe. And, straining credibility, during this second westward-moving phase Y-hap P has to team up with the several mtDNA R lines that Y-hap KMNOPS had originally left behind, and at the same time bypassing the many mtDNA M lines already present through India. Is that really likely?

If the pattern you postulate is correct surely we should be easily able to detect a trace of this two-phase movement of mtDNA R. But our present understanding of the haplogroup suggests a single star-like expansion. This pattern fits just the eventual Y-hap P expansion, not the previous KMNOPS movement into SE Asia.

On the other hand the distribution of mtDNA M fits the eastward migration of Y-hap KMNOPS rather well. M's star-like expansion seems based in India but roughly half the lines managed to move further east, especially into SE Asia, with a small selection of those eventually moving even further: into New Guinea/Melanesia and north as far as Japan.

Very few mtDNA M lines emerged westward beyond India. So mtDNA M is not closely associated with Y-hap P at all. Unlike mtDNA R. And Y-hap R is not closely associated with Y-hap KMNOPS. Unlike mtDNA M.

terryt said...

"Exploitation of marginal niches like Central Asia/Siberia and Europe? Whatever else we have not yet thought of?"

Or river bank and lakeside?

"They may not be Melanesian but found somewhere in the SE fringes of Greater Eurasia in any case".

Ebizur has shown that K3 is definitely Melanesian, with some making it as far as Western Polynesia.

"C3 did not penetrate too far inland"

What? Mongolia is not inland?

"C were originally boaters"

Maybe not originally, but I have claimed for years they were early, if not the first, boaters. They presumably reached Australia by 50,000 years ago, and must have used boats to get there. And I would argue that KMNOPS borrowed boating off them and away they went in turn. But by what route did Y-hap C reach Eastern Eurasia?

Maju said...

"So how can you persist in placing mtDNA R's origin in India?"

Because it cannot be any other way.

Anyhow, what I'm suggesting is that when pre-MNOPS migrated to SE Asia, it brought with it underived mtDNA R that would eventually (quite quickly) coalesce into the Eastern R sublineages (pre-B, pre-F, P and a few other small lineages).

"To make mtDNA R fit with both an original eastward-moving Y-hap KMNOPS and the subsequent westward Y-hap P back-migration you actually do need to postulate a two-stage movement for mtDNA R".

I'm totally detaching the back-migration of Y-DNA P from any mtDNA lineage.

One possibility that I have not yet considered in depth could be that P would have back-migrated through the steppes and not through South Asia, in which case it could have been initially associated to mtDNA X. There is some P(xR,Q) among NW Native Americans that is close by haplotype to P* among Mongolians (though this could well be R), so the case for the back-migration of P is somewhat unclear as of now. But mtDNA X seems to have a clear West Asian urheimat (highest diversity), so it doesn't look like in this case we can follow the mtDNA track: too thin.

If P back-migrated through South Asia though, it might be related to the presence of mtDNA B and F in the subcontinent. There seems to be also some support for this scenario and the archaeology of Siberia rather supports a West to East migration for Q or whatever lineage was involved in the spread of blade tech in the area. In this case, mtDNA X2 might have spread (thinly) with Y-DNA Q.

I really have more difficulties anyhow with the back-migration of mtDNA N to South Asia, where several N-derived lineages coalesced without doubt (among them R, but also N1'5 and N2). I'm considering that they might have back-migrated with Y-DNA C (C5) but the specific geographical destinies seem unrelated. So have to admit that I just don't know.

Maju said...

"On the other hand the distribution of mtDNA M fits the eastward migration of Y-hap KMNOPS rather well".

It's just too old: the first spark (and what a spark!) of the Great Eurasian Expansion. It should rather be coincident with the overall CDEF Y-DNA migration eastwards (CF and pre-D to be precise). And in fact Oriental M sublineages are often associated with Y-DNA C and D, which seem the early stars of the show in the East.

In South Asia the M explosion should correlate with the Y-DNA F node but I don't see clearly any F derived anything in the East before the MNOPS stage (or K if MNOPS ends up being a bluff).

So for me, there are two main flows into Eastern Eurasia (incl. Sahul):

1. MtDNA M and pre-N, with Y-DNA pre-D and pre-C.

2. MtDNA R with Y-DNA pre-MNOPS.

Maju said...

"What? Mongolia is not inland?"

Not too far inland. Not in comparison with the scope of Y-DNA N and Q. In Siberia (senso lato) you basically see these three Y-DNA lineages and you can easily conclude that, while N and Q migrated forth and back through the whole region, C3 remained constrained to the far East, not too far from the Pacific Ocean.

You can make all the noise you wish with petty details like Mongolia but that won't change anything. C3 is a NE Asian lineage and not a North Eurasian one, like N and Q.

"But by what route did Y-hap C reach Eastern Eurasia?"

Pre-C. Coastal. The fact that C's centroid is in SE Asia says it all, right? The fact that they left no traces in their way says also that they moved fast. The fact that all its sublineages have a roughly coastal distribution also says that they were the more specialized boaters in the OoA migration, though of course they were not the only ones.

terryt said...

"Is it? AFAIK Burma remains unsampled"

I apologise. I guess I've just come to accept D is present in Burma. Ebizur pointed out some time ago that the McDonald map shows that in SE Asia Y-hap D is found in Malaysia, Borneo, Sumatra and amoung the southern Han Chinese. And remember that most of the Burmese languages are classified as 'Tibeto-Burman'. Guess where the highest proportion of D in the world (almost 50%) is found. There's a very good chance that Y-hap D is very common in Burma.

"what I'm suggesting is that when pre-MNOPS migrated to SE Asia, it brought with it underived mtDNA R that would eventually (quite quickly) coalesce into the Eastern R sublineages (pre-B, pre-F, P and a few other small lineages)".

And almost at the same time coalesce into the western, or European, lineages. So how do you explain mtDNA R's star-like nature? The expansion must have been unbelievably rapid for the European mtDNA clades to be so close to the East Asian ones under the scenario you're proposing, especially considering you're claiming the mtDNAs entered Europe with Y-haps unknown (perhaps on their own?).

"If P back-migrated through South Asia though, it might be related to the presence of mtDNA B and F in the subcontinent".

MtDNA B and F are much more likely to be related overall to the expansion of Y-hap O. And also almost certainly entered India with that Y-hap.

"mtDNA N to South Asia, where several N-derived lineages coalesced without doubt (among them R, but also N1'5 and N2)".

I see you are now happy to place Y-haps G and IJ in SW Asia (on the Iranian Plateau) rather than in India. I quite agree with that placement. But for the Y-haps to survive they must have had an accompanying mtDNA. N1'5 provides a very likely candidate, and perhaps X. So N1'5 need never have been anywhere near India. Is there any compelling reason why this cannot be so?

"It's just too old"

No it's not. The distribution of the various mtDNA Ms through SE Asia and across Wallacea coincides almost exactly with the Y-hap Fs, and not with anything else.

"The fact that C's centroid is in SE Asia says it all, right?"

Sorry. C's centroid is in Inner Mongolia, enclosed within the Hwang Ho's northern loop.

"The fact that they left no traces in their way says also that they moved fast".

Hang on a minute. Weren't you, not very long ago, arguing that T could not have come from in SE Asia, no matter how rapidly, because there was no trace of it there? And concerning R8 in Orissa:

"But is in South Asia and unknown to exist further East".

Granted. But Y-haps M, S and K3 (and possibly K2 and K4), are not found west of Wallacea, yet for some unknown reason you're very happy to place their origin far to the west of that region. Did all these lines move together across Wallacea from Thailand that they left no trace along the way? And why the inconsistency? Is it an example of your earlier comment, 'You also have to consider research bias in these matters, right?'. You readily accept the evidence at face value when it coincides with your belief, but refuse to accept evidence that conflicts with those beliefs. I accept that Y-haps are more easily drifted out than are mtDNAs, but here we're dealing with just one mtDNA line and as many as half a dozen separate Y-haps.

Maju said...

Maybe I have not explained myself well enough. MtDNA R has a starlike nature but that doesn't mean that all R sublineages expanded at the same time (R did but not necesarily all and each of its sublineages, get the difference?) nor that all were coupled with MNOPS.

I'd rather think that R (and some other mtDNA lineages maybe) spread within the Y-DNA IJK expansion, of which MNOPS is the oriental branch (essentially). So maybe some of the Western R (and non-R) subclades spread with Y-DNA IJ and not with anything under K (unless it's T) and certainly MNOPS (P) played no or just a small and secondary role in the mtDNA R spread westward.

I just posted yesterday on a new control region mutation count exercise on mtDNA possible expansion scenarios. And what you see at the R+1 phase is seven R sublineages spreading in South Asia, SE/East Asia, Melanesia and West Asia (also 8 M and 1 N(xR) subclades - M was still expanding too but within each region: there's no sign that it moves across regions anymore, excepting M32 to Andaman).

At the R+2 stage, the spread to the East continues but at the R+3 and ulterior stages (not mapped yet) all the eastern R expansion is locally originated, derived from R9 (pre-F) and macro-B basically, which were already there.

So the spread of R to the East seems to be quite fast (two CR mutational steps, not more) and seems to be about the last migration from South Asia in that direction (excepting maybe some minor sublineage like the one suggested to link India and Australia at very minor levels in the Epipaleolithic).

The flow westwards lasts a bit more but not much more (by R+4 it seems all that matters has already arrived). So it feels like R is the swan song of the Great Eurasian Expansion and that the people who carried it eastward were closely related to those who carried it westward too, and that both processes happened at about the same time (50-40 Kya, I estimate).

But the Y-DNA lineages involved in the West (at least) are not just part of K, but a more variegated array that nevertheless belongs totally to IJK, except G. G is the only F(xIJK) lineage taking part, the same that M1 is the only M lineage in the Westward expansion. This suggests that this group of people were somehow distinct, differentiated from the C/D/F(xIJK)+M/N(xR) that starred the initial expansion. They are a subgroup of them, of course, but an specific one. And they are a group surely centered in South Asia.

Oversimplifying: Y-DNA IJK was coupled with R at this stage, and East of Bengal IJK is only present as MNOPS, though R does not show such phylogenetic distinction.

Maju said...

"MtDNA B and F are much more likely to be related overall to the expansion of Y-hap O".

Well, B at least is coupled with Q in America. So I feel this is too simplifying. From memory mtDNA F has been found in Siberia, at least in aDNA. Anyhow, O is the numerical bulk of NO in East Asia, with N being mostly a peripherical Northern clade. So I think arguing on this would be quite pointless anyhow.

"But for the Y-haps to survive they must have had an accompanying mtDNA. N1'5 provides a very likely candidate"...

Y-DNA IJK and G in West Eurasia would seem to couple essentially with mtDNA R, western R (R0, JT and U). Also M1, N1 and X. I think W represents a late flow from South Asia (16 CR mutations from L3, while X only has 10).

"The distribution of the various mtDNA Ms through SE Asia and across Wallacea coincides almost exactly with the Y-hap Fs, and not with anything else".

Hmmm. I don't think so. I think they spread with C in a previous episode. All Eastern F is K (and probably MNOPS): we don't see F(xIJK) over there (excepting minor clade F2 - "in East Asia", not Wallacea it seems).

"Sorry. C's centroid is in Inner Mongolia, enclosed within the Hwang Ho's northern loop".

That's by Chiaronia's paper but I can't agree. If you estimate the centroid of each of the five known sublineages (excepting America for both convenience and realism) and then trace the midpoint, it must fall in SE Asia, South China the furthest North. Whatever the case, with one lineage in South India, two in Sahul and two in NE Asia, SE Asia looks like the logical urheimat for C. Further north is simply absurd.

"Hang on a minute. Weren't you, not very long ago, arguing that T could not have come from in SE Asia, no matter how rapidly, because there was no trace of it there?"

There is no trace of any ancestor of T so far East. The ancestor of C was at some point in Africa. These two notions are so painfully obvious that I find highly annoying your mischievous question and distortion of what I said. You can't be ignoring these facts,, you know them very well, and you know why I stated that there is no reason to think that the ancestors of T ever went further east than Bengal.

This is the kind of manipulation that makes me really angry at you. Because either you think I am stupid enough so you can distort the reality and the facts right before my eyes or you are the stupid one who really can't understand anything even if basic as soon as it seems to contradict your Wallacea boating dogma (or both).

Maju said...

"But Y-haps M, S and K3 (and possibly K2 and K4), are not found west of Wallacea, yet for some unknown reason you're very happy to place their origin far to the west of that region".

F has highest basal diversity in South Asia. All those lineages are derived, several nodes downstream, from F. Again it's so basic that it's painful to have to explain.

"Did all these lines move together across Wallacea from Thailand that they left no trace along the way?"

I'm all the time assuming that there is a macro-lineage called MNOPS between K and all those lineages. So the trace in SE Asia is O and returning P (for instance) and the trace in South Asia is K, IJK and F (everything upstream).

Would MNOPS be a bluff, then I'd have to reconsider. But at this moment I'm assuming it's for real (because it's more parsimonious).

"... but here we're dealing with just one mtDNA line and as many as half a dozen separate Y-haps".

Not really:
- mtDNA P correlates with MNOPS in Melanesia
- mtDNA macro-F and macro B correlate with MNOPS in East Asia (NO in practical terms, sure)

It's tentative of course but makes some decent sense. And nope: it's not any "faith", just an idea I'm chewing on. I also do not pretend to find a strict, village by village correlation: I understand well that gender-biases affect migrations and drift to some extent.

I'd appreciate criticisms that make sense but I do not appreciate criticisms only intended to defend the (undefensible) dogma of the Wallacean invention of the raft, much less through mischievous means like distorting what I say or pretending not to know what you know very well.

I hope that in the future you have a straightforward and respectful attitude. We will get nowhere by means of denying or twisting the facts: that's for politicians and salesmen, not for people interested in science.

For example I am right now much more intrigued on how to explain the possible back-migration of N to South Asia, precisely because there seems to be no meaningful Y-DNA lineage that could be related with that process (C5 but doesn't correlate well at all). I am so puzzled that I'm even tempted of re-considering N as expanding in South Asia and also participating in the Y-DNA K/MNOPS migration eastward.

Maju said...

PS-

Also, please consider that Melanesian K/MNOPS may have taken other "brides" along the way, or even in Melanesia itself. MtDNA O (where is it found exactly?, anyone?) seems to expand by CR mutation count at the same time as P, while Q and M28 (derived from M lineages probably already localized in Melanesia) seem to expand just after them.

terryt said...

"I'd appreciate criticisms that make sense but I do not appreciate criticisms only intended to defend the (undefensible) dogma of the Wallacean invention of the raft"

I'm trying to get you to consider ALL the evidence. Not just the part of it that suits your belief.

"I'm all the time assuming that there is a macro-lineage called MNOPS between K and all those lineages".

Ah. Those Ks are a real pesky inconvenience for you aren't they. In spite of there being no reason at all for you to separate them from MNOPS you constantly claim such a separation.

"There is no trace of any ancestor of T so far East. The ancestor of C was at some point in Africa".

You're ignoring my main point. How do you explain the presence of the multiple KMNOPS Y-haps beyond Wallacea as being the result of all those separate haplogroups moving from much further west?

"MtDNA R has a starlike nature but that doesn't mean that all R sublineages expanded at the same time"

Probably not. But you've certainly used the starlike nature of mtDNA M to argue a single origin in India often enough. But here you're arguing for a multiple origin of Y-hap R. The scenario you're proposing requires some sort of 'pre mtDNA R' as having moved east from India to SE Asia with Y-hap KMNOPS, and then expand from there. Then you have another 'pre mtDNA R' moving west from India, at much the same time, with some unknown Y-hap. Both these separate 'pre mtDNA R' haplogroups then mutate into the same mtDNA R clade. Do you really think that's the most likely explanation for the evidence?

"So it feels like R is the swan song of the Great Eurasian Expansion"

I've argued all along that R expanded after N. So now you agree with that. The only disagreement we still have is where did R expand from?

"Well, B at least is coupled with Q in America".

Not coupled any closer to Y-hap Q than are mtDNAs C and D. All three mtDNA lines, and Y-hap Q, are widespread in America. As is A, but it seems accepted that A is later and so probably associated with Y-hap C3b. And mtDNA B is certainly not associated in any way with Q outside America.

"Y-DNA IJK and G in West Eurasia would seem to couple essentially with mtDNA R, western R (R0, JT and U)".

How could they be if R moved into the region from India when the Y-haps were already present outside India? And how did mtDNA R get from India?

"All Eastern F is K (and probably MNOPS): we don't see F(xIJK) over there (excepting minor clade F2 - 'in East Asia', not Wallacea it seems)".

Philippines evidently (according to the McDonald map). Close enough to Wallacea for me. The only reason F2 is a 'minor clade' is because Europeans are not part of it. I agree it's not that common, which is what we would expect if it had been drifted out by the multitude of later arrivals.

"please consider that Melanesian K/MNOPS may have taken other 'brides' along the way, or even in Melanesia itself".

As has every other Y-hap. I'm actually proposing basically that Y-hap F came into SE Asia with mtDNA M, then KMNOPS teamed up with the R branch of mtDNA N (which was either already there or came in afterwards) and the two of them spread in all directions.

"while Q and M28 (derived from M lineages probably already localized in Melanesia) seem to expand just after them".

Where's the problem with that?

"MtDNA O (where is it found exactly?, anyone?) seems to expand by CR mutation count at the same time as P"

It spreads all the way north to Japan and Mongolia and south to both mainland and island SE Asia, and even well out into the Pacific. The fact it expands at the same time as P is hardly surprising if a Wallacean origin is accepted for the two.

terryt said...

"That's by Chiaronia's paper but I can't agree".

You can't agree simply because it doesn't fit your belief. From an earlier comment:

"I think that macro-haplogroup C relates better with coastal exploitation in the eastern range of Greater Eurasia".

Strange. Just a few weeks ago you were arguing just as enthusiastically that there was no ancient connection between Australia and Japan. And another earlier comment:

"Instead Andamanese and proto-Ainu, both with loads of Y-DNA D, seem to have used them".

D is still an inconvenience for your belief. Y-hap D is found in both the Andaman Islands and in Japan, yet its distribution is in no way coastal. The connection between the two island regions is over land, via the hill country along the Tibet/China border. Besides, it's presence in Japan in no way necessitates any water crossing at all. At times of low sea level it was possible to walk to Japan from the mainland.

"Whatever the case, with one lineage in South India, two in Sahul and two in NE Asia, SE Asia looks like the logical urheimat for C. Further north is simply absurd".

The situation for Japanese Y-hap D applies to Japanese Y-hap C1: it could walk there. C1 does spread south into the Ryukyus. But so does Y-hap O2b, so Y-haps C1 and O2b could have moved together, later. C2, C4 and C6 necessitate a water crossing. But of the C3s only American C3b requires such an ability. All other members of C3's starlike expansion are mainland East Asian (although I don't know where C3e and C3f are found. Perhaps Ebizur can help?).

"It's notable that Northwest South Asia seems to be the homeland of the most important Eurasian Y-DNA lineages"

And what about C5 in India? Is it coastal? Or does it too originate in 'Northwest South Asia'?

"This still leaves the presence of N in South and West Asia ill-explained, asking for further refining of the model at the earliest stages".

I'll comment at your latest post some time.

Maju said...

Please:

"Ah. Those Ks are a real pesky inconvenience for you aren't they. In spite of there being no reason at all for you to separate them from MNOPS you constantly claim such a separation".

I am just not claiming any such separation, nor I did before. In fact I did the opposite all the time: they are MNOPS (probably). And that is precisely my point. And you know it. And now you misread my words to claim that I "always" said exactly the opposite of what I said and what you knew I said.

Please! It's well beyond what I consider reasonable margin of confusion and misunderstanding, really: either your brain is misfunctioning or you only want to see what you want to see, conveniently forgetting about everything else (or both).

"You're ignoring my main point. How do you explain the presence of the multiple KMNOPS Y-haps beyond Wallacea as being the result of all those separate haplogroups moving from much further west?"

That's why MNOPS is so convenient. It allows for a relatively small migration (one lineage) only multifurcating at SE Asia. :)

"But you've certainly used the starlike nature of mtDNA M to argue a single origin in India"...

No. What makes M original of South Asia is its huge basal diversity. The starlike structure what supports is a single demographic explosion, that logically happened in South Asia, but the reason for its geography is basal diversity, not the starlike structure.

That you don't seem to understand too many things, doesn't mean that I said things as you misunderstood them.

"But here you're arguing for a multiple origin of Y-hap R".

No. I am not arguing for any such "multiple origin" for R. Please! Not at all! Like M before it, R has a clear single origin at South Asia.

What I'm saying is that it would seem like Y-DNA pre-MNOPS may have migrated eastwards with early R, still undifferentiated. Call it R or call it pre-P, pre-R9, etc.: it's the same thing. Similarly you can call K to pre-MNOPS, again the same.

What's the problem with that? It all belongs to the same expansive process, specifically to the Eastern offshot.

Can you make an effort for being less irrationally confrontational and try to make a greater understanding effort?

"The scenario you're proposing requires some sort of 'pre mtDNA R' as having moved east from India to SE Asia with Y-hap KMNOPS".

Nope. Actually it requires pre-MNOPS (yet undifferentiated K) to have migrated with fully evolved R, that still was not P, R9 or any of the other oriental R sub-lineages.

"I've argued all along that R expanded after N".

And? That's as obvious as the daughter necessarily being younger than the mother. What's the mystery of that logic?

"So now you agree with that".

How could I not agree with that. When did I say the opposite?

"How could they be if R moved into the region from India when the Y-haps [IJ and G] were already present outside India?"

They were not. That's my whole point. IJ and G should have migrated westward along with T (and maybe R1(b)) and with mtDNA R, M1, N1, X. That there was some pre-IJ, for example, lurking for some time wherever, doesn't really matter.

You have to understand the difference between an upstream node (F, IJK in our case) and the downstream one (G, IJ). If F branched out in the year 70,000 BP (for instance), G can perfectly only have formed (branched out) in the year 10,000 BP (again just an example). Pre-G can perfectly have stayed for millennia and millennia lurking as a private not particularly successful lineage... until one day it got its opportunity (and became G). Otherwise we'd talk of "F5" or even F*.

Maju said...

"Philippines evidently (according to the McDonald map)".

I had not looked at that. There's a good array of minor F* over there following McDonalds. Not just in Philippines, but also in Malaysia, Australia and South China, it seems. I'd like some sort of confirmation anyhow. Is it all F2 or is F*?

"I'm actually proposing basically that Y-hap F came into SE Asia with mtDNA M, then KMNOPS teamed up with the R branch of mtDNA N (which was either already there or came in afterwards) and the two of them spread in all directions".

That doesn't make sense. Because:

1. R coalesced in South Asia.
2. MNOPS would have needed to wait for many many milennia until little red hood R showed up.
3. You forget of the IJK and K nodes.

"Where's the problem with that?"

No problem. Just suggestive that they expanded as MNOPS arrived to Melanesia.

"It spreads all the way north to Japan and Mongolia and south to both mainland and island SE Asia, and even well out into the Pacific".

That's Y-DNA O. I'm asking for mtDNA O, an N sublineage.

"You can't agree simply because it doesn't fit your belief".

No. I can't agree because of C2, C4 and C5.

"Strange. Just a few weeks ago you were arguing just as enthusiastically that there was no ancient connection between Australia and Japan".

Not sure what I said but whatever "connection" must be very old and be part of the Great Eurasian Expansion, not of any "Wallacean" holy grail.

Maju said...

"And another earlier comment:

"Instead Andamanese and proto-Ainu, both with loads of Y-DNA D, seem to have used them".

D is still an inconvenience for your belief".

What belief? Whatever the case D is not any "inconvenience" for me. Why would it be?

"Y-hap D is found in both the Andaman Islands and in Japan, yet its distribution is in no way coastal. The connection between the two island regions is over land, via the hill country along the Tibet/China border".

Maybe. But why would it be any inconvenience to me? It's you who claimed that such hill country could never be inhabited, not even transited through. For me it's just another land.

You need to check your mental abilities, seriously. You're going prematurely senile or something... :(

"Besides, it's presence in Japan in no way necessitates any water crossing at all. At times of low sea level it was possible to walk to Japan from the mainland".

Maybe or maybe not. I have read opposite reasonings on this regard. What I'm sure is that the early inhabitants of Japan used boats to gather raw materials, as was reported in some recent paper, arguing that this, along with the colonization of the low fringes of the Tibetan plateau (also a feat of "clan D" probably), constituted signs of "modern human behaviour".

So we have people boating and climbing to frosty plateaus at the same time. For me it's not any problem, because I assume by default that people can do both - the very same people can.

"And what about C5 in India? Is it coastal? Or does it too originate in 'Northwest South Asia'?"

And what about it? Is it a "most important Eurasian lineage"? Nope: it's a minor lineage even in its area. All the big Y-DNA Eurasian lineages are F derived. Of course you can argue what's big enough to enter this category, for example, why did I exclude C3 or D2, but when you look at Eurasia as a whole, you understand why I say that: H, IJ and K all seem to coalesce in NW South Asia. And maybe 90% of modern Eurasians belong to one of these lineages. The only region with some sizable population where something else is really big is Japan.

So whether you agree or not in emphasizing this fact, it's clear that is fundamentally correct: most Eurasians are direct patrilineal descendants from lineages that coalesced in NW South Asia.

And that is also true probably for the matrilineages.

terryt said...

"I am just not claiming any such separation, nor I did before".

Just a few earlier comments from you: 'we don't have any reason at this stage to think K1 had ever any relation whatsoever with the MNOPS proposed clade'. 'I really suspect that K2, K3 and K4 are likely to belong to some higher level subclade, not yet described'. 'I'm all the time assuming that there is a macro-lineage called MNOPS between K and all those lineages'.

So, whose 'brain is misfunctioning'? Or needs 'to check your mental abilities' or is 'going prematurely senile or something'?

"It allows for a relatively small migration (one lineage) only multifurcating at SE Asia".

But where in SE Asia? Even if we exclude all the Ks from MNOPS it only relieves the strain on Thailand a little. We still have two trans-Wallacean clades to accomodate. It's unlikely that Y-haps S and M came into Wallacea as fully-formed Y-hap S and Y-hap M. They presumably arose in situ from Y-hap MNOPS, probably in New Guinea, or somewhere very close to it. It's then easy to imagine the other two MNOPS haplogroups, NO and P, also forming near Wallacea. Prolonged and extensive to and fro movement on its western margin, especially by Y-hap O, has drifted out the basal haplogroups NO and P.

"IJ and G should have migrated westward along with T"

From your original article, 'It's notable that Northwest South Asia seems to be the homeland of the most important Eurasian Y-DNA lineages, not just of F but of several major nodes downstream of it (H, IJK, L, K, P, R)'. Was mtDNA R there as well? That again suggests a broad geographic region for mtDNA R's origin.

"Pre-G can perfectly have stayed for millennia and millennia lurking as a private not particularly successful lineage"

I'm pleased you understand that, because that is very close to the claim I've made for mtDNA N at your mtDNA post.

"Is it all F2 or is F*?"

My understanding is that everything east of India is F2. But we'll see what Spencer Wells discovered in the Philippines, when he publishes.

"That doesn't make sense. Because"

Number one is merely your assumption, not necessarily true. If your assumption proves to be incorrect then number two is a ridiculous statement. MNOPS could have arrived into the region where mtDNA was already present with some other Y-hap. And what does the basal haplogroup IJK have to do with mtDNA R in number three ?

"What belief? Whatever the case D is not any 'inconvenience' for me. Why would it be?"

You claimed Y-hap D must have had boats, whereas their distribution requires no such thing.

"What I'm sure is that the early inhabitants of Japan used boats to gather raw materials"

But how 'early' is that? I agree Jomons had boats by 15,000 years ago, but they didn't necessarily arrive in them.

"it's a minor lineage even in its area".

But you can't just go round ignoring 'minor' lineages just because they're minor. Their distribution can still tell us a great deal about human migration.

"The only region with some sizable population where something else is really big is Japan".

And Australia. It has virtually no mtDNA M and just a small proportion of Y-hap F. Surely that tells us something significant regarding early human migration. And let's not just use 'drift' as a convenient explanation.

Maju said...

Let's see:

"I am just not claiming any such separation, nor I did before".

"Just a few earlier comments from you: 'we don't have any reason at this stage to think K1 had ever any relation whatsoever with the MNOPS proposed clade'. 'I really suspect that K2, K3 and K4 are likely to belong to some higher level subclade, not yet described'. 'I'm all the time assuming that there is a macro-lineage called MNOPS between K and all those lineages'".

How does this contradict me? It does not: take pen and paper and draw a few lines and you'll see it all makes perfect sense.

I'll do it for you:

K -> MNOPS -> K2, K3, K4, M, NO, P, S
K -> L
K -> T
K -> K1

It may be wrong but that's what I'm hypothesizing all the time. Of these only MNOPS is oriental, hence, if this phylogeny stands, K coalesced in South Asia, simplifying things a lot.

If MNOPS does not exist after all or if some K# subclades would not belong to it, then the whole issue would need to be reviewed. But that's another matter.

"It's unlikely that Y-haps S and M came into Wallacea as fully-formed Y-hap S and Y-hap M. They presumably arose in situ from Y-hap MNOPS, probably in New Guinea, or somewhere very close to it".

Agree. Where's the problem? With the word "multifurcating"? Take a tea and relax: I fully agree that M and S only show signs of expansion in Melanesia and not in Thailand.

But MNOPS as a whole is a different matter. The MNOPS crossroads is rather at Thailand and not Wallacea: there is absolutely no reason to think that NO or P ever reached so far SE, even in stem stage.

Of course, if your only goal in life is to argue in favor of some conjectural Wallacean boaters' Eden, then you may force things so they fall to your side. But for the common of mortals, who don't care about such conjecture nor have any reason to think it's likely at all, the crossroads of MNOPS was surely in Indochina.

Also, I insist, there's nothing linking boats specifically to the success of some of the most numerous MNOPS subclades, like N, R or even Q before crossing to America. You are so extremely one-sided that you should write your own blog and stop bombarding people with your bias.

Maju said...

"Number one is merely your assumption, not necessarily true".

Number one is that mtDNA R coalesced in South Asia. Well, that's not at all my assumption but the most parsimonious conclusion on basal diversity: the vast majority of R sublineages are still South Asian and the rest are found both east and west of this region.

If you can't accept such basic premises, we just can't discuss anymore. Because you don't look at the phylogenetic data and draw conclusions from it, as I try to do, but you only want to manipulate the genetic data to your convenience.

I am not interested in such manipulations, which are nothing but pseudo-science.

"And what does the basal haplogroup IJK have to do with mtDNA R in number three?"

That's my basic hypothesis (of which MNOPS and oriental mtDNA R are just the eastern branch).

"You claimed Y-hap D must have had boats, whereas their distribution requires no such thing".

Crossing to Andaman and Japan requires no boats, crossing the Irrawaddy, the Mekong, the Yangtze, etc. requires no boats. Of course, who needs boats if one has faith in Wallacea, right Terry?

"But you can't just go round ignoring 'minor' lineages just because they're minor".

WTF! When I'm making a statement only about the MAJOR lineages, not just I can ignore the minor ones but in fact I must ignore them.

Re-contextualize, please! Arguing just because of arguing is not just pointless, it is confusing and annoying.

"The only region with some sizable population where something else is really big is Japan".

"And Australia".

517,000 Australian Aborigines are not a "sizable population". My urban area alone doubles that.

You really want to make me angry, don't you?

terryt said...

"How does this contradict me?"

Don't be ridiculous. It's what you said, so of course it won't contradict what you said.

"It may be wrong"

It is. This is the actual pattern:
TLKMNOPS - T, L, KMNOPS
KMNOPS - K1, K2, K3, K4, M, NO, P, S.

"Of these only MNOPS is oriental"

Wrong again. What about the Ks? You're peeling them off simply because they don't fit your belief.

"Number one is that mtDNA R coalesced in South Asia".

I repeat, although you obviously cannot read: 'that is your assumption'. There are reasons to question that assumption, although I'll follow that on the mtDNA post. Although 'If you can't accept such basic premises, we just can't discuss anymore'.

"crossing the Irrawaddy, the Mekong, the Yangtze, etc. requires no boats".

Quite possibly it doesn't. Certainly at times and in particular situations.

"When I'm making a statement only about the MAJOR lineages"

But you can't make a meaningful statement about the major lineages without considering their related minor lineages. Unless they don't fit the scheme you're proposing, in which case I quite understand why you'd ignore them.

"My urban area alone doubles that".

Is that so? Even way back in the Paleolithic?

Maju said...

"Don't be ridiculous. It's what you said, so of course it won't contradict what you said".

You posted that as if I was contradicting myself. O_O

"This is the actual pattern"...

And how come? You sound so sure... but have nothing to support it, not even the geographical parsimony that I do have.

Where I say "I think", "maybe", "looks like"... you come and say: "This is the actual pattern". Just because, of course.

Ahem!

"I repeat, although you obviously cannot read: 'that is your assumption'. There are reasons to question that assumption, although I'll follow that on the mtDNA post".

Well, if there are such reasons, please put them forward. I don't think there's any reason at all, with most R sublineages being found only in South Asia or shared with other regions.

I said (sarcastically): "crossing the Irrawaddy, the Mekong, the Yangtze, etc. requires no boats".

You said: "Quite possibly it doesn't. Certainly at times and in particular situations".

You're forcing it, bending reason in order to fit your faith. You're no different from any evangelist of those you so fiercely criticize when you write about evolution. Sad.

You need to close your Wallacean Bible and look at things as if it would have never existed. I know it's difficult, betraying one's faith is difficult. That's why it's best to have no faith at all.

"But you can't make a meaningful statement about the major lineages without considering their related minor lineages".

My statement was something like "all the major Y-DNA lineages of Eurasia coalesced ultimately in NW South Asia". You can run in circles around it like an angry imp for all long you wish but that was the statement. And it meant H, IJ and K essentially (and their ancestor F in general).

You may think it's unimportant but in that fact there are some major implications re. the colonization of Eurasia, specially at the late phase (the mtDNA R phase). And there are also some implication for your "wallacean boats" faith.

...

The first thing to face a problem is to admit it: so you have to admit that you have blind faith. Then you will be able to fight against it... but only then.

terryt said...

"You posted that as if I was contradicting myself".

I posted that you were persistently separating MNOPS from the remaining K Y-haps without any evidence whatsoever. You denied it, so I quoted you.

"And how come? You sound so sure... but have nothing to support it"

Sorry. I only have the original article separating Y-haps T and L from the K haplogroup. If that's not enough for you I'm sure it's only because you'd prefer it to be not so.

"Well, if there are such reasons, please put them forward".

I just have at the mtDNA post.

"You're forcing it, bending reason in order to fit your faith".

Have you never walked across a river?

"I know it's difficult, betraying one's faith is difficult".

I'm glad you realise it.

"That's why it's best to have no faith at all".

How would you know? Your faith in the Indian Garden of Eden is very touching.

"My statement was something like 'all the major Y-DNA lineages of Eurasia coalesced ultimately in NW South Asia'"

And that was soon after they'd first left Africa. Following that they moved back and forth round the world for some time until eventually a group emerged from India again and spread to Europe. You're concentrating on just the European phase and ignoring the rest.

"in that fact there are some major implications re. the colonization of Eurasia, specially at the late phase (the mtDNA R phase)".

Yes. In fact the implications for Europe lie only with that late phase.

"And there are also some implication for your 'wallacean boats' faith".

I don't see any of those. You'll have to spell it out I'm afraid. The Wallacean phase belongs to an earlier time than the migration into Europe, although it is fairly continuous in the region itself.

Ebizur said...

Manfred Kayser, Ying Choi, Mannis van Oven et al., "The Impact of the Austronesian Expansion: Evidence from mtDNA and Y Chromosome Diversity in the Admiralty Islands of Melanesia," Molecular Biology and Evolution 25(7):1362–1374 (2008).

Table 4
Bayesian-Based tmrca Dates of NRY Haplogroups with Assumed Melanesian origin, Estimated from all Melanesian Groups
and Separately from Admiralty Islands samples

C2-M38(xC2a-M208)
Melanesia: n=73, Mean=10,840, Median=10,180, 95% CI=5,061-17,320 years

Admiralty Islands: n=2, Mean=10,800, Median=9,880, 95% CI=4,126-19,950 years

C2a-M208
Melanesia: n=59, Mean=4,401, Median=4,318, 95% CI=2,778-6,583 years

Admiralty Islands: n=22, Mean=4,543, Median=4,197, 95% CI=1,973-7,746 years

M1a-P34
Melanesia: n=276, Mean=6,892, Median=6,667, 95% CI=4,450-9,763 years

Admiralty Islands: n=15, Mean=2,312, Median=2,103, 95% CI=818-4,370 years

S1d-M226
Melanesia: n=13, Mean=1,906, Median=1,804, 95% CI=668-3,129 years

Admiralty Islands: n=11, Mean=1,364, Median=1,198, 95% CI=397-2,706 years

S1-M254(xS1d-M226)
Melanesia: n=88, Mean=6,114, Median=5,802, 95% CI=3,407-8,874 years

Admiralty Islands: n=25, Mean=5,828, Median=5,409, 95% CI=2,726-9,876 years

K3-P79
Melanesia: n=12, Mean=4,061, Median=3,911, 95% CI=1,914-7,035 years

Admiralty Islands: n=6, Mean=3,405, Median=3,030, 95% CI=773-6,498 years

K-M9(xK3-P79, M1a-P34, O1a2-M110, O3a-M324, S1-M254)
Melanesia: n=101, Mean=15,260, Median=14,510, 95% CI=7,965-23,380 years

Admiralty Islands: n=40, Mean=14,480, Median=13,580, 95% CI=7,338-23,880 years

Note that none of the analyzed subclades of K-M9 (K3-P79, M1a-P34, S1-M254(xS1d-M226), and S1d-M226) has a TMRCA estimate for which the upper bound of the 95% confidence interval exceeds 10,000 years. All these haplogroups appear to be very young, despite the fact that they comprise an overwhelming majority of present Papuan Y-DNA (esp. M1a and S1; M1b1-M104/P22, which has not been distinguished from K*-M9 in this analysis, and K3-P79 are much more frequent in the small islands of Austronesian-speaking Melanesia and western Polynesia than they are in the mainland of New Guinea).

Second, note that there is no significant difference between the exclusively Austronesian-speaking populations of the Admiralty Islands and populations of Melanesia as a whole in regard to the TMRCA estimates for any haplogroup analyzed in this study except M1a-P34, which exhibits a significantly greater TMRCA in Melanesia as a whole.

Third, note that haplogroup K-M9 as a whole tends to exhibit a greater TMRCA estimate than haplogroup C (specifically, C2-M38) both in the Admiralty Islands and in Melanesia as a whole, though the 95% confidence intervals for the estimates do overlap. This fact does not support, though it also does not necessarily exclude, the hypothesis that would have haplogroup C2-M38 represent an older layer of population in Melanesia with K-M9 having been superimposed as a later adstrate.

Maju said...

When I said:

"I'm all the time assuming that there is a macro-lineage called MNOPS between K and all those lineages"...

I meant that K2, K3 and K4 would be derived from MNOPS and not directly from the K node (K -> MNOPS -> K2, K3, K4). It's not contradictory in any way. You misunderstood me.

"I only have the original article separating Y-haps T and L from the K haplogroup".

It does not. The paper (or at least what I have been able to know of it via Dienekes) does not specify which K# lineages belong to MNOPS, even if it does seem to suggest that some do: K -> MNOPS -> K, M, NO, P, S is what it said, where K is redundant and must refer in the second case to some K-number lineages but is not clear which ones.

Anyhow, if you have the full paper, what are you waiting to email it to me?

"I just have at the mtDNA post".

You basically said that mtDNA S would somehow magically force R to have coalesced in Wallacea (but not A or any other N sublineage). It's again a figment of your fantasy: nothing that makes any sense.

"Have you never walked across a river?"

Across a creek yes, not across a full major river. We are talking of some of the largest rivers on Earth, which are very wide and deep even near the mountains. I know you can swim across the stuary here... but also that it's impractical for daily life and to carry stuff across it. It's something that would only do the young people for mere amusement: a sport and not a practical activity. This is just a small river but it's dotted with bridges and, where these don't exist, ferry boats.

Similarly, I have often considered the better preserved landscape of my likely ancestors at Santimamiñe, which is just a small sandy/swampy stuary. You can cross swimming in low tide (with some risk because the current is really strong) but you really should not try that in high tide. But with a boat everything becomes a lot easier. You can of course walk 10 or 15 kilometers upstream until you reach the mountains, climb them and go around the obstacle... but it's plainly absurd and I'm 100% sure that if they lived where they did, overlooking the stuary, it was because they exploited it in full and not just one half of it.

So when we come to huge rivers like the Mekong or the Ganges... or the Lena in Siberia if you wish... well, I can't even imagine crossing them without some sort of boats. No: it makes no sense whatsoever.

You are just stubborn like a mule. But that doesn't give you any more reason.

"Your faith in the Indian Garden of Eden is very touching".

I don't say that there was any "Garden of Eden" anywhere except maybe in Copper Age Sumer. What I say is: follow the trail of the lineages as they are without any preconceptions.

"You're concentrating on just the European phase and ignoring the rest".

No. I'm not. NO is one of the major lineages of the World and together with the rest of the proposed MNOPS it's even larger.

"Yes. In fact the implications for Europe lie only with that late phase".

Again you're ignoring F and B, among others. So it's not just "Europe", though also.

Maju said...

Ebizur: is there a possible older source for those lineages out of Melanesia?

I mean: if they are so recent (who knows?) they must have originated in some older group. If so where? In Melanesia itself, Papua, Philippines, Wallacea, Taiwan?

By "Melanesia" does this paper includes or exclude New Guinea?

Ebizur said...

Maju asked,

"Ebizur: is there a possible older source for those lineages out of Melanesia?

I mean: if they are so recent (who knows?) they must have originated in some older group. If so where? In Melanesia itself, Papua, Philippines, Wallacea, Taiwan?"

The best candidate would be the place with the highest diversity of Y-DNA that belongs to the parent clade, i.e. MNOPS-M526. Of course, it is possible that Melanesia itself might be identified as the most likely place of origin according to this criterion. I am especially eager to know whether K3-P79 is a subclade of MNOPS-M526. Could someone brave inquire into this matter with Chiaroni, Underhill, or Cavalli-Sforza?

"By 'Melanesia' does this paper includes or exclude New Guinea?"
It includes New Guinea, and, in fact, most of the samples that have been pooled to form the set labeled as "Melanesia" in the TMRCA analysis of Kayser et al. (2008) have been obtained from populations resident on the island of New Guinea. The samples that are not from the mainland of New Guinea, but rather from outlying Melanesian islands, are Biak (n=10; Austronesians from the Biak Islands, located off the northern coast of western New Guinea; Mona et al. 2007), Tolai (n=19; Austronesians from New Britain, a large island located off the northeastern coast of New Guinea; Kayser et al. 2008), Trobriand Islanders (n=52; Austronesians from the Trobriand Islands, located near the southeastern tip of New Guinea and south of New Britain; Kayser et al. 2008), and Admiralty Islanders (n=147, including 11 Seimat–Wuvulu, 21 Titan, 17 Nyindrou, 4 Mokerang, 14 Ere–Kele, 23 Lele, 19 Nali, 20 Andra–Hus, and 18 Kurti; Austronesians from the Admiralty Islands, which are located off the northeastern coast of New Guinea and northwest of New Britain; Kayser et al. 2008).

Maju said...

"The best candidate would be the place with the highest diversity of Y-DNA that belongs to the parent clade, i.e. MNOPS-M526. Of course, it is possible that Melanesia itself might be identified as the most likely place of origin according to this criterion".

Looks like. But, if they are all that recent, then rather not but where then?

I also meant to ask about C2, though I guess it's the same case, right?

"Could someone brave inquire into this matter with Chiaroni, Underhill, or Cavalli-Sforza?"

Hehe. I'm not that blunt. Or rather, I'm so blunt at times that I fear to offend when doing such delicate dealings.

"Hey, Luigi. Your namesake here, could you tell me what sublineages exactly belong to MNOPS? Thanks". :D

Really, I'm more in the mood of waiting an see. It's intriguing but I'll sleep fine without knowing.

Thanks for your help in any case, Ebizur.

terryt said...

"I meant that K2, K3 and K4 would be derived from MNOPS and not directly from the K node (K -> MNOPS -> K2, K3, K4)".

But there's no evidence for that. I know exactly why you're keen to eliminate the Ks from consideration. I tried the same thing at remotecentral with my diagram comparing mtDNA and Y-haps. My motivation then was the same as yours is now: an attempt to make the evidence fit your prefered option, India. I was giving you the benefit of the doubt. With T and L still considered at the time to be part of the same clade it was sort of possible to accomodate an India origin. You criticised my separation on the grounds that L,T K1, K2, K3, K4 M1, M2, M3, NO, P and S were regional varieties of the same original haplogroup. Your stance has certainly changed now that T and L have been removed.

I have a visitor so I'll return later.

terryt said...

"This fact does not support, though it also does not necessarily exclude, the hypothesis that would have haplogroup C2-M38 represent an older layer of population in Melanesia with K-M9 having been superimposed as a later adstrate"

I'm pretty sure that C2 is accepted as entering Melanesia from Eastern Indonesia. It's usually accepted that C2 entered Melanesia with the Austronesians, so it makes sense that it is more recent than K there, rather than representing an older layer of population in Melanesia.

"The paper (or at least what I have been able to know of it via Dienekes) does not specify which K# lineages belong to MNOPS"

Perhaps not, but the associated diagram specifically claims (for the third time) 'and diamond: M525 that unifies KMNOPS'. I have a hard copy I printed off at the time, but I'll try to find the same thing on the net.

"We are talking of some of the largest rivers on Earth, which are very wide and deep even near the mountains".

They all have headwaters.

"I'm 100% sure that if they lived where they did, overlooking the stuary, it was because they exploited it in full and not just one half of it".

And how long ago was that?

"Of course, it is possible that Melanesia itself might be identified as the most likely place of origin according to this criterion"

Thank you.

terryt said...

"follow the trail of the lineages as they are without any preconceptions".

OK. Here we go:

As you've said, we have F* developing in NW South Asia (or perhaps further west). Somewhere on the Iranian Plateau anyhow: modern Iran, Afghanistan and the western half of Pakistan.

It's obvious that F* moved into India at some time, leaving behind what later became Y-hap G. In India Y-hap F* gave rise to several other regional haplogroups as its descendants spread around and across India. These became Y-haps F1, F2, F3 (this last a bit suspect), F4, H, and a clade we could call IJTKLMNOPS*. But this latter haplogroup may also have been left behind on the Iranian Plateau, because it splits into two: a Plateau variety (IJ) and an eastern haplogroup we could call TLKMNOPS*.

As this haplogroup moved further southeast it split into three: T, L and KMNOPS*. The three haplogroups are probably all Indian, although KMNOPS* may actually have moved east with F2. As may have Y-hap T, but we'll ignore that possibility for now.

It's obvious that KMNOPS* moved far to the southeast. Five of the eight haplogroups that coalesced from it are found east of Wallacea, three of them only so. And it's very easy to make a case for the other three, K1, NO and P, as having developed near there too.

The next point is that it is very difficult to support the case that Y-hap KMNOPS* entered SE Asia with mtDNA R. The expansion of the whole Y-hap F clade fits extremely well with the hypothesis that it and mtDNA M entered India together. And moved across India and into SE Asia together, as far as Wallacea.

On the other hand it is also very difficult to support the case that Y-hap C* brought mtDNA M into SE Asia. MtDNA M is a major clade through the whole region, whereas Y-hap C is a very minor component of the Indian Y-haps. To support the case you're forced to hypothesise major drift to reduce C to the required level.

But there's yet another problem. It seems likely that Y-hap C* reached Australia with mtDNA N*, so surely it's reasonable to assume that C came into SE Asia with N rather than with M.

Next. It's very easy to make the case for Y-haps NO and P having carried mtDNA R with them. This bunch of male and female haplogroups are spread together all the way from Melanesia to western Europe. By the time this paricular haplogroup combination emerged onto the steppes from NW South Asia it had become 'the first spark (and what a spark!) of the Great Eurasian Expansion'. But a lot had happened before then, as I've tried to point out several times.

terryt said...

Dienekes actually has the new haplogroup diagram up in his article:

http://dienekes.blogspot.com/2009/11/y-chromosome-diversity-human-expansion.html

terryt said...

Sorry. Me again:

"Looks like. But, if they are all that recent, then rather not but where then?"

Private lineages, whose expansion in the region was delayed?

Maju said...

"But there's no evidence for that".

Nor the opposite. I argued it on parsimony reasons.

I told you before and you agreed with it.

I'm not eliminating "the Ks": I am assuming they behave parsimoniously where there's no clear data.

"... the associated diagram specifically claims (for the third time) 'and diamond: M525 that unifies KMNOPS'"

Precisely. But still it's totally unclear because K includes T and L (and M, NO,P and S, of course, plust the smaller K-number clades and K*). So I undertstand that by using the terms "KMNOPS" and "K" under it they don't mean K as in ISOGG (defined by M9, P128, P131 and P132) but the K-number guys OR a fraction of them.

It's so extremely unclear that I had to draw my own provisional conclusions on the K-number lineages based on geographic parsimony.

When you know something more, tell me and we can continue this discussion properly.

"They all have headwaters".

And? The Ganges near the Himalaya is still wide and worse: it's chilly and and revolted. I know that you can have serious troubles in smaller rivers's headwaters like the James River of Virginia, where two friends almost died after their canoe capsized some 23 years ago, right at the Appalachians.

So it's clear that you are talking absolute nonsense: people won't go all the way up the river, having to cross more and more affluents as you walk, and then climb some mountains, like the Hymalayas or Hindu Kush, that then were pure glaciers all around, only to cross a single river.

You must stop bombarding me with those idiocies because I'm starting to believe you are not just a fanatic but also quite dumb.

"And how long ago was that?"

Since "always". Even Neanderthals lived there. And it's all like that everywhere: rivers are paths, not obstacles.

Maju said...

"... we could call IJTKLMNOPS*..."

Either you use standard ISOGG nomenclature or you can forget me. All your nomenclature around IJK-K-MNOPS is messy, confusing, slippery and annoying.

"On the other hand it is also very difficult to support the case that Y-hap C* brought mtDNA M into SE Asia. MtDNA M is a major clade through the whole region, whereas Y-hap C is a very minor component of the Indian Y-haps. To support the case you're forced to hypothesise major drift to reduce C to the required level".

I only need MNOPS sweeping after them, bringing some major mtDNAR (B, F, P) but also becoming associated with pre-existing oriental mtDNA M and N, depending where.

What I am considering is that the migration of a "tribe" with lineages (pre-)MNOPS (Y-DNA) and some R (evolving to B, F, P, etc. in situ) stuff (mtDNA) largely replaced, specially in SE Asia, China and Papua, the previous reality dominated by Y-DNA C and D and mtDNA M and N(xR). The effect was of course more marked in the Y-DNA side of the equation, as is logical if we consider a more or less male-dominated society.

"It's very easy to make the case for Y-haps NO and P having carried mtDNA R with them".

No with Y-DNA P.

But it's very easy to do the case for Y-DNA NO moving along mtDNA macro-F and macro B, and for Y-DNA M and S with mtDNA P. Hence MNOPS and oriental R seem coupled.

"Dienekes actually has the new haplogroup diagram up in his article"...

Thanks. Per ISOGG the haplogroup at the level of the star is called K. So placing K under a subclade of K is absurd. He probably means "some K-other" but it's not clear which.

terryt said...

"Per ISOGG the haplogroup at the level of the star is called K".

But ISSOGG hasn't got around to taking Y-haps T and L out yet. As far as I know the authors give no name for the clade at the star, but we can't 'use standard ISOGG nomenclature' for it, because they don't give it a name either. Call it K if you insist, but that gives a false impression of the makeup of the clade. T and L are pre-K so cannot be included in the name. With no 'standard ISOGG nomenclature' I've had to use TLKMNOPS* (call it whatever you like, but a rose etc.) and KMNOPS* (again call it what you like) to avoid confusion when refering to 'K'. But I'd tend to call just KMNOPS* Y-hap K if you want to shorten the nomenclature. What do you suggest for the clade that includes T and L?

"It's so extremely unclear that I had to draw my own provisional conclusions on the K-number lineages based on geographic parsimony".

Along with your assumptions, preconceptions and prejudices.

"I only need MNOPS sweeping after them, bringing some major mtDNAR (B, F, P)"

And that's probably what happened. Almost everywhere.

"What I am considering is that the migration of a 'tribe' with lineages (pre-)MNOPS (Y-DNA) and some R (evolving to B, F, P, etc. in situ) stuff (mtDNA) largely replaced, specially in SE Asia, China and Papua"

And probably starting out from round there, although I note you're still loath to include anything that smells at all of 'Y-hap K' in there.

"the previous reality dominated by Y-DNA C and D and mtDNA M and N(xR)".

Quite possibly. But the M must have come previously from India. And that makes Y-Hap F its most likely soulmate. Or perhaps your imagined new haplogroup K.

"No with Y-DNA P".

What? The descendants of Y-hap P (in the form of R and Q) spread right across Eurasia, into America and even down into Africa. And so do the descendants of mtDNA R. And you claim there's no connection between the two?

"But it's very easy to do the case for Y-DNA NO moving along mtDNA macro-F and macro B, and for Y-DNA M and S with mtDNA P".

Very easy indeed.

"Hence MNOPS and oriental R seem coupled".

But you don't want to include the P section of the haplogroup because? Is it because it doesn't fit a particular belief?

"So placing K under a subclade of K is absurd. He probably means 'some K-other' but it's not clear which".

You're obviously hoping so.

Maju said...

"But ISSOGG hasn't got around to taking Y-haps T and L out yet".

And will not. Haplogroup K is defined by certain SNPs that are upstream of whatever is MNOPS plus all other K, including L and T (of course).

That some people, including you, confuse K with the minor K-number clades is meaningless and only means that you are confused about the proper nomenclature.

"As far as I know the authors give no name for the clade at the star, but we can't 'use standard ISOGG nomenclature' for it, because they don't give it a name either".

It has a name: K.

Nobody is saying that K has suddenly stopped existing, just that there may be a large subclade downstream of it, grouping many but not all K sublineages.

"T and L are pre-K"...

No. L and T are part of K. This has never been questioned.

"What do you suggest for the clade that includes T and L?"

K. Its name is K and has been that way since 2001.

Maju said...

"But the M must have come previously from India. And that makes Y-Hap F its most likely soulmate. Or perhaps your imagined new haplogroup K".

I'm actually thinking of Y-DNA C and D, because F(xMNOPS) or at least F(xK) is extremely rare east of Bengal.

As there is a lot of old basal M in Eastern Eurasia, it must have arrived really soon, right at the beginning of the Great Eurasian Expansion. But there is no (or almost no) Y-DNA F that can be considered coupled with it, so we have to consider other options: notably the otherwise uncoupled Y.DNA lineages C and D. And anyhow, C is "brother" of F, while D is a "cousin", so it makes good sense.

"What? The descendants of Y-hap P (in the form of R and Q) spread right across Eurasia, into America and even down into Africa. And so do the descendants of mtDNA R. And you claim there's no connection between the two?"

Not a generic one. One that can be considered valid for every region. In West Eurasia Y-DNA P (R1) is only one of several F-derived lineages and in West Asia specifically, a large majority of mtDNA R-derived lineages is coupled with a majority of F(xK) Y-DNA ones (IJ notably).

On the other hand, in America and Siberia, Y-DNA Q is not significantly coupled with mtDNA R. Only in America with B but this lineage is not found in North Asia and hence its arrival to America must be considered quite a product of chance, of a rare "Beringian coincidence" (founder effect).

So I don't see any reason to support what you claim, unless R were a group of Amazons that picked their husbands from diverse groups. In West Eurasia mtDNA R is coupled most perfectly with F/IJK, not specifically K or MNOPS. In Eastern Eurasia this is different, with K or MNOPS being the only sublineage of IJK really common and that can be associated with mtDNA R (what you seem to agree with).

"But you don't want to include the P section of the haplogroup because?"

Because it's not consistent on its own with any mtDNA pattern. Only when you think in higher level clade IJK it begins to make sense.

Ebizur said...

Maju said,

"Looks like. But, if they are all that recent, then rather not but where then?

I also meant to ask about C2, though I guess it's the same case, right?"

terryt said,

"I'm pretty sure that C2 is accepted as entering Melanesia from Eastern Indonesia. It's usually accepted that C2 entered Melanesia with the Austronesians, so it makes sense that it is more recent than K there, rather than representing an older layer of population in Melanesia."

For at least the last decade, since the publication of Manfred Kayser, Silke Brauer, Gunter Weiss et al., "Melanesian origin of Polynesian Y chromosomes," Current Biology (2000) 10 : 1237–1246, Y-DNA haplogroup C in Oceanic Austronesians (e.g. Polynesians) has been assumed to have a proximal Melanesian origin. By the time Kayser et al. published "Reduced Y-Chromosome, but Not Mitochondrial DNA, Diversity in Human Populations from West New Guinea" in 2003, all the various forms of haplogroup K-M9(xO-M175) that had been found in Austronesian and Papuan populations likewise had been assigned a putative indigenous Melanesian/Papuan origin. (Note that Table 4 in the paper by Kayser et al. (2008) that I have referenced in one of my previous comments in this thread has been named "Bayesian-Based tmrca Dates of NRY Haplogroups with Assumed Melanesian origin, Estimated from all Melanesian Groups and Separately from Admiralty Islands samples," and it includes TMRCA estimates for C2*-M38, C2a-M208, and various subclades of K-M9.)

However, it has become quite clear that an overwhelming majority of haplogroup C Y-DNA in populations from Papua New Guinea eastward belongs to the subclade C2a-M208, which exhibits a rather young TMRCA:

C2a-M208
Melanesia: n=59, Mean=4,401, Median=4,318, 95% CI=2,778-6,583 years (Kayser et al. 2008)

Admiralty Islands: n=22, Mean=4,543, Median=4,197, 95% CI=1,973-7,746 years (Kayser et al. 2008)

Cox et al. (2007)
TMRCA for C2a-M208 based on variation in 12 Y-chromosome STRs: Mean=8,100 years, 95% confidence interval=3,900-12,300 years

Most Polynesian Y-DNA belongs to one particular subclade of C2a-M208, namely C2a1-P33, which exhibits an even shorter TMRCA.

Cox et al. (2007)
TMRCA of extant P33 lineages using variation in 23 Y-chromosome STRs under a coalescent-based geometric-geometric mutation model: Mean=4,500 years, 95% confidence interval=1,500-7,500 years

Ebizur said...

On the other hand, Y-DNA haplogroup C2*-M38(xC2a-M208) exhibits a rather long TMRCA, which suggests that it is either a paraphyletic assemblage of subclades that have expanded recently from immigrating founders with widely divergent STR haplotypes (parallel to the already identified subclade, C2a-M208), or else it is a truly indigenous haplogroup with a history of more than 10,000 years in Wallacea. (One can exclude the hypothesis that C2-M38 has originated in Sahul, since one should expect C2-M38 to appear also in Australia if it had reached New Guinea at such an early date, and C2-M38 has never been found in any Australian aborigine.)

Quoting Cox et al. 2007,
"The distribution of C-M38* (xM208, P33) is centered on eastern Indonesia and Melanesia (Underhill et al. 2001) and has been equated with the C-RPS4Y/DYS390.3del variant in earlier studies (Kayser et al. 2003). Bayesian dates for this mutation yielded upper confidence limits with considerable antiquity: 30,300
to 4,500 years b.p. (95% confidence interval; mean 10,600 years b.p.) (Kayser et al. 2003). Despite a reduced chronological range, we also infer a late Pleistocene ancestor for M38-derived lineages, between 37,200 and 28,800 years b.p. (95%
confidence interval; mean 33,000 years b.p.). Coupled with this marker’s spatial distribution, there seems little doubt that the ultimate paternal ancestors of Polynesian P33 carriers once lived in Melanesia or its immediate environs (Kayser et al. 2003)."

Kayser et al. (2008)
C2-M38(xC2a-M208)
Melanesia: n=73, Mean=10,840, Median=10,180, 95% CI=5,061-17,320 years

Taken together, these three TMRCA estimates for C2*-M38 suggest that it is most diverse in the Lesser Sunda Islands and the Maluku Islands of Wallacea in eastern Indonesia, because the upper bound for the TMRCA of C2*-M38 based on the "Melanesia" pool of Kayser et al. 2008, which includes samples from mainland Western New Guinea, mainland Papua New Guinea, and several outlying island groups, is much less than the upper bound of the earlier TMRCA estimates of Kayser et al. (2003) and Cox et al. (2007) that have been derived from pools of samples that include the Lesser Sundas and the Moluccas.

The frequency of C2*-M38 also shows a pattern opposite to that of its subclade, C2a-M208. C2*-M38 occurs most frequently in populations of Wallacea and Western New Guinea, whereas C2a-M208 occurs most frequently in populations of Remote Polynesia, but parts of the area around the midpoint between these two poles are practically devoid of both C2*-M38 and C2a-M208 (e.g. 0/32 C-M130 in the Solomon Islands according to Cox et al. 2006). So, in general, C2*-M38 is situated toward the west, C2a-M208 (and especially C2a1-P33) is situated toward the east, and K-M9 is particularly strong in the center (around Island Melanesia).

Ebizur said...

C2*-M38(xC2a-M208)
5/5 = 100% Tehit (Mona et al. 2007)
16/24 = 66.7% Maibrat (Mona et al. 2007)
15/24 = 62.5% Baham (NWNG; Mona et al. 2007)
202/352 = 57.4% Sumba (Cox et al. 2007)
5/10 = 50% Moskona (Mona et al. 2007)
4/12 = 33.3% Timor (Cox et al. 2007)
7/22 = 31.8% Karon (Mona et al. 2007)
10/36 = 27.8% Maluku Islands (Cox et al. 2007)
3/11 = 27.3% Mantion (Mona et al. 2007)
2/8 = 25% Moi (Mona et al. 2007)
5/22 = 22.7% C-M130 Banjarmasin, southern Borneo, Indonesia (Hurles et al. 2005)
7/33 = 21.2% Moluccas
9/44 = 20.5% Maewo, Vanuatu (Cox et al. 2007)
1/5 = 20% Irarutu (Mona et al. 2007)
41/234 = 17.5% C-M130 Vanuatu (Cox et al. 2006)
9/52 = 17.3% C-M130 Vanuatu (Hurles et al. 2005)
81/485 = 16.7% Flores (Cox et al. 2007)
2/12 = 16.7% Hatam (Mona et al. 2007)
5/31 = 16.1% Nusa Tenggaras
3/19 = 15.8% Tolai New Britain
10/65 = 15.4% C-M130 Kota Kinabalu, northern Borneo, Malaysia (Hurles et al. 2005)
5/34 = 14.7% Moluccas (Kayser et al. 2003)
6/48 = 12.5% West Province, Sulawesi (Cox et al. 2007)
4/33 = 12.1% PNG Coast
8/89 = 9.0% WNG Lowlands
9/105 = 8.6% Fiji
1/16 = 6.3% Micronesia (Cox et al. 2007)
2/37 = 5.4% New Guinea (Scheinfeldt et al. 2006; both C2*-M38 individuals belong to the "PNG Coast:North Coast" subset (2/25))
8/152 = 5.3% New Ireland (Scheinfeldt et al. 2006)
1/21 = 4.8% Coastal New Guinea (Cox et al. 2007)
2/44 = 4.5% C-M130 Papua New Guinea (Hurles et al. 2005)
1/29 = 3.4% Tonga
1/31 = 3.2% PNG Highlands
1/40 = 2.5% Southern Borneo
1/51 = 2.0% Highland New Guinea (Cox et al. 2007)
1/61 = 1.6% Samoa
6/395 = 1.5% New Britain (Scheinfeldt et al. 2006)
1/74 = 1.4% Mentawai (Cox et al. 2007)
1/75 = 1.3% Bougainville (Scheinfeldt et al. 2006)
1/77 = 1.3% Cook Islands
1/147 = 0.7% Admiralty Islands (cf. Table 1 of Kayser et al. 2008)
0/2 Onin (Mona et al. 2007)
0/6 Tokelau
0/7 Manus (Scheinfeldt et al. 2006)
0/9 Niue
0/9 Rapanui (Easter Island) (Cox et al. 2007)
0/10 Biak (Mona et al. 2007)
0/10 Nasioi (Bougainville) (Cox et al. 2007)
0/10 Wandamen (Mona et al. 2007)
0/11 Majuro (Hurles et al. 2005)
0/12 Tonga (Cox et al. 2007)
0/17 Malaysia
0/18 Samoa (Cox et al. 2007)
0/19 Ekari (Mona et al. 2007)
0/19 Mussau (Scheinfeldt et al. 2006)
0/25 Tahiti (Cox et al. 2007)
0/28 Philippines (Hurles et al. 2005)
0/32 Solomon Islands (Cox et al. 2007)
0/35 Bereina PNG
0/35 Malagasy (Hurles et al. 2005)
0/37 Philippines
0/39 Taiwan Aborigines (Hurles et al. 2005)
0/46 Kapuna PNG
0/50 East Futuna
0/50 New Britain (Cox et al. 2007)
0/52 Trobriand Islands
0/53 Java
0/56 Sumatra
0/94 WNG Highlands
0/100 Tuvalu

Ebizur said...

C2a-M208
9/9 = 100% C2a1-P33 Rapanui/Easter Island (Cox et al. 2007)
16/20 = 80.0% C-M130 Cook Islands (Hurles et al. 2005)
59/77 = 76.6% Cook Islands
13/18 = 72.2% Samoa (Cox et al. 2007) [includes 12/18 C2a1-P33]
16/25 = 64.0% Tahiti (Cox et al. 2007) [includes 15/25 C2a1-P33]
37/61 = 60.7% Samoa
15/25 = 60.0% C-M130 Western Samoa (Hurles et al. 2005)
23/54 = 42.6% C2-M38 Maori (Underhill et al. 2001)
7/17 = 41.2% C2-M38 non-Maori Polynesians (Underhill et al. 2001)
2/5 = 40% Irarutu (Mona et al. 2007)
4/12 = 33.3% C2a1-P33 Tonga (Cox et al. 2007)
5/16 = 31.3% C-RPS4Y Rapanui (or 5/10 = 50% when four P(xQ3, R1a) and two Q3 Y-chromosomes of putative European and Native American origins have been excluded; remaining five individuals belong to K*(xM,N,P), i.e. K(xM1, N1, P) in present nomenclature) (Hurles et al. 2003)
9/29 = 31.0% Tonga
15/50 = 30.0% East Futuna
6/21 = 28.6% C-M130 Kapingamarangi (Hurles et al. 2005)
23/94 = 24.5% WNG Highlands
8/37 = 21.6% New Guinea (Scheinfeldt et al. 2006; "The samples from New Guinea do not represent a true population
because they were men from a number of areas (primarily the Sepik) who had settled in the Northern Island
Melanesian region." All C2a-M208 individuals belong to the "PNG Coast" subset, including 7/25 "North Coast" and 1/2 "Rigo.")
17/100 = 17.0% Tuvalu
1/6 = 17% Tokelau
22/147 = 15.0% Admiralty Islands (Kayser et al. 2008)
1/7 = 14% Manus (Scheinfeldt et al. 2006)
14/105 = 13.3% Fiji
5/52 = 9.6% Trobriand Islands
2/21 = 9.5% C2a-M208(xC2a1-P33) Coastal New Guinea (Cox et al. 2007)
3/33 = 9.1% PNG Coast
2/35 = 5.7% Bereina PNG
1/19 = 5.3% Ekari (Mona et al. 2007)
1/19 = 5.3% Mussau (Scheinfeldt et al. 2006)
1/19 = 5.3% Tolai New Britain
1/22 = 4.5% Karon (Mona et al. 2007)
1/44 = 2.3% C2a-M208(xC2a1-P33) Maewo, Vanuatu (Cox et al. 2007)
3/395 = 0.76% New Britain (Scheinfeldt et al. 2006)
1/152 = 0.66% New Ireland (Scheinfeldt et al. 2006)
0/9 Niue
0/10 Nasioi (Bougainville) (Cox et al. 2007)
0/11 C-M130 Majuro (Hurles et al. 2005)
0/12 Timor (Cox et al. 2007)
0/16 Micronesia (Cox et al. 2007)
0/17 Malaysia
0/28 C-M130 Philippines (Hurles et al. 2005)
0/31 Nusa Tenggaras
0/31 PNG Highlands
0/32 Solomon Islands (Cox et al. 2007)
0/33 Moluccas
0/36 Maluku Islands (Cox et al. 2007)
0/37 Philippines
0/39 C-M130 Taiwan Aborigines (Hurles et al. 2005)
0/40 Southern Borneo
0/46 Kapuna PNG
0/48 West Province, Sulawesi (Cox et al. 2007)
0/50 New Britain (Cox et al. 2007)
0/51 Highland New Guinea (Cox et al. 2007)
0/53 Java
0/56 Sumatra
0/74 Mentawai (Cox et al. 2007)
0/89 WNG Lowlands
0/352 Sumba (Cox et al. 2007)
0/485 Flores (Cox et al. 2007)

Ebizur said...

All data are from Table S2 of Kayser et al. (2008) unless otherwise indicated.

One half of the Maori sample and approximately one quarter of the Other Polynesian sample of Underhill et al. (2001) belong to haplogroups that presumably reflect recent European admixture. When these Y-chromosomes have been excluded, C2-M38 (most of which may be presumed to be C2a1-P33 or at least C2a-M208) accounts for 85% (23/27) of the Maori and 54% (7/13) of the Other Polynesians.

Maju said...

Thanks Ebizur for your overwhelming yet very informative data.

I feel that the MRCA age estimates mentioned here and in Wikipedia (surely same sources) are just too low, sincerely. If C is, as claimed, some 60 Ka old, then C2 should be roughly 3/4 that age, that is: 45 Ka, C2a should be in the 30 Ka region and C2a1/2 in the 15 Ka area. Of course these are just very rough hunches but there is still a very large difference between 45 Ka and the 10 Ka estimated by Kayser for the various segments.

I guess that adding the pitiful state of MRCA age estimation methods, very specially for Y-DNA, and the fact that such a small area must have suffered from homogenization processes that naturally reduce diversity, can account for these differences.

Ebizur said...

I almost forgot to post about this interesting little error.

Kayser et al. (2003)
Moluccas
3/34 = 8.8% C-RPS4Y711(xC2-M38, C3-M217, C4-390.1del)
5/34 = 14.7% C2-M38(xC2a-M208)
6/34 = 17.6% K-M9(xS-M230, M1-M4, O-M175, P-M74)
7/34 = 20.6% S-M230
7/34 = 20.6% M1-M4(xM1b1-M104)
2/34 = 5.9% O1a-M119
4/34 = 11.8% O3-M122

Kayser et al. (2008)
Moluccas
1/33 = 3.0% C-RPS4Y(xC2-M38, C3-M217)
7/33 = 21.2% C2-M38(xC2a-M208)
5/33 = 15.2% K-M9(xK3-P79, M1-M4, M2-M353, M3-P117, NO-M214, P-M74, S-M230)
7/33 = 21.2% S1-M254(xS1d-M226)
7/33 = 21.2% M1a-P34
1/33 = 3.0% O1a-M119(xO1a2-M110)
1/33 = 3.0% O1a2-M110
4/33 = 12.1% O3a-M324(xO3a3b-M7, O3a3c-M134)

Note that one of the K*-M9 Y-chromosomes has been shaved off this team's "Moluccas" sample altogether between 2003 and 2008. This is not what I would call an error, but rather a consequence of the misfortunes that randomly may befall a sample in a lab; it may be tested to exhaustion, it may be misplaced and lost, and so forth. However, they have committed a glaring and undeniable error either in their genotyping or in their reporting of the results of their genotyping of two of the haplogroup C individuals from this sample set: these two individuals have miraculously mutated from C-RPS4Y711(xC2-M38, C3-M217, C4-390.1del) to C2-M38(xC2a-M208) in the course of five short years! At least Kayser et al. would have the reader believe so, since they have not published an erratum as far as I know.

Anyway, I will summarize the TMRCA estimates of haplogroup C2 and its subclades from all the studies available to me:

C2a1-P33
Cox et al. (2007):
"Given its Polynesian specificity, the P33 mutation probably arose just before, or during, the region’s initial settlement. Archeological remains of a founding Polynesian settlement at Nukuleke, Tonga, were radiocarbon dated to about 2,900 years b.p. cal. (Burley and Dickinson 2001). Although the Tongan and Samoan archipelagos of Western Polynesia were colonized rapidly, there was a hiatus before settlement farther eastward (Kirch 2000). The age of the P33 mutation should match this timeframe within the current limits of genetic dating. We inferred the TMRCA of extant P33 lineages using variation in 23 Y-chromosome STRs under a coalescent-based geometric-geometric mutation model (Watkins 2007). The P33 transition likely arose within a broad temporal window from 7,500 to 1,500 years b.p. (95% confidence interval; mean 4,500 years b.p.). Although encompassing a large chronological spectrum, this interval is consistent with radiocarbon estimates for the earliest Polynesian settlements."

(to be continued)

Maju said...

Curious, Ebizur. Is it possible that further and more modern testing of the samples has produced a refinement in the data?

Anyhow, I understand that even if C2a1 spread with Polynesian expansion, the mutation must have pre-existed most likely, being picked up by proto-Polynesians at the earliests stages, when the Melanesian islands group was apparently important in their cultural genesis.

Whatever the case I can take the 7000 years upper CI as reasonable. But 1500, LONG AFTER the colonization of Tonga, looks like nope.

Ebizur said...

Maju said,

"Curious, Ebizur. Is it possible that further and more modern testing of the samples has produced a refinement in the data?"

No, that is impossible. It is as simple as this: the field value at the intersection of the row "Moluccas" (n=34) and the column "C-RPS4Y711" (i.e. C-RPS4Y711(xC2-M38, C3-M217, C4-390.1del)) in Table 3 of Kayser et al. (2003) is 8.8 (i.e. 8.8%, or 3/34), and the intersection of the same row and the column "C-M38" (i.e. C2-M38(xC2a-M208)) in the same table contains the value 14.7 (i.e. 14.7%, or 5/34). In contrast, the "Moluccas" (n=33) row of Table S2 of Kayser et al. (2008) has a value of 1 at the intersection with the "C-RPS4Y*" column (i.e. 1/33 C-RPS4Y711(xC2-M38, C3-M217)) and a value of 7 at the intersection with the "C-M38*" column (i.e. 7/33 C2-M38(xC2a-M208)). There is no room for multiple interpretations here; Kayser et al. have made a flat-out error. C-RPS4Y711(xC2-M38, C3-M217) and C2-M38(xC2a-M208) are 100% mutually exclusive categories; any particular Y-chromosome cannot be both M38+ and M38- simultaneously.

terryt said...

"Thanks Ebizur for your overwhelming yet very informative data".

My thanks too. Especially the data on the Moluccas (or Maluku if you prefer). The presence of S is very interesting.

"it is most diverse in the Lesser Sunda Islands and the Maluku Islands of Wallacea in eastern Indonesia"

Ah. That place again. I see you've explained the significance of the dates yourself but I'll add my own comments.

"an overwhelming majority of haplogroup C Y-DNA in populations from Papua New Guinea eastward belongs to the subclade C2a-M208, which exhibits a rather young TMRCA"

All the dates you provide for the TMRCA for C2a-M208 easily fit the timing of the Austronesian expansion beyond New Britain.

"C2a1-P33, which exhibits an even shorter TMRCA".

Again that fits all the other data. The Polynesian expansion beyond Fiji/Tonga/Samoa dates back barely 2000 years.

"On the other hand, Y-DNA haplogroup C2*-M38(xC2a-M208) exhibits a rather long TMRCA"

And I'd go with your second option, 'it is a truly indigenous haplogroup with a history of more than 10,000 years in Wallacea'.

"One can exclude the hypothesis that C2-M38 has originated in Sahul"

True, for the reaons you propose. It's a separate Y-hap from either Australian C4 or New Guinea C6.

"The distribution of C-M38* (xM208, P33) is centered on eastern Indonesia and Melanesia"

Exactly as I suggested: 'I'm pretty sure that C2 is accepted as entering Melanesia from Eastern Indonesia'.

"we also infer a late Pleistocene ancestor for M38-derived lineages, between 37,200 and 28,800 years b.p."

I suspect that C3, C4 and C6 diverged in the region more than 50,000 years. Of course the M38 line may have been a 'private lineage' (as Maju terms it) until the more recent date.

"there seems little doubt that the ultimate paternal ancestors of Polynesian P33 carriers once lived in Melanesia or its immediate environs"

And what's really interesting is that Y-hap O is also fairly common in Polynesians, although it doesn't reach marginal East Polynesia. So the Polynesians were originally a mixed population: from both Melanesia and South China.

"parts of the area around the midpoint between these two poles are practically devoid of both C2*-M38 and C2a-M208"

The theory is that the first expansion of Austronesian-speaking people into the wider Pacific (Lapita pottery) was followed by a more Melanesian component. This would explain 'the K-M9 is particularly strong in the center (around Island Melanesia)'. Presumably specifically Y-hap M2.

terryt said...

Now. Maju.

"If C is, as claimed, some 60 Ka old, then C2 should be roughly 3/4 that age, that is: 45 Ka, C2a should be in the 30 Ka region and C2a1/2 in the 15 Ka area".

I hope my comments above help you fit it all together and make sense of the dates.

"K. Its name is K and has been that way since 2001".

OK then What do you suggest we call the clade that excludes T and L?

"As there is a lot of old basal M in Eastern Eurasia, it must have arrived really soon, right at the beginning of the Great Eurasian Expansion. But there is no (or almost no) Y-DNA F that can be considered coupled with it"

But there's rather a lot of what both of us would agree is F-derived Y-hap K, which seems to get all the way to, and across, Wallacea before it moves back again.

"so we have to consider other options: notably the otherwise uncoupled Y.DNA lineages C and D".

But there's no D at all in India, and very little C, so that idea doesn't really hold up. And they may not be uncoupled. Try mtDNA N, although I know you don't like to consider that possibility.

"unless R were a group of Amazons that picked their husbands from diverse groups".

It's more likely to be the other way round. Y-hap P picked up their wives from diverse groups. That certainly seems to be the case with Y-hap Q in America.

Ebizur said...

C2a-M208
Cox et al. (2007):
"The spatial and temporal distributions of ancestral haplogroups also enlighten the early history of ancestral Polynesians. A relative paucity of C-M208*(xP33) suggests that the marker may have developed not long before P33. Expanding Austronesian-speaking populations probably assimilated M208 individuals (Green 1991, 2003) as they moved eastward along New Guinea’s northern coast about 3,500 years b.p. (Spriggs 2003). Descendants of these admixed communities later swept Melanesian markers, such as the lineages considered here, to high frequency in the remote islands of Polynesia (Cox 2006a; Cox and Lahr 2006). Consequently, M208 must have arisen before (Kayser et al. 2003) or during this population expansion. Using variation in 12 Y-chromosome STRs in M208-derived individuals, we infer an origin for the M208 mutation between 12,300 and 3,900 years b.p. (95% confidence interval; mean 8,100 years b.p.), consistent with earlier Bayesian estimates (95% confidence interval 19,700 to 2,800 years b.p.; mean 6,900 years b.p.; Kayser et al. 2003)."

Kayser et al. (2003):
"Y STR haplotype diversity associated with C-M208 was 0.696 in the Dani and Lani (n = 23), 0.821 in the Trobriands/coastal PNG (n = 8), and 0.824 in the pooled Melanesians (n = 31), whereas it was 0.842 in the Polynesians from the Cook Islands (n = 23). No haplotype sharing was observed between WNG, PNG, and Polynesia, and differences between the three groups, based on RST values, were highly significant (P < .0001). The Y STR–based coalescence time of haplogroup C-M208 chromosomes was estimated to be ~6,900 years, with evidence for a population expansion starting ~2,200 years ago (table 5). When treated separately, the WNG haplotypes had a time back to the most recent common ancestor (TMRCA) of ~2,400 years, with expansion starting ~900 years ago; the Polynesian haplotypes had a TMRCA of ~3,500 years, with expansion starting ~1,600 years ago; and all Melanesian haplotypes together had a TMRCA of ~4,800 years, with expansion starting ~1,500 years ago."

Kayser et al. (2008):
Melanesia (includes samples from New Guinea and several surrounding Melanesian islands)
C2a-M208: n=59, Mean=4,401 years, Median=4,318 years, 95% CI=2,778-6,583 years, ESS=437

Admiralty Islands (includes only samples from the Admiralty Islands, located off the northeastern coast of New Guinea)
C2a-M208: n=22, Mean=4,543 years, Median=4,197 years, 95% CI=1,973-7,746 years, ESS=733

Mona et al. (2007)
C2a-M208
NW New Guinea: sample size insufficient (n=4, with 3 distinct haplotypes)

SW New Guinea: n=23, 5 distinct haplotypes, Mean estimate of haplogroup coalescence time (generation time of 25 years) = 1,800 years, 95% HPD low=725 years, 95% HPD up=3,225 years

Papua New Guinea: n=7, 4 distinct haplotypes, Mean=3,513 years, 95% HPD low=1,465 years, 95% HPD up=5,909 years

Scheinfeldt et al. (2006):
n=14, ASD calculated using the program Ytime (Behar et al. 2003)=1.276, TMRCA=46,200 [39,000- 57,000] years. Time of the most recent ancestor (TMRCA) estimates were calculated using ASD with the mutation rate of (2.8 ± 0.5) ×
10ˆ5 per locus per year (Zhivotosky et al. 2004). Includes C2a-M208 samples from the Papua New Guinea coast, Manus (the largest of the Admiralty Islands), New Britain, Mussau, and New Ireland.

terryt said...

"Expanding Austronesian-speaking populations probably assimilated M208 individuals (Green 1991, 2003) as they moved eastward along New Guinea’s northern coast about 3,500 years b.p."

That was basically the conclusion of the 2001 paper linked below. You may be interested in it. I corresponded for some time with one of the authors (Geoff Chambers) concerning Polynesian origins, so there is actually very little new for me here, except to note that conclusions drawn in 2001 have been confirmed with further research.

http://hpgl.stanford.edu/publications/HM_2001_v17_p271.pdf

Maju said...

"What do you suggest we call the clade that excludes T and L?"

MNOPS. I can accept KMNOPS but not K, which is an upstream node in fact.

"But there's rather a lot of what both of us would agree is F-derived Y-hap K, which seems to get all the way to, and across, Wallacea before it moves back again".

Which would be satisfactorily simplified if all or most of it belongs to MNOPS, as I am exploring in this note and you have said that agreed with.

"But there's no D at all in India, and very little C, so that idea doesn't really hold up".

Why not? M exploded just at the beginning of the Great Eurasian Expansion, I understand, expanding with CF and D (and pre-D or DE*) by means of a fast migration along coasts and rivers. It's a known fact that Y-DNA is less diverse than mtDNA and that it can suffer large displacements much more easly than mtDNA too, right?

The equivalent of the M explosion is the much more modest F node (just 7 basal lineages, plus F*) but pre-C and pre-D may have been around at the time, leading the handful of M* into SE Asia, where all would diversify.

Alternatively put the "Toba bottleneck" somewhere but I don't really see that clear. Just assign the would-be sublineages M* with pre-C and pre-D. There are some 15 basal M lineages in Eastern Eurasia, plus N (8 basal lineages in the region?), which is like 23 early Eurasian mtDNA lineages total. In the Y-DNA side (C and D only) that is like at least 7, which is about the same you get in South/West Eurasia (7 F and 1 C early lineages) for something like 32 early mtDNA lineages (mostly M but 4 N too).

Similar disproportions can be seen in other places and times and are just normal gender bias in drift. Essentially it seems to say that a female lineage has like three times greater chance of survival than a male one in the long run. This bias also makes mtDNA more a reliable reference for overall ancestry and for reconstruction of demographic history.

"It's more likely to be the other way round. Y-hap P picked up their wives from diverse groups. That certainly seems to be the case with Y-hap Q in America".

Then we are in agreement in this too. :)

Ebizur said...

C2*-M38(xM208, P33)
Cox et al. 2007
33,000 [95% CI 28,800-37,200] years. Data set includes C2*-M38 Y-DNA from Sundaland (1 Mentawai), Wallacea (202 Sumba, 81 Flores, 10 Maluku Islands, 6 West Province of Sulawesi, 4 Timor), New Guinea (1 Coastal New Guinea, 1 Highland New Guinea), Micronesia (1 Micronesia), and Island Melanesia (9 Maewo Island of Vanuatu).

Mona et al. (2007)
Northwest New Guinea
C-M38: 12,375 [6,050-21,325] years
n=56, number of distinct haplotypes=34

Southwest New Guinea
C-M38: 8,050 [1,825-16,900] years
n=8, number of distinct haplotypes=5

Papua New Guinea
C-M38: "Not estimated for sample size equal to or less than 5."
n=3, number of distinct haplotypes=3

Scheinfeldt et al. (2006)
C2-M38
n=31 (includes 17 C2*-M38(xM208): 2 PNG North Coast, 2 Nakanai/West New Britain, 4 Tolai/East New Britain, 2 Nalik/New Ireland, 2 Notsi/New Ireland, 2 Patpatar/New Ireland, 2 Kuot/New Ireland, 1 Teop/North Bougainville & 14 C2a-M208: 7 PNG North Coast, 1 PNG Rigo, 1 Manus, 1 Aneˆm/West New Britain, 2 Tolai/East New Britain, 1 Mussau, 1 Patpatar/New Ireland)
ASD=1.369
TMRCA(à la Zhivotovsky)=49,600 [42,000-61,000] years

Compare with the TMRCA estimate that Scheinfeldt et al. have obtained from their C2a-M208 samples separately from their C2*-M38 samples:
C2a-M208
n=14 (ethnic origins as detailed above)
ASD=1.276
TMRCA(à la Zhivotovsky)=46,200 [39,000-57,000] years

It seems that the inclusion of C2a-M208 samples in the calculation of a TMRCA for C2-M38 as a whole results in a figure that is much closer to the TMRCA for C2a-M208, whereas estimates of the TMRCA for C2*-M38(xC2a-M208) always have been much greater than estimates of the TMRCA for the C2a-M208 subclade. I am not quite sure how to explain this in a logical manner. Perhaps C2a-M208 might exhibit an unusually high variance in the populations of eastern New Guinea and western Island Melanesia studied by Scheinfeldt et al., thus increasing its TMRCA to a level that approaches the TMRCA of C2-M38 as a whole.

Ebizur said...

I should add that I am unsure whether the Mentawai Islands, located off the western coast of Sumatra in westernmost Indonesia, should be counted as part of "Sundaland." I am unfamiliar with the geological history of these small islands, and they may have remained isolated from Sumatra even at times of low sea level.

Also, I have mentioned before that Papua New Guinea marks the beginning of a zone of infrequent occurrence of C2*-M38(xC2a-M208) that extends eastward throughout Island Melanesia and Polynesia, with the apparent exception of at least the island of Maewo in Vanuatu. Most extant C2*-M38(xC2a-M208) seems to be confined to eastern Indonesia, including the islands of Wallacea and the western half of New Guinea. However, a comparison of the data sets of Mona et al. (2007) and Scheinfeldt et al. (2006) reveals that this is not because C2a-M208 "replaces" C2*-M38 throughout this broad zone; instead, it appears that both C2*-M38 and C2a-M208 are quite rare in Island Melanesia (especially the Solomon Islands, including Bougainville), but on the rare occasions when haplogroup C does occur in Island Melanesia, it is about equally likely (and in Vanuatu, perhaps even more likely) to belong to C2*-M38 as it is to belong to C2a-M208. C2a-M208 only starts to become notably frequent in the area around Fiji (14/105 = 13.3% C2a-M208 according to Kayser et al. 2008) and Tonga (9/29 = 31.0% C2a-M208 Kayser et al. 2008, or 4/12 = 33.3% C2a1-P33 Cox et al. 2007), which are, unsurprisingly, considered to be the "staging areas" for the initial settlement of Polynesia.

terryt said...

"Why not? M exploded just at the beginning of the Great Eurasian Expansion, I understand, expanding with CF and D (and pre-D or DE*) by means of a fast migration along coasts and rivers".

On what grounds do you include C and D with that M explosion? This diagram of Y-hap C is interesting:

http://4.bp.blogspot.com/_Ish7688voT0/SwWxkQGTZYI/AAAAAAAACCw/o3ls7yD1sig/s1600/maps-chiaroni.png

Note how extensively Y-hap C is distributed through India. It (presumably Y-hap C5) looks to have entered from the northwest, and stopped. Hardly convincing evidence for a 'fast migration along coasts and rivers'. And we still have the problem of D.

"It's a known fact that Y-DNA is less diverse than mtDNA and that it can suffer large displacements much more easly than mtDNA too, right?"

True, especially at a local level. However replacement at a regional level would be much slower. And that replacement is most likely to be simply by another regional version of the haplogroup being replaced. I understand that Australain Y-hap C4 dates back nowhere near as much as 50K, but C4 is much more likely to have replaced another version of Australian Y-hap C* anyway, rather than one from any other Y-hap clade.

"Essentially it seems to say that a female lineage has like three times greater chance of survival than a male one in the long run".

I'm glad you accept that. Do you still claim that the surviving modern mtDNA and Y-hap lines have to have left Africa together?

"C2-M38 n=31 ... TMRCA(à la Zhivotovsky)=49,600 [42,000-61,000] years"

That sounds pretty likely.

"estimates of the TMRCA for C2*-M38(xC2a-M208) always have been much greater than estimates of the TMRCA for the C2a-M208 subclade".

Makes sense that it would be greater, but the discrepancy in the differences is hard to explain. Perhaps we will understand the situation better once we can be more certain of where each haplogroup actually originated, and follow it from there.

"I am unfamiliar with the geological history of these small islands, and they may have remained isolated from Sumatra even at times of low sea level".

I'm fairly sure that is the case. They would not have been settled until the technology capable of reaching them had been developed.

"Most extant C2*-M38(xC2a-M208) seems to be confined to eastern Indonesia, including the islands of Wallacea and the western half of New Guinea".

Makes sense. After all the Australian C4 is related, and it has to have crossed Wallacea before entering Australia. So C2 presumably remained behind and developed in Wallacea from C*.

"instead, it appears that both C2*-M38 and C2a-M208 are quite rare in Island Melanesia (especially the Solomon Islands, including Bougainville)"

Those islands were already occupied by the time of the Austronesian expansion. So would tend to be bypassed for uninhabited islands beyond. Even in New Guinea offshore uninhabited islands would have been the first occupied. The C2 thread would be very thin along the New Guinea coast and through the Northern Solomons. That would account for the patchy distribution of C until it becomes established beyond Fiji. And some C2a may be the product of back movement from further east. It's known that the 'Polynesian outliers' are the product of such a movement back west, presumably from Samoa.

Maju said...

"It seems that the inclusion of C2a-M208 samples in the calculation of a TMRCA for C2-M38 as a whole results in a figure that is much closer to the TMRCA for C2a-M208, whereas estimates of the TMRCA for C2*-M38(xC2a-M208) always have been much greater than estimates of the TMRCA for the C2a-M208 subclade. I am not quite sure how to explain this in a logical manner".

It's simple, I think: MC age estimation methods always average distances, what obviously introduces distortions whenever you are comparing more than just two haplotypes.

Also STR haplotypes can perfectly cross the haplogroup barrier, depending what STRs you consider... but sometimes for very large STR blocks (as happens with the R1b modal, that is in nearly all R1b1b2a sub-levels).

The logical thing would be to adopt always the longest possible of all sample-to-sample results as the best possible estimate and not an average, which will always be distorted by overgrown subclades of all sorts. Eve then it'd be only a minimal estimate and not necessarily realistic (because you can always miss some even longer fork).

"... instead, it appears that both C2*-M38 and C2a-M208 are quite rare in Island Melanesia (especially the Solomon Islands, including Bougainville), but on the rare occasions when haplogroup C does occur in Island Melanesia, it is about equally likely (and in Vanuatu, perhaps even more likely) to belong to C2*-M38 as it is to belong to C2a-M208. C2a-M208 only starts to become notably frequent in the area around Fiji (...) and Tonga (...), which are, unsurprisingly, considered to be the "staging areas" for the initial settlement of Polynesia".

Makes total sense: initial C2/C2a colonization of Melanesia from Wallacea and secondary C2a (C2a1?) founder effect in core Polynesia.

It's very likely that the Polynesian founder effect distorts the MC perception of C2a as a whole, making it appear younger than it really is. Possibly exploring the ages by known clade and geographical region should give some refined answers.

Maju said...

"Note how extensively Y-hap C is distributed through India. It (presumably Y-hap C5) looks to have entered from the northwest, and stopped".

Well, a higher concentration in the West does not mean that C does not exist elsewhere.

From Trivedi:

West: 3.7%
South: 1.9%
East: 1.1%
North: 0%
Northeast: 0%

By ethnicity is also most frequent among Dravidians (2%) than among IEs (1.2%) and Austroasiatics (1.4%). It's also most important among tribals (2.2%), while among castes it's always under 1%.

So I'd say that Indian C (mostly C5) looks small but very old and has a coastal distribution through the Hindustan peninsula.

However it's interesting that you mention that high concentration towards the West because, if C5 back-migrated from SE Asia and did so bringing some mtDNA N (leading to R, N1'5 and N2), then it just makes total sense, though C5 had already been replaced by IJK, also in the NW, at the R explosion.

Curious.

"However [Y-DNA] replacement at a regional level would be much slower".

Agreed for most cases. But, if the fixation process happened not after expansion but before it... then it works well. It can still mean fixation for Y-DNA while various clades exist at mtDNA level.

"Do you still claim that the surviving modern mtDNA and Y-hap lines have to have left Africa together?"

Essentially yes: I don't think you can speak of several migration levels and I don't think you can have guys or girls going around the world through generations without partners of the opposite gender.

You just suggested that lineage replacement, even at the Y-DNA side can't really happen through large regions, so fixations must have happened at the very beginning.

I'd posit yet another law: that a migrating Y-DNA lineage always brings at least some residual mtDNA from its region of origin even if it also takes local partners.

If C5 backmigrated from SE Asia, then it surely brought early N with it.

Maju said...

I mean C leading to C5 (call it pre-C5), C5 as such surely only coalesced once in South Asia.

This basically means that the expansion of Y-DNA C and mtDNA N should have been simultaneous.

Ebizur said...

Kayser et al. (2003)
C2*-M38(xC2a-M208) TMRCA
10,600 [4,500–30,300] years

Note that the only haplogroups that have received older TMRCA estimates in this analysis are O1a-M119 (12,400 [5,000–34,000] years) and O3-M122 (11,100 [5,100–28,300] years).

"Haplogroup C-M38 Y chromosomes, which also carry the DYS390.3 deletion on the RPS4Y711T background (fig. 2) but lack the M208T mutation, occurred in 9% of the WNG lowland/coastal samples. This haplogroup was restricted to the Muyu, Mappi, Citak, and Asmat and was not found in the WNG highlands (tables 3 and 4; figs. 1 and 3). ... The TMRCA of all haplogroup C-M38 chromosomes was estimated to be ∼10,600 years, with a signal of population expansion beginning ∼4,700 years ago (table 5), although this was the smallest signal of population growth (∼0.005/generation) detected for any Y-chromosome mutation analyzed with this model in this or previous studies (Kayser et al. 2000a, 2001a, 2001b)." Sample set includes C2*-M38 Y-DNA from southern Borneo, the Moluccas, Nusa Tenggara, the Tolai people of New Britain, the Papua New Guinea coast and highlands, and the Muyu, Mappi, Asmat, and Citak peoples of the Western New Guinea lowlands/coast.

Is there any linguistic or other anthropological evidence to suggest that the Muyu (1/8 = 12.5% C2*-M38, 7/8 = 87.5% M1-M4), Mappi (1/10 = 10% C2*-M38, 1/10 = 10% S-M230, 7/10 = 70% M1-M4, 1/10 = 10% K-M9(xM1-M4, O1a-M119, O3-M122, P-M74, S-M230)), Citak (2/28 = 7.1% C2*-M38, 1/28 = 3.6% S-M230, 24/28 = 85.7% M1-M4, 1/28 = 3.6% K-M9(xM1-M4, O1a-M119, O3-M122, P-M74, S-M230)), and Asmat (4/20 = 20% C2*-M38, 15/20 = 75% M1-M4, 1/20 = 5% K-M9(xM1-M4, O1a-M119, O3-M122, P-M74, S-M230)) may have recent genetic links with populations of Wallacea? If we assume that haplogroup C2-M38 was not present in Sahul when it was a single continent because of the lack of this haplogroup in modern Australian aborigines, then we must also assume that some people who have immigrated into at least the coastal regions of New Guinea after the rising sea had separated that island from Australia have introduced C2-M38 to New Guinea, and the most likely source of such an immigrant population would seem to be somewhere in Wallacea, where haplogroup C2*-M38 is presently one of the most commonly occurring Y-DNA haplogroups.

Kayser et al. (2008)
C2*-M38(xC2a-M208) TMRCA for "Melanesia" sample set:
n=73, Mean=10,840, Median=10,180, 95% low=5,061, 95% up=17,320, ESS=336. "Melanesia" sample set includes the "Tolai New Britain," "PNG coast," "PNG highlands," and "WNG lowlands/coast" samples of Kayser et al. (2003), a couple C2*-M38 from the Admiralty Islands, and 56 C2*-M38 from northwestern New Guinea (cf. Mona et al. 2007).

A comparison of the C2*-M38 TMRCA estimates from Kayser et al. (2003) and Kayser et al. (2008) suggests that including C2*-M38 Y-DNA samples from Wallacea (as in the case of Kayser et al. (2003)) increases the upper bound of the 95% CI but does not increase the lower bound or the median relative to estimates of the TMRCA of C2*-M38 derived only from New Guinea, the Admiralty Islands, and New Britain (as in Kayser et al. (2008)).

Maju said...

"If we assume that haplogroup C2-M38 was not present in Sahul when it was a single continent because of the lack of this haplogroup in modern Australian aborigines"...

I don't think this is the case: I rather think that Sahul had various populations, tending to fixation and also that mountainous, wet and forested New Guinea was a different econiche than flat and arid Australia. Even animals are extremely different in both regions.

Plus there were probably two different migration routes: one by Malukku to New Guinea and another by Timor to Australia. These two migration routes surely diverged as far west as Sundaland (Java/Borneo) because there are in fact two "Wallaceas": the Sunda islands and the Sulawesi-Malukku group.

"Is there any linguistic or other anthropological evidence to suggest that the Muyu (...), Mappi (...), Citak (...), and Asmat (...) may have recent genetic links with populations of Wallacea?

Papuan languages are spoken in some islands of Wallacea, however this is a catch-all term. But Timorean languages are said to belong to a branch of the Trans-New Guinea phylum, which is the main phylum among Papuans. Also the West Papuan phylum joins the Birds' Head Peninsula (westernmost New Guinea) with Hahalmera linguistically.

However all these phyla are highly hypothetical.

From memory, I recall that some Sahullian mtDNA (Q?) was found also in some Wallacean islands. So the link surely exists but, IMO, looks like clearly pre-Austronesian, whatever the MC conjectures say.

Ebizur said...

North Bougainville (Scheinfeldt et al. 2006)
Saposa (Austronesian > Oceanic)
4/26 = 15.4% F-M89(xG-M201, H-M69, I-M170, K-M9)
14/26 = 53.8% K-M9(xK3-P79, M1-M4, M3-P117, O-M175, P-M74, S-M230)
2/26 = 7.7% K3-P79
1/26 = 3.8% M1b*-P87
2/26 = 7.7% M1b1-P22
3/26 = 11.5% O1a-M119

Teop (Austronesian > Oceanic)
1/18 = 5.6% C2*-M38(xC2a-M208)
6/18 = 33.3% K-M9(xK3-P79, M1-M4, M3-P117, O-M175, P-M74, S-M230)
1/18 = 5.6% K3-P79
2/18 = 11.1% M3-P117
4/18 = 22.2% M1b1-P22
1/18 = 5.6% O-M175(xO1a-M119, O3-M122)
3/18 = 16.7% O1a-M119

Buka Island (a small island located just off the northern tip of Bougainville)

Buka (Austronesian > Oceanic)
1/10 = 10% S-M230
1/10 = 10% M3-P117
4/10 = 40% M1b1-P22
1/10 = 10% O-M175(xO1a-M119, O3-M122)
2/10 = 20% O1a-M119
1/10 = 10% O3-M122

North Bougainville & Buka total (Austronesian > Oceanic)
1/54 = 1.9% C2*-M38(xC2a-M208) [only in Teop]
4/54 = 7.4% F-M89(xG-M201, H-M69, I-M170, K-M9) [only in Saposa]
20/54 = 37.0% K-M9(xK3-P79, M1-M4, M3-P117, O-M175, P-M74, S-M230) [only in Saposa and Teop]
3/54 = 5.6% K3-P79 [only in Saposa and Teop]
1/54 = 1.9% M1b*-P87 [only in Saposa]
10/54 = 18.5% M1b1-P22
3/54 = 5.6% M3-P117 [only in Teop and Buka]
2/54 = 3.7% O-M175(xO1a-M119, O3-M122) [only in Teop and Buka]
8/54 = 14.8% O1a-M119
1/54 = 1.9% O3-M122 [only in Buka]
1/54 = 1.9% S-M230 [only in Buka]

Central Bougainville
Aita (Non-Austronesian)
15/18 = 83.3% M1b1-P22
3/18 = 16.7% K-M9(xK3-P79, M1-M4, M3-P117, O-M175, P-M74, S-M230)

South Bougainville
2/3 M1b1-P22
1/3 K-M9(xK3-P79, M1-M4, M3-P117, O-M175, P-M74, S-M230)

Western Province, Solomon Islands (Cox et al. 2006; linguistically approx. 83% Austronesian and approx. 17% non-Austronesian/Central Solomons Papuan as of 1999 census)
11/20 = 55.0% K-M9(xM1-M4/M106, O-M175, P-M74/92R7, S-M230)
9/20 = 45.0% O-M175

Malaita Province, Solomon Islands (Cox et al. 2006; linguistically approx. 100% Austronesian)
8/12 = 66.7% K-M9(xM1-M4/M106, O-M175, P-M74/92R7, S-M230)
1/12 = 8.3% S-M230
3/12 = 25.0% M1-M4/M106
(The authors have noted: "O-M175 was not identified in the sample from Malaita Province, where only Austronesian languages are currently spoken. Conversely, nearly half the men sampled from the Western Province are O-M175 carriers (45%), despite speaking a Papuan language. It is unclear whether this effect is an artefact of small sample size or an accurate representation of Y-chromosome diversity in Solomon Islands, but the finding again emphasizes that community studies can discern levels of variability that may be missed at higher scales of analysis.")

Although the available data are woefully insufficient, it appears that the most common haplogroup in the Solomon Islands archipelago is K-M9(xM1-M4/M106, O-M175, P-M74/92R7, S-M230). Judging from the Bougainville data, this K-M9 type that is common in the Solomon Islands is probably not K3-P79 or M3-P117. It could be M2-M353, but this haplogroup has been reported so far only in Fiji (4/105 = 3.8%) and East Futuna (1/50 = 2%). It is also interesting that populations of the Solomon Islands archipelago possess significant amounts of haplogroup O-M175 despite their almost total lack of any sort of haplogroup C-M130 (only 1/107 = 0.9% C2*-M38(xC2a-M208) so far, and that one case of C2*-M38 in the Solomon Islands archipelago has been found in an Austronesian-speaking Teop individual from the northern tip of Bougainville Island at the northern end of the archipelago), and the comments of Cox et al. (2006) suggest that many of these O-M175-bearing individuals in the Solomon Islands are not speakers of an Austronesian language, but rather speakers of either Bilua or Touo, the two Papuan/non-Austronesian languages of Western Province of the independent state of Solomon Islands.

Ebizur said...

Maju said,

"Plus there were probably two different migration routes: one by Malukku to New Guinea and another by Timor to Australia. These two migration routes surely diverged as far west as Sundaland (Java/Borneo) because there are in fact two "Wallaceas": the Sunda islands and the Sulawesi-Malukku group."

On what basis do you claim that Australia has been settled via Timor (and presumably the rest of the Lesser Sunda/Nusa Tenggara Islands), while New Guinea has been settled via the Maluku Islands?

What do you mean by "two Wallaceas"? Which of the Sulawesi-Maluku Islands (let's call them "North Wallacean islands") were joined together during the era when Sahul is thought to have been settled by anatomically modern humans? Which of the Lesser Sunda Islands ("South Wallacean islands") were joined together during that era? (Note that the "Sunda Islands" that you have mentioned are not all part of Wallacea; the largest of the Sunda Islands, which are known as the "Greater Sunda Islands," are considered to have been part of a hypothetical "Sundaland," a southeastern extension of the Asian continent, at times of low sea level, and it is these Greater Sunda Islands that are the namesake of "Sundaland." Only the Lesser Sunda Islands (minus Bali) are part of Wallacea, having remained separate from both Asia/Sundaland and Sahul even during the ice ages.)

According to the Wikipedia article on the Lesser Sunda Islands, "Only Bali was part of the Ice Age continent of Sundaland, separated by the Lombok Strait from a 400-mile-long island including present-day Lombok, Sumbawa, Komodo, Flores, Solor, Adonara, and Lembata. Pantar and Alor formed a small island just to the east, while Sumba and Timor remained separate islands to the south." So, it looks like there were at least four separate South Wallacean islands: a major island that comprised modern Lombok, Sumbawa, Komodo, Flores, Solor, Adonara, and Lembata, plus three smaller islands (Pantar-Alor, Sumba, and Timor).

terryt said...

Fascinating information Ebizur. I'm still digesting it.

"Is there any linguistic or other anthropological evidence to suggest that ... may have recent genetic links with populations of Wallacea?"

I think the only real possibility is the Hoabinhian. It's mainly mainland SE Asia but elements do seem to reach New Guinea, so presumably arrived there via Wallacea.

"we must also assume that some people who have immigrated into at least the coastal regions of New Guinea after the rising sea had separated that island from Australia have introduced C2-M38 to New Guinea"

Almost certainly so. That's why C2 in New Guinea is relatively young.

"O-M175 was not identified in the sample from Malaita Province, where only Austronesian languages are currently spoken. Conversely, nearly half the men sampled from the Western Province are O-M175 carriers (45%), despite speaking a Papuan language. It is unclear whether this effect is an artefact of small sample size or an accurate representation of Y-chromosome diversity in Solomon Islands"

I'm reasonably sure that the reason for the many discrepancies is that the Austronesian-speaking people have mixed with the pre-existing populations through the islands. Small population sizes on each island have led to confusing combinations. I'd be fairly certain that C and O are markers of Austronesian presence and the Ks and Ms of pre-Austronesian populations, originally from New Guinea and Northern Melanesia.

"The logical thing would be to adopt always the longest possible of all sample-to-sample results as the best possible estimate and not an average"

True.

"It's very likely that the Polynesian founder effect distorts the MC perception of C2a as a whole, making it appear younger than it really is".

It could be just two or three thousand years old. Derived from C2* beyond Fiji. Its presence west of Fiji would be explained by backmigration to the Polynesian outliers.

"So I'd say that Indian C (mostly C5) looks small but very old and has a coastal distribution through the Hindustan peninsula".

Quite possibly more than 50K, explaining its distribution through a number of cultural groups.

"Curious".

It doesn't make sense that 'C5 back-migrated from SE Asia'. Its distribution suggests it came in from the northwest rather than from SE Asia. As far as I know its not found anywhere near so far east. So it couldn't have brought 'mtDNA N (leading to R, N1'5 and N2)' from there.

"This basically means that the expansion of Y-DNA C and mtDNA N should have been simultaneous".

I agree with that, although it's got nothing to do with India.

"wet and forested New Guinea was a different econiche than flat and arid Australia".

I agree. Hopefully I'll get a chance to explain later.

Maju said...

"On what basis do you claim that Australia has been settled via Timor (and presumably the rest of the Lesser Sunda/Nusa Tenggara Islands), while New Guinea has been settled via the Maluku Islands?"

I don't "claim" it. I know it's a valid hypothesis and I suspect it's likely, considering the differences in genetics we see in both regions (Australia and Melanesia).

"What do you mean by "two Wallaceas"? Which of the Sulawesi-Maluku Islands (let's call them "North Wallacean islands") were joined together during the era when Sahul is thought to have been settled by anatomically modern humans?"

Afaik they were also separate islands back then, though a few might have been united in larger islands as well. What I mean is that for migrational purposes, the Lesser Sunda chain (ok, taking note of the naming detail) and the Sulawesi-Malukku groups are two different "island-hoping" bridges, the latter leading to New Guinea and the former either to Australia or to the now submerged Sahul shelf.

terryt said...

"On what basis do you claim that Australia has been settled via Timor"

I've got my doubts about that too. I see someone has put Y-hap C as being New Guinea Highlands at Wiki. That gives us C6 in New Guinea, C4 in Australia and C2 in Wallacea. Suggests an ancient presence in the region. As well as that we have C1 in Japan, C3 across Central Asia and C5 in Northwest India, also presumably all ancient.

As threatened here's what I think happened round Wallacea:

First off, there was no single movement across it.

Y-hap C* moved across Wallacea before any other Y-hap (c. 50K), but did not occupy New Guinea to any great extent (although it was probably connected to Australia). As Maju said, 'mountainous, wet and forested New Guinea was a different econiche than flat and arid Australia'.

Y-hap MNOPS* may have already occupied Sundaland at the time, but had not yet crossed Wallacea. Y-hap MNOPS* then adopted the necessary boating technology and crossed Wallacea. Being pre-adapted to jungle-clad hill country members of the haplogroup rapidly spread through New Guinea, eventually making it to a few islands immediately to the north, becoming Y-haps M and S. Y-haps P and NO headed back in the other direction, through Eurasia (by 40K).

Next the Y-hap Ks became active in Wallacea and, with improved boating technology, spread further. I suspect associated somehow with the Hoabinhian (say 10-20K).

Then Y-hap O moved back south into Wallacea, mixed with Y-hap C2* and, with yet more improved boating technology, out into the Pacific they went (c. 5K). Y-hap K3 and M2 followed along behind, after they too had adopted the improved boating.

Then some Y-hap C2a moved back into parts of Melanesia (2K) forming the Polynesian Outliers.

Fits the genetic, linguistic and archeological evidence. Now, we just have to apply the same sort of analysis to the rest of the world.

Maju said...

"It could be just two or three thousand years old. Derived from C2* beyond Fiji. Its presence west of Fiji would be explained by backmigration to the Polynesian outliers".

I don't think so: C2 is absent in the Austronesian homelands of Taiwan and, secondarily, Philippines. Making any association of Melanesian C2 with Austronesian peoples, other than these borrowing lineages from the Melanesian natives, makes no sense to me.

"It doesn't make sense that 'C5 back-migrated from SE Asia'. Its distribution suggests it came in from the northwest rather than from SE Asia".

Frequency or diversity? AFAIK no research has dealt with diversity within C5, which could give us a more clear answer. Frequency may be distorted by sample bias (for example oversampling tribals).And may also be distorted by ulterior migrations.

However if, as you claim C5 spread from West India, then we would have to consider a South Asian origin for C as a whole, right? I feel it doesn't make much sense, really.

What I'd suggest is that, whatever the exact spot of radiation within South Asia, C leading to C5 (pre-C5) back-migrated from SE Asia. Considering its distribution, I'd say they did not follow the Ganges, so the coastal route or coastal+Krishna river route is what the implacable Razor of Brother Occam demands.

"As far as I know its not found anywhere near so far east".

I'm telling you it's found at levels of 1.1% in East India, meaning Bengal, Bihar, Jharkhand and Orissa. It is not found in NE India (Assam and surroundings) nor in North India (the mid-upper Ganges basin), but it's found in the South and the West regions. That tells of a "coastal" distribution.

""This basically means that the expansion of Y-DNA C and mtDNA N should have been simultaneous".

I agree with that, although it's got nothing to do with India".

So how did 1/4 of mtDNA N sublineages got into South Asia then? O_O

terryt said...

I'd also like to mention that this 'Wallacean genetic expansion' Maju keeps complaining about is no random idea I've simply made up. Get yourself along to your local library or your nearest university library and look at Cavalli-Sforza's 1994 book 'History and Geography of Human Genes'. And then look at the map of the second principal component for Asia and the sixth principal component for the world.

Sorry. The maps don't seem to be on the net, but a review of the book:

http://www3.interscience.wiley.com/journal/110504485/abstract?CRETRY=1&SRETRY=0

terryt said...

"I know it's a valid hypothesis and I suspect it's likely, considering the differences in genetics we see in both regions (Australia and Melanesia)".

Two separate Wallacean crossings is more likely.

"C2 is absent in the Austronesian homelands of Taiwan and, secondarily, Philippines".

Maju. That's the interesting thing. It's not from the Austronesian homeland. It joined up along the way.

"Making any association of Melanesian C2 with Austronesian peoples, other than these borrowing lineages from the Melanesian natives, makes no sense to me".

Maybe it makes no sense to you, but that's what happened.

"then we would have to consider a South Asian origin for C as a whole, right?"

Not necessarily. Iranian Plateau, like most other haplogroups (including probably CDEF*), would suffice.

"That tells of a 'coastal' distribution".

No it doesn't. It tells us it moved across India in samll numbers as far as Bengal, Bihar, Jharkhand and Orissa.

"So how did 1/4 of mtDNA N sublineages got into South Asia then?"

It would have been just as easy to move into India as out of it. And what about the other 3/4 not found in India?

"Considering its distribution, I'd say they did not follow the Ganges, so the coastal route or coastal+Krishna river route is what the implacable Razor of Brother Occam demands".

I'll leave you to invent any hypothesis you wish.

Maju said...

"It joined up along the way".

There isn't any "way" between Philippines and Melanesia, except the small archipelago of the Spice Islands (Malukku). You give too much importance to those islands and I can't just take it for granted just because you say so: they are paths, not homelands... unless you can prove the opposite somehow (not with mere speculative reasoning again, please, something like diversity data).

"Iranian Plateau, like most other haplogroups (including probably CDEF*), would suffice".

Less speculation and more evidence, please. The Iranian plateau (a non-entity in Paleolithic terms, unlike the regions west or, specially, east of it) looks totally meaningless, is, like all semiarid plateaus, rather hostile to human settlement (compare with Anatolian or Iberian plateaus) and has no hard data like archaeological sites or genetic diversity to support your claims.

Plus, all C lineages are clearly east of it.

"It would have been just as easy to move into India as out of it. And what about the other 3/4 not found in India?"

What are you arguing against this time without comprehension? I am saying that some apportion of N, leading to those lineages, back-migrated TO South Asia, probably with pre-C5.

"I'll leave you to invent any hypothesis you wish".

So I appeal to Brother Occam and you avoid reasoning with this evasive maneuver. I'm tired of your fetishist non-reasoning, really.

Ebizur said...

In South Asia, haplogroup C Y-DNA has been found most frequently in the north, particularly among Hazaras (C3*-M217), Burushos (C3*-M217), Nepalis (C3, C5, and C(xC3, C5)), and Muslims and upper caste Hindus of Uttar Pradesh (subclade(s) not specified). Haplogroup C Y-DNA of any sort is very rare in the south of the Subcontinent, where F(xK) clades (especially H1a-M82 and J2-M172) and K clades (especially R1a1a-M17, L1-M76, and R2-M124) predominate. According to my weighting of the data of Watkins et al. (2008), haplogroup C is only about as common as haplogroup Q in southeastern India (Andhra Pradesh & Tamil Nadu), where each accounts for approximately 2% of all Y-DNA in the region.

Zhao et al. (2008)
Uttar Pradesh Muslim and high-caste Hindu total (96 Bhargavas + 88 Chaturvedis + 118 other Brahmins + 154 Shia + 104 Sunni):
48/560 = 8.6% C-RPS4Y/M216 (highest in non-Bhargava, non-Chaturvedi Brahmins: 12/118 = 10.2% C-RPS4Y/M216, lowest in Bhargavas: 6/96 = 6.3% C-RPS4Y/M216)
17/560 = 3.0% E1b1b1-M35
17/560 = 3.0% CF-M168(xC, DE, G, H(1?), I, J(2?), K)
23/560 = 4.1% G-M201
59/560 = 10.5% H1-M52
77/560 = 13.8% J2-M172
17/560 = 3.0% K-M9(xL-M11/M20, M1-M4, NO-M214, P-92R7/M45/M74, T-M70) (I wonder what these might be. Could K1-M147 really be so common in Uttar Pradesh?)
15/560 = 2.7% L-M11/M20
15/560 = 2.7% O-M175
16/560 = 2.9% P-92R7/M45/M74(xQ1a3a-M3, R1-M173, R2-M124)
130/560 = 23.2% R1a1a-M17
3/560 = 0.54% R1b-M343
123/560 = 22.0% R2-M124

Gayden et al. (2007)
Nepal
Tamang (Tibeto-Burman people of the Himalayan foothills around and especially north of the Kathmandu Valley):
2/45 = 4.4% F-M213(xH-M69, J-M304, K-M9)
1/45 = 2.2% O3a3c-M134(xO3a3c1-M117)
38/45 = 84.4% O3a3c1-M117
2/45 = 4.4% R2-M124
2/45 = 4.4% R1a1a-M198

Newar (Tibeto-Burman-speaking people of the Kathmandu Valley):
2/66 = 3.0% C5-M356
4/66 = 6.1% H1a-M82
4/66 = 6.1% J2a-M410
1/66 = 1.5% J2b2-M241
14/66 = 21.2% O3a3c1-M117
17/66 = 25.8% R2-M124
17/66 = 25.8% R1a1a-M198
7/66 = 10.6% R1b1b2-M269

Kathmandu (mainly Nepali-speaking Hindus):
3/77 = 3.9% C-M216(xC3-M217, C5-M356)
2/77 = 2.6% C3-M217
1/77 = 1.3% C5-M356
(C-M216 total: 6/77 = 7.8%)
1/77 = 1.3% H*-M69(xH1-M52, H2-Apt)
4/77 = 5.2% H1*-M52(xH1a-M82)
4/77 = 5.2% H1a-M82
3/77 = 3.9% J2a-M410
5/77 = 6.5% J2b2-M241
1/77 = 1.3% N-M231(xN1c-Tat)
1/77 = 1.3% N1c1a-P21
1/77 = 1.3% O3a-M324(xO3a3c-M134)
2/77 = 2.6% O3a3c-M134(xO3a3c1-M117)
13/77 = 16.9% O3a3c1-M117
1/77 = 1.3% Q-M242(xQ1a1-M120, Q1a3-M346)
8/77 = 10.4% R2-M124
27/77 = 35.1% R1a1a-M198

terryt said...

"Could K1-M147 really be so common in Uttar Pradesh?"

Easily, if it came by boat from Wallacea and then up the Ganges!

"There isn't any 'way' between Philippines and Melanesia, except the small archipelago of the Spice Islands (Malukku). You give too much importance to those islands"

So poor Ebizur did all that hard work for nothing? 'Taken together, these three TMRCA estimates for C2*-M38 suggest that it is most diverse in the Lesser Sunda Islands and the Maluku Islands of Wallacea in eastern Indonesia, because the upper bound for the TMRCA of C2*-M38 based on the 'Melanesia' pool of Kayser et al. 2008, which includes samples from mainland Western New Guinea, mainland Papua New Guinea, and several outlying island groups, is much less than the upper bound of the earlier TMRCA estimates of Kayser et al. (2003) and Cox et al. (2007) that have been derived from pools of samples that include the Lesser Sundas and the Moluccas'. So that's probably where C2-M38 originated. So what's your problem?

"Plus, all C lineages are clearly east of it".

True. But from the information Ebizur has just provided it looks extremely unlikely they actually moved east through India: 'haplogroup C is only about as common as haplogroup Q in southeastern India'. Perhaps Ebizur could provide us with information re. age and distribution of the various C3 clades?

"In South Asia, haplogroup C Y-DNA has been found most frequently in the north, particularly among Hazaras (C3*-M217), Burushos (C3*-M217)"

I'd guess it's almost certain that C3 in India is a product of migration from Central Asia, perhaps relatively recently. The C-M216(xC3-M217, C5-M356) in Kathmandu is intruiging though. Wonder what it is? Even if it's C* it certainly argues against a 'coastal route'.

"So I appeal to Brother Occam"

You very seldom actually 'appeal to Brother Occam'. You make up all sorts of excuses as to why your belief is the correct interpretation of the evidence: increased mutation in particular lines when it suits your belief, extreme drift when it suits your belief, improbable founder effects when it suits your belief, unjustified bottlenecks, and so on.

terryt said...

I'll just remind Maju of some of Ebizur's information. When I look back through it I see,as I mentioned before, that C2* probably became fixed in the Maluku archipelago 40-50 Kya.

"Scheinfeldt et al. (2006):
n=14, ASD calculated using the program Ytime (Behar et al. 2003)=1.276, TMRCA=46,200 [39,000- 57,000] years".

By 10 Kya. C2* had Managed to reach islands north of New Guinea:

"Kayser et al. (2008)
C2-M38(xC2a-M208)
Melanesia: n=73, Mean=10,840, Median=10,180, 95% CI=5,061-17,320 years

Admiralty Islands: n=2, Mean=10,800, Median=9,880, 95% CI=4,126-19,950 years"

as had some Ks (probably from somewhere on the New Guinea Mainland):

"K-M9(xK3-P79, M1a-P34, O1a2-M110, O3a-M324, S1-M254)
Melanesia: n=101, Mean=15,260, Median=14,510, 95% CI=7,965-23,380 years

Admiralty Islands: n=40, Mean=14,480, Median=13,580, 95% CI=7,338-23,880 years"

Presumably it was from these islands north of New Guinea that C2* joined the Austronesian expansion of Y-hap O and mtDNA B. In fact it's generally assumed that the Lapita ware expansion into the Pacific started from New Britain.

C2a* may have originated either there or somewhere on the route out into the pacific, or perhaps as far east as Tonga/Samoa. Some C2a moved back west into some of the smaller, more isolated Melanesian islands, explaining at least some of the westerly presence of the haplogroup.

By then K3 had also spread in turn from the Admiralty Islands through Melanesia as far as Tonga and Samoa:

"K3-P79
10/19 = 52.6% Mussau (Scheinfeldt et al. 2006)
3/6 = 50% Tokelau
28/100 = 28.0% Tuvalu
92/395 = 23.3% New Britain (Scheinfeldt et al. 2006)
4/19 = 21.1% Tolai New Britain
1/7 = 14% Manus (Scheinfeldt et al. 2006)
12/105 = 11.4% Fiji
16/152 = 10.5% New Ireland (Scheinfeldt et al. 2006)
5/50 = 10.0% East Futuna
2/37 = 5.4% New Guinea (Scheinfeldt et al. 2006)
6/147 = 4.1% Admiralty Islands
3/75 = 4.0% Bougainville (Scheinfeldt et al. 2006)
3/77 = 3.9% Cook Islands
1/29 = 3.4% Tonga
1/31 = 3.2% PNG Highlands
1/33 = 3.0% PNG Coast
1/61 = 1.6% Samoa"

This last explains how the generally accepted later more Melanesian look developed in western regions of the original Lapita ware people. And the patchy distribution of C2* and C2a between wallacea and Polynesia.

Maju said...

Very interesting as always, Ebizur. The presence of C5 in North India is interesting, as is that apportion of K*.

However sampling "Muslim and high-caste Hindu" people in Uttar Pradesh is like sampling "Aristocrats and Jews" in England. It's like it doesn't need to match at all with what the real people has, right?

"I wonder what these might be. Could K1-M147 really be so common in Uttar Pradesh?"

I have always suspected Y-DNA K had a North Indian homeland, so it's the kind of data I find unsurprising, yet informative. I can't say if it's K1 because nobody seems to know much about it but probably not only.

Notice also the 2.9% P* a few lines below (other Q?). And what about the 3% CF* a few lines above: it's all F-number clades or what? Certainly they are right wehn they say that the devil is in the details. In the overlooked details in fact.

Maju said...

"Easily, if it came by boat from Wallacea and then up the Ganges!"

Your naivety is moving... but bad for a serious discussion.

"You very seldom actually 'appeal to Brother Occam'. You make up all sorts of excuses as to why your belief is the correct interpretation of the evidence: increased mutation in particular lines when it suits your belief, extreme drift when it suits your belief, improbable founder effects when it suits your belief, unjustified bottlenecks, and so on".

You are making this accusation, full of falsehoods, only to avoid the original issue, which was about your claim of C5 existing only in West India (Gujarat-Mumbai-Goa region) and then about them migrating there from SE Asia by some mysterious overland route and back to the Bay of Bengal also by some mysterious overland route.

So you sure keep avoiding the issues by diverting the matter. Do you do that because it suits your belief? Partly yes (no canoes before Terry says so) and partly just because of what I can't describe but as dumbness, because the overall pesudo-explanation you give doesn't make sense at all: you could not even persuade yourself if you tried seriously.

But you don't care, do you?

"I'll just remind Maju of some of Ebizur's information. When I look back through it I see,as I mentioned before, that C2* probably became fixed in the Maluku archipelago 40-50 Kya".

Well, I don't really want to make too subtle distinctions between Wallacea and Sahul: it's all the same dead end of the Eurasian migratory routes.

Also, as you know, I can't believe in the molecular clock.

So you think that C2 is from the lowlands?, fine with me.

terryt said...

"I have always suspected Y-DNA K had a North Indian homeland"

And that's why haplogroups K2, K3 and K4 are so common there.

"to avoid the original issue, which was about your claim of C5 existing only in West India (Gujarat-Mumbai-Goa region) and then about them migrating there from SE Asia by some mysterious overland route and back to the Bay of Bengal also by some mysterious overland route".

On what grounds do you claim that C5 was ever in SE Asia? I've never so much as even hinted that C5 came from there. It is certainly far from common in eastern India.

"So you think that C2 is from the lowlands?"

You've got serious problems with your geography if you're calling the Malukus 'lowlands'. I suppose mostly they are close to sea level.

"you could not even persuade yourself if you tried seriously".

Unfortuantely you seem all to easily able to persuade yourself, mostly by avoiding inconvenient evidence.

"Well, I don't really want to make too subtle distinctions between Wallacea and Sahul"

Of course not. 'Do you do that because it suits your belief?'

Ebizur said...

terryt said,

"True. But from the information Ebizur has just provided it looks extremely unlikely they actually moved east through India: 'haplogroup C is only about as common as haplogroup Q in southeastern India'. Perhaps Ebizur could provide us with information re. age and distribution of the various C3 clades?"

Haplogroup C3c:
3,500 [95% CI 300–19,700] years (Katoh et al. 2004)

2,750 ± 1,370 years (Karafet et al. 2002)

Haplogroup C3c seems to be a very young clade, limited in its distribution to Yukagir, Chukotko-Kamchatkan, Nivkh, Tungusic, Mongolic, and the more strongly "Oriental"-influenced Turkic (e.g. Kazakhs) and Iranic peoples (e.g. Tajiks). Pakendorf et al. (2006) write, "Two Evenks from southern Yakutia (unpublished data), one Yakut-speaking Evenk, and one Yukaghir had the derived state at M48, but the ancestral state at M86; this was further confirmed by sequencing the relevant fragments. This indicates that the M86 mutation occurred after the M48 mutation, as shown in Fig. 5; following the YCC nomenclature (Y_Chromosome_Consortium 2002), the branch defined by M48 alone would belong to haplogroup C3c*, while the branch defined by mutations at both M48 and M86 would belong to haplogroup C3c1." Pakendorf's finding has never been replicated by another researcher, nor has her refinement of C3c into C3c*-M48 and C3c1-M86 been adopted by the ISOGG. Also, Hammer et al. (2006) have reported finding haplogroup C3c-M86 Y-DNA in three Japanese individuals (two from Kyushu and one from Tokushima in eastern Shikoku), amounting to approximately 1% of all their samples from Japan, but their finding of haplogroup C3c in Japan has never been replicated by any other team of researchers. Karafet et al. (2002) have proposed "Mongolia or the Lake Baikal region" as the most likely region of origin of haplogroup C3c, but I would prefer the Amur River basin because of the strong representation of this subclade among the so-called Palaeo-Siberians of the Russian Far East (including the Nivkhs of Sakhalin Island) and among both Northern (e.g. Negidals) and Southern Tungusic (e.g. Udegeys) peoples of the Amur region. Haplogroup C3-M217 among modern Mongols is mostly C3(xC3c), and some populations that supposedly have been influenced by Mongols historically (e.g. Hazaras) have plenty of C3(xC3c) but no C3c, so a hypothesis of an origin of haplogroup C3c in a hypothetical contact zone between early Tungusic and Palaeo-Siberian populations in far eastern Siberia several thousand years ago seems much more plausible to me.

Age estimate for the Native American-specific C3b-P39 (Zegura et al. 2004): 2,550 ± 1,910 years

This TMRCA estimate for Native American C3b-P39 seems suspiciously low in my opinion. I hope another researcher or team of researchers will attempt an estimate of the TMRCA of C3b-P39.

As far as I recall, I have never seen any TMRCA estimate for C3a-M93 or C3d-M407. In the published literature, C3a-M93 has been found only in a couple Japanese individuals, but I recall reading on DNA Forums about one year ago that one of the rare C3-M217 carriers from Europe (in this case, the individual's recorded ancestry was from somewhere in the north of the British Isles, I think either Scotland or Northern Ireland) has tested positive for M93, so it seems that it might be a fairly old subclade. C3d-M407 has been reported by Sengupta et al. (2006) in one Han Chinese and one Yakut if I remember correctly, so this subclade might also be fairly old, considering the distance between these two populations. However, one should keep in mind that the haplogroup identifications of Sengupta et al. (2006) are notoriously unreliable.

In any case, Karafet et al. (2002) have presented a TMRCA estimate of 11,900 ± 4,800 years for C3-M217 as a whole.

Ebizur said...

By the way, a review of the list of SNPs tested by Gayden et al. (2007) has allowed me to make the following refinements:

Tamang
0/45 C-M216
2/45 = 4.4% F-M213(xF2-M427/M428, G-M201, H-M69, J-M304, K-M9) <-Is this Indian F*, or rather some sort of haplogroup I-M170? If the latter, it would fit nicely with the finding of R1b1b2-M269 in the neighboring Newars.

Newar
2/66 = 3.0% C5-M356

Kathmandu:
3/77 = 3.9% C-M216(xC1-M8, C2-M38, C3-M217, C4a-M210, C5-M356)
2/77 = 2.6% C3-M217(xC3a-M93, C3c-M77, C3d-M407)
1/77 = 1.3% C5-M356
1/77 = 1.3% N-M231(xN1a-M128, N1b-P43, N1c-Tat)
1/77 = 1.3% N1c1a-P21
1/77 = 1.3% Q-M242(xQ1a1-M120, Q1a3-M346)

Tibet:
4/156 = 2.6% C3-M217(xC3a-M93, C3c-M77, C3d-M407)
7/156 = 4.5% N-M231(xN1a-M128, N1b-P43, N1c-Tat)
2/156 = 1.3% Q1a1-M120
3/156 = 1.9% Q1a3-M346(xQ1a3a-M3)

Maju said...

"And that's why haplogroups K2, K3 and K4 are so common there".

L, T, P, K1 and maybe others (K*) are. If K2, K3 and K4 are part of MNOPS, then we would not need to consider them anymore as part of the basal diversity of K.

This is an IF but all related to MNOPS at the moment is.

"On what grounds do you claim that C5 was ever in SE Asia?"

I don't. C was. C5 is part of C, isn't it?

"You've got serious problems with your geography if you're calling the Malukus 'lowlands'".

They are small islands, right?

""Well, I don't really want to make too subtle distinctions between Wallacea and Sahul"

Of course not. 'Do you do that because it suits your belief?'"

Because they are obviously tightly knit in a single region. You can dig in the details but from the viewpoint of Eurasia, they are just like any other region, and not one that seems central, really.

Ebizur said...

The Q-M242 individual from Kathmandu is actually a confirmed Q-M242(xQ1a1-M120, Q1a2-M25/M143, Q1a3-M346, Q1a6-M323, Q1b-M378). Please excuse my clumsiness tonight.

Ebizur said...

I have compared the Japanese data of Sengupta et al. (2006) and Underhill et al. (2000), and I am now unsure whether C3a-M93 has ever been detected in more than one Japanese individual. Have both these studies made use of the HGDP cell lines? The haplogroup identifications of Sengupta et al. appear to be problematic as I have mentioned previously; for example, they have reported all DE-YAP individuals in the Japanese sample as "D1-M015," whereas Underhill et al. have reported only 1/23 D1-M15 in the Japanese sample, which seems much more reasonable in light of the results of other published studies of Japanese Y-DNA. Anyway, both Underhill et al. (2000) and Sengupta et al. (2006) have reported 1/23 C3a-M93 in a sample of Japanese, which I suspect to be none other than the Japanese sample of the HGDP.

Furthermore, Sengupta et al. (2006) appear to have been the first to report the M407 mutation that marks haplogroup C3d, finding it in 1/24 Han and 2/18 Yakut, also apparently from the HGDP cell lines.

terryt said...

Thanks for all that work Ebizur.

"Haplogroup C3c seems to be a very young clade, limited in its distribution to Yukagir, Chukotko-Kamchatkan, Nivkh, Tungusic, Mongolic, and the more strongly "Oriental"-influenced Turkic (e.g. Kazakhs) and Iranic peoples (e.g. Tajiks)".

Interesting that the Mongolic are included. As you say, 'Haplogroup C3-M217 among modern Mongols is mostly C3(xC3c), and some populations that supposedly have been influenced by Mongols historically (e.g. Hazaras) have plenty of C3(xC3c) but no C3c'. The map in the link to East Asia at Dienekes even has most Mogolian Y-haps as C-M130, presumably x the whole of C3. I'll make further comments there.

"This TMRCA estimate for Native American C3b-P39 seems suspiciously low in my opinion".

Maybe. I'm fairly convinced it's reasonably recent, long after Q anyway.

"In any case, Karafet et al. (2002) have presented a TMRCA estimate of 11,900 ± 4,800 years for C3-M217 as a whole".

That could be too tecent too.

terryt said...

"Because they are obviously tightly knit in a single region".

Where we know that 'C was. C5 is part of C, isn't it?' As is C2. So we have C5 in India, C2 in Wallacea, C6 in New Guinea, C4 in Australia, C1 in Japan and C3 in Central Eurasia. So they must at some time have been part of a single related population. But did it enter SE Asia via India? I doubt it.

"and not one that seems central, really".

but we still must take them into account.

"If K2, K3 and K4 are part of MNOPS, then we would not need to consider them anymore as part of the basal diversity of K".

No, of course not. They'd be part of the basal diversity of KMNOPS. I'm sure I remember you claiming that diversity usually coincides with origin. Three quarters of the Y-hap Ks are concentrated round and beyond Wallacea.

Include M, N, OP and S and what do you get? At least five eighths still concentrated round Wallacea and beyond. We still have greater diversity there.

So we're forced to include L and T to reach 50%. But one of those 50%s is widespread. The other 50% is still concentrated round Wallacea and beyond.

Maju said...

"C5 is part of C, isn't it?' As is C2. So we have C5 in India, C2 in Wallacea, C6 in New Guinea, C4 in Australia, C1 in Japan and C3 in Central Eurasia".

C6 in New Guinea! AFAIK "C6 is a recently recognized group whose geographical associations were not reported". Do you have a source for your claim?

For the rest I agree and I get from that an expansion knot for C at Indochina. It's mere euclidean maths.

"But did it enter SE Asia via India? I doubt it".

UFO abductions are always the best explanation, aren't they?

"They'd be part of the basal diversity of KMNOPS".

We agree on this. But immediately after, in another show of your self-contradictory irrationality you say:

"Three quarters of the Y-hap Ks are concentrated round and beyond Wallacea".

If they are part of MNOPS then they are not anymore basal K. What results in only one basal K lineage (out of four) being located anywhere east of Bangladesh.

If they are MNOPS then they are not anymore basal K sublineages. Crystal clear.

"Include M, N, OP and S and what do you get? At least five eighths still concentrated round Wallacea and beyond".

All MNOPS it would seem. It makes 1/4 of all basal lineages of K.

Btw, there's no "OP" haplogroup: there are NO and P, the latter almost totally unrelated to SE Asia.

Ebizur said...

Maju said,

"C6 in New Guinea! AFAIK 'C6 is a recently recognized group whose geographical associations were not reported'. Do you have a source for your claim?"

It seems that terry has been following my recent updates to the Wikipedia entry on Y-DNA haplogroup C. There I have referenced the article by Laura Scheinfeldt, Françoise Friedlaender, Jonathan Friedlaender et al., "Unexpected NRY Chromosome Variation in Northern Island Melanesia," Molecular Biology and Evolution 23(8), p. 1633:

"Lineage C4-P55
This branch was found in one sample in our series
from the New Guinea highlands (not shown, but included
in diversity calculations)."

Karafet et al. (2008) have named the Y-DNA lineage marked by the P55 mutation "haplogroup C6," and this nomenclature has been adopted by ISOGG.

Ebizur said...

C3c (M48, M77, M86)
28/40 = 70.0% Stony Tunguska Evenks (Western Evenks from Vanavara, Strelka-Chunya, Baykit, and Surinda villages in the Evenk National District; Pakendorf et al. 2006 & 2007)
15/22 = 68.2% Oroqen (Hammer et al. 2006)
6/9 = 66.7% Iengra Evenks (Pakendorf et al. 2007; includes 2/9 C-M48(xM86) and 4/9 C-M86)
24/38 = 63.2% Kazaks/Kazakstan: Almaty, Katon-Karagay, Karatutuk, Rachmanovsky Kluchi (Zerjal et al. 2002)
10/16 = 62.5% Okhotsk Evenk (Lell et al. 2002)
19/31 = 61.3% Eastern Evens (Karafet et al. 2002; Hammer et al. 2006; Pakendorf et al. 2007)
12/20 = 60.0% Udegey/village of Gvaysugi in the middle of the Khor River basin (Lell et al. 2002; Torroni et al. 1993)
18/31 = 58.1% Yenisey Evenk (Lell et al. 2002)
31/54 = 57.4% Kazak/Kazakstan (Wells et al. 2001)
52/95 = 54.7% Siberian Evenk (Hammer et al. 2006; I think this sample should include the "Eastern Evenk"/"Nyukzha Evenk" sample of Karafet et al. 2002 and Pakendorf et al. 2007)
42/78 = 53.8% Nyukzha Evenks (Karafet et al. 2002; Pakendorf et al. 2007)
9/17 = 52.9% Negidal (Lell et al. 2002; the original paper has at least one error in the Negidal entries of its data table)
12/24 = 50.0% Central Evens (Pakendorf et al. 2007)
11/24 = 45.8% Mongolian (Wells et al. 2001)
13/31 = 41.9% Oroqen (Xue et al. 2006)
7/18 = 38.9% Itel’men (Lell et al. 2002)
20/53 = 37.7% Ulchi/Nanai (Lell et al. 2002)
37/99 = 37.4% Kalmykians/Elista, Republic of Kalmykia, Russian Federation (Nasidze et al. 2005)
6/17 = 35.3% Nivkh/Rybnovsk and Nekrasovka villages in northern Sakhalin Island (Lell et al. 2002; Torroni et al. 1993)
14/41 = 34.1% Manchurian Evenk (Hammer et al. 2006)
9/27 = 33.3% Koryak (Lell et al. 2002)
20/60 = 33.3% Uriankhai (Katoh et al. 2004)
18/60 = 30.0% Zakhchin (Katoh et al. 2004)
7/26 = 26.9% Ewenki/PRC (Xue et al. 2006)
3/13 = 23.1% Yukaghir (Pakendorf et al. 2006; includes 1/13 C-M48(xM86) and 2/13 C-M86)
13/65 = 20.0% Mongolians/Mongolia: Ulaanbaatar (Zerjal et al. 2002)
8/40 = 20.0% Tuvan (Lell et al. 2002)
27/149 = 18.1% Mongolia (Hammer et al. 2006)
13/85 = 15.3% Khalkh (Katoh et al. 2004)
5/33 = 15.2% Yakut-speaking Evenks (Pakendorf et al. 2007; includes 1/33 C-M48(xM86) and 4/33 C-M86)
5/41 = 12.2% Kyrgyz/Kyrgyzstan: central Kyrgyzstan (mixed) (Zerjal et al. 2002)
5/45 = 11.1% Hezhe/PRC (Xue et al. 2006)
4/40 = 10.0% Khoton (Katoh et al. 2004)
2/22 = 9.1% Tajiks/Tajikistan: Penjikent (Zerjal et al. 2002)
5/55 = 9.1% Tuvan (Pakendorf et al. 2006)
4/45 = 8.9% Inner Mongolian (Xue et al. 2006)
8/101 = 7.9% Manchu/northeastern China (Katoh et al. 2004)
1/13 = 7.7% Buryat (Lell et al. 2002)
4/52 = 7.7% Kyrgyz/Kyrgyzstan (Wells et al. 2001)
3/39 = 7.7% Uygur/Yili (Xue et al. 2006)
3/42 = 7.1% Tuvinian (Wells et al. 2001)
5/98 = 5.1% Altai (Hammer et al. 2006)
4/81 = 4.9% Buryat (Hammer et al. 2006)
2/41 = 4.9% Xibe (Xue et al. 2006)
2/42 = 4.8% Arab/Bukhara, Uzbekistan (Wells et al. 2001)
1/22 = 4.5% Western Evens (Pakendorf et al. 2007)
1/22 = 4.5% Tajik/Khojant, Tajikistan (Wells et al. 2001)
3/70 = 4.3% Uzbek/Khorezm, Uzbekistan (Wells et al. 2001)
1/24 = 4.2% Chukchi (Lell et al. 2002)
2/53 = 3.8% Kyushu, Japan (Hammer et al. 2006)
1/32 = 3.1% NE Yakut (Pakendorf et al. 2006)
1/33 = 3.0% Uyghurs/Kazakstan: Almaty, Lavar (Zerjal et al. 2002)
1/35 = 2.9% Manchu (Xue et al. 2006)
1/39 = 2.6% Daur (Xue et al. 2006)
2/78 = 2.6% Tajik total/Tajikistan & Uzbekistan (Wells et al. 2001)
1/40 = 2.5% Tajik/Samarkand, Uzbekistan (Wells et al. 2001)
1/44 = 2.3% Karakalpak/Uzbekistan (Wells et al. 2001)
1/45 = 2.2% Uzbek/Samarkand, Uzbekistan (Wells et al. 2001)
1/67 = 1.5% Uygur/Xinjiang (Hammer et al. 2006)
1/70 = 1.4% Tokushima, Japan (Hammer et al. 2006)
3/259 = 1.2% Japan total (Hammer et al. 2006)
4/366 = 1.1% Uzbek total/Uzbekistan (Wells et al. 2001)
1/96 = 1.0% Southern Altaians (Kharkov et al. 2007)
1/184 = 0.54% Yakuts total (Pakendorf et al. 2006)

Ebizur said...

0/4 Ainu (Hammer et al. 2006)
0/12 Iranians/Shiraz (Wells et al. 2001)
0/12 Koryaks (Karafet et al. 2002)
0/12 Lezgi/Azerbaijan: Azerbaijan/Dagestan border (Zerjal et al. 2002)
0/15 Ossetians/Georgia: southern Ossetia (Zerjal et al. 2002)
0/15 Sinte Romani/Uzbekistan (Wells et al. 2001)
0/16 Iranians/Esfahan (Wells et al. 2001)
0/16 Tajik/Dushanbe (Wells et al. 2001)
0/17 Micronesia (Hammer et al. 2006)
0/19 Azeri/Azerbaijan: Baku (Zerjal et al. 2002)
0/19 Tofalar (Lell et al. 2002)
0/19 Uzbek/Kashkadarya, Uzbekistan (Wells et al. 2001)
0/20 Kurds/Turkmenistan: Bagyr (Zerjal et al. 2002)
0/20 Zhuang/Guangxi (Hammer et al. 2006)
0/21 Armenians/Armenia: Yerevan (Zerjal et al. 2002)
0/21 Turkmens/Turkmenistan: Ashgabat (Zerjal et al. 2002)
0/22 Crimean Tatar/Uzbekistan (Wells et al. 2001)
0/22 Dungans/Kyrgyzstan: Alexandrovka, Osh (Zerjal et al. 2002)
0/23 Saami/Russia (Wells et al. 2001)
0/24 Iranians/Tehran (Wells et al. 2001)
0/25 British (Wells et al. 2001)
0/25 Indonesia, West (Hammer et al. 2006)
0/25 Ishkashimi (Pamiri)/Tajikistan (Wells et al. 2001)
0/25 Korean/PRC (Xue et al. 2006)
0/25 Svans/Georgia: Svanetia (Zerjal et al. 2002)
0/26 Aomori, Japan (Hammer et al. 2006)
0/26 Georgians/Georgia: Kazbegi (Zerjal et al. 2002)
0/26 Orkney (Wells et al. 2001)
0/28 Pomor/Russia (Wells et al. 2001)
0/28 Uzbeks/Uzbekistan: Kashkadarya region (Zerjal et al. 2002)
0/30 Bartangi (Pamiri)/Tajikistan (Wells et al. 2001)
0/30 Han/Lanzhou, Gansu (Xue et al. 2006)
0/30 Turkmens/Turkmenistan (Wells et al. 2001)
0/31 Uygur/Urumqi (Xue et al. 2006)
0/31 Yagnobi/Tajikistan (Wells et al. 2001)
0/32 Han/Yili, Xinjiang (Xue et al. 2006)
0/32 Malay (Hammer et al. 2006)
0/33 Australian Aboriginal People (Hammer et al. 2006)
0/33 Siberian Eskimo (Lell et al. 2002)
0/33 North Iran (Regueiro et al. 2006)
0/33 Qiang (Xue et al. 2006)
0/34 Han/Chengdu, Sichuan (Xue et al. 2006)
0/34 Hani (Xue et al. 2006)
0/34 Li/Hainan (Xue et al. 2006)
0/34 She (Xue et al. 2006)
0/35 Buyi (Xue et al. 2006)
0/35 Han/Harbin, Heilongjiang (Xue et al. 2006)
0/35 Han/Meixian, Guangdong (Xue et al. 2006)
0/35 Hui/PRC (Xue et al. 2006)
0/35 Tibetans (Xue et al. 2006)
0/35 Yao/Bama, Guangxi (Xue et al. 2006)
0/35 Yao/Liannan, Guangdong (Xue et al. 2006)
0/38 Kazan Tatar/Russia (Wells et al. 2001)
0/40 Dungan/Kyrgyzstan (Wells et al. 2001)
0/40 Han/Guangdong (Hammer et al. 2006)
0/41 Uighur/Kazakstan (Wells et al. 2001) [Does this not include the Kazakstani Uyghur sample tested by Zerjal et al. (2002)?]
0/42 Han/northeastern China (Katoh et al. 2004)
0/43 Korean/Korea (Xue et al. 2006)
0/43 Uzbek/Tashkent, Uzbekistan (Wells et al. 2001)
0/43 Yizu/Sichuan (Hammer et al. 2006)
0/44 Han/Shaanxi (Hammer et al. 2006)
0/44 Shugnan (Pamiri)/Tajikistan (Wells et al. 2001)
0/45 Korean (Wells et al. 2001)
0/45 Okinawa, Japan (Hammer et al. 2006)
0/45 Tamang (Gayden et al. 2007)
0/46 Papua New Guinea (Hammer et al. 2006)
0/46 Sourashtran/South India (Wells et al. 2001)
0/47 Armenians/Armenia (Wells et al. 2001)
0/47 Japanese (Xue et al. 2006)
0/47 Central Yakut/"okayushie" dialect (Pakendorf et al. 2006)
0/48 Philippines (Hammer et al. 2006)
0/48 Taiwan Aborigines (Hammer et al. 2006)
0/49 Russian/North Russia (Wells et al. 2001)
0/49 Tujia/Hunan (Hammer et al. 2006)
0/49 Central Yakut/"akayushie" dialect (Pakendorf et al. 2006)
0/50 Northern Altaians (Kharkov et al. 2007)
0/50 Lebanese/Lebanon (Wells et al. 2001)
0/51 She (Hammer et al. 2006)
0/52 Manchu/Liaoning (Hammer et al. 2006)
0/53 Iranians/Samarkand, Uzbekistan (Wells et al. 2001)
0/53 Melanesia (Hammer et al. 2006)
0/54 Nenets/Russia (Wells et al. 2001)
0/55 Indonesia, East (Hammer et al. 2006)
0/56 Vilyuy Yakut (Pakendorf et al. 2006)
0/58 Miao (Hammer et al. 2006)
0/58 Uzbek/Bukhara, Uzbekistan (Wells et al. 2001)
0/60 Polynesia (Hammer et al. 2006)
0/60 Yao/Guangxi (Hammer et al. 2006)

Ebizur said...

0/61 Shizuoka, Japan (Hammer et al. 2006)
0/63 Uzbek/Fergana Valley, Uzbekistan (Wells et al. 2001)
0/66 Newar (Gayden et al. 2007)
0/68 Uzbek/Surkhandarya, Uzbekistan (Wells et al. 2001)
0/70 Vietnam (Hammer et al. 2006)
0/75 Korea (Hammer et al. 2006)
0/77 Kathmandu, Nepal (Gayden et al. 2007)
0/79 Korean/northeastern China (Katoh et al. 2004)
0/84 Han/Taiwan (Hammer et al. 2006)
0/84 Kallar/South India (Wells et al. 2001)
0/85 Korean/Seoul, South Korea (Katoh et al. 2004)
0/89 Russian/Tashkent, Uzbekistan (Wells et al. 2001)
0/91 Sri Lanka (Hammer et al. 2006)
0/105 Tibet (Hammer et al. 2006)
0/117 South Iran (Regueiro et al. 2006)
0/117 Japanese/Kanto region (Katoh et al. 2004)
0/129 Yadhava/South India (Wells et al. 2001)
0/156 Tibet (Gayden et al. 2007)
0/405 India (Hammer et al. 2006)

terryt said...

"AFAIK 'C6 is a recently recognized group whose geographical associations were not reported'. Do you have a source for your claim?"

Just that really unreliable source, Wikipedia:

http://en.wikipedia.org/wiki/Haplogroup_C_(Y-DNA)

Quote, 'C6 (P55) Found in the highlands of New Guinea[20]' I didn't realise that was your work Ebizur, so belated thanks.

"UFO abductions are always the best explanation, aren't they?"

I realise that you regularly need to postualte such, but evidence Dienekes has recently put up at his blog suggests very strongly that Y-hap C originated in Mongolia. I agree that there could well have been a secondary 'expansion knot for C at Indochina', specifically Wallacea. But Y-hap C5 was not part of it.

"If they are part of MNOPS then they are not anymore basal K".

Perhaps not. But make up any theory you like, they're still extremely unlikely to be just a single clade within the wider haplogroup. They are equal to either M, NO, P and S, or they are equal to T, L, MNOPS. Take your pick. Even in the latter case you still have half the clade round Wallacea.

"Btw, there's no 'OP' haplogroup: there are NO and P"

Sorry. Obviously I put the comma in the wrong place. I thought that with your knowledge you would have been able to work that out without my help.

"C3c (M48, M77, M86)
28/40 = 70.0% Stony Tunguska Evenks"

Seems, as you suggested somewhere, that C3c is a far northeastern haplogroup that has spread west over time. Possibly with Turkish and Mongol expansion, although perhaps that expansion is just a continuation of pre-historical processes. I'm fairly sure it was you who mentioned C3c probably arose after C3b had entered America. Makes sense. Have you seen the map Dienekes links to? It places most Mongolian Y-haps in C(xC3), which is not really supported by some of the information you provide here: '11/24 = 45.8% Mongolian (Wells et al. 2001).

Maju said...

Re. C6:

This is odd. The claim is sourced to this paper (PDF) where C-P55 is called "C4".

The data for this clade is not shown in table 1. I'll check the supplemental material if possible later.

Maju said...

"... evidence Dienekes has recently put up at his blog suggests very strongly that Y-hap C originated in Mongolia".

Which "evidence"? I read that discussion and your obsession on emphasizing Mongols of all ethnicities (ignoring the fact that Mongols are original from Eastern Mongolia and Manchuria, btw, insisting that the Hazara have anything to do with Mongols and other weak pseudo-logic so typical of you) without intervening because I have more than enough with rebuking you here.

""If they are part of MNOPS then they are not anymore basal K".

Perhaps not. But make up any theory you like, they're still extremely unlikely to be just a single clade within the wider haplogroup".

You said before several times that you agreed with the hypothesis. Of course, if, out of convenience, you change your mind now, we have nothing to discuss, because we can only wait for further research. I though we were discussing within that hypothesis (i.e.: MNOPS does exist and it does include all or most K lineages in SE Asia/Oceania) but if it's not the case, we better leave the debate on hold until the matter is clarified.

"They are equal to either M, NO, P and S, or they are equal to T, L, MNOPS. Take your pick. Even in the latter case you still have half the clade round Wallacea".

It'd be two different clades: either K or MNOPS. Anyhow remember that much K-other was eventually found to be part of M. All those Oceanian lineages seem to be ill-researched and may well be part of larger sublineages. Caution is highly recommended.

Maju said...

The supplementary material doesn't say anything. It's the least informatve supp. material I have ever read. :(

terryt said...

"You said before several times that you agreed with the hypothesis".

Ages ago. When I put up my connections between mtDNA and Y-haps. And I did it then in an attempt to make the evidence fit your 'out of India' theory. You disagreed with me then, but I see now it's you turn to claim a separate lineage to make the evidence fit your theory. Interesting.

"I though we were discussing within that hypothesis (i.e.: MNOPS does exist and it does include all or most K lineages in SE Asia/Oceania)"

We are. And the majority of those lineages are found around Wallacea: greatest diversity.

"your obsession on emphasizing Mongols of all ethnicities"

OK. Mongolians then, to be more specific. The map has most Y-haps in Mongolia belonging to C*, not C3. Although Ebizur has countered this evidence with his own research.

"Caution is highly recommended".

Quite. Including assuming an Indian origin for mtDNA R and a Thailand origin for MNOPS.

Maju said...

"Ages ago".

No. Within the timeframe of this post, which was published in February 11. And you did agree with that working hypothesis several times, not just once (review the comments you made).

Now it seems you have realized how inconvenient it is for your Wallacean mythology and so you seem to want to change your mind but you don't do it clearly so you can keep playing with both sides.

"We are. And the majority of those lineages are found around Wallacea: greatest diversity".

Only within MNOPS, not within K (within this hypothesis, of course).

"Although Ebizur has countered this evidence with his own research".

Hurray for Ebizur again. Just that calling your own faulty reasoning "evidence" is... well...

"Quite. Including assuming an Indian origin for mtDNA R and a Thailand origin for MNOPS".

Not sure about Y-DNA MNOPS, after all it's by the moment just hypothetical, but mtDNA R coalescing and expanding from South Asia has nothing to discuss: it's crystal clear.

Ebizur said...

Haplogroup C3c addendum:

BATWING estimate of TMRCA
3,500 [95% CI 300–19,700] years

Age of STR variation by Td approach
w=0.0032 : 1,100 ± 300 years
w=0.00069 : 5,000 ± 1,500 years
(Katoh et al. 2004. C3c-M48 sample set includes 13 Khalkh, 20 Uriankhai, 18 Zakhchin, and 4 Khoton from Mongolia plus 8 Manchu from northeastern PRC.)

2,750 ± 1,370 years (Karafet et al. 2002. Total sample set includes 89 Forest Nentsi, 59 Tundra Nentsi, 28 Komi, 47 Khants, 131 Selkups, 38 Nganasans, 9 Entsi, 67 Dolgans, 48 Kets, 18 Western Evenks, 78 Eastern Evenks, 35 Yakuts, 81 Buryats, 145 Mongolian Khalkhs, 40 PRC Evenks, 23 Oroqens, 11 Yukaghirs, 31 Evens, 12 Koryaks, 22 Siberian Eskimos, 52 Manchu, 44 Shaanxi Han, 98 Altais, 54 Uzbeks, 13 Kirghiz, 30 Kazakhs, 68 Uygurs, and 61 Russians. Haplogroup C3c-M86 occurs most frequently in this study's "NE Siberia" (Yukaghirs, Evens, Koryaks, Eskimos) and "Central-South Siberia" (Eastern Evenks, Yakuts, Buryats, Mongolian Khalkhs, PRC Evenks, Oroqens) groups, and rarely in this study's "NW Siberia" (Forest Nentsi, Tundra Nentsi, Komi, Khants, Selkups, Nganasans, Entsi, Dolgans, Kets, Western Evenks) and "Altai/Central Asia" (Altais, Uzbeks, Kirghiz, Kazakhs, Uygurs) groups. Judging from the data in other papers that have quoted Karafet et al. 2002, the C3c-M86 Y-chromosomes of the "Altai/Central Asia" group are concentrated in the Kazakhs and Kirghiz, those pf the "NW Siberia" group are concentrated in the Western Evenks and Dolgans, those of the "NE Siberia" group are concentrated in the (Eastern) Evens and Yukaghirs, and those of the "Central-South Siberia" group are concentrated in the Evenks, Oroqens, and Khalkhs, but four C3c-M86 Buryats, five C3c-M86 Altais, and one C3c-M86 Xinjiang Uygur also should have been included in the data set from which the TMRCA estimate has been calculated. None of this study's 52 Liaoning Manchus and 44 Shaanxi Han belongs to C3c-M86.)

Age estimate for the Native American-specific C3b-P39 (Zegura et al. 2004): 2,550 ± 1,910 years
(C3b-P39 sample set consists of 14 Apache, 7 Cheyenne, 5 Tanana, 5 Sioux, and 1 Navajo.)

Bashkir (Lobov et al. 2009)
Saratovsky & Samarsky
0/50 C3c-M48

Permsky
0/43 C3c-M48

Sterlibashevsky
2/52 = 3.8% C3c-M48

Western Orenburgia
5/43 = 11.6% C3c-M48

Eastern Orenburgia
0/34 C3c-M48

Abzelilovsky
0/80 C3c-M48

Baymaksky
0/89 C3c-M48

Burzyansky
0/80 C3c-M48

Bashkir total
7/471 = 1.5% C3c-M48 [0/89 = 0.0% Baymaksky - 5/43 = 11.6% Western Orenburgia]

The Bashkir total of Lobov et al. 2009 would fall between 1/67 = 1.5% Uygur/Xinjiang (Hammer et al. 2006) and 1/70 = 1.4% Tokushima, Japan (Hammer et al. 2006) in my haplogroup C3c frequency list.

terryt said...

"No. Within the timeframe of this post, which was published in February 11. And you did agree with that working hypothesis several times, not just once (review the comments you made)".

I have, and I can't find anywhere that I've suggested the Y-hap Ks as a group are separate from MNOPS as a group. They are four more members of the group: K1, K2, K3, K4, M, NO, P and S.

"Now it seems you have realized how inconvenient it is for your Wallacean mythology and so you seem to want to change your mind"

It is you who have changed your mind, because it's inconvenient for the belief that MNOPS* originated anywhere other than somewhere very near Wallacea.

"Only within MNOPS, not within K (within this hypothesis, of course)".

So now you're agreeing that MNOPS originated round Wallacea?

"Just that calling your own faulty reasoning 'evidence' is... well..."

Maju. Your 'reasoning' puts haplogroups in all sorts of unrealistic places simply because such placement suits your belief. My reasoning at least has haplogroup members moving at a realistic pace in realistic directions from realistic regions of origin. I'll comment further at your Khoi-San post.

"Age estimate for the Native American-specific C3b-P39 (Zegura et al. 2004): 2,550 ± 1,910 years"

Doesn't that seem a little too recent to represent Y-hap C's arrival in America?

Maju said...

It's absolutely crazy to discuss with you: look at the map of this post: MNOPS has always been placed in SE Asia.

I have pancakes on the pan

terryt said...

"MNOPS has always been placed in SE Asia".

Yes. You placed it there, ignoring it's brother clades. Once you include them it pushes MNOPS* further south and east.

Maju said...

Well, it's a possibility.

But I don't think that there is any clear genetic nor archaeological evidence that would support migrations from Sundaland or Wallacea, much less those involving the lineages NO and P.

And we have not the slightest idea on how K2, K3 and K4 fit in.

So I choose to place the node in Indochina though guess that South China could do too.

terryt said...

"Well, it's a possibility".

Thank you. And it's raining here at last (I thought you may be interested, unlikely).

"And we have not the slightest idea on how K2, K3 and K4 fit in".

Or K1. But they're related, probably closely, and most are found beyond Wallacea, as are M and S.

"I don't think that there is any clear genetic nor archaeological evidence that would support migrations from Sundaland or Wallacea"

Cavalli-Sforza has some quite clear genetic evidence in his principal component maps. And we can't use blade technology as a marker for H.sapiens in the east. The Australian Aborigines didn't aquire blade technology until about 5000 years ago (count the zeros, I've not made a mistake).

Maju said...

"And it's raining here at last"...

Gratz. I don't think NZ will be affected by drought as seriously as Australia: too much ocean and too little land for that. But Australia... :(

"And we can't use blade technology as a marker for H.sapiens in the east".

Here you're talking seriously, finally!

Of course blade tech is a useless ref. east of Bengal before the Hoabinhian. Or so it seems at the present stage of research. But, would there have been a massive backmigration from SE Asia as you propose, we should see some kind of tech or other cultural items backmigrating with them, flake industries maybe. I am not aware of any such archaeological trail, really.

While blade tech seems to have existed in some parts of South Asia since c. 103 Ka ago. its spread seems to be a Western Eurasian phenomenon mostly, at least in the first 20-30 Ka or so. So it's something that began in South Asia and spread only to West Eurasia (Central Asia and North Africa included).

This makes total sense in the context of South Asians expanding westward but doesn't seem to be related to SE Asia in any way. So archaeology seems to deny your hypothesis (or at least does not support it at all).

terryt said...

"I don't think NZ will be affected by drought as seriously as Australia"

You're correct there. But we've had a bad one for us.

"Here you're talking seriously, finally!"

I've always claimed that we can't use blade technology as a marker for H.sapiens, especially in the east. It was you who maintained we should use it as a marker, specifically across Central Asia and into East Asia.

"its spread seems to be a Western Eurasian phenomenon mostly, at least in the first 20-30 Ka or so. So it's something that began in South Asia and spread only to West Eurasia"

So we can't use blade technology as a marker for any migration into India from further east either. Humans reached New Guinea/Australia at least 50 Kya, so any back migration would date to that time or slightly more recently. It's quite possible that blade technology in India actually developed out of a combined technology.

Maju said...

"I've always claimed that we can't use blade technology as a marker for H.sapiens, especially in the east. It was you who maintained we should use it as a marker, specifically across Central Asia and into East Asia".

Not in the East. Only in Altai, because the older layer is not MSA derived tech but Mousterian and Neanderthal remains are found as well. It seems that in most places where Neanderthal existed, there blade tech marks the transition more or less.

But this only applies to West Eurasia, and Altai is exactly the NE limit of this prehistorical province.

Maju said...

"So we can't use blade technology as a marker for any migration into India from further east either".

No, I know. But the R girls were "a minute later" using blade tech in West Eurasia. So your hypothesis is inconsistent because the only distinctiveness of the R bunch was precisely blade tech... in West Eurasia, of course. They did not bring anything we know of from further east.

"Humans reached New Guinea/Australia at least 50 Kya, so any back migration would date to that time or slightly more recently. It's quite possible that blade technology in India actually developed out of a combined technology".

103,000 and then 75,000 BP there was blade tech in South Asia. So it evolved before any back migration. Not only that, some blade tools also exist in Mousterian (Neanderthal) contexts of Palestine c. 60 Ka ago (not sure of the exact dates: they vanish later anyhow).

terryt said...

"But the R girls were 'a minute later' using blade tech in West Eurasia. So your hypothesis is inconsistent because the only distinctiveness of the R bunch was precisely blade tech... in West Eurasia, of course".

But the R girls have nothing to do with blade technology in the east. So it wasn't blade technology that they carried west.

"They did not bring anything we know of from further east".

Seems the accompanying Y-haplogroups are a possibility. The East Asian maps at Dienekes show Y-hap NO present in several countries just a little west of Wallacea: Borneo, Vietnam, Thailand. The haplogroup stretches further north into Taiwan, Korea, Mongolia and Tibet. Perhaps Ebizur can enlighten us as to whether the haplogroup is wrongly labeled.

Furthermore the map has Y-hap P in Tenggara, Borneo, Java, Sumatra, Malaysia and the Philippines. Again perhaps Ebizur can tell us if the haplogroup should be either R or Q (as it almost certainly should be in regions further north). However if the haplogroups are labeled correctly it may tell us something about their movement from immediately west of Wallacea.

Maju said...

"But the R girls have nothing to do with blade technology in the east. So it wasn't blade technology that they carried west".

They could carry blades to the west and something else to the east...

I have already mentioned before that at this phase, the migration westward associated with mtDNA R includes other haplogroups (mtDNA M1, N1, maybe X; Y-DNA IJ, T and G probably). While related, these two populations have a somewhat different composition, with less diversity in the one leading to SE Asia (MNOPS and R only, it seems).

I don't make any strict association blade-mtDNA R. Instead I make a quite strict association blade-West Eurasian H. sapiens, which is largely mtDNA R and Y-DNA IJK.

Maju said...

"The East Asian maps at Dienekes show Y-hap NO present in several countries just a little west of Wallacea: Borneo, Vietnam, Thailand. The haplogroup stretches further north into Taiwan, Korea, Mongolia and Tibet".

Do you mean NO(xN,O)? I know that more or less. Though surely Ebizur is much better at the exact details.

I think that Ebizur and I (and you probably, as this is no inconvenient for your pet theory) essentially agree with NO having spread from south to north in East Asia. I think it's a pretty much uncontrovertible fact. However where exactly in the south is probably more difficult to say.

"Furthermore the map has Y-hap P in Tenggara, Borneo, Java, Sumatra, Malaysia and the Philippines".

The problem with P is that there's a lot of P in East Asia (and sometimes America too) that has not been properly tested for both the markers of Q and R, so it can always be something else within the clade. I'd love to know the details of it but what I'd really would like to know is where is there true P(xQ,R) and what haplotype structure has. At the current stage of knowledge I fear we can't clarify this matter.

"... immediately west of Wallacea".

Could you make a sentence or two that does not include the word "Wallacea" in it for a change? Mainland SE Asia would do in this case.

terryt said...

"Mainland SE Asia would do in this case".

No it wouldn't. 'Mainland SE Asia' is the ancient 'Sunda'. The combined Australia/New Guinea is ancient 'Sahul' and the region between the two is 'Wallacea'. Call it something else if you like, but it's a separate region.

"At the current stage of knowledge I fear we can't clarify this matter".

But at the present stage of knowledge we'd have to assume it's P(xQ,R) because neither R or Q are at all common near the region, although tested for occasionally. So in this case P is definitely found in (dare I say it?) Wallacea. But of course that's inconvenient for your belief so we have to desperately come up with some other explanation, like questioning the evidence entirely: 'so it can always be something else within the clade'. A new one perhaps? U?

"Do you mean NO(xN,O)? I know that more or less".

I'd presume so. And guess where it's found? It fits well with the distribution of all the related haplogroups M, S and the various Ks.

terryt said...

Sorry. I missed the earlier comments:

"I make a quite strict association blade-West Eurasian H. sapiens, which is largely mtDNA R and Y-DNA IJK".

Yes. Once they'd left India, or they probably carried it from there. But I'm sure that Y-haps Q and R were included as well as IJK.

"I have already mentioned before that at this phase, the migration westward associated with mtDNA R includes other haplogroups (mtDNA M1, N1, maybe X; Y-DNA IJ, T and G probably)".

I'd agree that the later phases do, but I strongly suspect that the earlier phases don't. And I suspect that mtDNA N1 and X and Y-haps IJ and G were already outside India, if they'd ever actually been in it. Of course you're still assuming that mtDNA R arose in India.

"I don't make any strict association blade-mtDNA R".

Almost certainly no connection at all. But I'm sure you know exactly what technology I believe Y-haps NO and P, along with mtDNA R, originally carried from that unmentionable place between mainland SE Asia and New Guinea.

Maju said...

"'Mainland SE Asia' is the ancient 'Sunda'".

No. In everyday language it's Indochina (Burma, Thailand, Camboya, Laos, Vietnam and sometimes also Southern China).

And that's what you said: "Borneo, Vietnam, Thailand". Ok, Borneo is not Indochina (was I thinking in Burma?) but still it's far fetched to link them magically with your mythical Wallacea.

"But at the present stage of knowledge we'd have to assume it's P(xQ,R) because neither R or Q are at all common near the region"...

Do you even know what SNPs were tested? A map is just a map after all.

I am not going to assume what you want me to assume "just because". I need the data. And then I will assume what the data says (and nothing more). If it says P(xR1) then it's that (i.e. Q, R*, R2 or P*), for instance.

terryt said...

"Burma, Thailand, Camboya, Laos, Vietnam and sometimes also Southern China)".

All connected at times of lowered sea lavel, and so collectively called 'Sunda'.

"Ok, Borneo is not Indochina (was I thinking in Burma?) but still it's far fetched to link them magically with your mythical Wallacea".

Really? Borneo forms the West of Wallacean coastline.

"Do you even know what SNPs were tested? A map is just a map after all".

I don't know what SNPs were tested but I presume the authors had a reason to refer to it as P.

"I need the data".

Perhaps Ebizur can help.

"And then I will assume what the data says (and nothing more). If it says P(xR1) then it's that (i.e. Q, R*, R2 or P*), for instance".

But preferably not P*. Certainly not in that region.

Maju said...

From Wikipedia: "Sundaland is a biogeographical region of Southeastern Asia that comprises the Malay Peninsula and Maritime Southeast Asia islands of Sumatra, Java, Borneo, and surrounding smaller islands".

It is not Thailand or Vietnam. That's further north.

"Really? Borneo forms the West of Wallacean coastline".

No. Borneo is part of Sundaland. Wallacea begins at Sulawesi. Following Wikipedia again, it includes the following Indonesian provinces: Sulawesi, Maluku, North Maluku (except Aru islands), West Nusa Tenggara, East Nusa Tengara and West Timor, as well as East Timor, naturally.

You should know what exactly is Wallacea and what is Sundaland, really. You have been talking of them for more than a year.

"I don't know what SNPs were tested but I presume the authors had a reason to refer to it as P".

I would not presume so much, really.

"But preferably not P*. Certainly not in that region".

Why not? I am considering that MNOPS spread from Indochina, there could be P(xQ,R) in Indochina without being any contradiction.

Whatever the case, I can't judge without knowing the data.

terryt said...

"No. Borneo is part of Sundaland. Wallacea begins at Sulawesi".

You've got it wrong Maju. How far is Suluwesi from the east Borneo coast? Wallacea is the region between Sunda and Sahul. Between Borneo and New Guinea/Australia, with those two regions being the western and eastern boundaries.

terryt said...

Do you really believe that humans would have occupied all the islands of Wallacea yet never have had anything to do with the mainland coast on either side?

Maju said...

"How far is Suluwesi from the east Borneo coast?"

Wallace line, nothing more, nothing less: a conceptual divide.

I got it right: Borneo is NOT Wallacea.

"Do you really believe that humans would have occupied all the islands of Wallacea yet never have had anything to do with the mainland coast on either side?"

I don't know. Probably not... unless they did not have boats (irony meant).

Whatever the case, neither Borneo nor New Guinea are part of Wallacea. Get your mythical country straight first of all. If you want to say "Indonesia", "the Malay Archipelago" say so but don't mess concepts at whim because from confusion only confusion can be born.

terryt said...

"Whatever the case, neither Borneo nor New Guinea are part of Wallacea. Get your mythical country straight first of all".

But they certainly form the western and eastern boundaries, so stop being stupid (unless you can't help it).

"I don't know. Probably not... unless they did not have boats (irony meant)".

You know very well that to enter Wallacea they had to have boats. So, again, stop being stupid.

"If you want to say 'Indonesia', 'the Malay Archipelago' say so"

I certainly do not mean either place. Malaysia does not reach as far as Wallacea and much of Indonesia is also outside Wallacea. You're obviously very disappointed that Y-haps NO and P do actually show evidence of having first appeared in Wallacea. Get over it.

Maju said...

Hopefully you have read the latest news from about Wallace line being a real barrier for Y-DNA in Indonesia (mentioned at Dienekes' and here). The details are not too clear (pay per view paper) but the abstract is very clear about a divide at Wallace line within Indonesia: Y-DNA O to the west and Y-DNA C, M and S to the East.

Karafet argues, with good sense for what I understand for almost all Y-DNA being Paleolithic, however she also seems to suggest a four wave pattern (not sure of what she means: maybe C, M and S, pre-Austronesian O and Austronesian O).

Whatever the case, what are you arguing for: a homeland at Wallacea or a homeland at Sundaland? Wallace line is in between and, even if they had some boating capability (of course), they could not sail the open seas every other day with just a mere canoe.

terryt said...

"what are you arguing for: a homeland at Wallacea or a homeland at Sundaland?"

Who-ever first entered Wallacea had to have done so from Sundaland. The only real question involved there is, 'by which route did they reach Sundaland?'

"she also seems to suggest a four wave pattern (not sure of what she means: maybe C, M and S, pre-Austronesian O and Austronesian O)".

I'm not sure of a fourth but I see three easily enough. I'd suspect it was C first. Then some sort of K or MNOPS. Then O. Perhaps she sees the KMNOPS as involving two crossings. It certainly seems to have been spread over quite some time.

"Wallace line is in between"

Wallace's line is usually drawn beween Lombok and Flores, then north along the east Borneo coastline, then either south of the Phillipines or straight north between Palwan and the Phillipines. So it basically hugs the East Borneo coastline. Wallacea is the region east of this line, excluding Australia and New Guinea and all islands north and east of there. I tend to include the Phillipines in Wallacea because the presence of a few monkeys there is insufficient to include it as part of the SE Asian ecological region. And elephants and rhinos didn't get there. Besides which Flores is east of Wallace's Line but elephants did get there. Wallace's Line has therefore not always been entirely impassable.

"but the abstract is very clear about a divide at Wallace line within Indonesia: Y-DNA O to the west and Y-DNA C, M and S to the East".

Strange idea, because O definitely crossed it. As must have C, M and S, or at least their ancestors. And C2 is present in Wallacea, and seems to have originated there, before moving further east into the wider Pacific.

"they could not sail the open seas every other day with just a mere canoe".

But they didn't have to every day. I'd guess that coastal exploitation, with occasional voyages to the closer islands, would have been normal though.

Maju said...

"The only real question involved there is, 'by which route did they reach Sundaland?'"

Uh? By the Kraa isthmus, then much wider than today. Probably by the coasts but if you think they had to walk over the top of the mountains in dense jungle just because they could (by your dogma) not have any boats at all yet (and hence not be able to cross any mid-sized river or mangrove), they still had to go through the Kraa isthmus.

"So it basically hugs the East Borneo coastline".

Draw it the way you want: it runs between Borneo and Sulawesi, that's the fact. It's not a dotted line on a map (the map is never the reality) but a system of channels that never went dry nor became too narrow to be crossed by mere swimming. That's the real Wallace Line: wide spans of persistent salty water.

"Wallace's Line has therefore not always been entirely impassable".

Absolutes are not real either. Nobody says it was "entirely" impassable, just a major barrier. You can bypass barriers and obstacles but it's a lot harder than if they are not there.

"Strange idea, because O definitely crossed it".

Not sure in what apportions. Karafet suggests that Austronesian colonization had some impact but a minor one. Maybe that's it.

And probably, after the early Austronesian colonization there would be further gradual flows from the largest islands eastward. Even today that is a real trend.

"But they didn't have to every day. I'd guess that coastal exploitation, with occasional voyages to the closer islands, would have been normal though".

Crossing Wallace Line is not visiting closer islands, probably most people living at its shores (either one) in Paleolithic times never did it in their whole lives, even if they knew there was a land at the other side.

And anyhow, what I just said at Dienekes': fluids' dynamics: the barrier would generally prevent significative migration because mounting enough pressure to not just overcome it but also the demic pressure at the the other side is quite unrealistic.