December 09, 2008

Reduction of human cranial diversity and distance from Africa

It is important to note what this paper has actually discovered, which is not clear from the abstract, and is only briefly discovered in the paper itself.

The paper does not prove that climate has not played an important role in the shaping of human cranial variation. Rather it shows that it does not play an important role in the level of diversity found in different populations. But, I really can't think why anyone would propose that.

Let's take a particular trait, say nose breadth or cranial breadth. It has been demonstrated that the former tends to be narrower and the latter broader in more northerly compared to more equatorial regions. But, this is a difference in the means and not a difference in the variance of the traits in question.

The authors write:
All the most informative traits were distributed in the anterior region of the cranium (yet not all anterior traits were highly informative). This came as a surprise as the anterior regions of the cranium are generally assumed to be more affected by natural selection (i.e. facial area and cranial breadth), while the temporal bone and some traits of the neurocranium are considered to reflect population history (Beals et al. 1984; Franciscus & Long 1991; Roseman 2004; Harvati & Weaver 2006).

Diversity is reduced in either of two ways:
  • Serial bottlenecks, when a subset of the variation present in the ancestral population migrates to new regions
  • Selection, which reduces diversity in the traits under selection
Under the scenario of bottlenecks, it is expected that all traits should be equally affected by the reduction of diversity.

Under the scenario of selection, however, it is expected that traits under selection will be more affected by the new ecological conditions that occur away from the cradle.

The finding that the anterior region of the cranium is more informative in terms of reduced diversity is actually consistent with selection and not with bottlenecks, since it is in this part, in nasal features, facial flatness, prognathism, etc. that populations have developed strongly differentiated types under selection.

Furthermore, in this paper the study is limited to skulls from the last 2,000 years. But, if bottlenecks are indeed responsible for the reduction of diversity, then this reduction of diversity would be visible in the earliest Homo sapiens skulls from the various regions, as these would be descended from the few bottlenecked migrants. Are Upper Paleolithic Europeans, for example, more or less diverse than Africans of equivalent age, or even modern Africans?

On the contrary, recent skulls may be less diverse than the earliest ones, due to a longer period of selection, i.e., the tens of thousands of years between the earliest Homo sapiens in Europe or Asia and the ones of the last 2,000 years. Moreover, this selection ought to have been strongest in regions further from Africa, as this is correlated with different environments (although not necessarily the 3 climate variables considered here).

In conclusion:
  • Within-population variance decreases with distance from Africa, but this does not measure between-population difference in mean trait values
  • Both bottlenecks and selection result in reduced variance
  • A stronger variance-reduction signal in the anterior cranium is more consistent with selection than with bottlenecks
  • The bottlenecks theory should be more visible in early, not recent skulls as those studied in this paper. Recent skulls should have reduced variance compared to their more ancient counterparts due to a long period of selection.

Proceedings of the Royal Society B doi:10.1098/rspb.2008.1563

Distance from Africa, not climate, explains within-population phenotypic diversity in humans

Lia Betti et al.

Abstract

The relative importance of ancient demography and climate in determining worldwide patterns of human within-population phenotypic diversity is still open to debate. Several morphometric traits have been argued to be under selection by climatic factors, but it is unclear whether climate affects the global decline in morphological diversity with increasing geographical distance from sub-Saharan Africa. Using a large database of male and female skull measurements, we apply an explicit framework to quantify the relative role of climate and distance from Africa. We show that distance from sub-Saharan Africa is the sole determinant of human within-population phenotypic diversity, while climate plays no role. By selecting the most informative set of traits, it was possible to explain over half of the worldwide variation in phenotypic diversity. These results mirror those previously obtained for genetic markers and show that ‘bones and molecules’ are in perfect agreement for humans.

Link

143 comments:

Unknown said...

So only distance from Africa played an important role in cranial shaping, ah?
And how do the researchers explain the existance, IN THE SAME SPOT, of populations with different phenotypes?
And also why in other Homo species like the Erecti and Habilines diversity did not decrease as they moved out of Africa?
Neanderthals also were significantly different!
Finally if climate and distance from Africa play a big role in the cranial formation and perhaps the general phenotype, why in specific areas of the World DIFFERENT Homo species have some curious common features?
For example, the Indonesian Homo Erectus have some similarities with Australasian Homo Sapiens skulls. Why?
The climate was not the same 200000 years ago with that of 40000 years ago.
Their diet and the general flora and fauna were not the same too!

Maju said...

I have to, largely, agree with the authors. A lot of what is claimed as "adaptative" evolution is not: it's just founder effects and more or less rigid fixation of phenotypes in once small homogeneous populations.

Said that, some particular traits may have been favored by biological adaptation as well. But they could only be selected for because they were in the genetic pool to begin with.

It's interesting anyhow that the overlapping male/female area of most likely origin is south of The Horn: Middle and Southern Africa only.

eurologist said...

To Antigonos' questions:

Largely, because there is a huge difference between what happens in 50,000 years (with newly-found cultural innovations that reward wide-range travel and contact) versus 500,000 years with a model that encourages close-range existence.

terryt said...

"A lot of what is claimed as "adaptative" evolution is not".

But Maju we all know that you believe the environment has had a minimal effect of the human phenotype, except for some complex process giving rise to pale skin after clothes had been invented that covered all of the skin anyway.

"they could only be selected for because they were in the genetic pool to begin with".

Not heard of mutations then?

"the overlapping male/female area of most likely origin".

There is absolutely no reason at all why the basal male and female haplogroups must have left Africa at the same time or, indeed, to have originated in the same region.

"because there is a huge difference between what happens in 50,000 years (with newly-found cultural innovations that reward wide-range travel and contact) versus 500,000 years with a model that encourages close-range existence".

I'd say there's not a "huge" difference, just a speeding up of basically the same processes.

Maju said...

But Maju we all know that you believe the environment has had a minimal effect of the human phenotype, except for some complex process giving rise to pale skin after clothes had been invented that covered all of the skin anyway.


Precisely more reason for pale faces and, very specially, rosy cheeks (also found in NE Asia, btw). Rosy cheeks maximize the absorption of UV and the metabolization of vitamin D.

There is absolutely no reason at all why the basal male and female haplogroups must have left Africa at the same time or, indeed, to have originated in the same region.

There is a simple reason: common sense. ^^

terryt said...

"There is a simple reason: common sense".

Common sense has never been proof of any scientific theory. We know that Y-chromosome and mtDNA haplogroups can be remarkably independent. Dienekes comes up with endless examples on this blog. Besides, the timing of the basal mutations for the male and female haplogroups could be as far apart as 50,000 years or more. And mtDNA lines are usually much more long-lived in particular regions. Y-chromosomes tend to move in over them.

Common sense is far from proof in this case.

Maju said...

And mtDNA lines are usually much more long-lived in particular regions. Y-chromosomes tend to move in over them.


We don't know for sure but, in any case, that doesn't mean population replacement but just stronger drift in the male side. The origins of the clades seem to be concentrated in certain areas too (for example all high level Eurasian branches directly coalesced from African ones, both mtDNA L3 or Y-DNA CT).

There is a wide area of overlap in highest cranial variability in those maps and it is very suggestive that it's that way. Moreso when more and more early advanced cultural and technological elements are appearing in Southern Africa and when Indian MP appears to correlate best with Southern than Eastern African MSA contemproary industries.

And to all that you add the common sense of thinking that men and women surely migrated together first of all, as they do modernly in similar societies.

terryt said...

"the common sense of thinking that men and women surely migrated together first of all, as they do modernly in similar societies".

No they don't. Men travel much further than women, and are often first into a new region where they mate with women already there. Sometimes no women from their own culture arrive at all.

"all high level Eurasian branches directly coalesced from African ones, both mtDNA L3 or Y-DNA CT".

This argument suggests that you believe the woman who had the mutation in her mitochondrial DNA that lies at the root of all modern mtDNA haplogroups was a completely different species from her parents, let alone her brothers and sisters.

What's more you believe this woman had the absolutely unbelievable good fortune to meet and mate with the man who had the mutation that lies at the root of all modern Y-chromosome haplogroups. Who, again, was a totally different species from his parents etc.

I'd say, Maju, that you are hopelessly entangled in the Jewish origin myths you have been exposed to over your lifetime, especially during your childhood.

moneduloides said...

It's neutral theory; think genetic drift, not selection.

Maju said...

No they don't. Men travel much further than women, and are often first into a new region where they mate with women already there. Sometimes no women from their own culture arrive at all.

Examples? We are talking hunter-gatherers here, all right?

This argument suggests that you believe the woman who had the mutation in her mitochondrial DNA that lies at the root of all modern mtDNA haplogroups was a completely different species from her parents, let alone her brothers and sisters.

What's more you believe this woman had the absolutely unbelievable good fortune to meet and mate with the man who had the mutation that lies at the root of all modern Y-chromosome haplogroups. Who, again, was a totally different species from his parents etc.

I'd say, Maju, that you are hopelessly entangled in the Jewish origin myths you have been exposed to over your lifetime, especially during your childhood.


That's nonsense. And I certainly would not sustain such ideas (different species, Jewish creation mythology, etc.) You are just avoiding to discuss the real matter by throwing arbitrary accusations against me. That's pitiful to say the least and it's all only in your mind in any case.

terryt said...

"You are just avoiding to discuss the real matter by throwing arbitrary accusations against me".

That's rubbish, Maju. You certainly seem to believe there is something totally different between humans and all other species, both genetically and in range expansion.

"Examples?"

Several regions of South America have European Y-chromosomes and mtDNA from Indigenous people. Men were certainly the first Europeans to arrive in New Zealand and immediately began to have children with the local women. If European women had never reached here we would have the same situation as the previous example. Polynesian haplogrouops suggest very strongly that the men and women came from basically separate sources, as does similar evidence from America. Genghis Khan and his mates don't seem to have taken women with them but certainly have left their Y-chromosome widely spread.

"We are talking hunter-gatherers here, all right?"

OK. So you believe that hunter-gatherers behave totally differently to the rest of us, and to all other species, and always have. Refer to my first comment above.

Maju said...

You certainly seem to believe there is something totally different between humans and all other species, both genetically and in range expansion.

Your belief about my supposed beliefs is, like most beliefs, wrong.

So you believe that hunter-gatherers behave totally differently to the rest of us, and to all other species, and always have.

Not at all. You're just rambling here.

I said hunter-gatherers because your (quite predictable) colonial examples are obviously not adequate to explain their migrations.

Why don't you go to the grain and show me when modern or historically documented huntergatherers behave as you claim they do?

The first documented apparent evidence of sex-biased migration belongs already to the Neolithic period. Additionally, it's very much clear that Y-DNA tends to drift much more than mtDNA, with no need at all to look for separate origins for men and women (that would be really anti-natural). It was clear before but now it's even more. I'm not aware of anyone dealing with this biology news item but the first thing that came to my mind after understanding the process was precisely the strong Y-DNA drift that this gender-generation phenomenon implies.

terryt said...

"The first documented apparent evidence of sex-biased migration belongs already to the Neolithic period".

Wouldn't that simply be because we have more evidence for that period?

"Why don't you go to the grain and show me when modern or historically documented huntergatherers behave as you claim they do?"

All 'modern and historically documented huntergatherers' behave much the same as the rest of us do.

terryt said...

Interesting link Maju. Thanks. Other species too seem able to regulate the ratio of male to female in some way.

Maju said...

Wouldn't that simply be because we have more evidence for that period?

Maybe.

All 'modern and historically documented huntergatherers' behave much the same as the rest of us do.

Not in many aspects. Certainly the huntergatherers I have read about don't seem to part away from their women and go elsewhere in search of new females. This instead may have been somewhat more common since the concept of horde/army (a mobile unit almost exclusively made up of men in their most fertile age) was created. This kind of male conqueror migratory entity is not conceviable within the huntergatherer way of life, IMO, and I have never ever read of anything like that among them.

Other species too seem able to regulate the ratio of male to female in some way.

Don't know. But I imagine that the mechanism is the same for all mammals.

I don't really see how this mechanism is "regulating": the same effect should be achieved by mere randomness in meiosis and reproduction (mechanism that is in fact dominant overall). As I see it, it's not a regulating mechanism as much as a caprice of nature that happen to be self-balancing too.

terryt said...

"the huntergatherers I have read about don't seem to part away from their women and go elsewhere in search of new females".

True. But the Australian Aborigines had a complicated system of outbreeding which had people moving between different groups, often over quite huge distances. And some just got annoyed with people in their birth group and moved off to join another group.

They certainly didn't have anything like armies. However some North American groups seem to have had something similar to armies, which used to invade other people's territories.

terryt said...

By the way. In what way are "Polynesian haplogrouops suggest very strongly that the men and women came from basically separate sources, as does similar evidence from America" my "(quite predictable) colonial examples". Both examples are from prehistory.

Maju said...

True. But the Australian Aborigines had a complicated system of outbreeding which had people moving between different groups, often over quite huge distances. And some just got annoyed with people in their birth group and moved off to join another group.

Ok. All that sounds like normal genetic flow within an ethnic group or neighbour ethnicities.

They certainly didn't have anything like armies. However some North American groups seem to have had something similar to armies, which used to invade other people's territories.

North American natives were already in transition to the Neolithic if not a full-fledged Neolithic of marginal areas (with hunter-gathering still as major activity, much as we can see in many areas of Neolithic Europe). They also soon got horses (a very late Neolithic, actually Chalcolithic, developement in the Old World) which increased their mobility a whole lot.

The typical HG book cases: Australians, Bushmen, Pygmy, etc. do not appear to behave the way you have suggested in this thread.

By the way. In what way are "Polynesian haplogrouops suggest very strongly that the men and women came from basically separate sources, as does similar evidence from America" my "(quite predictable) colonial examples". Both examples are from prehistory.

Again we are talking of Neolithic peoples. In any case, one thing is the simplistic interpretation of autosomal genetic data and another, possibly very different thing, the real events that unfolded.

Remember that study that found that British male and female lineages had different sources. A most simplistic explanation suggested that Basque seamen had kidnapped Frisian women to settle the islands. But a more dedicated look at the matter shows clearly that in fact they all came from Middle Europe, where "Basque" (R1b) Y-DNA was once much more prevalent than it's today. In fact, at higher resolutions, the British Y-DNA is not Basque but just ancient Western/Central European ("Magdalenian"). Similarly, it's possible that a deeper look at Austronesian genesis may clarify that matter without any need of "men being from Mars, women from Venus" or whatever. Nevertheless, being a more modern late Neolithic Viking-like kind of expansion I'm open to consider that you're right in this case.

Ebizur said...

Just in case anyone has forgotten or did not know in the first place:

Prior to European colonization, Polynesians are thought to have belonged almost exclusively to Y-DNA haplogroups C2a1-P33 and O3*-M122(xO3a3c-M134). Haplogroup C2a1-P33 is a derivative of haplogroup C2-M38, which is known to be common among some populations of central and eastern Indonesia, but subclade C2a1-P33 seems to occur exclusively in Polynesians. Haplogroup O3(xO3a3c) is found most frequently in populations of the Philippines and southern China (e.g. Hmong/Miao).

Native Americans are thought to have belonged exclusively to haplogroups C3-M217(xC3c-M86), Q1a3a-M3, and Q-M242(xQ1a3a-M3). Haplogroup Q1a3a-M3 is especially predominant among indigenous South Americans. Haplogroups Q-M242(xQ1a3a-M3) and C3-M217(xC3c-M86) are relatively common among indigenous North Americans. Worldwide, haplogroup C3-M217(xC3c-M86) occurs most frequently among Mongols, and it is also frequent among Koreans, Manchus, Tungus (Evenks, Evens, Oroqens, Negidals), Altayans, Tuvans, and Han Chinese in some localities. The subclade C3c-M86 is extremely common among Tungus, Kazakhs, and Palaeo-Siberians (especially Nivkhs, Koryaks, and Yukaghirs), but it is apparently not found in the Americas; the low diversity of haplotypes within haplogroup C3c suggests that it might have originated and spread after the Americas had already been colonized. Outside of North America, haplogroup Q-M242(xQ1a3a-M3) occurs most frequently in Siberia, among Kets, Selkups, Chukchis, Yukaghirs, and Altayans (i.e. mostly Palaeo-Siberians, but also having a strong presence among the Samoyedic-speaking Selkups and the Turkic-speaking Altayans).

However, both Native Americans and Polynesians have experienced significant gene flow (mostly from Europeans) since colonial times. Depending on the locality and ethnic group, the Y-DNA of modern Native Americans and Polynesians may range from nearly purely native to at least 88% foreign (mostly European).

terryt said...

Thanks Ebizur. That's really great.

Maju commented earlier, "the common sense of thinking that men and women surely migrated together". Take haplogroup C in North America for a start. No mtDNA haplogroup appears to be at all closely associated with it. So, Maju, what haplogroup did it arrive in North America with? Or did it simply move in and have children with women who were already there? Anyway, any mtDNA haplogroup it might have come in with spread much further than the C Y-chromosome haplogroup. Men and women migrating together?

Y-chromosome Q. Descends from P and almost certainly evolved somewhere near the Central Asian steppes. No American mtDNA haplogroups, except possibly X, originate in that region. They all come from further east. Therefore Y-chromosome Q almost certainly teamed up with the local women as it moved east, eventually entering America with a mixture of haplogroups. Men and women moving together? I don't think so.

The Polynesians. Two Y-chromosomes C and O, but almost exclusively mtDNA B. One of the Y-chromosomes was not originally associated with that mtDNA haplogroup. Once again, do we have men and women moving together?

A final interesting little quote from a book I'm currently reading, "Australia's mammal Extinctions", concernig Australian Aborigines about 10,000 years ago.

"And at around the same time they [Tacon and Blackwell(1995)] find depictions of fighting in the region's rock art, with people 'flailing boomerangs, dodging spears and chasing each other with weapons raised'. Frightening animal-headed beings also appear, and at some sites there are paintings of people arrayed in 'great battle scenes'. These are among the earliest depictions of warfare in any world art tradition."

Australian Aborigines "in transition to the Neolithic if not a full-fledged Neolithic of marginal areas"?

Maju said...

Your prerception of how haplogroups may come together is overly simplistic and flawed, Terry.

You have to understand, first of all, that people has no idea (and certainly not in the past) of their haplogroup "identity". They just were a group of men and women living together and having offspring. They surely knew who their mother and greatmother was and maybe they had a vague idea of some mythical ancestor but that's about all. Even when they are fascinated by genealogies, they can only reach so far in the past (memory is limited and volatile) and can hold misconceptions.

You also have to understand that along the generations, specially in small populations, there is a phenomenon called drif, which causes some clades to vanish and others to become fixated or nearly so. This drift clearly appears to be much more marked for male clades than for female ones (an obvious reason being that men can potentially father many more children than even the most fertile woman - there may be others).

So you have, let's say, a Q dominated clan somewhere in Siberia and they (let's assume patrilocality) incorporate women from neighbouring groups who have different clades (four surviving to our day). Then they cross Beringia and so on.

C3? C3 may have been a more recent input in America (Clovis?) but should have coalesced along the rest in Beringia anyhow. Maybe they incorporated women as they moved but maybe they just carried the same generic mtDNA haplogroups (roughly) as did the Q clan before them, because they came from the same origin (just different founder effects in the Y-DNA lineage).

You can think of many other alternative explantions but what you should not do is to think of haplogroups as something as "races" or different populations necesarily. They may mean that or not. This difference between Y-DNA and mtDNA is just everywhere and, worst, when you look at other variables like autosomal DNA or X-DNA, you find other different patterns. What does all this imply? To me that ancestry is complex and cannot be simplified only to the father and mother gender lineages, that each person has many (tending to infinite) different ancestors and that the fact that some gender lineages have survived and others not is probably just a matter of random accidents (drift, founder effects) and not of different origins.

So, sure, if Q correlates with X, they certainly must have incorporated other women. But actually they probably also incorporated other men. Maybe Q were the ones incorporated to some other tribe, who knows?, the case is that once fused, this lineage survived the rigors of drift and the rest did not - at least not in that group. But that was probably just an accident not any pseudo-epic "drang nach osten" by Western Eurasian males into Eastern Eurasian territory, killing the guys and raping the girls. It was surely something much more amiable and complex with the final result being probably just an accident imposible to predict.

You must embrace Chaos. That a population now is 90% Q or whatever doesnt mean they always were that way. In periods of demographic stagnation (Beringia in the Ice Age for instance) haplogroups drifted almost necesarily, while in moments of migration, some were lucky enough to suceed and expand while others were not. The overall effect of all those accumulative accidents is not predictable: they are much more like Mandelbrot's butterfly.

In Mandelbrot's words: "Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line". He could well have added a comment on genetics, I guess.

terryt said...

I'm pleased to see you now agree with me: "when you look at other variables like autosomal DNA or X-DNA, you find other different patterns. What does all this imply? To me that ancestry is complex and cannot be simplified only to the father and mother gender lineages, that each person has many (tending to infinite) different ancestors and that the fact that some gender lineages have survived and others not is probably just a matter of random accidents (drift, founder effects)".

My original comment was: "There is absolutely no reason at all why the basal male and female haplogroups must have left Africa at the same time or, indeed, to have originated in the same region".

I have never, ever believed, nor claimed, that haplogroup distribution was the result of "any pseudo-epic 'drang nach osten' by Western Eurasian males into Eastern Eurasian territory, killing the guys and raping the girls". Far from it. I agree that it results from "something much more amiable and complex with the final result being probably just an accident imposible to predict".

The difference between us is simply that I believe the same process extends much further back in time than just since 'modern' humans left Africa. The reason we find so little evidence for interbreeding between modern humans and Neanderthals is because of the same "random accidents (drift, founder effects)" you use to account for modern haplogroup distribution.

Maju said...

My original comment was: "There is absolutely no reason at all why the basal male and female haplogroups must have left Africa at the same time or, indeed, to have originated in the same region".


It appears as they did exactly that anyhow. There is a quite clear parallel between Y-DNA and mtDNA branching in Africa: Y-DNA A corresponds quite well with L0 and L1, Y-DNA B with L2 and Y-DNA CT (mostly E in Africa) with L3. It's not absolute of course but it does seem to be a parallelism.

All Eurasian clades belong to this last pair of African DNA. So they did not spawn directly from the A-L0,1 population nor from the B-L2 one, but from the CT-L3 one, which once upon a time probably was relatively homogeneous and geographically circunscribed.

Once in Eurasia the parallel is broken though, what to me just means that all Eurasians were more or less together before expanding in all directions (and that probably was in South Asia).

I have never, ever believed, nor claimed, that haplogroup distribution was the result of "any pseudo-epic 'drang nach osten'...

It seemed to me you were claiming that when you suggested that a presumptly homogeneous group of haplogroup Q males took East/NE Asian females as they marched into America. The truth is surely far from that anyhow.

The difference between us is simply that I believe the same process extends much further back in time than just since 'modern' humans left Africa. The reason we find so little evidence for interbreeding between modern humans and Neanderthals is because of the same "random accidents (drift, founder effects)" you use to account for modern haplogroup distribution.

Look: I could not care less about the supposed Neanderthal admixture but the case is that we can't find any single Neanderthal gene in us, not just haplogroups (curious enough in itself) but neither any single piece of nuclear DNA at all. I find therefore that claiming admixture with no evidence is unreasonable and fanatic.

For me it's like listening to a creationist: no evidence, pseudoscience, beliefs before facts.

Ebizur said...

I should add that somewhere between 19.0% and 35.3% of Nivkhs, indigenous inhabitants of the lower Amur River basin and northern Sakhalin Island, belong to haplogroup P(xR1a). These Nivkhs probably also belong to haplogroup Q, but we may never know for certain, because the researchers are too lazy to test for any markers of haplogroup Q.

So, you might want to add Nivkhs to the list of Palaeo-Siberian populations that possess a substantial haplogroup Q component (but it might as well be haplogroup R1b for all we will ever know).

Two other Palaeo-Siberian populations, the Koryaks and Itelmens, seem to have much lower frequencies of haplogroup Q, despite being linguistically related to the Chukchis. Lell et al. (2002) have found haplogroup P-M45(xR1a1-M17, Q1a3a-M3) in only 5/27 = 18.5% of a sample of Koryaks, and haplogroup P(xR1a1, Q1a3a) is completely absent from their sample of Itelmens (0/18). It appears that most (60%+) Koryaks and Itelmens belong to haplogroups C3c-M48 and C-RPS4Y(xC3c-M48). The Nivkhs, who are usually considered to be linguistically unrelated to the Chukotko-Kamchatkans (i.e. Chukchis, Koryaks, Itelmens, etc.), also belong mainly to haplogroup C (38.1% C-RPS4Y according to Tajima et al.; 2/17 = 11.8% C(xC3c) and 6/17 = 35.3% C3c-M48 for a total of 8/17 = 47.1% C according to Lell et al.) and haplogroup P(xR1a). It is plausible that these Palaeo-Siberian populations may be remnants of the group(s) that originally colonized the Americas.

terryt said...

Ebizur. Thanks for that expansion. Seems we have yet more evidence for independence of mtDNA and Y-chromosome haplogroups. The information I have claims most mtDNA of Nivkhs is Y, not found in America as far as I'm aware although distantly related to A. Most mtDNA of Itelman is G, again not found in America although related to C and D. Y-chromosome Q's distribution gives the impression of a formerly widespread haplogroup that has been largely replaced by other haplogroups' more recent expansions.

Maju. "It appears as they did exactly that anyhow".

This from someone who's just said, "To me that ancestry is complex and cannot be simplified only to the father and mother gender lineages". So in fact you believe humans behaved at that early stage totally differently from the way they have behaved ever since. Sounds suspiciously like belief in Adam and Eve to me.

"There is a quite clear parallel between Y-DNA and mtDNA branching in Africa".

No there's not. Almost any measure you like to apply to mutation rates puts at least several thousand years between the two haplogroups' origin.

The "silly argument" we had at another Dienekes post developed because you are convinced, in spite of "no evidence, pseudoscience, beliefs before facts" that for the last 200,000 years the Northern Hemisphere's atmosphere has been completely different from that of the Southern Hemisphere. Your wish to believe this arises simply so that you can say that white skin evolved in response to this strangely different atmosphere. As you say, "For me it's like listening to a creationist".

terryt said...

Oh. I forgot.

"There is a quite clear parallel between Y-DNA and mtDNA branching in Africa".

Africa's quite a big continent, even if we just consider the stretch from South Africa to east Africa.

Maju said...

I should add that somewhere between 19.0% and 35.3% of Nivkhs, indigenous inhabitants of the lower Amur River basin and northern Sakhalin Island, belong to haplogroup P(xR1a). These Nivkhs probably also belong to haplogroup Q, but we may never know for certain, because the researchers are too lazy to test for any markers of haplogroup Q.

So, you might want to add Nivkhs to the list of Palaeo-Siberian populations that possess a substantial haplogroup Q component (but it might as well be haplogroup R1b for all we will ever know).


You're probably right Ebizur. Anyhow, P(xQ,R1) has been detected in strangely high ammounts (3/200) among Natives of NE North America (Bolnick 2006). It is way too high a percentile to be claimed as part of the post-colonial European ancestry and it's extremely unlikely it arrived there from South Asia (specially as there's no other South Asian exclusive clade in that sample - or any other that I know of). There are many incognites regarding all the tree of haplogroup P and it's possible that Siberian peoples have some answers in their genes. Sadly, as you say, there's not a big interest in studying them properly.

...

So in fact you believe humans behaved at that early stage totally differently from the way they have behaved ever since. Sounds suspiciously like belief in Adam and Eve to me.

Terry: I don't believe anything. It's you who is holding belief before and even against the hard data. I'm just saying that your hypothesis, no matter how natural and logical it may sound to you, is clashing once and again against a wall of, I presume frustrating, negative data.

No there's not. Almost any measure you like to apply to mutation rates puts at least several thousand years between the two haplogroups' origin.

I don't believe in TRMCA either. In fact I find terribly suspicious that the ages appear to be so different. I am 99% sure that they are missing something essential - if not just building on thin air.

you are convinced, in spite of "no evidence, pseudoscience, beliefs before facts" that for the last 200,000 years the Northern Hemisphere's atmosphere has been completely different from that of the Southern Hemisphere.

I have never said that. It's about the same ammounts of N, O2, CO2 and H2O, right? It is you who is making unreal claims on how the atmosphere is here, based on whatever your personal impresions that indepedent observations (mine but surely also many others) cannot confirm.

Your wish to believe this arises simply so that you can say that white skin evolved in response to this strangely different atmosphere.

Nope. I have always argued that light pygmentation is a common phenomenon of loss of function once people left the Tropics (UV radiation became much less menacing, as it arrives in a steep angle around the year).

I only say (along many others, I did not invent this theory, believe me) that the particular climatic conditions of Europe, specially NW Europe may have favored the radicalization of such trend giving notably higher apportions of extremely pale skin, sometimes with little or no ability to tan (quasi-albinism) and rosy cheeks, all them very conevenient to capture UV rays in pretty dark climates like those of Atlantic or Northern Europe.

It is a geographical fact that Glasgow or St. Petersburg are at the latitude of Nunavut and Yakutia (and "sunny" Spain or Turkey are at the latitude of New York or Tokyo) and that no other continent on Earth supports such relatively warm cimates at such extremely high latitudes. This alone is already a good explanation. But it's not just latitude: it's also cloud cover, wich is rather persistent in all Oceanic Europe throughout the year.

Is this enough to explain extreme depygmentation in some Caucasoids, notably those living in such extreme latitudes? Surely yes, specially as it affects such a decisive building block as vitamin D, whose lack does not just cause rickets but also damages the brain.

But I have no particular faith in that: it's just a very good explanation that works very well however you look at it.

Africa's quite a big continent, even if we just consider the stretch from South Africa to east Africa.

Sure, Africa is a continent in the full sense of the word, unlike Europe, which is just a large peninsula of Asia. So?

terryt said...

"In fact I find terribly suspicious that the ages appear to be so different".

Is there any reason at all why you should you be suspicious? Or do you not understand how genes move through populations? Even after all Ebizur's hard work.

"Sure, Africa is a continent in the full sense of the word, unlike Europe, which is just a large peninsula of Asia. So?"

Way back you claimed: "the overlapping male/female area of most likely origin is south of The Horn: Middle and Southern Africa only".

You're unnecessarily greatly oversimplifying the situation if you're narrowing the region of origin down to such a small region. How many individuals do you believe there were in this 'original' population?

Maju said...

Is there any reason at all why you should you be suspicious? Or do you not understand how genes move through populations? Even after all Ebizur's hard work.

You are basically claiming that the female half of humankind spread all around first and that the male half arrived a lot later. It would not apply to this or that case but to all the planet, even Bushmen. I make no sense of it, really.

Also the datations for Y-DNA, particularly, often appear to make no sense with archaeological data at all.

I have discussed this matter many times before. Use google for a more extended explantion(s).

You're unnecessarily greatly oversimplifying the situation if you're narrowing the region of origin down to such a small region.

So Africa is "a big continent" but half Africa is "a small region". Can you at least be consistent with yourself?

terryt said...

"You are basically claiming that the female half of humankind spread all around first and that the male half arrived a lot later".

No I'm not. If you still can't see how mtDNA haplogroup L0 and Y-chromosome haplogroup A need not have been near neighbours, or even contemporaries, there's obviously some basic aspect of population genetics you are yet to understand. Until I work out what that aspect is, and then explain it to you, we're wasting our time.

Afriqash said...

I think any reasonable person who contemplates the complexity of human behaviour and the multiple forces often unpredictable that are at play in making human history will agree that that Y-DNA dating is more of a compass than a clock.

Maju said...

I think any reasonable person who contemplates the complexity of human behaviour and the multiple forces often unpredictable that are at play in making human history will agree that that Y-DNA dating is more of a compass than a clock.

Great, beautiful sentence. And an excellent synthesis. :-)

terryt said...

And so is mtDNA dating, but there seems to be a relative consistency in all dates published so far. And my point is merely that there is no reason why the two haplogroups should correspond in either time or point of origin, because of "the complexity of human behaviour and the multiple forces often unpredictable that are at play in making human history".

Maju said...

But do you agree that the coalescence event (or period) of mtDNA M and N and of Y-DNA C, D and F should be roughly the same? Guess you don't but most people do.

It's hard to explain OOA in terms of several migrations. Even if someone can think that the time of the OOA epysode may need to be moved back in time (I suspect that, for example) and that the hyper-simplicity of the fast coastal migration model may need some fine tuning, it's hard to see two or more migrational events either in genetics or archaeology. In the case of genetics it's quite clear that all Eurasian+ clades spawn from a single African lineage (L3 and CT respectively) and that would not probably be the case if there would have been several OOA epysodes.

That basically means (with extremely high probability) that all people that then crossed to Eurasia (no matter if they did by Bab el Mandeb or by the Sinai peninsula, no matter if they did before or after Toba) must come from the same ancestral African group. They do not belong to Y-DNA A nor B and they do not belong to any other of the many mtDNA L lineages present in Africa.

So, as far as I can see, a single origin in Africa for all the Eurasian (and some African too) lineages is self-evident and hardly questionable. It is very likely that in that ancestral population L3 and CT were fixated, though maybe had already began to branch out. It is most unlikely that other lineages were present at more than anecdotical apportions as well.

terryt said...

"But do you agree that the coalescence event (or period) of mtDNA M and N and of Y-DNA C, D and F should be roughly the same?"

Not necessarily. And what about Y-chromosome E? I agree with your comment, "that the time of the OOA epysode may need to be moved back in time" but that doesn't necessarily mean "two or more migrational events". The population simply expanded its geographic range: it spread across virtually all of Africa, across the Middle East and into India. One continuous interbreeding population.

Agreed the mtDNA of the element of this population outside Africa must have involved at least two variations of L3 (probably more than two, just selected down to M and N in the early days).

But the Y-chromosome that went out with the first mtDNA was not necessarily some ancestor of C-T. The C-T expansion is extremely complicated and looks to be independent of the early mtDNA, although originating in broadly the same region.

The evidence shows that C-T broke into two lines which then each broke again into two more lines. Try to sort out a simple explanation for the distribution of these four lines! Or try to correlate their distribution with any mtDNA expansion. The Y-chromosome expansion certainly was no simple single migration. It was long and complicated, and could well have occurred after the mtDNA lines M and N had left Africa. It simply spread through the already expanded population, replacing (not necessarily rapidly) any pre-existing Y-chromosome variants in the, by then, widespread population.

The DNA evidence indicates that the expansion beyond the Africa/India arc, on the other hand, does involve the surviving modern mtDNA and Y-chromosome lines largely moving together.

Maju said...

... but that doesn't necessarily mean "two or more migrational events".

What is exactly what I am saying.

And what about Y-chromosome E?

It's just the remaining of DE, actually of all CT, that coalesced in Africa into a single haplogroup (with very few located upstream exceptions). It also means, IMO, that the founder group was small and geographically homogeneous untill E began expanding.

But the Y-chromosome that went out with the first mtDNA was not necessarily some ancestor of C-T.

Could prefectly be CT* ("*" meaning close to the root here) or even DE* and CF*. Based in modern distribution, I'd say that CT evolved into CF already in Asia (no native CF south of the Sahara, right?) but CT evolved into DE in Africa (DE* only found there). So the most likely migration event in the Y-DNA side was something like many CT* (ancestors of CF clades) and a few DE* (ancestors of D). While the remaining group was like many DE* (ancestors of E) and a few CT* (that were drifted out).

C and D are clear "living fossils" of the early "fast coastal migration". They were drifted out by F elsewhere. Maybe you can speculate with a second wave "out of South Asia" because of that reason but not sure (K and C appear together in so many places). In any case be sure to include K in the equation. And K is the main genetic reason I need to push back the dates of Y-DNA, cause K must have been there c. 70-50,000 BP: in northern India probably.

Of course this is necesarily a model: such processes happened along many generations and can hardly be described in simple terms.

The C-T expansion is extremely complicated and looks to be independent of the early mtDNA, although originating in broadly the same region.

Not necesarily. Just stop thinking in M and N (and even R) as separate groups of people. In al places they are mixed, albeit in different apportions. West Eurasians have M1 and all other Eurasians have some N clades. They were together all the time. M appears dominant or balanced though, except in the particular case of West Eurasia, that obviously implies a major founder effect. A founder effect where M was rare and R aboundant. It certainly means a relatively late moment in Eurasian colonization (because R appears developed and even already branched out).

The evidence shows that C-T broke into two lines which then each broke again into two more lines.

It's that way: when one group emigrates, "lottery" causes (normally) one clade to get fixated in the emigrant group and one to do the same in the one remaining behind. Unless the groups are very large, what was not the case in Paleolithic bands. But these groups can remix all along, they can be just a few hundred kilometers away... or tens of thousands. The short distance case was surely much more frequent but sometimes dissimulated by further drift and fixation.

But you must mean CF, not CT. CF may well have split in India, with F being more important in the North (Narmada-Son-Ganges axis) and C maybe in the south (Krishna river and purely coastal axis) - but not sure. D was more common if anywhere in some avant-guard group. In any case, at least C and F-derived K met at the Ganges Delta (possibly) and migrated together to many places in Eastern Eurasia c. 60-50,000 BP. Later C (and the even rarer D) was outdrifted in South Asia (with some exceptions remaining till today) but that was "after K".

The Y-chromosome expansion certainly was no simple single migration.

In my model it was pretty simple, as described above. We just tend to see C and F as parallel but, due to drift or maybe to some more dynamic group that caused a serial founder effect, F is mostly K (much like N is mostly R, though they are not totally parallel). In any case, for me, K and C migrated together to SE Asia and beyond. D may have been marching ahead of them though (not really sure about D). And they did it with mtDNA M and N (both).

Whatever other model would need some archaeological support that does not exist. This model fits best with the archaeological data we do have: single wave roughly along the coasts (not necesarily excluding parallel inland shortcuts) and it's the one that best explains the survival of C and D specially in the East (and its near absence in south and west Eurasia - due to drift and/or internal migrations).

I don't understand the final lines of your reasoning. Yet the "evidence" indicates often what we want or are able to see. Beware of optical illusions.

terryt said...

"In al places they are mixed, albeit in different apportions".

No they're not.

"They were together all the time".

If their present distribution is a result totally of drift it would be patchy with each haplogroup widely randomly scattered, not related to well-defined geographical regions as the present distribution demonstrates. And you contradict your own statement later: "when one group emigrates, 'lottery' causes (normally) one clade to get fixated in the emigrant group".

"actually of all CT, that coalesced in Africa into a single haplogroup".

All we can say about CT is that it is extinct, as far as we know. Later Y-chromosome expansions have replaced it completely. It may never have been numerous but this raises the question of what men were those contemporary mtDNA haplogroups having their children with?

Likewise the CF Y-chromosome is extinct and the DE one virtually so. I accept that haplogroups C, D, E and F belonged to a single population. But it was a widespread one, stretching from North Africa to Northern India and including at least much of Central Asia. And the population probably included other Y-chromosome haplogroups as well.

Population expansions carry Y-chromosome and mtDNA haplogroups with them, not the other way round: haplogroups leading the population expansion. But men, especially, often move from their natal group.

They have children with neighbouring groups. Their grandchildren could well be fathered by men from these neighbouring groups. Their Y-chromosome is no longer in the group but their grandchildren still have one quarter of the incoming male's genetic material, his nuclear DNA. If there is an advantage in any of the genes he introduced then selection acts on that gene. It, too, spreads into neighbouring groups. The expansion of the gene will then tend to carry along any women (and men) who already have it. The mtDNA will follow along behind the gene, but not necessarily to the extreme geographic extent of that gene. Mountains, deserts, forests and open water have always slowed down and deflected human genetic expansions.

If we examine the evidence from this perspective we can easily understand our evolution right from back before Australopithecus.

By the way. Including Y-chromosome K and his descendants along with mtDNA R and her descendants is to introduce a red herring. They are both the product of a secondary set of expansions after those than gave rise to distibution of earlier haplogroups. As you concede, "It certainly means a relatively late moment in Eurasian colonization".

Maju said...

No they're not.

Where not? Except maybe in some random corner subject to extreme drift?

If their present distribution is a result totally of drift it would be patchy with each haplogroup widely randomly scattered...

Not just of drift but of a process in which drift was definitely involved, specially at the beginning.

All we can say about CT is that it is extinct, as far as we know.

And that there is none of it in Africa that cannot be attributed to "recent" backflow (specially North Africa).

Later Y-chromosome expansions have replaced it completely.

It just drifted into C and F. C and F are just CT with some extra SNP mutations, these may well have happened at different spots (for example Karnataka and Gujarat - just a guess). Every haplogroup has several defining SNP mutations, that means drift towards one of the many possible (and maybe at some moment extant) alternative downstream branches that are not anymore. It can indicate that the population remained together and did not split, while the presence of several survivor branches may mean a real split, even if maybe just temporary.

Take Y-DNA P for example. It's defined by some half dozen SNP mutations if not more. Q and R are also defined by a long stream of mutations each. A lot of possible new clades that never got consolidated in all that. Or better: that they did got consolidated but only in the form of Q and R, in fact mostly in the subclades of these haplogroups (as there is little R* and Q*).

Among the many "daughter" lineages only a few, if any, became consolidated in the mid run. The rest may have existed but were "reabsorbed" ("outdrifted").

It may never have been numerous but this raises the question of what men were those contemporary mtDNA haplogroups having their children with?

Varied. They did not make genetic tests back then, you know. All women were similarly valid potential mothers, it just depended on who was available.

Likewise the CF Y-chromosome is extinct and the DE one virtually so.

In fact the original DE is also extinct most likely. The found DE* is just another or several other downstream haplogroups, just that they have not been studied with sufficient depth. What DE* tell us is about diversitiy in that macro-clade, not that the actual root haplogroup of maybe 80,000 years ago (or more) is still alive with no further mutations. That interpretation would be simplistic and wrong.

DE is nevertheless alive in DE*, D and E, and all their subclades. You cannot expect haplogroups to remain frozen in time. They are all the time evolving in downstream direction (at least for the time spans we manage here).

I accept that haplogroups C, D, E and F belonged to a single population. But it was a widespread one, stretching from North Africa to Northern India and including at least much of Central Asia.

Do you think that such a huge territory can be described as "a single population"? No way! Much less in the Paleolithic.

Population expansions carry Y-chromosome and mtDNA haplogroups with them, not the other way round: haplogroups leading the population expansion.

What I probably said/meant was that haplogroup diversification can mean population split. In a sufficiently large population maybe not but in small hunter-gatherer type societies almost necesarily.

We see the diversification (or lack of it) and can suppose in consequence a split or stability.

But men, especially, often move from their natal group.

Or women, this varies (patrilocality and matrilocality).

If we examine the evidence from this perspective we can easily understand our evolution right from back before Australopithecus.

That's a very daring claim specially with such slippery foundations as your "theories".

By the way. Including Y-chromosome K and his descendants along with mtDNA R and her descendants is to introduce a red herring. They are both the product of a secondary set of expansions after those than gave rise to distibution of earlier haplogroups.

I don't see that they are, could be or much less must have been a second expansion. All would seem to stem from South Asia and their distribution rather suggests that they mostly expanded coupled with other clades. Also we do not have any archaeology that would appear to demand, much less support, such hypothetical second expnsions.

As you concede, "It certainly means a relatively late moment in Eurasian colonization".

Not sure when I said that but I believe it was in relation only with the colonization West Eurasia. Colonization that, following archaeology is not much later than in the rest anyhow (and could even pre-date some East Asian areas, for instance).

What I mean, what I think... is that after the arrival to South Asia, there was a long coalescing period and then a expansion from there. And that the expansion into SE Asia may have been the first one. But the differences between East and West Eurasia may just have been caused by where they stemmed from in South Asia - something like Pakistan or Bangla Desh, so to say: opposite extremes in the E-W axis, and possibly holding then already different "alchemy": more M in the East, as well as Y-DNA C and D , more R in the west, and no C nor D Y-DNA.

The East-West Eurasian "duality" in haploid genetics (quite relative, as there are many elements in common too) may well just mean a similar "duality" in South Asia prior to expansion. But the shared elements (Y-DNA F and K, all major mtDNA branches) also imply certain genetic unity still at that stage anyhow.

Only a handful of derived branches (basically those derived from P and NO) mean "secondary expansions" clearly. Secondary expansions that happened mostly through the steppary belt and Siberia.

terryt said...

"Only a handful of derived branches (basically those derived from P and NO) mean 'secondary expansions' clearly".

Why would these branches suddenly represent some different process? Why aren't earlier apparently 'derived' haplogroups such as mtDNA C,Z,D, G,B,F,HV, etc and Y-chromosomes L-T also products of 'secondary expansions'? Or is there some reason why we must regard these earlier branches as being different?

"Do you think that such a huge territory can be described as 'a single population'?"

Aren't we talking about a species here?

terryt said...

And I liked this recent comment from John Hawks' blog,

"We've got people out there who are talking about biblical models of human migration, like Noah-and-the-Flood level bottlenecks".

Maju said...

Why would these branches suddenly represent some different process?

Cause they are very young in the whole tree and most of their distribution is rather in the vast, once semidesertic, northlands.

Besides they are extremely widespread to mean otherwise: P from Patagonia... well... to Patagonia again (but from the other side), NO from New Zealand to Finland (across continental Asia mostly).

They do not follow the pattern of the rest, not even remotely. Also they tend to be alone, not mixed with other clades or mixed with different, locally older, ones (like IJ in Western Eurasia... but C3 in America, things like that).

They are just too young to be so widespread unless you follow this model of secondary expansion. It just cannot make sense otherwise, even if they would happen to be 30 or 40,000 years old. They should not be old enough to be in the first wave as such.

Don't know how to explain better but I can't make sense of them otherwise.

Why aren't earlier apparently 'derived' haplogroups such as mtDNA C,Z,D, G,B,F,HV, etc and Y-chromosomes L-T also products of 'secondary expansions'?

Not sure what you mean, CZ and D certainly took part in secondary expansions, but mostly within the Native American and later Turco-Mongol ones (not sure right now about all the details). HV too, within the West Eurasian processes, specially Gravettian and derived cultures (probably already as H and V as separated claes). So, yes, I may have skipped the mtDNA clades.

But what's "L-T"? There's no such clade. L is a localized Southern Asian clade (ok, it may have expanded with Neolithic and post-Neolithic waves but within South Asia basically), T (aka K2) is only West Eurasian and is rare - it doesn't really seem to matter much.

What are you trying to say?

I don't make much sense.

Or is there some reason why we must regard these earlier branches as being different?

They are not "earlier branches". They are "late" in the overall tree. We were discusing the big macro-haplos and now you come with lower tier clades that have basically regional distributions. What's your point. Can you compare them with N and M? No. Can you compare their regional distributions with the global ones of P and NO? Nope again.

Mixing apples and oranges, I think.

Aren't we talking about a species here?

I'm just flippant with these kind of questions. Species and population are different things. Population means group that interacts on regular basis, it's not a strict term nor a strictly biological one but rather a practical one. Probably New Zealanders and British share a lot genetically, yet they are two different populations because of geography. Hope you get the point.

Maju said...

And I liked this recent comment from John Hawks' blog,

"We've got people out there who are talking about biblical models of human migration, like Noah-and-the-Flood level bottlenecks".


That would be ridiculous. But certainly, each time a group parted to colonize a new territory, they had a bottleneck-like effect. It's called founder effect in fact but the results are very similar: greatly decreased genetic diversity.

That happened with the OOA epysode but surely also later. For instance the colonization of West Eurasia appears as one of those pseudo-bottlenecks.

Drift also causes bottleneck-like effects in haploid lineages. But true bottlenecks... one would have to think of somthing of the dimensions of Toba mega-explosion - at least.

terryt said...

"But certainly, each time a group parted to colonize a new territory, they had a bottleneck-like effect".

And isn't that precisely what we see with the various haplogroups? They reach fixation in some group that had parted from some other group. surely this process was going on right back to way beyond the time when existing haplogroups first appeared?

"We were discusing the big macro-haplos and now you come with lower tier clades that have basically regional distributions. What's your point. Can you compare them with N and M?"

Yes. Surely the distribution of the 'big macro-haplos' was a product of the same process as that of the 'mini-haplos'. It's just that the macro-haplos have been around for so much longer and so they have become more widespread.

"Cause they are very young".

Again, aren't more ancient divisions caused through the same process of diversification and spread?

"They do not follow the pattern of the rest, not even remotely".

No, because they are more recent. On the other hand more ancient haplogroups from the same regions may have moved with them at times.

"Also they tend to be alone".

Earlier you asked, "Where not? Except maybe in some random corner subject to extreme drift?" I think you've just answered your own question.

"They are just too young to be so widespread unless you follow this model of secondary expansion".

That's what I've been trying to tell you. The haplogroups did not all spread as one. And there have not been only two expansions. There have been a whole series from a whole lot of different regions. It wasn't a simple process, but when you realise that the present haplogroup distribution is simple to undersatand.

"But what's 'L-T'?"

It's my shorthand way of grouping all K's descendant haplogroups. All of them the product of 'secondary' expansions, some widespread, some relatively local.

"Species and population are different things".

But species are made up of populations. Therefore the same rules apply.

Maju said...

It's just that the macro-haplos have been around for so much longer and so they have become more widespread.

Depending what you mean. You are making strange comparisons that only confuse the picture. And this tendency of you to ask captious questions on only vaguely related matters to then turn back with something that appears to be but is not anything similar is kind of annoying, sincerely. That kind of tactics are normal I guess in the Parliament where means are secondary to goals, but, when discussing seriously over facts with scientific attitude, means are all and goals don't matter.

MtDNA M and N (and R) are very much mixed everywhere. Y-DNA NO and P very rarely. That's why NO and P appear to have originated and expanded separately, while N and M appear to have expanded together from the beginning. Certainly much more side by side than the other example. Same applies to C and K for instance (together, mixed nearly everywhere) but not to H and B for instance.

It's my shorthand way of grouping all K's descendant haplogroups.

Just K would be much better.

All of them the product of 'secondary' expansions, some widespread, some relatively local.

Nope. L is local, M is local, S is local and T is local. Only P and NO are widespread and we can talk properly of independent expansions. The distribution of K-derived clades is part of the K expansion which must have gone hand by hand with that of C in the east and of other F clades in the West, IMO. Maybe I'm wrong but in any case the process corresponds to undifferentiated K, as all derived clades are local, except P and NO.

What archaeological process would you associate with the supposed independent expansion of K? I can think of no one but maybe I'm missing something.

But species are made up of populations. Therefore the same rules apply.

Again "mareando la perdiz" (untranslatable Spanish expresion that means going forth and back without any sense, getting the rest dizzy but deeply unsatisfied).

You said that the Africa to India arch were a single population. I said "imposible". You shrugged and said "species". I said "it's not the same at all". And now you come with that happy sentence that means absolutely nothing, like claiming that, as galaxies are made of atoms, the same rules apply... totally senseless.

Stop that game or I'll have to stop discussing with you totally.

terryt said...

"MtDNA M and N (and R) are very much mixed everywhere".

Rubbish. There is no reason at all why you can include R with any original N expansion. Sure, R is a product of N's expansion but R, along with lines B, F, HV, P, JT and UK, are part of a separate, and later, expansion. Once you exclude these haplogroups M and N have virtually no overlap at all, except in SE Asia.

"L is local, M is local, S is local and T is local. Only P and NO are widespread and we can talk properly of independent expansions".

But all derive locally from K's expansion. OK, some of them didn't expand very far from the region the mutation first appeared in, but NO and P sure did.

"I can think of no one but maybe I'm missing something".

Invention of the boating necessary to cross Wallace's line is a possibility but you seem to 'believe' that humans left Africa with that technology, so the possibility doesn't come into your conciousness.

It seems you are fixed on India as being a major region of dispersal, even though all the evidence suggests the route east from India to SE Asia has always been extremely difficult. It has always been easier for humans to move from SE Asia to India, the opposite direction.

I believe the reason you are fixed on India is because, indeed, virtually all European haplogroups originated there. But these are universally a product of migrations long after any 'original' expansion from Africa. A lot had happened before they'd been able to even think about entering Europe.

Incidently, why is it impossible "that the Africa to India arch were a single population"?

Maju said...

Rubbish. There is no reason at all why you can include R with any original N expansion.

Same as K in relation to F. They both appear to have sprung from South Asia along the expansion of the other clades.

Once you exclude these haplogroups M and N have virtually no overlap at all, except in SE Asia.

· Haplogroup N(xR)

· Haplorgoup M

· Haplogroup R

You can see they do overlap in many other areas too (West and East Eurasia overall, Oceania, America - not shown). Basically everywhere.

But all derive locally from K's expansion. OK, some of them didn't expand very far from the region the mutation first appeared in, but NO and P sure did.

We agree in this (glad about it).

NO and P nevertheless may well have coalesced not so far from South Asia anyhow, before re-expanding in their own separate process. A likely area of coalescence for P would be around Afghanistan (Pakistan, Central Asia...), while the area of coalescence of NO would be in southern China possibly or not too far away.

Invention of the boating necessary to cross Wallace's line is a possibility but you seem to 'believe' that humans left Africa with that technology, so the possibility doesn't come into your conciousness.

That is, IMO, your personal bulb-over-head syndrome: you build your whole model on that assumption of yours, what is like recosntructing biological history based on religious beliefs.

I see no particular reason to argue for a fast coastal migration without boats. And I see absolutely no reason to think that basic boating was not among the many abilities humans already had before leaving Africa. In fact, even H. erectus (generically speaking) may have known something about it, as southern Iberia was colonized across Gibraltar strait in the Acheulean period (while the main migration arrived overland from West Asia - what anyhow also implied crossing the Bosphorus).

This does not deny that crossing from Indonesia to Sahul implies a greater feat. But precisely for that reason, and because Australia appears to have been colonized so early, I need to think that humans were already pretty much good boaters before doing that. Much like Red Eric, da Gama, Colombus and Elkano are the late product of many milennia of European sailing, not the origins of it.

It seems you are fixed on India as being a major region of dispersal...

It seems quite obvious from DNA. Also what we know of Indian Prehistory can perfectly fit with that (close typological connections with Africa, very early stone blade tools, likely continuity before and after Toba...). Finally it's the perfect geographical intermediate zone.

...even though all the evidence suggests the route east from India to SE Asia has always been extremely difficult. It has always been easier for humans to move from SE Asia to India, the opposite direction.

If you can walk in one direction, you can in the opposite. That's silly.

The route is not that bad anyhow: we have the famous coastal route (boating was surely needed to cross the Ganges delta, the main barrier) and we have the Brahmaputra route as well. We even have genetic "living fossils" in the Andaman islands (right in between) telling us that Y-DNA D went by that route most likely.

Said that, I imagine that SE Asia was an important secondary node that is sometimes overlooked.

I believe the reason you are fixed on India is because, indeed, virtually all European haplogroups originated there.

No. I'm not that much Eurocentric, I beleive. I have looked at the whole picture before agreeing with this model (pretty much mainstream). I don't agree with some of the mainstream assumptions though (datation for instance).

Incidently, why is it impossible "that the Africa to India arch were a single population"?

Because such a long distance is not transited in a day, nor in a week nor probably in months. Certainly not with the technology they had.

I see that it is pretty much evident that once Bab-elMandeb was crossed (and moreso after arriving to South Asia, crossing mountains and rivers, deserts or semidserts, and many thousands of kilometers) the Eurasians became a different population than their African relatives. This does not exclude further interaction totally but implies that it was very much limited. It's not any "a priori" belief but a logical conclusion that fits with all the data I know.

It's like Britain and New Zealand. In fact with the technology available then it was infinitely more difficult to transit the distance between India and Somalia than now it's for you to travel to London. It could be done, sure, but it needed several generations probably (plus it's hard to see how they could know about each other at all after time had erased collective memories).

Again I find incredible that you even ponder that. Why do you think they could even have any relation at all?

terryt said...

"And I see absolutely no reason to think that basic boating was not among the many abilities humans already had before leaving Africa".

You accept, then, that Europeans were hopelessly primitive. It took them until just 10,000 years ago to occupy islands in the Mediterranean, even though they had possessed boats for at least 50,000 years before this.

"We agree in this".

But it's obvious from haplogroup diagrams that mtDNA haplogroup R's expansion is, like Y-chromosome K's, a later expansion than any 'original' expansion. And yet you still insist on including this later expansion within N's distribution, simply so that you can say M and N's distribution overlap. I invite you to examine N's distribution once you have taken out R's. You can then put R's origin within any limited region you wish to but you will still see the same pattern, virtually no overlap with M except for the Far East.

"I see no particular reason to argue for a fast coastal migration without boats".

Experience has shown me I cannot assume I know what you mean by any comment, but this contradicts your statement, "imposible" in reply to "You said that the Africa to India arch were a single population". Why could they not be if the expansion from Africa to Australia was as rapid as you seem to 'believe'. Besides which you admit, in relation to similarity between African and Indian populations, "It seems quite obvious from DNA. Also what we know of Indian Prehistory can perfectly fit with that (close typological connections with Africa, very early stone blade tools, likely continuity before and after Toba...)". But they're not the same population?

"I'm not that much Eurocentric, I beleive".

Maju, you are totally Eurocentric. You seem to have no idea whatsoever of the geography of SE Asia.

"If you can walk in one direction, you can in the opposite".

I'm absolutely stunned that you still believe at this stage that we are talking about walking.

terryt said...

More on "We agree in this (glad about it)".

It's obvious that Y-chromosome lines L, M, N/O and P developed in various regions after K's expansion. They were spread from India to New Guinea and probably each developed at extremes of K's distribution: P in Pakistan, L in India, N/O in SE Asia and M in New Guinea. Some of these subsequently expanded again and, in turn, broke into regional variations.

Similarly for mtDNA haplogroup R. Daughter lines presumably originally formed locally within the confines of R's expansion: P, F and B in the east and H/V, U/K and J/T in the west. Again, some of these subsequently expanded once more.

Now, just carry the same process back in time.

Maju said...

You accept, then, that Europeans were hopelessly primitive. It took them until just 10,000 years ago to occupy islands in the Mediterranean, even though they had possessed boats for at least 50,000 years before this.

Yes, we Europeans are hopelessly primitive (if that makes you happy).

Seriously: we are talking of mere canoes or rafts, not high-seas ships. It's hard to see how you can navigate to these islands with a simple canoe. Once sailing was invented these islands were eventually colonized one after the other.

Also it's very possible that the colonizers of Europe (and West and Central Asia before it) were much less interested/knowledgeable of boating than those that followed the famed "coastal route" to Eastern Eurasia. After all, it's very likely that their migration happened via inland routes (excepting the Bosphorus/Dardanellos, both extremely narrow straits).

We have indirect evidence that they were able to cross the straits of Messina and Gibraltar but these are now not wider than 15 km, while reaching Cyprus or Corsica would have needed to navigate very much larger open seas areas. Basically it seems Western Eurasians did not dare to get into unknow waters with wahtever means they had, that they only crossed where the opposite shore was visible and relatively easy to reach (yes, Morocco can be seen from Spain and vice versa, at least in good weather and the strait can be crossed even with a mere surfing board if the weather allows and the swimmer is in good shape).

So they did have boats, even if they never adventured into the high seas.

"We agree in this".

But...


Couldn't be so good. Ah!

But it's obvious from haplogroup diagrams that mtDNA haplogroup R's expansion is, like Y-chromosome K's, a later expansion than any 'original' expansion.

Can you add the "Y-DNA"/"mtDNA" tags? For a moment I thought you were talking of Y-DNA R.

And yet you still insist on including this later expansion within N's distribution...

I showed you the distribution of N(xR) and it's also widespread through all Eurasia, Oceania and America. Hence there's no particular reason to think that the expansion of R is something necesarily detached, moreso as both appear to spawn out of South Asia.

If you paid attention this paragraph above would have been unnecesary.

I invite you to examine N's distribution once you have taken out R's. You can then put R's origin within any limited region you wish to but you will still see the same pattern, virtually no overlap with M except for the Far East.

I have already invited you to do the same, yet you seem blind. N(xR) exists in all Eurasian regions, Oceania and America.

Just to put things straight:
· N* is found (at least) in India, Australia and Japan.
· N1, N2, I and W are found in West Eurasia.
· S is found in Australia.
· A is found in East Asia and America.
· Y is found in East Asia.
· X is found in West Eurasia, Altai and America.

Only SE Asia and New Guinea are odd in this regard but, in the case of SE Asia surely because of limited research only (in remote New Guinea is surely because of founder effects).

Experience has shown me I cannot assume I know what you mean by any comment, but this contradicts your statement, "imposible" in reply to "You said that the Africa to India arch were a single population". Why could they not be if the expansion from Africa to Australia was as rapid as you seem to 'believe'.

First I think that the "rapid" expansion counts only from India. And, second, no matter how "fast" it was, it surely implied several milennia and many many generations. "Fast" here is more like the snail concept of "fast", not like a rocket.

Besides which you admit, in relation to similarity between African and Indian populations, "It seems quite obvious from DNA. Also what we know of Indian Prehistory can perfectly fit with that (close typological connections with Africa, very early stone blade tools, likely continuity before and after Toba...)". But they're not the same population?

They cannot be. New Zealanders also have (roughly) British culture yet they are not the same population anymore. Even with modern technology the rate of genetic exchange between both antipodic islands must be now very limited. Two distinct populations therefore, and this is much more clear for early Eurasians and their African cousins.

"If you can walk in one direction, you can in the opposite".

I'm absolutely stunned that you still believe at this stage that we are talking about walking.


Walking or boating (or probably both) it doesn't really matter. We are talking about coastal travel, not high-seas navigation.

More on "We agree in this (glad about it)".

It's obvious that Y-chromosome lines L, M, N/O and P developed in various regions after K's expansion. They were spread from India to New Guinea and probably each developed at extremes of K's distribution: P in Pakistan, L in India, N/O in SE Asia and M in New Guinea. Some of these subsequently expanded again and, in turn, broke into regional variations.


More like L in Pakistan or Afghanistan (if we look at its modern distribution at least). NO maybe more like in Southern China.

The problem is not this anyhow but where and when did K originated and how it expanded. I argue for expansion along with C, at least to the East, as both haplogroups are commonly coupled and because it's hard to find an archaeological process for two waves. Also it would be extremely hard to explain two crossings into Sahul prior to the developements of high seas sailing (one is already hard to explain, two would be such an unlikely coincidence that we can rule it as impossible).

Similarly for mtDNA haplogroup R. Daughter lines presumably originally formed locally within the confines of R's expansion: P, F and B in the east and H/V, U/K and J/T in the west. Again, some of these subsequently expanded once more.

Yes but this happened along the expansion of N and M most likely.

Now, just carry the same process back in time.

Why? Notice that very few mutations separate these clades (R from N, both N and M from L3-root). It means that all happened in a short evolutionary time. There is just not much "before". And there is no archaeological evidence for serial expansions "out of India" (though in some cases there may have been parallel ones).

Finally, if you think with me that OOA pre-dates Toba and that the coalescence of the major clades happened in South Asia before and after Toba, there is more than enough before for all the high level differentiation to belong to internal South Asian processes prior to the expansion elsewhere. South Asia is big, not in vain it's known as "the subcontinent".

terryt said...

This is completely ridiculous.

"Basically it seems Western Eurasians did not dare to get into unknow waters with wahtever means they had, that they only crossed where the opposite shore was visible and relatively easy to reach".

Perhaps, yet by at least 50,000 years ago in SE Asia humans had crossed several expanses of water where they couldn't see the opposite shore . Doesn't make sense.

"Hence there's no particular reason to think that the expansion of R is something necesarily detached".

Why then do all diagrams of mtDNA haplogroups show R
diversifying separately from N. Does that mean nothing? If you paid attention you would understand what it means.

"Just to put things straight".

If you were prepared to look you would see these haplogroups are basically spread from West Eurasia right across Central Asia to the Far East and thence to both Australia and America. The few variations found in India are most obviously accounted for by relatively recent immigration from east or west.

"Only SE Asia and New Guinea are odd in this regard".

Why? Surely R fills that space and so exactly fits the above scenario.

"New Zealanders also have (roughly) British culture yet they are not the same population anymore".

They still share many Y-chromosome and mtDNA haplogroups though.

"We are talking about coastal travel".

How do you walk along the coast?

"Also it would be extremely hard to explain two crossings into Sahul prior to the developements of high seas sailing".

It's a problem with your scenario but not with mine.

"Yes but this happened along the expansion of N and M most likely".

Not necessarily so at all. Quite the contrary in fact.

"Why?"

Why would ancient processes be completely different from what happens today?

Maju said...

Perhaps, yet by at least 50,000 years ago in SE Asia humans had crossed several expanses of water where they couldn't see the opposite shore . Doesn't make sense.


They were two different populations to begin with. Maybe Eastern Eurasians were more seagoing. Anyhow this basically applies to proto-Australians, as the rest of the East Eurasian endeavours did not require such crossings.

So you are taking the strange case of Australians (and Papuans) and making of it a matter that should affect the whole world. We can't know how it happened. Maybe even there were then more islands that later sunk Krakatoa style. It's a very unstable area. Maybe the crossing required less than what you claim and in any case it was just one.

Why then do all diagrams of mtDNA haplogroups show R
diversifying separately from N. Does that mean nothing? If you paid attention you would understand what it means.


Separately or parallely? Maybe they just put them separate because of clarity.

I see what it means: that your reading of those graphs is shallow.

If you were prepared to look you would see these haplogroups are basically spread from West Eurasia right across Central Asia to the Far East and thence to both Australia and America.

That is just your opinion. There's absolutely zero evidence to support that theory.

Going to Australia in any case always demands travelling through SE Asia.

The few variations found in India are most obviously accounted for by relatively recent immigration from east or west.

Not sure what you're talking about exactly (very imprecise) but in any case I don't see it. R and M are both most diverse in South Asia, holding many rare exclusive high level ramifications. N is less clear but to go to Australia, it must have gone via South and SE Asia.

Why? Surely R fills that space and so exactly fits the above scenario.

You are the one insisting on N(xR). Don't be hypocrite.

[New Zealanders and Britons] still share many Y-chromosome and mtDNA haplogroups though.

But they are not the same population. Not anymore.

How do you walk along the coast?

*Eyes wide open* Walking. Coast means where land meets water, you walk by the land of course and navigate through the water. Unless you are now having great new ideas about amphibious boats and Jesus-like sea walking... (really no idea where are you going to come now).

"Also it would be extremely hard to explain two crossings into Sahul prior to the developements of high seas sailing".

It's a problem with your scenario but not with mine.


No. It's a problem in itself. Do you now that story about the donkey that accidentaly played a flute... Accidents like those don't happen twice.

Why would ancient processes be completely different from what happens today?

Technology, population density, civilization, social sturcture, economy...

Many reasons. Hunter-gatherer societies, cosmogony, technology... everything is very different from modern lifestyle. Do you think that if the white colonists of North America would not have got guns and railroads, not to mention the ability to move millions across the ocean and very developed farming allowing high population densities, they would have displaced the natives so easily? Impossible, unthinkable, ridiculous.

The processes of modernity are proper of this age and developement, the process of Neolithic and Metal Ages are proper of those stages of social organization and densities, and the processes of hunter-gatherers are those that belong to that socio-technological stage as well.

Get real.

terryt said...

"Coast means where land meets water, you walk by the land of course and navigate through the water".

So it's completely simple matter to walk right along the coast from Bilbao to Northern Portugal? Never leaving the seaside. I challenge you, or anyone else, to do it.

It continues to amaze me. You have reminded us all many times that the Paleolithic human population probably consisted of small, relatively isolated groups. And at such times drift and founder effect have their strongest influence. Yet you still insist that expanding groups contained a multitude of different Y-chromosome and mtDNA haplogroups.

It is most unlikely that any expanding group contained more than just one or two mtDNA and Y-chromosome haplogroups. Therefore it is most likely that mtDNA haplogroups H/V, U/K, J/T, B, F and P were yet to differentiate when mtDNA haplogroup R began its expansion. Those haplogroups almost certainly arose regionally from R after its initial expansion. Some of them, in turn, later expanded still further, and some stayed pretty much where they originated. If you can't see that then I've been wasting my time.

Maju said...

So it's completely simple matter to walk right along the coast from Bilbao to Northern Portugal? Never leaving the seaside. I challenge you, or anyone else, to do it.

With the reasonable margin (some kilometers maybe in some spots) sure. In fact there is a road all along the coast. Further inland mountains become hardly passable, though historically the main route was south of the mountains (main road of the Way of St. James, which actually pre-dates Christianity - but this route also has a coastal variant).

Anyhow you've chosen a very difficult coast, with many mountains looking right upon the sea. But yes: coastal travelling on foot is very much possible and surely common historically, specially when sailing was not available yet. I guess that boats would be used to cross estuaries and rivers instead of taking long deviations to the interior.

Notice anyhow that these migrations did not take place in one year not even one generation. They were moving step by step as need and curiosity demanded.

The most important feature of the coastal route is not anyhow its practicability (though it's not impracticable at all, specially using boats to cross estuaries and swamps) but its many resources in form of seafood. As long as the sea provides, people will be able to survive and continue their long transgenerational journey.

The GIS study anyhow showed that the coastal route around India was/is as good as the Krishna and Narmada inland riverine routes. I understand anyhow, even if the GIS study did not consider boating, that they did use boats both at the coast and along the rivers.

Yet you still insist that expanding groups contained a multitude of different Y-chromosome and mtDNA haplogroups.

Groups split and merge. I see no difficulty in this matter at all. Maybe one group went via the Narmada-Son route and another along the coast, meeting in Bengal and remixing before adventuring towards SE Asia. Or maybe both groups took the same route at different moments and still met at Bengal anyhow before daring to cross the huge barrier that is the Ganges delta. Or maybe the whole migration process was fast enough not to allow fixation of a single clade (drift happens but needs time). I can't really choose the most likely option but I do think that they left South Asia together and/or remixed in SE Asia (before jumping to Sahul and NE Asia).

It is most unlikely that any expanding group contained more than just one or two mtDNA and Y-chromosome haplogroups.

If the arrival to South Asia was followed by a demic expansion, then all clades should have expanded first of all. Maybe at different rythms but the expansion should have been general.

I do not take bands or tribes as isolated units, as you seem to do. Even if drift, in the long run, can cause fixation in one or few haplogroups, this doesnt mean that new "remix" bands would not again carry several clades within them. People have to make deals with their neighbours and, while sometimes the deal is mere war, more often the deal is peace and ethnic fussion. Otherwise how do you explain that all known historical demic expansions absorb so many pre-existent clades as they go?

Therefore it is most likely that mtDNA haplogroups H/V, U/K, J/T, B, F and P were yet to differentiate when mtDNA haplogroup R began its expansion.

That's pretty obvious to me. Even if Indian U2 is a back-migration, the "founding mothers" of West Eurasians must have been at least three (N, M and R). Wether they all arrived together or in different, yet close in time, moments is unknown to me. For practical purposes we can consider them a single migration event, the same that we consider that all Native American mtDNA haplogroups (five!!! - not counting ancient DNA M*, located in one fossil person) probably arrived at the same time (the exception might be X if anything, which could be more strictly associated with the only secondary expansion we know of: Clovis).

Following your method, Native Americans must have arrived in nothing less than five different waves! Instead it seems to me that Beringia allowed for all those clades to persist before the colonization process began.

Notice also that mtDNA appears much less susceptible to drift and fixation than Y-DNA. After all each woman can only have so many children. Otherwise we should not find so many different mtDNA clades in Europe (and in each of its regions), for example. Even such a homogeneous Y-DNA ethnicity as Basques, have many native mtDNA haplogroups, including extremely rare clades of U that must have survived since Aurignacian probably).

terryt said...

"Otherwise we should not find so many different mtDNA clades in Europe".

Following your method Europeans must have arrived in just one wave.

"Anyhow you've chosen a very difficult coast, with many mountains looking right upon the sea".

And you believe that's unusual?

"In fact there is a road all along the coast".

I'll bet you anything you like that there wasn't one in Paleolithic times. I'm also sure that there are stretches of the coastline where even experienced modern yachties, let alone Paleolithic mariners, are very reluctant to approach too closely to the shore. These same stretches of coastline are usually also those most difficult to pass along by land. Anyway unless someone is absolutely sure of what lies ahead it is most unlikely they would move simply along the coast. Perhaps Paleolithic man knew that ahead of them lay the lucky country, Australia.

"still met at Bengal anyhow before daring to cross the huge barrier that is the Ganges delta".

But that's just the first of many huge barriers. I'm sure you will benefit from a brief overview of the geography of SE Asia.

The Malay Peninsula is simply a southern extension of the extremely mountainous region that fans out south and east from the Himalayas, separating China from Vietnam, Laos and Thailand. These same mountains also provide a very formidable obstacle to any land migration between these regions and India.

And this is not the only formidable mountain range in Burma. Another in Western Burma, the Arakan Yoma, separates Burma from Assam. This range too connects to the Himalayas. At its southern end lowered sea level connects this range to the Andaman Islands, although the islands may never have actually been connected to the mainland. We know that Y-chromosome D is common in Tibet, a high mountainous region to the north. D could easily have entered the Andamans along this route rather than through India. Certainly there is evidence of many ancient migrations in that direction,from Tibet to India, and none in the other.

During times of lowered sea level the Malay Peninsula mountain range simply extends further south into and around Indonesia, virtually enclosing the South China Sea. And at such times both the East China and Yellow Seas become dry land, or at least a series of closely spaced islands, connecting the South China Sea with the Sea of Japan. An ideal region to perfect island-hopping boating technology. Movement has occurred in both directions several times.

There is no reason at all why all the people in this eastern region must have come originally through India.

"For practical purposes we can consider them a single migration event".

You seem to believe the whole world was populated through 'a single migration event'.

Maju said...

Following your method Europeans must have arrived in just one wave.

In one or two waves most likely, not counting the easily identifiable Neolithic layer. Aurignacian and Gravettian, that's it. And maybe Gravettian was formed in Central Europe anyhow - not sure.

And you believe that's unusual?

Never mind, it doesn't really matter. But many coasts are not mountainous at all (most of Atlantic Europe, virtually all Africa, Arabia, India, China, Australia, most of America...). They may be swampy though.

I'll bet you anything you like that there wasn't one in Paleolithic times.

Not a road surely but migratory routes certainly followed the coast here, in spite of the mountains.

I'm also sure that there are stretches of the coastline where even experienced modern yachties, let alone Paleolithic mariners, are very reluctant to approach too closely to the shore. These same stretches of coastline are usually also those most difficult to pass along by land.

Well, you don't want to get in the middle of reefs, do you? But I don't think these are the most difficult to pass by land or even by sea if you take the logical precautions (i.e. keeping some distance). In fact risky waters normally mean shallow waters, what implies that the land route may be easier (no water barriers).

Anyway unless someone is absolutely sure of what lies ahead it is most unlikely they would move simply along the coast.

Why not? Specially if it's the easiest route (or one of them). Here, for instance it's pretty clear that migrations followed the coast basically and that inland areas were just too harsh to be densely or continuously inhabited. Further into the continent, they followed the rivers specially.

Coasts and rivers anyhow normally offer better resources than other areas (no water: no life).

Perhaps Paleolithic man knew that ahead of them lay the lucky country, Australia.

They had an acid trip and saw it divinely inspired visions... c'mon!

But that's just the first of many huge barriers. I'm sure you will benefit from a brief overview of the geography of SE Asia.

What? Rivers, some not so high mountains, what that really made up an impassable barrier?

The Malay Peninsula is simply a southern extension of the extremely mountainous region that fans out south and east from the Himalayas, separating China from Vietnam, Laos and Thailand. These same mountains also provide a very formidable obstacle to any land migration between these regions and India.

And? Once you start boating at the Ganges delta, you continue boating through the Burmese coast, the Irawady delta... and then either keep navigating southwards or just cross the itshmus by land. They had all the time of the world ahead of them, as they moved only a little stretch maybe each generation (if anything at all).

Those mountains are maybe harsh but not impassable at all. So I guess there could have been more than one route (though the coast all the way to the isthmus seems the easiest one to me, specially for fishing communities).

We know that Y-chromosome D is common in Tibet, a high mountainous region to the north. D could easily have entered the Andamans along this route rather than through India.

Or vice versa. Tibet was not inhabited until the Ice Age was over in any case and D is also common in Japan and found at low frequencies all through East Asia (as well as in Australia).

The mountain chains you mention are not and never were impassable barriers. Please: if Papuans can travel up to 100 km in a single journey through the mountains (and some western mountaineers can as well), you must understand that mountains, as long as they have no permanent ice, are not any impassable wall, just a relative obstacle.

Certainly there is evidence of many ancient migrations in that direction,from Tibet to India, and none in the other.

Of "recent" migrations you must mean. Tibet was desert in the Paleolithic and the Himalayas are even now virtually impossible to cross. The overall East Asian genetic influence into South Asia is limited to Assam (more properly part of SE Asia maybe), some Himlayan valleys and the Austroasiatic tribals of Orissa. That's all.

There is no reason at all why all the people in this eastern region must have come originally through India.

"Originally" is a big word. "Originally" we all come from amoebas...

But the only possible alternative route would be through the steppes and Siberia. A most difficult route you must admit. A route that certainly should have depygmented people to some extent (so no Negritos, Papuans, Melanesians... as result). A route that shows scattered human evidence only long after Australia was colonized.

Also it would imply an origin in West Asia, an area that was then Neanderthal territory (and was colonized by humans only later on).

MtDNA M and R are quite clearly of Southern Asian origin, F and C too. Only N(xR) and D may be more questionable but still both seem to demand a southern coastal route for their distribution. I am willing to consider a secondary node in SE Asia but certainly there's no reason to think of any such node in the Altai (for instance) at least for the high tier clades, much less to explain the colonization of Sahul from such a remote source, nor any archaeological evidence suggesting a N->S route in Eastern Asia.

The southern coastal route certainly remains as the only serious candidate for the colonization of the East. And South Asia must have been the previous stop, logically.

Ebizur said...

Maju said,

"Or vice versa. Tibet was not inhabited until the Ice Age was over in any case and D is also common in Japan and found at low frequencies all through East Asia (as well as in Australia)."

Y-DNA haplogroup D is not found in Australia. All studies of Australian Y-DNA have noted that Australian aboriginal Y-DNA is all derived from haplogroups C and F, with their haplogroup F Y-DNA being almost totally derived from its subclade K-M9. Most Australian aboriginal Y-DNA is derived from haplogroup K-M9; if I remember correctly, it consists of about 60% haplogroup K derivatives and 40% haplogroup C derivatives.

Their mtDNA is mostly haplogroup N(xR), with some haplogroup R derivatives (e.g. haplogroup P) and a very few (approx. 5% or less) haplogroup M derivatives (e.g. haplogroup Q).

Ebizur said...

OK, here's the breakdown of the Y-DNA haplogroups in a sample of 33 Australian aboriginals from the supplementary material of Hammer et al. (2005):

16/33 = 48.5% C-RPS4Y(xC3-M217, C2-M38, C1-M8)
14/33 = 42.4% K-M9(xS-M230, L-M20, M1-M5, NO-M214, P-P27)
3/33 = 9.1% R-M207

I know I have seen other data on Australian Y-DNA somewhere, so I'll look around for that and post it when I have found it. In the data of Hammer et al., Australian aborigines are split almost 50%-50% between haplogroups C(xC1, C2, C3) and K(xS, L, M1, NO, P).

Maju said...

The 2005 McDonald maps divulgative paper reported Y-DNA D in low ammounts among Australian Aborigines and also in SE Asia, specially Sumatra.

I don't know where he got that information from but, following it, haplogroup D appears to have basically an Eastern Eurasian distribution from Australia to NE Siberia, and also inland into Central Asia.

Ebizur said...

Maju said,

"The 2005 McDonald maps divulgative paper reported Y-DNA D in low ammounts among Australian Aborigines and also in SE Asia, specially Sumatra.

I don't know where he got that information from but, following it, haplogroup D appears to have basically an Eastern Eurasian distribution from Australia to NE Siberia, and also inland into Central Asia."

Well, that's bullshit. Haplogroup D Y-DNA has never been reported in any sample of Australian aborigines or NE Siberians. Haplogroup D is found in small numbers of Central Asians and East Asians besides the groups in Japan and southwestern China (Pumizu/Prmi, Zangzu/Tibetans, Qiangzu, etc.) that have very high frequencies of this haplogroup. The Y-DNA of most NE Siberian natives belongs to haplogroup C3 or haplogroup Q, with some haplogroup R1a1, N1c1, I, etc. that have probably come from recent foreign (mostly Russian in this case) influence. They do not have any haplogroup D Y-DNA, although they do have plenty of haplogroup D mtDNA.

By the way, you have probably confused the haplogroup F-M89 Y-DNA in the pies for Evenks (who should properly be described as "East Siberians," not "NE Siberians") and Australian aborigines for haplogroup D. Please zoom in (125% magnification should be sufficient) and notice that that small slice of the pies for Evenks and Australians is filled with the beige color that represents haplogroup F, not the yellow-green color that represents haplogroup D.

Ebizur said...

By the way, the only populations that are shown as including any haplogroup D Y-DNA on the McDonald map are as follows:

Uzbeks (UZ)
Kazakhs (KZ)
Altayans (AL)
Mongols (MO) [Does not include Buryats (BU)]
Uyghurs (UG)
Tibetans (TB)
Southern Hans (HA)
Japan (JP)
Malaysia (MY)
Borneo (BO)
Sumatra (SU)

There is no haplogroup D represented in any of the other pies. Absolutely none, anywhere.

Ebizur said...

I've been looking around for more Australian Y-DNA data, and it looks like they have more haplogroup C than I have previously thought. It's the Papuans who are mostly haplogroup K-M9; the Australians seem to be mostly haplogroup C4, with a significant presence of some mysterious subclade(s) of haplogroup K-M9.

According to the data table at http://www.pnas.org/content/suppl/2007/05/10/0702928104.DC1/02928Fig5.pdf:

Australia:
6/6 = 100% C4a-M347+DYS390.1del

Australia/Desert:
24/35 = 68.6% C4a-DYS390.1del
1/35 = 2.9% F-M89(xK-M9)
6/35 = 17.1% K-M9(xNO-M214, M2-M353, K1-M147, S-M230, M1-M4, P-M45)
1/35 = 2.9% O-M175
3/35 = 8.6% R1-M173

Australia/Arnhem:
6/60 = 10.0% C-M130(xC2-M38, C3-M217, C4a-DYS390.1del)
32/60 = 53.3% C4a-DYS390.1del
1/60 = 1.7% F-M89(xK-M9)
18/60 = 30.0% K-M9(xNO-M214, M2-M353, K1-M147, S-M230, M1-M4, P-M45)
3/60 = 5.0% R1-M173

Australia:
3/7 C4b-M210
1/7 F-M89(xK-M9)
3/7 R1-M173

Pooling all five (including Hammer et al. 2005) Australian aboriginal data sets:
87/141 = 61.7% C(xC2, C3)
3/141 = 2.1% F(xK)
38/141 = 27.0% K(xM1, NO, P, S)
12/141 = 8.5% R-M207
1/141 = 0.7% O-M175

It looks most likely that the haplogroup R Y-DNA found among Australian aborigines actually all belongs to the subclade R1-M173. This probably reflects recent genetic influence from Europe.

Incredibly, I have found that none of the published studies of Australian aboriginal Y-DNA has tested its haplogroup K-M9 Australians for any mutation that indicates membership in haplogroup T-M70. This means that there is a possibility that aboriginal Australians might represent the densest concentration of haplogroup T-M70 Y-DNA in the world.

Maju said...

you have probably confused the haplogroup F-M89 Y-DNA in the pies for Evenks (who should properly be described as "East Siberians," not "NE Siberians") and Australian aborigines for haplogroup D.

Darn, you are right. Thanks for the clarification. It really looks yellow and not beige... until you zoom.

So D is a SE/East asian clade only, including Andaman.

Incredibly, I have found that none of the published studies of Australian aboriginal Y-DNA has tested its haplogroup K-M9 Australians for any mutation that indicates membership in haplogroup T-M70. This means that there is a possibility that aboriginal Australians might represent the densest concentration of haplogroup T-M70 Y-DNA in the world.

Why T (former K2)? This clade is only found in Western Eurasia. It's only logical that nobody expects it at all. As they have tested for S (former K5, common in New Guinea), then it should be some other not yet described Australian-specific haplogroup/s.

On second thought, I notice among the data you are posting that they tested for specific subclades of M, not for M-P256, so it could well be M(xM1,M2). M3 (former K7) has not been tested either, right?

Ebizur said...
This comment has been removed by the author.
Ebizur said...

Maju said,

"So D is a SE/East asian clade only, including Andaman."

Yes, that is correct. The distribution of Y-DNA haplogroup D seems to correlate quite closely with that of Y-DNA haplogroup O (very strong in Japan and Tibet, fading away into northern East Asia and Central Asia). The Austronesians and other Southeast Asians are notable for mostly having only haplogroup O without any haplogroup D; northern East Asians and Central Asians have O plus D in combination. Siberians essentially have neither O nor D, although there are a few strays that are probably recent introductions from the south.

"Why T (former K2)? This clade is only found in Western Eurasia. It's only logical that nobody expects it at all. As they have tested for S (former K5, common in New Guinea), then it should be some other not yet described Australian-specific haplogroup/s."

I was simply noting the fact that there is no published study of Australian Y-DNA in which the authors have reported testing their samples for any marker of haplogroup T. It just struck me as being sloppy and lacking thoroughness.

"On second thought, I notice among the data you are posting that they tested for specific subclades of M, not for M-P256, so it could well be M(xM1,M2). M3 (former K7) has not been tested either, right?"

Yes, that is correct; these studies were performed prior to Karafet et al. 2008's publishing a phylogenetic tree of haplogroup M that has been redefined by the P256 mutation. Before Karafet et al. 2008, "haplogroup M" referred to what is now known as haplogroup M1, and the current M2 and M3 were known as K1 and K7, respectively.

The Australia/Desert and Australia/Arnhem K-M9 samples have been tested for haplogroup M2 (former K1) in addition to haplogroup M1 (former M), so they could only belong to K*, L, M*, M3, or T. I guess that they probably belong to haplogroup K* (i.e. an Australian-specific subclade or subclades of haplogroup K-M9), but it certainly would be nice to find out for sure.

terryt said...

Ebizur. Thanks for all that information regarding Australia and Southeast and East Asia.

Perhaps you can help further. I thought the haplogroups from the Khasi Hills might be revealing regarding ancient human movement through Eastern India. The people certainly look East Asian and their languages are mostly East Asian. As Maju said concerning Assam, it might be "more properly part of SE Asia". But this just throws further doubt on any sort of movement east through the region. It was probably uninhabited before these East Asian people moved in.

Maju wrote: "But many coasts are not mountainous at all".

But much of the Hadramawt and Makran coast is. And surely once humans had the boating ability necessary to cross the Bab al Mandeb a migration north up the Red Sea would be so much easier than the uninviting South Arabian coast. It is therefore absolutely stunning that this boating ability failed to cross the relatively short distance between the Red and Mediterranean Seas until many thousands of years later. The only possible conclusion is that boating is not as ancient as you believe. Your comment, "Once you start boating at the Ganges delta, you continue boating through the Burmese coast, the Irawady delta..." is therefore totally irrelevant.

"But the only possible alternative route would be through the steppes and Siberia. A most difficult route you must admit. A route that certainly should have depygmented people to some extent".

Perhaps difficult, but at times of more moderate climate so much easier than any 'southern coastal route'. Also likely to be rapid because of open savannah-like vegetation, accounting for both lack of depygmentation and lack of evidence.

Maju said...

Maju wrote: "But many coasts are not mountainous at all".

But much of the Hadramawt and Makran coast is.


I also said that mountains did not really matter after all.

And surely once humans had the boating ability necessary to cross the Bab al Mandeb a migration north up the Red Sea would be so much easier than the uninviting South Arabian coast. It is therefore absolutely stunning that this boating ability failed to cross the relatively short distance between the Red and Mediterranean Seas until many thousands of years later.

Not correct. H. sapiens are first found in the Mediterranean coasts (Levant, North Africa) before any other place (except Tropical/Southern Africa). Just that in that area they appear to have disappeared (?) later because of Neanderthal expansion and desertization.

The only possible conclusion is that boating is not as ancient as you believe.

Ok, I admit that your divinely inspired vision of "Viking-like" Papuans colonizing the world on dug-out canoes is the only possible one. Surely the mountains of Hadramaut, the disappearence of AMHs from the Mediterranean and the extinction of mammoths have something to do with that, even if my feeble mind fails to see any correlation whatsoever. ^^

Your comment, "Once you start boating at the Ganges delta, you continue boating through the Burmese coast, the Irawady delta..." is therefore totally irrelevant.

Yes, common sense is totally irrelevant when you have bright lightbulb over your head. I know.

[In Siberia/Central Asia] Also likely to be rapid because of open savannah-like vegetation...

Most of India was savannah some 60,000 years ago.

...accounting for both lack of depygmentation and lack of evidence.

Sure: deeply-black skinned people roamed the narrow steppary/desertic corridor north of 45º latitude. Then they boated to Australia through the Amur river and the open Pacific Ocean in a dug-out canoe. They also colonized Mars in the meantime (dug-out canoes are flippant for interplanetary travels, you know) but they definitively ignored the tropical areas of Asia because coastal routes are bad for boating.

Sorry to be sarcastic but I can't amke any sense of it all.

Maju said...

Terry: check this abstract mentioned at Mathilda's Anthropology Blog. While the evidence is contradictory, it seems that southern Indian/Lankan tribals may have migrated to Australia in a second wave.

So maybe you're right after all about Y-DNA C and K (for instance) being spread in two different migrations, but South Asia still seems to be at the origin.

terryt said...

"H. sapiens are first found in the Mediterranean coasts (Levant, North Africa) before any other place (except Tropical/Southern Africa)".

What on earth has that got to do with boats? Just because they were able to reach coastal Africa (from within Africa) doesn't mean boats were involved. They probably didn't even migrate along the coast.

"Most of India was savannah some 60,000 years ago".

Very true. But a significant part of it wasn't. Precisely the region in question: Northeat India. Tropical forest, mountains and coastal mangrove swamp forest have always been significant barriers to human movement.

I'll ignore your stupid light bulb comments. Your own ideas are much more the result of blindly ignoring reality.

terryt said...

This on haplogroups in the Khasi region of NE India:

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2065843

However I have a major reservation concerning the comment:

"However, Austro-Asiatic speakers, hypothesized as probably the earliest settlers in the Indian subcontinent".

It is very difficult to make a case for an ancient origin for either Tibeto-Burman or Austro-Asiatic in India. If they originated some 50,000 years ago we would be unlikely to be able to discern any relationship within the groups at all today. And Northeast India contains just a very few of the many branches of the Austro-Asiatic and Tibeto-Burman language families. This makes it unlikely the two families developed there. This is not to say the two families don't have some ancient relationship with Dravidian. It's possible, if they are related, that the common origin was in India. But as they exist today it's much more likely that both language families have been introduced to Northeast India from outside it.

And surely such a variety of haplogroups in a small isolated region, where drift should have been a major factor, screams out in favour of relatively recent, and prolonged, admixture. And the comment, "Overall, the populations of the same linguistic family seem to cluster together" is revealing. Not what we'd expect in an ancient population.

Maju said...

What on earth has that got to do with boats? Just because they were able to reach coastal Africa (from within Africa) doesn't mean boats were involved. They probably didn't even migrate along the coast.

It's possible.

Whatever the case I think that you have built a daring hypothesis that basically states that boats were invented in Indonesia, while most people would contend that they existed since long before island-hoping (and maybe accidents) brought humans to Sahul.

We may have no evidence from H. sapiens but certainly Acheulean H. erectus crossed the Gibraltar strait. If H. erectus could boat, H. sapiens could too. Anc actually it was surely a necesary condition for all the migrations across the Bosphorus/Dardanellos as well (H. erectus into Europe, H. neaderthalensis into Asia, H. sapiens into Europe.

Very true. But a significant part of it wasn't. Precisely the region in question: Northeat India. Tropical forest, mountains and coastal mangrove swamp forest have always been significant barriers to human movement.

I'd say mangrove coastal jungles/swamps are pretty good enviroments for humans to exploit. Rich in fish and other resources. You need boats though.

I'll ignore your stupid light bulb comments. Your own ideas are much more the result of blindly ignoring reality.

I do understand that you have apreconcieved idea and are trying to bend reality to it. That's anti-science.

Anyhow, I just come from discussing pygmentation genetics in my blog and one of the findings that have arosen recently (Kittles et al. 2006) is that tropical Asians/Oceanians do retain the ancestral African alelles of dark-pygmentation. Instead West and East Eurasians show two separate (yet convergent) evolutions to low insolation climates. This alone appears to support the mainstream model of "coastal" migration along southern Asia and the formation of two stocks in the west and east of the continent that would eventually evolved into lighter pygmentations separately.

Your over-complicated model, based only on the absurd fantasy that people did not know anything about boats until they reached Indonesia, would actually imply the opposite: joint (and not distinct) pygmentation lightening in Central Asia for Western and Eastern Eurasians and "de novo" evolution into dark types in Indonesia/Oceania later on.

And pygmentation genetics is just one piece of evidence. There's much more: haploid and nuclear genetics, archaeology...

So for me the coastal model is just way too solid, at least since the arrival to South Asia.

It is very difficult to make a case for an ancient origin for either Tibeto-Burman or Austro-Asiatic in India. If they originated some 50,000 years ago we would be unlikely to be able to discern any relationship within the groups at all today.

In this I fully agree with you. They are "back-migrants" from Eastern Asia.

terryt said...

"certainly Acheulean H. erectus crossed the Gibraltar strait". "a necesary condition for all the migrations across the Bosphorus/Dardanellos".

It is possible that any level of boating is not actually necessary in neither of these cases. Lowered sea level has connected Africa and Europe across Gibraltar at times, although I don't think during the Acheulian, but just very brief contact is required. And the same goes for the islands through the Aegean Sea, where land movement has been as great as in SE Asia.

"If H. erectus could boat, H. sapiens could too".

Ability to boat and actually having boats are two different things. Homo erectus presumably didn't spread far from SE Asia with their boating technology, extremely primitive as it presumably was. However once Home sapiens had picked up the technology they spread it far and wide.

"tropical Asians/Oceanians do retain the ancestral African alelles of dark-pygmentation".

There's no selection pressure for de-pygmentation if humans, while crossing Asia, moved no further north than about 45 degrees. It was you who first claimed de-pygmentastion was a requirement: "A route that certainly should have depygmented people to some extent (so no Negritos, Papuans, Melanesians... as result)".

"Your over-complicated model".

Far less complicated than your one, requiring a complete lack of drift and founder effect during early stages of human existence somewhere in India, followed by extremely complicated levels that still manage to leave modern haplogroups regionally distributed rather than the random distribution we would expect under your model. At least my 'complicated model' accounts easily for this distribution.

Maju said...

Lowered sea level has connected Africa and Europe across Gibraltar at times...

No, it has not. The passage was narrower but still open and required boats or at least rafts.

And the same goes for the islands through the Aegean Sea, where land movement has been as great as in SE Asia.

No. Aegean islands for the most part were not colonized before Neolithic. Human travel between Asia and Europe happened across the Marmara Sea straits.

There's no selection pressure for de-pygmentation if humans...

In relatively high latitude, yes. People need vitamin D for correct developement. Lack of vit. D may perfectly make a whole population to die off out of wrongly developed brains and feeble bones, among other problems.

Far less complicated than your one, requiring a complete lack of drift and founder effect during early stages of human existence somewhere in India...

How come? India is huge and fertile. Early human presence in the subcontinent would mean expansion, not stability or contraction (hence no or very low drift initially).

But anyhow, I don't have clear why you claim that the coastal model has no founder effects or drift. L3 could have diverged in many hundreds of separate lineages without drift, yet we only see two in Eurasia. Same for male lineages: only three. This implies some marked founder effect or drift, maye even a massive bottleneck at the Toba event.

... followed by extremely complicated levels that still manage to leave modern haplogroups regionally distributed rather than the random distribution we would expect under your model.

You should also expect founder effects in any migration, by definition. The problem is not to explain how haplorgoups appear to be regionally distibuted but why at old stages (major macro-lineages) they do not appear to be regionally concentrated but rather the opposite (M, N and R nearly everywhere, for instance).

At least my 'complicated model' accounts easily for this distribution

I fail to see how.

terryt said...

"In relatively high latitude, yes. People need vitamin D for correct developement".

45 degrees is hardly a "relatively high latitude".

"Aegean islands for the most part were not colonized before Neolithic".

So even though humans had boats in the Mediterranean for 50,000 years it was only in the Neolithic they were able to reach the Aegean islands. There's something drastically wrong with your theory, Maju.

"I fail to see how".

Yet you previously wrote: "In this I fully agree with you. They are 'back-migrants' from Eastern Asia". So do you believe that no new haplogroups came in with the languages? Come on.

It used to be widely accepted that before about 10,000 years ago the human inhabitants of SE Asia and Southern China looked much more like modern Papuans and Melanesians than like the East Asian phenotype found pretty much throughout the region today. This change in appearance is completely downplayed today, perhaps deliberately.

The change must have been the product of an immigrant population, probably from much further north. The same movement is probably responsible for the East Asian appearance of many Northeast Indian groups. It is inconceivable that no new Y-chromosome or mtDNA haplogroups came in with this population. What are your nominations? These haplogroups certainly didn't originate in India.

terryt said...

By the way. I suspect that the Marmara Sea straits region was a lot easier to cross when just a river connected the Black and Aegean Seas before sea level rose (about 10,000 years ago many seem to believe).

And I'm pretty sure the Mediterranean Basin was a series of shallow lakes and salt flats around five million years ago. Long before the Acheulian I'll admit.

Maju said...

So even though humans had boats in the Mediterranean for 50,000 years it was only in the Neolithic they were able to reach the Aegean islands. There's something drastically wrong with your theory, Maju.

I don't see why. If you can't see the opposite shore in a clear day, it's "open seas": a place not adventure to with a mere canoe. But with all likehood humans who migrated into Europe knew how to make boats of some sort, though they probably used them mostly in rivers (and coastal waters at most).

It's the same in Asia anyhow, though maybe some of the migrating communities were more active in this activity, as it allowed them to fish in the rich tropical waters and maybe was a tradition transmited from their ancestors at the Red Sea. This would be the primary groups making the intergenerational voyage along the coasts.

The only exception is the "jump" to Sahul, that clearly requires some high seas adventuring. It is this daunting crossing which requires an explanation, not the use of canoes by the rest.

45 degrees is hardly a "relatively high latitude".

Outside Europe, the vast majority of people lives south of it. Even in Europe that was the case before the end of the Ice Age (i.e. the restoration of the warm Atlantic Conveyor). It is an extreme latitude for tropically adapted species as ours in any case.

Latitude 45ºN crosses Vermont, Kazakhstan and Sakhalin (so you get an idea of how far north that is). Only in the special context of Europe it appears relatively to the south, crossing at middle France. It is also doubly distant from the Equator than the tropics are, so in summer it gets the radiation that either tropic (the lines, not the climatic area) get in winter, and in winter gets only a fraction of that.

Yet you previously wrote: "In this I fully agree with you. They are 'back-migrants' from Eastern Asia". So do you believe that no new haplogroups came in with the languages? Come on.

You're mixing apples and oranges again: Austroasiatics are high in Y-DNA O, which is surely original from East or SE Asia. They probably arrived in the Neolithic though.

It used to be widely accepted that before about 10,000 years ago the human inhabitants of SE Asia and Southern China looked much more like modern Papuans and Melanesians than like the East Asian phenotype found pretty much throughout the region today. This change in appearance is completely downplayed today, perhaps deliberately.

It's not clear. The so-called "australoid" type is often extremely ambiguous when dealing with Paleolithic remains. The Ainu for instance have been described as such and yet are very different from any Oceanian native population, both in appearence and haplogroups. Some Indians have also been described as such and again the differences are sometimes striking, not to mention the supposed "australoid" early Native Americans like Lucía. It's a catch-all term, too ambiguous to be really meaningful.

Said that, it's possible that a more modern East Asian phenotype spread in a second wave, an internal one of that region, together with Y-DNA O (or NO) maybe. When? Some say that in Neolithic but Paleolithic can't be excluded anyhow. The archaeology of the region (or at least my knowledge of it) is not enough to reach to solid conclusions regarding phenotype changes and possible migrations associated with them. Whichever migrations anyhow were regional process exclusive of Eastern (and Northern) Eurasia. Even Native Americans are somewhat apart of the typical Mongoloid phenotype (and many SE Asians are too, as well as historical Ainu), what suggests a relatively recent movement or phenotype fixation process.

The change must have been the product of an immigrant population, probably from much further north.

NO appears to be original of middle or southern China.

These haplogroups certainly didn't originate in India.

NO as such did not but its ancestor, K, probably did.

By the way. I suspect that the Marmara Sea straits region was a lot easier to cross when just a river connected the Black and Aegean Seas before sea level rose (about 10,000 years ago many seem to believe).

I don't really believe in that hypothesis, to be honest. But if true, they still needed to cross a river, what requires canoing.

And I'm pretty sure the Mediterranean Basin was a series of shallow lakes and salt flats around five million years ago.

You're wrong in that. The Mediterranean Sea has never been a lake. Africa is pushing against Europe, not separating from it. The Mediterranean will become a lake or series of it some time into the distant future - but not ever in the past. It the main remnant of the sea that separated Gondwana from Laurasia (Thetys Sea?).

Ebizur said...

Maju said,

"Said that, it's possible that a more modern East Asian phenotype spread in a second wave, an internal one of that region, together with Y-DNA O (or NO) maybe. When? Some say that in Neolithic but Paleolithic can't be excluded anyhow. The archaeology of the region (or at least my knowledge of it) is not enough to reach to solid conclusions regarding phenotype changes and possible migrations associated with them. Whichever migrations anyhow were regional process exclusive of Eastern (and Northern) Eurasia. Even Native Americans are somewhat apart of the typical Mongoloid phenotype (and many SE Asians are too, as well as historical Ainu), what suggests a relatively recent movement or phenotype fixation process."

That's just stupid, Maju. Native Americans are the prototypical Mongoloid phenotype, along with their relatives in eastern Siberia, Mongolia, and vicinity. Populations of southern China, Japan, Indochina, Austronesia, Austroasiatic peoples of India, etc. are definitely not "prototypical Mongoloids."

It is ridiculous to claim that the Mongoloid phenotype spread with Y-DNA haplogroup NO-M214 (altogether ignoring the problem of the mtDNA) when this haplogroup is completely absent from prototypical Mongoloids, such as Native Americans, and is also practically absent from most native East Siberian populations. Even haplogroup P-M45 would be a better candidate for representing the spread of the "prototypical Mongoloid phenotype" than NO-M214 would, but the truth is that neither of them is directly related; the original Y-DNA of "prototypical Mongoloids" is obviously the subset of haplogroup C that includes at least the C3-M217 clade, and this early population absorbed influence from haplogroup P-M45 and eventually haplogroup NO-M214 carriers (which, you should recall, are derived from haplogroup K-M9 and ultimately F-M89).

terryt said...

"If you can't see the opposite shore in a clear day, it's 'open seas'".

From what I understand, though I've never been there, through much of the Aegean it's impossible to be out of sight of land. End of your hypothesis.

"It's the same in Asia anyhow".

No it's not. Boating is long-established along the East Asian coast from the Sea of Japan to the south China Sea and across to Melanesia.

"a tradition transmited from their ancestors at the Red Sea".

You're still obsessed with this Adam and Eve in the Garden of Eden story aren't you. There's absolutely no evidence for boating in the Red Sea till long after humans had reached Australia.

"The only exception is the "jump" to Sahul, that clearly requires some high seas adventuring".

So the technology was carried just one way, across Wallace's Line, and it didn't move back through the existing population? I find that difficult to believe.

"NO as such did not but its ancestor, K, probably did".

Seems pretty right. But doesn't that mean O originated outside India? From K certainly, but from a branch of K that had moved outside India.

"The Ainu for instance have been described as such and yet are very different from any Oceanian native population".

But that's probably because Oceanian native populations are not Australoid. It amazes me that many Europeans who see even such people as the Croats and Serbs as different 'races' lump all Chinese, Polynesians, Mongolians, Melanesians, SE Asians and Australian Aborigines into one relatively homogeneous race. They are each easily as different from each other as any two European groups you care to nominate.

"But if true, they still needed to cross a river, what requires canoing".

No it doesn't. I've personally walked across many rivers, sometimes with difficulty I'll admit. But at other times it has been fairly easy, especially if there is a log or some such over it.

Maju said...

That's just stupid, Maju. Native Americans are the prototypical Mongoloid phenotype...

No. Some Native Americans fit pretty well but others are quite distant, with prominent noses (even more prominent than most Caucasoids) and other traits that are clearly anomalous in the archetype.

It is ridiculous to claim that the Mongoloid phenotype spread with Y-DNA haplogroup NO-M214 (altogether ignoring the problem of the mtDNA) when this haplogroup is completely absent from prototypical Mongoloids, such as Native Americans, and is also practically absent from most native East Siberian populations.

Maybe you're right. It'd be easier certainly if the type would have spread from North to South, what doesn't seem to match what NO appears to have done.

(altogether ignoring the problem of the mtDNA)

Sure. I decided to avoid that labyrinth. I admit my guilt.

Any ideas?

but the truth is that neither of them is directly related; the original Y-DNA of "prototypical Mongoloids" is obviously the subset of haplogroup C that includes at least the C3-M217 clade

Hmmm. Both lineages could share the same phenotype. Phenotype is obviously not defined by Y-DNA.

But I find easier to think of radical drift and founder effects in the far North, where the population was always very small, than in the much more densely populated south (south of the Great Wall, not SE Asia).

There's also some badly dated skull from Southern China, near Indochina, (claimed as very old but uncertain) that is typically Mongoloid, unlike other skulls found in the region that are much more ambiguous.

But well... whatever. A problem wuld not be anymore a problem if we had the solution.

terryt said...

"Both lineages could share the same phenotype. Phenotype is obviously not defined by Y-DNA".

Couldn't agree more. Here's what I reckon has happened.

I agree with you that Y-hap O at least helped carry the East Asian phenotype into Southern China and Northeast India, from the north. It, along with certain subsets of C, is certainly associated with the Austronesian expansion. But how did it get up north?

O's brother group, N, is widely spread through Northern Eurasia, especially north of 45 degrees. This implies that O originated south of 45 degrees. Yet north of 25 degrees if it came from the north and partially replaced the Melanesian-looking people who then lived south of 25 degrees.

Perhaps O developed around the Yellow and East China Seas. Haplogroup N/O had presumably introgressed into East Asian-looking people as it expanded north up the coast.

We can easily go back further, to Y-chromosome K. Ultimately it broke into 11 different lines. Many of these have a relatively restricted distribution. This suggests they evolved in situ rather than being the reult of an extremely comlex pattern of drift after an original combined expansion. Y-hap K and immediately descendant lines are spread from, say, Northwest frontier Province in Pakistan (P/Q/R and T), through India (L and various Ks), up the East Coast of China (N/O) and through SE Asia as far as New Guinea and Australia (M, S and various Ks). This suggests K expanded rapidly through its geographic range from some small region within it, but not necessarily from India.

Where are most of these 11 early branches within K found? That's correct. Around Wallace's Line. Could it possibly be that K expanded from there? I accept totally that its own parent line, F, had come through India. It's just that I have trouble reconciling C and D's distribution with their having come via the same route.

The link regarding Northeast India does show very clearly that pre-existing Indian haplogroups, especially mtDNA, survived as a substrate to the Austro-Asiatic and Tibeto-Burman immigrants. As I keep reminding you, Y-chromosome and mtDNA do not always move together. The people of Northeast India are, as are all groups, a product of various waves of migration. And if we're prepared to look we can untangle these waves of migration. However what we will never find is simplicity. The whole process was, and still is, far more complex than what many comment-posters here seem prepared to accept.

Maju said...

This suggests K expanded rapidly through its geographic range from some small region within it, but not necessarily from India.

What do you have against India? LOL It's prefectly placed in the middle of that geography.

Where are most of these 11 early branches within K found? That's correct. Around Wallace's Line.

Geographical distribution of core zones:

- West: K4, T, maybe P
- South: K1, L, maybe P
- East: NO
- Oceania: M, S, K3

Nobody seems to know where K2 is found.

This alone strongly suggests a continental Eurasian (and therefore possibly South Asian) origin for the macro-haplogroup, even if New Guinea has some odd diversity that may well be caused by lack of good understanding of the internal genealogy of those clades (not imposible that they end all up grouped in a single clade whith more research).

But, also, I just "discovered" that the newest draft of the YSOGG tree joins K and IJ in a macro-haplogroup called IJK. IJ, the parallel branch of K, is a Western Eurasian haplogroup almost exclusively.

This has brought some to suggest K's origins are West Asian - though, considering the relative lack of research in South Asia, I still think of a more eastern urheimat - but not as far east as SE Asia, much less New Guinea, in any case.

In any case, it seems that the formation and spread of IJ and K cannot be treated as separate phenomena anymore.

The link regarding Northeast India does show very clearly that pre-existing Indian haplogroups, especially mtDNA, survived as a substrate to the Austro-Asiatic and Tibeto-Burman immigrants

Can you repost that link? I browsed all 79 comments and could not find it. It does not surprise me at all but I wonder which are the "pre-existing Indian" Y-DNA haplogroups in this case.

The whole process was, and still is, far more complex than what many comment-posters here seem prepared to accept.

I think you are the one being simplistic here. The "great expansion" happened some 60-50 thousand years ago but humans were in the subcontinent (and maybe adjacent regions like parts of West and SE Asia) since long before, maybe since c. 105,000 BP and surely before Toba in any case (at least that's what I think). This gives more than enough time for all widely distributed haplogroups (except NO and P) to have coalesced by then. Maybe there was more than one migratory wave, but all appear to have gone through South Asia (or originated there).

terryt said...

Here's the link:

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2065843

"What do you have against India?"

Nothing personal, although I've never been there. But even when we include Pakistan and Bangladesh Australia is still nearly twice as large. And we know how diversified Australian haplogroups are. Sure, much of India's haplogroup diversity is a product of inward migration, and Australia has not all been inhabitable all the time, but it seems unlikely that a region as relativel small as the Indian subcontinent could, on its own, produce such an early haplogroup diversity, no matter how long humans had remained there before spreading out.

"it seems that the formation and spread of IJ and K cannot be treated as separate phenomena anymore".

Thanks for posting that link. But joinig IJ to K simply expands K's distribution to the Iranian Plateau, doesn't affect the eastern end.

We should perhaps stop this discusion here, or after your next comment. But you say:

"I think you are the one being simplistic here".

And the very next sentence is:

"The 'great expansion' happened some 60-50 thousand years ago".

What could be a simpler interpretation of the evidence than that? You redeem yourself next with:

"but humans were in the subcontinent (and maybe adjacent regions like parts of West and SE Asia) since long before, maybe since c. 105,000 BP and surely before Toba in any case".

Agree totally. But this also gives more than enough time for them to reach SE Asia and to take advantage of any amelioration of climate to move through Central Asia. And you also concede:

"Maybe there was more than one migratory wave".

Maju said...

...to move through Central Asia

You're obsessed with Central Asia. Archaeology is now suggesting rather old dates with "proto-Aurignacian" over there (and also in Iran and Turkey) but there's nothing to suggest that they migrated Eatsward from there (exception made of the Y-DNA Q lineage, quite scattered outside America). The R/Q division may have happened then (i.e. some 50,000 years ago) but that means that the K division and expansion must be older - and that means that Central Asia was not the route humans followed to Eastern Eurasia.

But even when we include Pakistan and Bangladesh Australia is still nearly twice as large.

But mostly a desert. It's also a geographic cul-de-sac, extremely isolated from the rest of the world. It's only logical to think that Australia was in the recieving end of the human migrations all the time, and not any exporter. But additionally neither archaeology nor genetics appear to suggest an Australian origin of anything known in Eurasia (the opposite instead stands true).

Thanks for posting that link. But joinig IJ to K simply expands K's distribution to the Iranian Plateau, doesn't affect the eastern end.

It does suggest a more western origin for K as well as for its mediate ancestor IJK. That "western" origin could well be in South Asia but hardly any more to the east. Wikipedia claims West Asia in fact, though I'd say that is also rather unlikely (at most Iran/Afghanistan).

To me it suggests that IJ, G and K migrated westward in a rather small window of time, maybe in different specific movements but all in the time that Neanderthals were being pushed back already. K would then diversify into P, T and K4 (at least), already in West (or Central in the case of P probably) Asia.

Meanwhile other clades (at least K, but probably also C and D) migrated eastward. Again it's virtually impossible to determine if they did together or in different but sequential migrations but K would then diversify in NO, M, S and maybe other still undefined clades in SE Asia or Sahul (equally C appears to have diversified also in Eastern Eurasia mostly, and D too).

IMO, all these expansions happened in a "short" timeframe c. 60-50,000 years ago.

terryt said...

I'm sorry about this. I did promise you the last word but I can't resist.

Surely under your scenario ancient India simply replaces Australia as "a geographic cul-de-sac, extremely isolated from the rest of the world". And Australia has not always been "mostly a desert". At times Central Australia has been quite moist.

As a result of this surely what we know about Australia in relatively more recent times should hold true for a much earlier India. When we actually look at all the evidence we find this is probably in fact the case.

Y-hap F and mt-hap M certainly seem to have expanded from India but Y-haps C and D and mt-hap N's distribution is most parsimoniously explained as a movement across Central Asia at some climatically appropriate moment. But when the climate cooled again N became isolated into small surviving groups scattered along its migration route. Selection limited surviving lines to I and W in warmer parts of the Zagros Mountains and Iranian Plateau, A and X somewhere round the Mongolian Plateau (probably along south-facing mountain ranges), Y round the Sea of Japan, R round the South China Sea and S across Wallace's Line in Australia. Y-haps C and D's distribution easily fits the same pattern.

Meanwhile mt-hap M had gradually worked its way across the Ganges Valley, through the Burmese Mountains and so to Mainland Southeast Asia. Rather than indicating a subsequent migration north, then east through Central Asia, mt-hap M's distribution seems to simply demonstrate similar, though lesser, regional selection. C and Z somewhere at the western end of its distribution, probably on the Iranian Plateau, and other haplogroups spread across M's whole distribution: the triangle formed by India, New Guinea and the Great Plain of China (or the Yellow Sea, take your pick). Y-hap F's distribution, especially that of K, again easily fits the same pattern.

Mt-haps M and N became mixed in East Asia. However it was mainly mt-hap N's descendant line R that next spread over top of these already established haplogroups. Mt-haps D and G may have been carried north through China along with R's expansion though. Mt-haps HV, JT and UK ultimately emerged from Northwest India to spread around Northwest Eurasia. Yet again, Y-hap K and descendants easily fits this pattern. By the way, I think Y-hap K4 is also eastern, confined to Indonesia, not "the west". But I'll check it out.

Although, for various reasons, I consider it unlikely, I'll concede for now that mt-hap R may have originated in India. But it certainly evolved and expanded after mt-haps N and M had already mixed through East Asia.

terryt said...

And this is vry interesting:

http://johnhawks.net/node/1798

"Sweden may have largely been ice-free between 59,000 and 40,000 years ago". Just the sort of climate amelioration that could permit a Central Asian migration. But we could be considering an even earlier one. Before Toba presumably the climate was relatively warm.

Maju said...

...but Y-haps C and D and mt-hap N's distribution is most parsimoniously explained as a movement across Central Asia...

They are not inf fact.

Y-DNA C is found in South, SE, East and NE Asia, as well as in America and Sahul, mostly in regioanlly differentiated subclades. All Central Asian (and West Eurasian) C is best explained as a recent Turco-Mongol input. So why would Central Asia had anything to do with the spread of a haplogroup that is distributed primarily across the Indo-Pacific area?

Y-DNA D, admittedly, could be more ambiguous but its presence in Andaman islands certainly would appear to at least leave open the coastal route as a serious prossibility.

As for mtDNA, N(xR) is again somewhat ambiguous but some Australian N* subclades best relate with West and South Eurasian W, what basically excludes the Central Asian route.

Also you are ignoring R, which must be of South Asian origin too.

In brief, South Asia (and maybe SE Asia, and even possibly some areas of West/Central Asia) appear to be original for all high level Eurasian clades that are somewhat clear. The ones that are maybe less clear do not exclude South Asia as possible origin at all.

Hence the most parsimonious result is South Asia as urheimat of ALL top level Eurasian lineages. I can accept as possible a "greater South Asia" model, extending into SE Asia and Iran... but the overall concept of mainly coastal migration is pretty clear.

Sweden may have largely been ice-free between 59,000 and 40,000 years ago

In fact that's the time when humans may have colonized Central Asia as far north as the Altai (and also, in the last moment of this window, when Aurignacians took over most of Europe). But in archaeological terms this colonization of Central Asia had no apparent implications for East Asia (even today deserts are a major barrier between Central and East Asia, only overcome with the domestication of horse it seems). And in genetic terms, I understand that this process is rather related to the formation of Y-DNA P or rather its division into R and Q. So it's actually "after K".

terryt said...

Fascinating. Sorry, again I can't resist. You seem totally unaware of the inconsistencies and contradictions in your theory.

You seem prepared to concede that India's haplogroup diversity as compared with Australia's is partly a product of immigrant haplogroups to India yet, in contradiction, you claim most haplogroups originated in India.

You concede that Y-hap P/Q/R's expansion is part of a later one than K's yet, inconsistently, you cannot see that K's expansion is later than F's, and mt-hap R's expansion is later than N's.

By the way, the evidence indicates that Y-hap Q's expansion was largely west of the Altai Mountains and then north of the Central Asian mountains, via the upper Ob, Yenisey, and Lena river valleys, rather than through Mongolia (not always desert). The region south of the mountains was already occupied. Guess who by? Along the way Y-hap Q picked up mt-haps A, X, B and D from south of the mountains. However mt-hap C may have gone all the way with Q.

Lastly, would you like to tell us all here how common mt-hap Y is in India? Or S, or X, or A. Yet you believe they all originated there.

Ebizur said...

terryt said,

"The evidence indicates that Y-hap Q's expansion was largely west of the Altai Mountains and then north of the Central Asian mountains, via the upper Ob, Yenisey, and Lena river valleys, rather than through Mongolia (not always desert). The region south of the mountains was already occupied. Guess who by? Along the way Y-hap Q picked up mt-haps A, X, B and D from south of the mountains. However mt-hap C may have gone all the way with Q."

That's nonsense. Y-DNA haplogroup Q is found throughout Asia, including Pakistan, Vietnam, China, Mongolia, and Korea, not only in Central Asia and Siberia. In Asia and the Americas, it is almost always found mixed up with Y-DNA haplogroup C3-M217; the Q+C3 Y-DNA mix is obviously the original Y-DNA of northern Asians (and Native Americans, which are a derived subset of prehistoric northern Asians). The only part of Asia in which Y-DNA haplogroup Q is clearly not found alongside Y-DNA haplogroup C3 is India (Q, C5, and C*, but no C3).

Maju said...

You seem prepared to concede that India's haplogroup diversity as compared with Australia's is partly a product of immigrant haplogroups to India yet, in contradiction, you claim most haplogroups originated in India.

This is not contradictory. Just means back-migration. Some haplogroups of South Asia, notably Y-DNA ones (and notably R1a among them) are with all likehood recent back-migrations dating to the Holocene. MtDNA-wise there is also something of that but much less.

Re. Y-DNA Q. I think Ebizur has responded to you pretty well.

Lastly, would you like to tell us all here how common mt-hap Y is in India? Or S, or X, or A. Yet you believe they all originated there.

I do not claim that mtDNA Y, S, X or A originated in India. That would be against all logic. Only their ancestors did, probably.

Why do you have to twist my discourse like that? Is that you did not truly understand or just that you are looking desperately for any kind of "weakness" in any form, even where they do not exist at all?

terryt said...

"I think Ebizur has responded to you pretty well".

Earlier he wrote:

"Outside of North America, haplogroup Q-M242(xQ1a3a-M3) occurs most frequently in Siberia, among Kets, Selkups, Chukchis, Yukaghirs, and Altayans". And "I should add that somewhere between 19.0% and 35.3% of Nivkhs, indigenous inhabitants of the lower Amur River basin and northern Sakhalin Island, belong to haplogroup P(xR1a). These Nivkhs probably also belong to haplogroup Q".

There are a couple of good maps showing the distribution of various ethnic groups' distribution, during the early 17th and late 20th centuries, in Yuri Slezkines (1995) book "Arctic Mirrors" (Cornell University Press, USA). Once you check these maps you will see that none of the above groups are actually found in Mongolia.

Therefore, although "Y-DNA haplogroup Q is found throughout Asia, including Pakistan, Vietnam, China, Mongolia, and Korea" this in no way precludes it from being an immigrant into these regions. Therefore "the Q+C3 Y-DNA mix is obviously" [not necessarily] "the original Y-DNA of northern Asians". I'll grant that Q and C3 were near neighbours. Of course if you insist on using the magic words 'bottleneck', 'drift' and 'founder effect' you may be able to make a very complicated case in support of your theory.

" Only their ancestors did, probably".

OK. We still have India as a centre for the vast majority of haplogroups' origin. So let's go even bigger: to Africa, specifically the Sahel region. Australia fits into the Sahara Desert with plenty of room to spare. And remember that the Sahara has not always been desert, and also don't forget the Thar Desert of Northern India.

Birkina Faso, representative of the Sahel if any single country can be, has almost entirely Y-hap E (McDonald map, perhaps too old for you to take seriously being from 2005). Even the whole of North Africa is mostly versions of E. Obviously E has diversified considerably, and the defining letters are merely conventions, but it's still basically just a single haplogroup. Apart from a few early surivors (A and B) the main variation in the E pattern occurrs where it's fairly obvious there has been an influx from outside Africa. So even a region as large and ancient as the Sahel has basically been reduced to just one haplogroup. And yet you still believe that a relatively large diversification of Y-chromosomes occurred in little old India?

I'll grant Y-haps H and L, and later the ancestor of P evolved there. All others could easily have come in from outside. Even F probably moved into India, unless it evolved there after being carried in as C/F.

I'll also grant that Y-hap F and mt-hap M could well have entered India together. Drift probably then reduced them to being the only haps there before Y-hap K and mt-hap R began their expansion.

Oh, and I earlier left out another example of your inconsistencies and contradictions: your insistence that people came out of Africa with boats capable of negotiating Wallacea but incapable of reaching any Agean Islands. Difficult to explain that.

Maju said...

Therefore, although "Y-DNA haplogroup Q is found throughout Asia, including Pakistan, Vietnam, China, Mongolia, and Korea" this in no way precludes it from being an immigrant into these regions. Therefore "the Q+C3 Y-DNA mix is obviously" [not necessarily] "the original Y-DNA of northern Asians".

Ok. Agreed. Along the milenni in those cold high latitudes there were surely many small nomadic groups, each one once and again tending to fixation in a single clade until remix with their neighbours. It's I believe very hard to estabilish a simple model in such context of extremely low density and maxed nomadism. But I do think that C3 and Q are among the oldest Y-DNA clades in Siberia and nearby lands, and that's why these two clades are the ones that were in Berigia soon enough to parcipate in the colonization of America.

Obviously E has diversified considerably, and the defining letters are merely conventions, but it's still basically just a single haplogroup.

As much as F (or C or D). These clades are syblings/cousins.

The concept of "single haplogroup" means nothing except the existence of a common paternal (or maternal for mtDNA) ancestor some time far away in the remote past. It is obviously not the same a shared ancestor of level zero, like the "Y-DNA Adam", shared by all living humans than a shared ancestor at Nth level like the one shared by all people with Y-DNA R1b1b2ba1b4c1, who are rather few people mostly restricted to some narrow geographical area.

Well, C, D, E and F nodes are closer to "Adam" than to the present.

Approximatively E1 and E2 are comprable to the different subclades of F (F1, F2, F3, F4, G, H and IJK), E1a and E1b are comparable to IJ and K, for instance, and E1b1b (the "Mediterranean" variant of it) would be roughly comparable to J2 or R. Of course the comparison is just a rough approximation, specially because the different clades are not researched at similar depth (not at all) and I'm just considering the (known) branching nodes' sequence and not the possible age of each clade, but I warn you not to get confused by the "Eurocentric" nomenclature that uses simple letter names for the derivates of F and K and not for the other branches. K could well be called "F7a" and R1b would then be something like "F7a8b1b" and that would give a fairer impression of their temporal correlation with clades like E1b1b or C3a - always very rough anyhow.

So even a region as large and ancient as the Sahel has basically been reduced to just one haplogroup. And yet you still believe that a relatively large diversification of Y-chromosomes occurred in little old India?

The other survivors also count, their mere existence yields a very different diversity measure than if they would have gone extinct or never existed at all. E is not more a "single haplogroup" than F is, which extends over an even much larger area and numbers of people. And South Asia is anything but "little".

South Asia (SAARC member countries) has 5 million Km2 (mostly inhabitable and fertile), Africa has 30 million Km2 (but a big deal are deserts), Europe has 10 million Km2 (but half of it was uninhabited in the Ice Age). For your reference, New Zealand has only 270,000 Km2 and Australia 7.7 million Km2 (mostly desertic).

South Asia (SAARC) is large and fertile enough to host a lot of people, even with the very low densities of Paleolithic. It has other advantages, like a mostly tropical climate and few deserts. I would also argue that when a new population arrives to a virtually "empty" large area, they expand first of all, what will naturally lead to genetic diversification in due time, specially as the subcontinent is large enough for the scattered peoples to live mostly apart from each other in different areas.

But I do not argue for South Asia alone. I do not exclude that areas of SE Asia and even West Asia were also involved to some extent. What I do argue for is for the early colonization of Eurasia happening almost exclusively along the tropical/subtropical southern strip, in which South Asia necesarily played a central pivotal role.

I'll grant Y-haps H and L, and later the ancestor of P evolved there.

L and P are "syblings", H would be their "uncle", so to say.

All others could easily have come in from outside.

What others? O? Yes, O probably came from SE Asia in the Neolithic. But where was the common ancestor of L, P, O and so many others? Where did K, and IJK before it, coalesce? I'd say that in South Asia too (though the discovery of IJK may suggest rather towards Iran than towards Myanmar - unsure).

I'll also grant that Y-hap F and mt-hap M could well have entered India together. Drift probably then reduced them to being the only haps there before Y-hap K and mt-hap R began their expansion.

Then how do you explain:

- The presence of Y-DNA C* and C5 in South Asia
- The widespread distribution of mtDNA N both east and west of the subcontinent, as well as within it
- The fact that mtDNA R (derived from N, not M) strongly appears to have coalesced in South Asia

And where did the omnipresent Y-DNA K arose according to you? In New Guinea's highlands? How then did L, P and T make it to West Eurasia? How then do you explain that K is derived from IJK and its sybling branch is almost only present in West Eurasia?

It is much simpler to think of a more or less long period of diversification and coalescence into the different clades in "Greater South Asia" and then one or more migrations out of it, followed by local processes of homogeneization (drift) and also some back-migrations (but hardly C). That "out-of-India" moment/s is when Eastern and Western Eurasians (as well as Southern Asians, Melanesians and Australian Aboriginals) diverged, but not before having shared a good deal of prehistory more or less together in the subcontinent or near it. This sharing is what caused that C, D and F (K included), as well as M and N (R included) are so much mixed in Eastern Eurasia.

There's no reason to think in many different migrations towards the East (with the likely exception of Y-DNA Q, and possibly related mtDNA X2, that do seem to have gone via the north) or at least not migrations that would be too spaced in time, but rather peoples that would "remix" in the process of expanding to SE Asia and then to Eastern Asia and Sahul.

Oh, and I earlier left out another example of your inconsistencies and contradictions: your insistence that people came out of Africa with boats capable of negotiating Wallacea but incapable of reaching any Agean Islands. Difficult to explain that.

First Most of Indonesia was part of the continent then. Second the peoples who went to SE Asia are not the same as those who went to Europe, many many generations separated them already. This migration may be called "fast" but it's only "fast" if we think of 10-20,000 years, something like 500-1000 generations, as "fast". This kind of "fast" is five or ten times slower than all our historical age, 50 or 100 times slower than Modernity and 100 or 200 times slower than the European colonization of most of the USA. Their relatedness would be equivalent to that of modern people descending from whoever developed the Solutrean retouch. You know what this means, right? Lost of memory, very diferent histories in all that time, very different lifestyles and even technologies. Even in such a conservative context as the Paelolithic, things change a lot in 10 or 20 milennia. They could be as related as a Tuareg and a Pole. And obviously Tuaregs have forgotten how to use canoes, while the Polish are probably now building huge transcontinental ships (or at least were a few years ago).

"Fast" but not so much, ok?

Maju said...

In other words, I presume that the peoples that headed to the west had a more inland way of life mostly, while those who went to the east, crossing the mangroves of Bangla Desh and Burma, had a more seagoing lifestyle overall, ability that was increased as they reached the Wallacean islands by mere need.

But I still don't know how they managed to reach Australia in a mere canoe. And that's the part that I'd like to see explained, there's absolutely no mystery in refusing to adventure into the deep seas without knowing what is beyond, moreso with a mere canoe, the mystery (and the feat) is to do it, as apaprently the ancesors of Melanesians and Australian Aborigines did.

Obviously they did it somehow but do you have an explanation?

terryt said...

"Obviously they did it somehow but do you have an explanation?"

I doubt that it was a canoe, especially not a dugout one. Humans almost certainly developed that later, perhaps more recently than 10,000 years ago. Much more likely, and pretty much accepted in this part of the world, is rafts of some sort. Perhaps just bundles of reeds, bark or bamboo tied together.

As for not knowing what was beyond. Many islands are within sight of each other and so the beginning of the technology is not hard to understand. However it seems that the much later Polynesians came to believe that their hero/god Maui would fish up islands ahead of them. The legend would survive because those people for whom he failed to fish up an island failed to survive to prove the myth wrong. I think we both agree that mythology is a fascinating subject.

"But I do think that C3 and Q are among the oldest Y-DNA clades in Siberia and nearby lands, and that's why these two clades are the ones that were in Berigia soon enough to parcipate in the colonization of America".

I totally agree that Q derives from P and so ultimately came from somewhere in the North of the Indian subcontinent. However I strongly suspect that C is the older clade in Siberia, although originally not as far north as Q later moved. And I don't accept for a moment that Q and C moved into America together. C moved into Northern Siberia,and so to America, after Q had zoomed past. As far as I know there is no C in South America, not what we'd expect if the haplogroups had simply been subject to drift there. Besides which Q and C3 merely overlap in places, not coincide.

"K could well be called 'F7a' and R1b would then be something like 'F7a8b1b'".

Exactly. I totally agree with your further analysis of haplogroup formation. But I think Y-hap E has actually been almost as well researched as other European haplogroups. After all it is a common Jewish haplogroup and many people of Jewish extraction are involved in the research. I presume you've seen Hammer's E haplogroup diagram. If not I'll track it down for you.

"But where was the common ancestor of L, P, O".

Obviously they all share a few mutations with K, and K is derived from one of the Fs. The diagram I can see shows 4 mutations at the root of K. Obviously they needn't have all happened at the same time but they must each have happened in single individuals. These individuals could have lived anywhere within F's huge geographic spread, possibly India but not necessarily so. However we have to have an explanation for K's huge expansion. The men must have possesed some extremely useful advantage to be able to spread so widely and rapidly. Have you any nominations for what this advantage may have been? I think you already know my answer.

"The presence of Y-DNA C* and C5 in South Asia".

C* is especially common around what would have been at time a landlocked South China Sea. C* could easily have moved to India from there, and C5 derived from it in India.

"The widespread distribution of mtDNA N both east and west of the subcontinent, as well as within it".

And north of it. If you take out all the R-derived haplogroups N is not very common at all in India. And of its earlier divisions haps I and W are much more likely to be relatively recent arrivals, perhaps with the Neolithic. Within India N itself possibly also belongs to this time.

"The fact that mtDNA R (derived from N, not M) strongly appears to have coalesced in South Asia".

Possibly, but not necessarily.

"How then did L, P and T make it to West Eurasia?"

Ah. Now we agree. Through India.

"How then do you explain that K is derived from IJK".

IJ is a 'brother' clade to all the Ks including L, M, NO, PQR, S and T. They each developed locally after the indivuals with the basal mutations had already become widely spread, from New Guinea, through India to at least the Iranian Plateau.

"It is much simpler to think of a more or less long period of diversification and coalescence into the different clades in 'Greater South Asia'".

It's actually much simpler to think of a long period of back and forth migration rather than diversification in a single region. The most common Y-hap in New Zealand would be some variation of R, European. However a substantial minority cary versions of C, Polynesian. I doubt very nuch that either you or Ebizur would vehemently maintain the two haplogroups had formed within some single population before entering New Zealand. Yet something similar seems to form the basis of your interpretations of ancient haplogroup diversity.

"the peoples who went to SE Asia are not the same as those who went to Europe, many many generations separated them already".

But they were the same when they left Africa. Surely if They'd had boating technology it would have been useful for crossing slow-moving rivers, so why would they have lost it?

Maju said...

I doubt that it was a canoe, especially not a dugout one. Humans almost certainly developed that later, perhaps more recently than 10,000 years ago.

That's absurd. All "primitive" peoples around the world use canoes of some sort, most often dug out ones.

Much more likely, and pretty much accepted in this part of the world, is rafts of some sort. Perhaps just bundles of reeds, bark or bamboo tied together.

I accept rafts as alternative but you won't persuade me that people crossed wide rivers and swamps in mere improvised poorly built artifacts. Much less that they crossed the c. 60 km that separated Timor and Australia in the Ice Age. They must have already been somewhat seagoers by then, even if the first discovery of Australia was probably by accident (why else would anyone boat so far away from the coast?).

As for not knowing what was beyond. Many islands are within sight of each other and so the beginning of the technology is not hard to understand. However it seems that the much later Polynesians came to believe that their hero/god Maui would fish up islands ahead of them.

Same for the Ganges Delta or many other places where canoing would be at least strongly advisable. You are obsessed with the mere fact of boating being invented in Indonesia but by then humans had already some 150 or maybe as much 200,000 years of (pre-)history behind them, and humans are not dumb, you know. If they need to cross a river, a lake or a swamp, they would soon figure out one and many ways of doing it. And just to arrive as far east they must have boated more than once.

You are obsessed with a different idea and it seems it's not possible to persuade you. Fine.

Polynesians are different: they were a much more advanced seagoing people who knew how to sail the oceans.

However I strongly suspect that C is the older clade in Siberia

C3? Maybe.

These individuals could have lived anywhere within F's huge geographic spread, possibly India but not necessarily so.

Not just that. In fact F and K could well have been drifted out in their place of origin in all that time (though not too likely) but also K must have originated in a place from where reaching all the rest of areas where it's found historically would be realistic (and preferably without contradicting other genetic or archaelogical info available). So far the only reasonable candidate is Southern Asia, because it's the only thing that stands between East and West Eurasia (except the narrow, cold and arid steppary corridor), because Eastern and Western Eurasians have different evolutionary adaptations to high latitudes (pygmentation but not only) and because the archaeoloical chronology for arrival to Australia, for example, pre-dates that of arrival to middle East Asia (so hardly any northern route).

You are building on thin air just because you are obsessed with the idea that boating is not ancestral to Humankind since Africa but that somehow it must have been developed in Indonesia. That's just a hypothesis you fail to provide any evidence for and to which you try to make the rest of the data to fit even if it's impossible.

You're building the house beginning with the roof and that doesn't work. First the cimentations, first make sure that your *belief* in canoing only being invented upon arrival to Indonesia is true and not just a trick of your imagination. Once you dispel that obsession you can see the rest unfolding naturally and don't make us all waste so much time arguing once and again on the same issues.

However we have to have an explanation for K's huge expansion. The men must have possesed some extremely useful advantage to be able to spread so widely and rapidly. Have you any nominations for what this advantage may have been? I think you already know my answer.

No I have no idea which is your answer nor I'm persuaded that the "K men" had any particular advantageous attribute. Drift happens. When I look at high K Papuans, for example, I see not what could be such advantage in comparison to, say, low K Mongols, Jews or whatever. But I must have some prejudice against Papuans, I guess - their customs kinda irk me.

C* is especially common around what would have been at time a landlocked South China Sea. C* could easily have moved to India from there, and C5 derived from it in India.

C* is not a haplogroup nor is C-root (i.e. before diverging into C1, C2, C3, C4, C5...) C* is just what has not been classified within C - it could well have many SNP defined haplogroups.

Anyhow, was not there in this lenghty discussion some other piece of info that supported migration from India to Australia (and not vice versa)?

N is not very common at all in India.

Let's see: N is not very common anywhere - except as derived R and A. Nevertheless (excluding R):

· West Eurasia: two clades, one shared with Australia (some of the Australian N* is more related to W than to any other clade, including the rest of the Australian N*)
· South Asia: unknown number of clades: N* only
· Australia: S and an unknown number of clades known as N*, some of which are closerly related to West Eurasian W
· East Asia: two clades (A and N9/Y)

So nowhere you have more than two defined clades of N (excluding R) and in some areas it's really unknown (N*).

"The fact that mtDNA R (derived from N, not M) strongly appears to have coalesced in South Asia".

Possibly, but not necessarily.


R is the more clear of all mtDNA macrolineages to call for a South Asian homeland. Not just geography but South Asia also has the largest number of R subclades anywhere.

IJ is a 'brother' clade to all the Ks including L, M, NO, PQR, S and T.

I think you don't get the concept: IJ and K separated first, then K divided and spread (the same that IJ split into I and J). IJ is not sybling of L, M, etc. but their "uncle". There's no trace of any sort linking IJ to Eastern Eurasia, the highest diversity of IJK for that geneaogical level is in West Asia. I can accept South Asia (and in fact I think it's more likely, with IJ being the product of an early branch prior to the split of K) but cannot really accept Eastern Eurasia: it just makes no sense.

They each developed locally after the indivuals with the basal mutations had already become widely spread, from New Guinea, through India to at least the Iranian Plateau.

And hey all evolved into K from IJK in all those separate places simultaneously? Sorry, no way. The possibilities of an SNP happening twice are much lower than getting the big prize in lottery. Happening in maybe seven or ten different places at the same time would really be a miracle.

It's actually much simpler to think of a long period of back and forth migration rather than diversification in a single region.

It's not. Though maybe it depends on what you consider "simple".

Anyhow, back and forth migration when there's already other people like yourself (roughly same technology, same intellectual capability, similar societies and economy) in the place of destination implies either being rejected or being absorbed. It's not as simple as moving into "brand new" areas.

The normal result for a small band moving into the territory of consolidated peoples would be either expulsion or assimilation by the natives. You would need real hard evidence of such migrations happening (something clear like Aurignacian, you know) to really have some solid ground. But you're building all in your mind, with no evidence of any sort.

It's tiresome after 92 posts (only in this thread), really.

New Zealand

You have to understand that the mass migrations of modernity or imply a radical technological advantage that allows (almost demands) massive genocide. Those of modernity are the most impressive, yet the ones that are more distant from the conditions of the Paleolithic.

If instead of modern peoples the newcomers would have got a similar tech to Polynesians, they would have faced many bigger difficulties. Think maybe in the Vikings in Ireland... most Irish have little or no Viking ancestry nowadays - because there was no such technological abyss.

In brief you're comparing apples an washing machines: they have nothing to do with each other.

And I'm strating to detect some deep colonialist bias in your discourse anyhow.

But they were the same when they left Africa. Surely if They'd had boating technology it would have been useful for crossing slow-moving rivers, so why would they have lost it?

They never lost it I think. Just that some peoples apear to have become even more seagoers. Again the problem is not in Mediterraneans not rowing like crazy towards the horizon until they probably got lost... but in Wallaceans doing it against all apparent logic. That's why I think that (a) they were more familiar in general with the sea and (b) they were driven to Sahul accidentally first of all (but were able to return and tell the story, maybe long before anyone ever settled permanently Australia or New Guinea).

Ebizur said...

Maju said,
"But I do think that C3 and Q are among the oldest Y-DNA clades in Siberia and nearby lands, and that's why these two clades are the ones that were in Berigia soon enough to parcipate in the colonization of America".

terryt said,
"I totally agree that Q derives from P and so ultimately came from somewhere in the North of the Indian subcontinent. However I strongly suspect that C is the older clade in Siberia, although originally not as far north as Q later moved. And I don't accept for a moment that Q and C moved into America together. C moved into Northern Siberia,and so to America, after Q had zoomed past. As far as I know there is no C in South America, not what we'd expect if the haplogroups had simply been subject to drift there. Besides which Q and C3 merely overlap in places, not coincide."

Several studies, including that by Karafet et al. (2002), have found that haplogroup Q is the most diverse Y-DNA haplogroup in indigenous Siberian populations, followed closely by haplogroup C3. To me, at least, this suggests that haplogroup Q-bearing groups "dragged along" some haplogroup C3 people during the initial populating of Siberia. These haplogroup C3 people also reached the Americas; haplogroup C3 has been found at low frequency even in South America, such as among the Wayuus, but it is more commonly found among populations of North America, especially those which speak an Algonquian, Siouan, or Dene (AKA Athabaskan) language. In any case, I think the mtDNA landscape of the Americas and northern Asia is best associated with the haplogroup C3 component of their Y-DNA pools; haplogroup Q is too closely related to haplogroups that are found in populations whose mtDNA has nothing in common with indigenous populations of northern Asia and the Americas.

Maju said...

In any case, I think the mtDNA landscape of the Americas and northern Asia is best associated with the haplogroup C3 component of their Y-DNA pools; haplogroup Q is too closely related to haplogroups that are found in populations whose mtDNA has nothing in common with indigenous populations of northern Asia and the Americas.

Hmmm. What do you mean?

It does seem that some Q, always as minority clade, is found in West and South Asia. And while Q1a3a is the typical Native American clade, its closest relative, Q1a3*, is found only in South Asia (albeit at low frequency).

Would you think with me that even if Q (as P-derived and also because minor presence in West and South Asia) had a relatively western origin, Q1a3a should have migrated in NE direction along with other Q1a clades, which are mostly found in Northern and East Asia?

terryt said...

"All 'primitive' peoples around the world use canoes of some sort, most often dug out ones".

Far from true. Do you know anything at all about boating?

"They must have already been somewhat seagoers by then".

Quite. And where is the best place to learn that skill?

"You are obsessed with a different idea and it seems it's not possible to persuade you".

The same is far more true of you. You seem completely unable to understand that technology has always been an accumulating skill. I give up. I'm wasting my time with you.

terryt said...

"And hey all evolved into K from IJK in all those separate places simultaneously?"

Deliberate misinterpretation? Mind you your earlier comment:

"And hey all evolved into K from IJK in all those separate places simultaneously?"

Shows you don't even begin to understand how evolution actually happens. You are obsessed with some 'single point of origin' theory. Based on you childhood brainwashing?

My last comment. Promise.

terryt said...

However there's still someone here worth communicating with.

Ebizur wrote:

"I think the mtDNA landscape of the Americas and northern Asia is best associated with the haplogroup C3 component of their Y-DNA pools; haplogroup Q is too closely related to haplogroups that are found in populations whose mtDNA has nothing in common with indigenous populations of northern Asia and the Americas".

Exactly. Y-haps and mt-haps are often surprisingly independent. As Y-hap Q moved east across Asia it accumulated mt-haps from the south of their route, originally associated with Y-hap C. All the evidence points that way.

"haplogroup Q is the most diverse Y-DNA haplogroup in indigenous Siberian populations".

Suggests an already well diversified and reasonably large migrating population. Less diversity in the local C suggests a longer period spent as a reasonably small population under selection, or drift if you prefer.

"it is more commonly found among populations of North America, especially those which speak an Algonquian, Siouan, or Dene (AKA Athabaskan) language".

There's a strong possibility these language groups are slightly later arrivals. Certainly it's accepted that the southerly versions, Apache and Navajo, are recent arrivals in their present locations.

I wasn't aware of Y-hap C in South America. Thanks.

Ebizur said...

Maju said,

"Hmmm. What do you mean?"

I think you know what I mean. Haplogroup Q-M242 and haplogroup R-M207 are both subclades of haplogroup P-M45 (or QR-M45 or whatever you may call it). Haplogroup P-M45 is a subclade of haplogroup K-M9, which is a subclade of haplogroup IJK-L15, which is a subclade of haplogroup F-M89.

Maju said,
"It does seem that some Q, always as minority clade, is found in West and South Asia. And while Q1a3a is the typical Native American clade, its closest relative, Q1a3*, is found only in South Asia (albeit at low frequency)."
The problem is how to distinguish original "Western" haplogroup Q individuals from recent Central Asian migrant haplogroup Q individuals. I know that haplogroup Q1a3-M346(xQ1a3a-M3) has been found in a couple of Pashtuns in Pakistan, but I don't know where in India this haplogroup has been found. Anyway, haplogroup C3-M217 has also been found in northwestern South Asia (including at least one Pashtun in Pakistan from the same sample that found the Q1a3(xQ1a3a) individuals), so it is possible that the haplogroup Q1a3(xQ1a3a) found in this region might also have been introduced by an already admixed C3+Q population from Siberia or northern East Asia. We really need to see more deep clade testing of haplogroup Q and haplogroup C3 individuals from other parts of Asia.

Maju said,
"Would you think with me that even if Q (as P-derived and also because minor presence in West and South Asia) had a relatively western origin, Q1a3a should have migrated in NE direction along with other Q1a clades, which are mostly found in Northern and East Asia?"
I can't wholeheartedly agree with this hypothesis yet, because the resolution of haplogroup Q-M242 in most previously published studies of Asian populations is so poor. Someone needs to perform more thorough testing of haplogroup Q people from Central Asia, Siberia, Mongolia, Vietnam, etc. (though I might speculate that the haplogroup Q in Vietnam will turn out to be mostly Q1a1-M120).

Ebizur said...

Wow, this is interesting. Sharma et al. (2007) write, "Recently, a novel subgroup Q4 has been identified which is defined by bi-allelic marker M346, representing HG Q (0.41%, 3/728) in Indian population." This appears to be a reference to Sengupta et al. (2006), who have reported finding "Q4-M346" (now Q1a3-M346(xQ1a3a-M3)) in 1/14 Uttar Pradesh Brahmins (North India), 1/31 Vellalars (Tamil, South India), 1/21 Halbas (IE-speaking tribe, Central India), 2/20 Pathans (SE Iranic-speaking tribes, North Pakistan), and 1/20 Makrani (NW Iranic, South Pakistan). The only one of these samples to contain any haplogroup C3 (or haplogroup C in general) is the Pathan sample: 1/20 C3-M217.

Ebizur said...

Here are some data I have collected in regard to the frequencies of haplogroups C(3) and Q among Asian populations (I have tried to avoid using samples in which haplogroup C or haplogroup Q was found in only a single individual, but I made exceptions for a few samples, like the Anhui, China one, whose total sample sizes were very small):

Kets (Tambets et al. 2004):
3/48 = 6.25% C
45/48 = 93.75% Q
C+Q total: 48/48 = 100%

Selkups (Tambets et al. 2004):
2/131 = 1.5% C
87/131 = 66.4% Q
8/131 = 6.1% R1(xR1a)
25/131 = 19.1% R1a
C+Q total: 89/131 = 67.9%

Northern Han (Hammer et al. 2005):
2/44 = 4.5% C3-M217(xC3c-M86)
2/44 = 4.5% Q-P36
1/44 = 2.3% R-M207
C3+Q total: 4/44 = 9.1%

Han/Gansu (Wen Bo et al. 2004):
7/60 = 11.7% C-M130
3/60 = 5.0% Q1-M120
1/60 = 1.7% P-M45(xQ1-M120)
C+Q1 total: 10/60 = 16.7%


Han/Zhengzhou, Henan (Wen Bo et al. 2004):
2/50 = 4.0% C-M130
4/50 = 8.0% Q1-M120
1/50 = 2.0% P-M45(xQ1-M120)
C+Q1 total: 6/50 = 12.0%

Han/Chifeng, Inner Mongolia (Wen Bo et al. 2004):
12/60 = 20.0% C-M130
2/60 = 3.3% Q1-M120
C+Q1 total: 14/60 = 23.3%

Han/Laizhou, Shandong (Wen Bo et al. 2004):
14/85 = 16.5% C-M130
4/85 = 4.7% Q1-M120
C+Q1 total: 18/85 = 21.2%

Han/Jinan, Shandong (Wen Bo et al. 2004):
4/100 = 4.0% C-M130
3/100 = 3.0% Q1-M120
C+Q1 total: 7/100 = 7.0%

Han/Xi'an, Shannxi (Wen Bo et al. 2004):
2/63 = 3.2% C-M130
1/63 = 1.6% Q1-M120
2/63 = 3.2% P-M45(xQ1-M120)
C+Q1 total: 3/63 = 4.8%

Han/Urumqi, Xinjiang (Wen Bo et al. 2004):
2/51 = 3.9% C-M130
2/51 = 3.9% Q1-M120
2/51 = 3.9% P-M45(xQ1-M120)
C+Q1 total: 4/51 = 7.8%

Han/Hefei, Anhui (Wen Bo et al. 2004):
3/22 = 13.6% C-M130
1/22 = 4.5% Q1-M120
C+Q1 total: 4/22 = 18.2%

Han/Nanjing, Jiangsu (Wen Bo et al. 2004):
6/100 = 6.0% C-M130
2/100 = 2.0% Q1-M120
C+Q1 total: 8/100 = 8.0%

Han/Shanghai (Wen Bo et al. 2004):
4/55 = 7.3% C-M130
2/55 = 3.6% Q1-M120
C+Q1 total: 6/55 = 10.9%

Han/Huize, Yunnan (Wen Bo et al. 2004):
4/66 = 6.1% C-M130
2/66 = 3.0% Q1-M120
C+Q1 total: 6/66 = 9.1%

Han/Hangzhou, Zhejiang (Wen Bo et al. 2004):
10/106 = 9.4% C-M130
2/106 = 1.9% Q1-M120
C+Q1 total: 12/106 = 11.3%

Tujia (Hammer et al. 2005):
3/49 = 6.1% C-RPS4Y(xC1-M8, C2-M38, C3-M217)
9/49 = 18.4% C3-M217(xC3c-M86)
2/49 = 4.1% R-M207
(This Tujia sample does not contain any haplogroup Q individuals, but it does contain two types of haplogroup C in addition to a couple haplogroup R individuals)

Manchurian Evenk (Hammer et al. 2005):
4/41 = 9.8% C3-M217(xC3c-M86)
14/41 = 34.1% C3c-M86
4/41 = 9.8% Q-P36
2/41 = 4.9% R-M207
C3+Q total: 22/41 = 53.7%

Evenk (Hammer et al. 2005):
13/95 = 13.7% C3-M217(xC3c-M86)
52/95 = 54.7% C3c-M86
4/95 = 4.2% Q-P36
1/95 = 1.1% R1a
C3+Q total: 69/95 = 72.6%

Even (Hammer et al. 2005):
4/31 = 12.9% C3-M217(xC3c-M86)
19/31 = 61.3% C3c-M86
1/31 = 3.2% Q-P36
2/31 = 6.5% R1a
C3+Q total: 24/31 = 77.4%

Vietnam (Hammer et al. 2005):
3/70 = 4.3% C3-M217(XC3c-M86)
5/70 = 7.1% Q-P36
1/70 = 1.4% R-M207
C3+Q total: 8/70 = 11.4%

Uygur (Hammer et al. 2005):
1/67 = 1.5% C-RPS4Y(xC1-M8, C2-M38, C3-M217)
3/67 = 4.5% C3-M217(xC3c-M86)
1/67 = 1.5% C3c-M86
2/67 = 3.0% Q-P36
31/67 = 46.3% R-M207
C+Q total: 7/67 = 10.4%
C3+Q total: 6/67 = 9.0%

Mongolia (Hammer et al. 2005):
51/149 = 34.2% C3-M217(xC3c-M86)
27/149 = 18.1% C3c-M86
4/149 = 2.7% Q-P36
6/149 = 4.0% R-M207
C3+Q total: 82/149 = 55.0%

Mongolia (Xue et al. 2006):
22/65 = 33.8% C3(xC3c)
13/65 = 20.0% C3c
3/65 = 4.6% P(xR1a)
6/65 = 9.2% R1a
(a different sample of Mongolians for comparison)

Altai (Hammer et al. 2005/Karafet et al. 2002/Tambets et al. 2004):
17/98 = 17.3% C3-M217(xC3c-M86)
5/98 = 5.1% C3c-M86
17/98 = 17.3% Q-P36
46/98 = 46.9% R1a
C3+Q total: 39/98 = 39.8%

Sri Lanka (Hammer et al. 2005):
3/91 = 3.3% C-RPS4Y(xC1-M8, C2-M38, C3-M217)
3/91 = 3.3% Q-P36
15/91 = 16.5% R-M207
C+Q total: 6/91 = 6.6%

Yukaghir (Pakendorf et al. 2006):
4/13 = 30.8% Q-P36
1/13 = 7.7% C3-M217(xC3c-M48/M86)
3/13 = 23.1% C3c-M48/M86
C3+Q total: 8/13 = 61.5%

Tuvan (Pakendorf et al. 2006):
5/55 = 9.1% C3c-M86
3/55 = 5.5% C3-M217(xC3c-M48/M86)
9/55 = 16.4% Q-P36
4/55 = 7.3% R1b1b1-M73
12/55 = 21.8% R1a1-M17
C3+Q total: 17/55 = 30.9%

Ebizur said...

By the way, Bo Wen et al. (2004) did not test for C3-M217. I assume that practically all the haplogroup C Y-DNA found among Han Chinese should be haplogroup C3-M217(xC3c-M48), because Hammer et al. (2005) and Xue et al. (2006) have not found any other sort of haplogroup C Y-DNA among Hans.

I have more data on populations in North Asia, East Asia, and Central Asia, but, unfortunately, most researchers have not tested for a marker of haplogroup Q, let alone any of its Asian subclades. I don't know when researchers will quit publishing figures for "P(xR1a)" and "Y*(xA, C, DE, J, K)"!

Anyway, one fact that has, unfortunately, been swept under the rug is that haplogroup C3 and haplogroup Q are found throughout China and even into at least Vietnam in Southeast Asia. In fact, the world's greatest number of haplogroup C3 (and haplogroup C in general) individuals must be found in China. Someone needs to explain how all these haplogroup C3 and haplogroup Q individuals have ended up in China and Vietnam.

terryt said...

"Someone needs to explain how all these haplogroup C3 and haplogroup Q individuals have ended up in China and Vietnam".

Thanks for your interesting data. What is reasonably consistent is that, apart from some far north groups, both C and Q form a minor percentage in most of the groups. What's more the proportion basically decreases as we move south. This, and the geographic distribution, hints that both may have been introduced during Ghengis Khan's rule. After all one branch of Y-hap C is often called the Ghengis Khan Y-chromosome, and it's unlikely he was the only member of his army to leave sons scattered around the place. And his army contained more than just Mongols. Interesting that Q is found in Sri Lanka though.

Maju said...

"All 'primitive' peoples around the world use canoes of some sort, most often dug out ones".

Far from true.


Evidence, data? Maybe those living in the desert do not... but that's about all.

Do you know anything at all about boating?

I'm not a "sea dog" at all but I can drive a boat, I'm licensed to do it and have used it profesionally in the past. My family has a long history of direct relation with sea activities as well. I have never lived far from the sea (except when I studied in the USA). Is that enough for you or you need a gold medal in the Olympics to allow me to give an opinion?

Quite. And where is the best place to learn that skill?

Quiet waters to begin with, not in the middle of nowhere with only salty and possibly stormy water around you. Yes or yes?

Doing heroic mariner feats belongs only to the very advanced state. Crossing into Australia from Timor probably compares to what Red Erik, Colombus or Elkano acomplished in later times. Obviously these were only the apogee of a much longer tradition of marinerism, not the first sailors of Earth.

You seem completely unable to understand that technology has always been an accumulating skill.

Well, it's you who claim that archaic Indonesians not only crossed into Sahul (fact) but that they just invented the boat the day before to do such a daunting trip (or almost). Your claim compares to make Neil Armstrong (or von Braun) the inventor of fireworks. We all know that rockets and fireworks were invented by the Chinese long before, yet they have just begun to get into the space race. Equally the boat could well have been (and surely was) developed in Africa prior to the OOA epysode; the people who crossed Wallace line were at the high end, not the humble beginnings, of this art.

"And hey all evolved into K from IJK in all those separate places simultaneously?"

Deliberate misinterpretation? Mind you your earlier comment


Not deliberate certainly but maybe a misread. You said:

IJ is a 'brother' clade to all the Ks including L, M, NO, PQR, S and T. They each developed locally after the indivuals with the basal mutations had already become widely spread...

I understood that you meant that L, M, NO, P, S, etc. evolved directly from IJK basal mutations, each one developing the K-specific mutations separately, what obviously makes no sense.

But if you meant that they developed after K expansion (what I agree with), your reply was of no use because my question was about the origin of K, not of its subclades.

The question then stands: IJK has two branches West Eurasian IJ and extremely widespread K. What does this suggest about the coalescence of K? I'd say not far from West Eurasia, though South Asia is a serious possibility anyhow. Hardly New Guinea in any case.

You are obsessed with some 'single point of origin' theory. Based on you childhood brainwashing?

Man: I quitted church at 11, was a quite active atheist at 13, a political activist at 15, and have never been worried at all about descendancy issues (I have no children nor I plan to have them nor I have ever been really interested in them - I believe that the really important stuff happens in the mind, not in the genes). I have no idea of what you're talking about. Were your raised on Zen Buddhism or something? Don't throw stones on your own roof, don't project on me which are probably your own issues (sooo typical).

What is clear is that each haplogroup, each SNP, means a single point of origin (some unique man for Y-DNA, and some unique woman for mtDNA). Nevertheless the ones spreading it may have well been relatively distant descendants anyhow and not the first person carying the mutation.

... but I don't know where in India this haplogroup has been found.

I just said "South Asia" but (without further reference) Wikipedia says "Pakistan and India". Maybe it's wrong in the latter.

Wow, this is interesting. Sharma et al. (2007)...

Was just reading the same paper. He uses his own simplified non-YSOGG (2008) nomenclature, calling "Q" to Q1a so probably his "Q5" is Q1a5 (>1 only among Kashmiri and the Yadava, a Dravidian non-tribal ethnicity from the south). He also reports "Q*" (i.e. Q1a* in India, >1 only among Maharastra IEs).

it is possible that the haplogroup Q1a3(xQ1a3a) found in this region might also have been introduced by an already admixed C3+Q population from Siberia or northern East Asia.

I'd rather think in Central Asia but anyhow it may well be exogenous. What is less clear is in regard to Q or Q1a as a whole. Q1a6 is, oddly enough, found among Yemeni Jews at significative numbers (Jews are known for having been rather mobile historically and it would well be a "Khazar" clade or something of the like, though it's odd that it's not found in Europe, Askenazim, or elsewhere, would this be the case).

You have two Q1a subclades (out of 6), plus some Q1a*, that doesn't seem to fit the East/North Asian pattern of the rest of the clade.

Anyhow I've always thought of the Altai (or somewhere nearby) as a likely urheimat for this clade. But some branches appear to have moved in a southern/western direction, even if the bulk moved to the east.

Someone needs to perform more thorough testing of haplogroup Q people from Central Asia, Siberia, Mongolia, Vietnam, etc.

Of course. The more we know, the better we can think about it.

The only one of these samples to contain any haplogroup C3 (or haplogroup C in general) is the Pathan sample: 1/20 C3-M217.

It reminds me of the relative (very relative) aboundance of Q in Anatolia, where C3 is virtually lacking. It's possible that migrating peoples from Central Asia (be them Turkic or IE) may have once been higher in Q than C3 or that the ones migrating were that way. Again it does suggest a somewhat more westernly preference for Q than C3. All very relative and subject to interpretation, of course.

Here are some data I have collected in regard to the frequencies of haplogroups C(3) and Q among Asian populations...

Very interesting, thanks.

Someone needs to explain how all these haplogroup C3 and haplogroup Q individuals have ended up in China and Vietnam.

In principle I'd think of migration from the North.even if not necesarily (nor probably) as distinct populations, but rather within the mostly O peoples.

...may have been introduced during Ghengis Khan's rule.

Unlikely, IMO. Genghis Khan is hyped by the media and simplistc "recentist" thought. A small deal may correspond to this historical event but Mongols were just the proverbial drop in the Chinese (and Vietnamese) sea; even if they were extremely succesful in their reproductive biological roles, they would hardly have changed the genetic landscape so much. Additionally haplogroup Q is much rarer than C3 among Mongols and Tungus (another exogenous ruling dynasty of China) while the apportions found in China and Vietnam for these two clades are in general much more balanced.

Ebizur said...

Wen Bo et al. (2004):
Northern Han (Gansu n=60; Shijiazhuang, Hebei n=14; Zhengzhou, Henan n=50; Dalian, Liaoning n=48; Chifeng, Inner Mongolia n=60; Laizhou, Shandong n=85; Jinan, Shandong n=100; Xi'an, Shaanxi n=63; Liquan, Shaanxi n=27; Urumqi, Xinjiang n=51)
C-M130: 44/558 = 7.9% [0%-20.0%]
DE-YAP(xD1-M15): 14/558 = 2.5% [0%-8.3%]
D1-M15: 2/558 = 0.36% [0%-2.35%]
F-M89(xK-M9): 49/558 = 8.8% [3.5%-22.9%]
K-M9(xP-M45, O1a-M119, O2a-M95, O3-M122): 92/558 = 16.5% [7.1%-33.3%]
O3-M122(xM7, M134): 160/558 = 28.7% [18.3%-42.4%]
O3a3c-M134: 140/558 = 25.1% [14.1%-50.0%]
O1a-M119: 23/558 = 4.1% [0%-8.3%]
O2a-M95(xO2a1-M88): 6/558 = 1.1% [0%-2.1%]
Q1a1-M120: 22/558 = 3.9% [0%-8.0%]
P-M45(xQ1a1-M120): 6/558 = 1.1% [0%-3.9%]

Hammer et al. (2005):
Northern Han
2/44 = 4.5% C3-M217(xC3c-M86)
1/44 = 2.3% G-M201
1/44 = 2.3% J-12f2
1/44 = 2.3% NO-M214(xN1-LLY22g, O-M175)
3/44 = 6.8% N1-LLY22g(xN1a-M128, N1b-P43, N1c1-M178)
1/44 = 2.3% N1a-M128
6/44 = 13.6% O3-M122(xM134, LINE1)
18/44 = 40.9% O3a3c-M134
5/44 = 11.4% O+LINE1 (O3...)
2/44 = 4.5% O2*-P31(xO2a-M95, O2b-SRY465)
1/44 = 2.3% O2a-M95(xO2a1-M111)
2/44 = 4.5% Q1-P36
1/44 = 2.3% R-M207

Xue et al. (2006):
Non-Muslim Hans from Harbin (Heilongjiang), Yili (Xinjiang), Chengdu (Sichuan), Lanzhou (Gansu), and Meixian (Guangdong)
Y*(xA, C, DE, J, K): 1/166 = 0.6% [1/32 = 3.1% Yili]
C3*(xC3c): 20/166 = 12.05% [2/32 = 6.25% Yili - 6/30 = 20.0% Lanzhou]
DE(xE): 3/166 = 1.8% [0/35 Harbin & Meixian - 2/30 = 6.7% Lanzhou]
J: 4/166 = 2.4% [0/35 Meixian - 3/30 = 10.0% Lanzhou]
K(xNO, P): 4/166 = 2.4% [0/35 Harbin - 3/32 = 9.4% Yili]
N1*(xN1a, N1b, N1c): 5/166 = 3.0% [0/32 Yili - 2/30 = 6.7% Lanzhou]
N1c1: 1/166 = 0.6% [1/35 = 2.9% Harbin]
O(xO1a, O2, O3): 1/166 = 0.6% [1/30 = 3.3% Lanzhou]
O1a: 18/166 = 10.8% [1/35 = 2.9% Harbin - 7/35 = 20.0% Meixian]
O2*(xO2a, O2b): 12/166 = 7.2% [0/30 Lanzhou - 5/35 = 14.3% Meixian]
O2a(xO2a1): 3/166 = 1.8% [0/35 Harbin & Meixian - 1/30 = 3.3% Lanzhou]
O2a1: 4/166 = 2.4% [4/34 = 11.8% Chengdu]
O3*: 19/166 = 11.4% [1/35 = 2.9% Meixian - 7/32 = 21.9% Yili]
O3+LY1: 25/166 = 15.1% [3/35 = 8.6% Harbin - 9/35 = 25.7% Meixian]
O3a3a-M159: 1/166 = 0.6% [1/35 = 2.9% Meixian]
O3a3b-M7(xLY1rev): 3/166 = 1.8% [0/35 Harbin - 2/35 = 5.7% Meixian]
O3a3b-M7+LY1rev: 1/166 = 0.6% [1/32 = 3.1% Yili]
O3a3c-M134(xO3a3c1-M117): 17/166 = 10.2% [1/35 = 2.9% Meixian - 8/35 = 22.9% Harbin]
O3a3c1-M117: 19/166 = 11.4% [2/32 = 6.25% Yili - 5/34 = 14.7% Chengdu]
P(xR1a): 1/166 = 0.6% [1/32 = 3.1% Yili]
R1a: 4/166 = 2.4% [0/35 Harbin & Meixian - 2/30 = 6.7% Lanzhou]

It seems that haplogroup O3-M122 and all its subclades account for only about 50% of present Han Chinese Y-DNA. So, how does one explain the Y-DNA of the other half of the world's most populous ethnic group?

Furthermore, the composition of the Y-DNA of Han samples has varied considerably from study to study; this suggests to me that the Han population is still terribly undersampled.

terryt said...

OK. So I break a promise. Maju. You are geting more and more ridiculous.

"it's you who claim that archaic Indonesians ... just invented the boat the day before to do such a daunting trip (or almost)".

Either your understanding of English is very bad or you are being deliberately stupid. Humans would have been hanging around Indonesia long before they were able to cross Wallace's line. And there would have been plenty of regions with "Quiet waters to begin with" to spend years gaining experience. Let me know when you (with as many friends as you care to include) have made a dugout canoe using handaxes. Then we can discuss ancient canoeing technology seriously.

"Man: I quitted church at 11".

What was that someone said about a child before seven years old?

You do seem to have a basic understanding of genetics, though, when you write, "What is clear is that each haplogroup, each SNP, means a single point of origin (some unique man for Y-DNA, and some unique woman for mtDNA). Nevertheless the ones spreading it may have well been relatively distant descendants anyhow and not the first person carying the mutation".

But I should point out that even in a region as hugely populated as modern India the people do not form a single interbreeding population. Geography, caste and religion divides them into much smaller interbreeding subpopulations or tribes. And you have even admitted this tendency was probably more extreme during the Paleolithic.

In any population not infinitely large inbreeding acts to reduce genetic diversity, not increase it. As a result each separate subpopulation drifts towards a reduced number of haplogroups rather than a greater number. Once a new haplogroup has formed it has two options: increase its proportion in the subpopulation or become extinct. Either of these may take a long time to come to pass though.

Diversification happens once a subpopulation containing a high proportion of a perticular haplogroup is able to expand through other subpopulations for some reason or other. Only once this newly expanded population breaks up in turn into a new set of subpopulations do you get diversification. This happens throughout nature and is why we have grey ducks in this part of the world and you have mallards in yours. It's why African, Indian and European cattle look different from each other, and why Africa has gorillas and chimpanzees and SE Asia has gibbons and orang utans.

If we take mt-hap M to start with we find a huge variety just within the India subcontinent. This suggests M has been in India a long time. The variations presumably formed as some 'original' population broke into subpopulations. They have now become mixed to some extent. A few of these subpopulations managed to move out of India but they almost certainly further differentiated into the haplogroups we know as C, Z, D, G, E and Q as regional variants of their ancestral haplogroups. They didn't come out of India already fully differentiated.

Same with mt-hap N. Its haplogroups also almost certainly developed locally as regional variants after some original expansion of N. They didn't expand with it in some already differentiated form. The distribtion of these regional variants can tell us the extent of the expansion. We have problems reconciling N's expansion to an Indian origin.

Same pattern with Y-chromosomes. Some fit India, some don't.

Once you accept the above you can see there is no simple single point of origin for all the haplogroups we find outside Africa today. Unless it's the Iranian Plateau, which humans have to move through in order to reach India from Africa anyway.

So in answer to you comment, "IJK has two branches West Eurasian IJ and extremely widespread K. What does this suggest about the coalescence of K?" Perhaps the mutation at the base of both K and IJ occurred in a man living on the Iranian Plateau, but K itself had certainly reached New Guinea, and probably well into China, by the time it diversified further. Therefore the final mutation leading to basal K need not have occurred in India.

Maju said...

It seems that haplogroup O3-M122 and all its subclades account for only about 50% of present Han Chinese Y-DNA. So, how does one explain the Y-DNA of the other half of the world's most populous ethnic group?

Mostly remnants of a larger diversity from before O and O3 expansions, I guess. Some of it could be attributed to immigration from Mongolia/Manchuria/Uyghuristan/Altai. But what would surprise me would be the opposite: that Chinese would be as terribly homogeneous as to have only one clade or almost. MtDNA-wise the area is also pretty diverse. And historically we know that China (quite obviously) was not any single monolithic bloc. Han identity is based mostly in language also (as most ethnicities), so assimilating others who adopt the language and identity doesn't appear to pose any big deal.

Let's see in detail: N is so close to O that surely belongs to the same demic process, at least largely. D, C3 and Q also belong clearly to Eastern Asian diversity (regardles of what I said of Q being somewhat more westerner by origin, this refers to the Paleolithic). The real outliers are "Silk Road" clades like J, G and R1, which may have arrived at many times in late Prehistory or even historical times, but these ammount to something like 5% overall (Wen's study mentions some 16% K* but a large deal should be N or rare O clades).

Furthermore, the composition of the Y-DNA of Han samples has varied considerably from study to study; this suggests to me that the Han population is still terribly undersampled.

They are a huge bunch: one of every five people alive today. In the past surely less but still should have been relatively a lot of people.

Maju said...

Humans would have been hanging around Indonesia long before they were able to cross Wallace's line.

That's not clear at all. Australia is one of the oldest "Eurasian" areas to be colononized by H. sapiens.

Whatever the case there's absolutely no reason to think that the boat was invented in Indonesia. It's just your speculation.

Let me know when you (with as many friends as you care to include) have made a dugout canoe using handaxes. Then we can discuss ancient canoeing technology seriously.

LOL

Stop making ridiculous demands and go to the facts, the hard data, the evidence.

In any population not infinitely large inbreeding acts to reduce genetic diversity, not increase it. As a result each separate subpopulation drifts towards a reduced number of haplogroups rather than a greater number. Once a new haplogroup has formed it has two options: increase its proportion in the subpopulation or become extinct. Either of these may take a long time to come to pass though.

I understand that perfectly. But I reject that in the long run populations are just closed inbreeding pools. In any case, your reasoning (that is not different from mine) actually favors the formation of several, maybe many, distinct populations where different clades would have become more or less fixated.

So there you have F, C and D. And there you have H and IJK, and maybe G farther NW (unsure) in a second moment, surely still sharing the subcontinent with some C-only groups. Etc.

So where's the problem? Specially if you're flexible in regards to the exact borders of South Asia.

Diversification happens once a subpopulation containing a high proportion of a perticular haplogroup is able to expand through other subpopulations for some reason or other.

Or in "empty" land, or along with other clades (bands, tribes and nations do form alliances all the time and they don't ask for pedigree normally), etc.

We have problems reconciling N's expansion to an Indian origin.

We have problems to determine any specific likely origin of N, Indian or not, at this stage of research. Whatever the case it's so widespread that it must ahve formed soon after the OOA epysode (or maybe even before). But the geography of its known distribution does suggest South Asia as a major geographical node, so why not the origin?

Same pattern with Y-chromosomes. Some fit India, some don't.

All fit India if you accept that the subcontinent would have later suffered internal processes of homogeneization, when many clades vanished or became very rare.

They would not fit India if you try to equate modern castes and their strict rules (surely post-Neolithic in origin) with the clans and tribes that existed in the Middle Paleolithic. But obviously that's not the case.

Once you accept the above you can see there is no simple single point of origin for all the haplogroups we find outside Africa today. Unless it's the Iranian Plateau, which humans have to move through in order to reach India from Africa anyway.

The Iranian Plateau is quite arid, is badly comunicated with SE Asia and Australia (must go via India), we know of no archaeological evidence that would suggest it played any major role and nowadays shows a very clear Western genetic pattern. It's an even more unlikely spot to look for the origins of Y-DNA C or D than India (though guess it would serve for K, except for its bad connection to the sea and hence with Sahul).

And also its considered part of South Asia in some readings (and I have always said that my concept of South Asia is not rigid, that may expand somewhat to East and West - but not to Siberia).

Perhaps the mutation at the base of both K and IJ occurred in a man living on the Iranian Plateau, but K itself had certainly reached New Guinea, and probably well into China, by the time it diversified further. Therefore the final mutation leading to basal K need not have occurred in India.

It's not a must but a reasonable possibility. What's between West Asia and East Eurasia? India.

And I would not emphasize the "first" or the "final" mutation in K. When such series of SNPs happen (and they do all the time, at least in well researched clades), it surely means that the population carrying the "proto-clade" was in a stable period, when all alternative branches were reabsorbed by mere drift. That's the coalescence period and that's what we are discussing here: where did K coalesce? And I think it must have coalesced either in West or South Asia, West maybe because of IJ but South because its more central to its present distribution.

Of the described 10 subclades of K:
- 2 are clearly West Eurasian (T and K4)
- 3 are clearly from Sahul (M, S and K3)
- 1 is clearly from SE/East/North Asia (NO)
- 2 are clearly from South Asia (L and K1)
- 1 is not really known where it's been found (K2)
- 1 must have originated somewhere between NW South Asia and southern Central Asia (P) and is now extremely widespread.

The obvious most diverse areas are: Sahul (Melanesia basically), South Asia and West Eurasia. Join the dots and there's only one of the three dots that can be said to be in the middle. That one is South Asia.

Maybe I'm wrong. But don't dismiss what is so apparent.

Have you even stopped to ponder this obviousness at all?

terryt said...

"Have you even stopped to ponder this obviousness at all?"

Many times, and for many years. I've always found it inadequate.

"I have always said that my concept of South Asia is not rigid, that may expand somewhat to East and West - but not to Siberia".

Agreed. But it may include some other parts of Central Asia to the south of Siberia. You write:

"Of the described 10 subclades of K:
- 2 are clearly West Eurasian (T and K4)
- 3 are clearly from Sahul (M, S and K3)
- 1 is clearly from SE/East/North Asia (NO)
- 2 are clearly from South Asia (L and K1)
- 1 is not really known where it's been found (K2)
- 1 must have originated somewhere between NW South Asia and southern Central Asia (P) and is now extremely widespread".

Now that's just Y-hap K, and it's a pretty widespread region. My argument is that the mutations defining most of these individual K-derived haplogroups must have happened somewhere in the region where they are now found, because many of them are confined just to relatively small regions. Therefore K itself had become rapidly, and widely spread: from West Eurasia through South Asia and on to Sahul (probably including New Guinea) and into SE/East/North Asia (although its northerly extremity may be the result of a more recent expansion). I agree that that doesn't tell us exactly where K originated, just somewhere in that wide region. However we have to explain why K was so easliy able to expand so widely and rapidly.

Same with P and its descendants. It's distribution almost certainly represents a more recent expansion that further mutations broke into basically eastern (Q) and western (R) branches. But we can easily explain its rapid spread: development of the ability to survive the cold of the Far North, probably sewn skin clothing and efficient mammoth hunting.

Now. It seems to me that K's diversification is more recent than that of C. It would be a brave person who claimed that the first haplogroup into a region always becomes the most common in that region, however unless other factors intervene it seems to be most likley true. Australia is predominantly just one Y-hap: its own version of C. Any K derived haplogroups are closely related to New Guinea versions. Suggests they didn't come in together. Same with mt-haps. There is virtually no M or M derived haplogroups in Australia, just a very small amount of the New Guinea mt-hap Q. Predominantly N, along with a haplogroup derived from N's descendant R (P). Again this P mt-hap is shared with New Guinea and, along with mt-hap Q, could derive from a later immigration. This strongly suggests that all these different haplogroups were not part of a single poplulation in any particular region, let alone India.

Maju said...

Agreed. But it may include some other parts of Central Asia to the south of Siberia.

But K does not appear to have migrated (as such pre-derived K) through Central Asia into East Asia.

My understanding is that the first K that reached Central Asia was probably P already (as P is the only haplogroup that appears to have a Central Asian center of gravity). But I may be wrong in that.

Now that's just Y-hap K, and it's a pretty widespread region.

All Eurasia, but not with the same diversity everywhere. I consider the areas with lower diversity, as much more unlikely origins than those that have higher diversity, which are all in the south of the continent.

Therefore K itself had become rapidly, and widely spread: from West Eurasia through South Asia and on to Sahul (probably including New Guinea) and into SE/East/North Asia (although its northerly extremity may be the result of a more recent expansion).

Agreed.

I agree that that doesn't tell us exactly where K originated, just somewhere in that wide region.

But look at high level diversity and also archaelogy for some clues.

However we have to explain why K was so easliy able to expand so widely and rapidly.

IMO because it was largely part of the first expansion out of India. Maybe even its main driver, or maybe just the last wave. But in any case it should have participated of this migratory drive.

Unless you can suggest a more or less clear archaeologically documented process by which it may have expanded over other clades, not in some specific region, but nearly everywhere.

Same with P and its descendants. It's distribution almost certainly represents a more recent expansion that further mutations broke into basically eastern (Q) and western (R) branches. But we can easily explain its rapid spread: development of the ability to survive the cold of the Far North, probably sewn skin clothing and efficient mammoth hunting.

That may apply for Q but R seems to have spread by SW Asia into Europe and elsewhere (R2 in India, R1* in Africa...). It's not like we have detected any migration going through the steppes into Europe, except arguably some marginal earliest settlers of Eastern Europe that do not appear to have much continuity. Later on, in Gravettian times, from Europe into the Asian steppes the opposite may have happened though.

One thing I wonder about is that Gravettian seems so terribly important, with likely offshots as far as Altai (UP), North Africa (UP) and West Asia (maybe UP too but specially Mesolithic) and instead no clear genetic pattern is clearly associated with this "superculture". I would certainly think of R1 (and mtDNA H, V and various U subclades) as associated to Gravettian, though R and P probably belong to earlier Asian-centered stages, as does HV and U as such.

Now. It seems to me that K's diversification is more recent than that of C.

Hard to say. C and K certainly belong to different levels in the genealogical tree. But it's possible that C coalesced for long in a small population and lack of research alone explains the relatively low number of SNPs known. Certainly C's list of known SNPs appears really short considered how longeve the "sybling" of F must be.

It's also possible that, if K originated in NW South Asia, C may have been coalescing already in NE South Asia and been pulled (or even pushed) by K in its migration eastward.

It would be a brave person who claimed that the first haplogroup into a region always becomes the most common in that region, however unless other factors intervene it seems to be most likley true.

That's known as founder effect. But what if the founder effect implies several clades? Then we should consider also drift and that means that the several subpopulations created after the migration, would then drift into random clades within their original diversity (or derived from those). Some of the original diversity may be lost alltogether but at the begining there's no way to predict which will survive and which will be gone: that's basically random, unpredictable.

Australia is predominantly just one Y-hap: its own version of C. Any K derived haplogroups are closely related to New Guinea versions. Suggests they didn't come in together. Same with mt-haps.

I would not totally rule out two separate migrations but it's also possible that drift selected them randomly after both groups separated, maybe once in Sahul. Both scenarios are possible.

The single migration model would say: several clades of K, one of C and surely some F(xK) too, in the Y-DNA side; four M clades, several distinct N(xR) lineages and a couple of R lines or so, in the mtDNA side, participated in the process.

The double migration model would instead say:

- for New Guinea (basically): several clades of M and at least some R particpated in the mtDNA side, while the Y-DNA side would be only K (three clades)

- for Australia (minimal): several lines of N (one of them closely related to West Eurasian W, but not the others) in the mtDNA, at least C (C4) on the Y-DNA

Several clades are shared though: Y-DNA K, mtDNA Q (subclade of M) and mtDNA P (subclade of R)

You could maybe hypothesize a different migration for each of those clades but the simplest model is that of a single migratory epysode, by one or two routes, soon after the OOI epysode, some 10,000 years after Toba or so, bringing people that carried diverse clades already and then concentrating them in the different areas by natural drift.

Even the "purest", most isolated and tiny populations you may think of, are seldom 100% a single haplogroup. Drift homogenizes genetic pools locally but some diversity normally persists. The wave that carried people to Sahul was that of an expansive population, not a shrinking or stable one, so some diversity existed with all likehood.

There is virtually no M or M derived haplogroups in Australia

Q.

There is virtually no M or M derived haplogroups in Australia, just a very small amount of the New Guinea mt-hap Q. Predominantly N, along with a haplogroup derived from N's descendant R (P). Again this P mt-hap is shared with New Guinea and, along with mt-hap Q, could derive from a later immigration

It could. Or it could not. I can accept a dual parallel migration to Australia and New Guinea, that's not a big deal (I just don't know). But when you put it in a global context, where do you find C/N or K/M pairings alone elsewhere? These couplings are rather rare and suggest local accidents, no matter if they happened already in Sahul or in Indonesia, prior to the crossing.

This strongly suggests that all these different haplogroups were not part of a single poplulation in any particular region, let alone India.


As you see above to me it suggests rather the opposite.

Ebizur said...

By the way, I just noticed that Gayden et al. (2007) have reported finding 2/156 = 1.3% Q1a1-M120 and 3/156 = 1.9% Q1a3-M346 in a sample from Tibet. So, Q1a1-M120 has been found in at least Han Chinese, Koreans, Chinese-speaking Central Asian Muslims (Dungans), Hazaras, and Tibetans. Q1a3-M346 has been found in at least Pashtuns, Makrani Balochs, Uttar Pradesh Brahmins, Vellalar Tamils, Halba tribals, Native Americans, and now Tibetans.

terryt said...

I'm going away for a few days so before I reply to some of your points I think I should explain why I believe that SE Asia has long been a centre for boating technology.

Humans had almost certainly reached Australia by 50,000 years ago, and possibly 10,000 years before that. Presumably they reached New Guinea at the same time.

However it's not until 30,000 years ago that several lines of evidence prove human presence beyond there, in the Northern Solomon Islands. This suggests to me an improvement in boating technology allowed that presence. Versions of Y-hap K are common through the region and may have entered Australia at the same time.

About 8000 years ago another improvement in boating technology let humans move even further out into the Pacific: as far as Fiji, Tonga and Samoa. Indonesian versions of Y-hap C seem to lead the expansion.

Somewhere between 300 BC and 300 AD a further improvement let humans expand right across the Pacific and be first to New Zealand, Hawai'i, Easter Island and all islands within the triangle. Y-hap O probably caught up at that time.

This ancient gradual improvement in boating technology in SE Asia contrasts sharply with what we observe in the Mediterranean, where effective salt-water boating seems to appear suddenly around 10,000 years ago. From then on Mediterranean open water voyaging itself demonstrates progressive improvement. To me this suggests that the technology was introduced from somewhere else.

Ebizur asked, "So, how does one explain the Y-DNA of the other half of the world's most populous ethnic group?"

I suggest that movement back and forth between north and south has a long and complex history through China and along the East Asian coast. Certainly it has been much more common than any movement between India and SE Asia or across Central Asia. I'm sure you'll have a lot of fun unraveeling the strands of those movements and look forward to reading your conclusions.

Maju wrote, "But K does not appear to have migrated (as such pre-derived K) through Central Asia into East Asia".

I'm sorry for not making myself clear. I was not refering to Y-hap K then. I agree "that the first K that reached Central Asia was probably P".

"But look at high level diversity and also archaelogy for some clues".

But a high level of diversity can easily be caused by varying levels of research. Once you realise all K Y-haps in Melanesia have a common origin there's not much incentive to subdivide further. In fact I've seen some research that indicates each separate group of islands basically has a separate haplotype.

"But what if the founder effect implies several clades?"

I'd guess most founder populations contain several clades. But the problem here is that, as you say, "the several subpopulations created after the migration, would then drift into random clades within their original diversity". If drift was responsible for the present distribution of clades we would find each clade randomly distributed rather than being confined to particular geographical regions. Admittedly several clades do demonstrate partly random and isolated distributions, but many do not.

"where do you find C/N or K/M pairings alone elsewhere?"

After all this time, and the various posts Dienekes has put up, you still believe Y-haps and mt-haps always remain together? Anyway C and N derived haplogroups often occur together in Northern Asia and K derived and M derived haps are found scattered round the world. Not alone admittedly because migration has mixed most human populations considerably whereas migration across Wallace's Line has been much more limited.

Maju said...

If drift was responsible for the present distribution of clades we would find each clade randomly distributed rather than being confined to particular geographical regions. Admittedly several clades do demonstrate partly random and isolated distributions, but many do not.


At certain, older, levels they are "randomly distributed" for the greatest part: clades like Y-DNA F or K, or mtDNA N, M and R are quite "chaotically" distributed all around Eurasia. Others show some greater geographical structure but these are certainly "chaotic", what suggest migrations involving several lineages.

you still believe Y-haps and mt-haps always remain together?

No, but Paleolithic men and women did tend to migrate together (and still do in the few cases of pervivence we know about) - even if they outmarry, they migrate all together, not just the men.

Anyway C and N derived haplogroups often occur together in Northern Asia and K derived and M derived haps are found scattered round the world. Not alone admittedly because migration has mixed most human populations considerably whereas migration across Wallace's Line has been much more limited.

The case is that if you compare with other populations where the high level haploid diversty is low you see different couplings: in West Eurasia K is combined mostly with R (an N subclade), in America K combines with five mtDNA "high clades" but only one is M-derived. Only in India you find that K-M combo (or the F-M one) as dominant.

On the C/N pairing I'm willing to concede just out of researching laziness.

But this means that the Papuan K-M alleged founder combo connects best with India (and maybe other spots, unsure).

Instead the Australian alleged founder combo (though C is also important in Melanesia: C2) appears to connect best with Eastern Asians by the paternal side and with West Eurasians (if anything) by the maternal one (a Australian N* lineage is directly linked to W, I'll call this "pre-W" for lack of a better name). I really don't gather how these two would have met if they were not then in Southern or SE Asia.

If the Sahulian clades would represent a "fossil" of early Eurasians, they tell me a story of diversity, including at least Y-DNA C, F(xK) and K, and mtDNA M, N(xR,pre-W), pre-W and R. Some of these macro-clades' derivates are concentrated in either part of old Sahul but others are shared.

These macro-clades (pre-W excepted) are also the most common ones in mainland Eurasia, so my conclusion is that they must have been more or less together, mixed or living not too far from each other, in the time of the early expansion within Eurasia, wave that reached Sahul as well.

Ebizur said...

Maju said,

"The case is that if you compare with other populations where the high level haploid diversty is low you see different couplings: in West Eurasia K is combined mostly with R (an N subclade), in America K combines with five mtDNA "high clades" but only one is M-derived. Only in India you find that K-M combo (or the F-M one) as dominant."

The four major pan-American mtDNA haplogroups are A2, B2, C1(xC1a), and D1. C1 and D1 are both derived from haplogroup M. A2 is a subclade of A, which is a subclade of N(xR), and this haplogroup is predominant among most indigenous populations of North America (except for some populations of California and the Southwest, which tend to have high frequencies of haplogroup B). Haplogroup B2 is the only Native American mtDNA haplogroup that is derived from macrohaplogroup R.

So, the major Native American mtDNA haplogroups include two subclades of macrohaplogroup M (C1(xC1a) and D1), one subclade of macrohaplogroup N(xR) (A2), and one subclade of macrohaplogroup R (B2). Haplogroups A, C, and D are all typical indigenous North Asian haplogroups, so there is no difficulty imagining that they might have simply spread across the area that is now the Bering Strait and into North America. However, haplogroup B, besides being somewhat distinctive from the other Native American mtDNA haplogroups due to its being derived from macrohaplogroup R, is also odd for being more commonly found among East/Southeast Asians and Oceanians (e.g. Japanese, Polynesians, Thais, Indonesians) than among North Asians, but it has been found at relatively low frequency among some "southern North Asian"/transitionally Central/East Asian groups (e.g. Altayans, Tuvans, Mongols).

Maju said...

The four major pan-American mtDNA haplogroups are A2, B2, C1(xC1a), and D1. C1 and D1 are both derived from haplogroup M.

You're right in this.

Guess there may be some connection between the two macro-clades in regard to the Indo-Pacific arch.

I can only see a more or less coastal migration along the Indo-Pacific arch in all that, including all what you say in your last post (and what is being dscussed in a more recent post on Japanese and Korean DNA).

You say that you want to be able "I told you" but would you allow us to tell you, whenever your realize that the coastal migration model is very consistent, "we told you"? ;-)

Maju said...

Caveat: but the coincidence of Y-DNA K (Q) in America with mtDNA M (D and C) can hardly be the product of a long shared origin. There at least Q appears to have arrived to Berigina via the mainland, while mtDNA D and C, as well as A and B, appear to be original from East or NE Asia. Only mtDNA X2 could be maybe related to the migration of Q through North Asia but, ironically, its spread in America approaches more that of NE Asian C3 than the one of Q.

What again suggests that all these clades were just scrambled together in Beringia, prior to the migrations, even if some had arrived from Central Asia and most (specially in the mtDNA side) from East/NE Asia instead.

But does not suggest a similar kind of K/M paring as the one we find in New Guinea.

Btw, why don't you start writing your own blog or website where you explain your hypothesis in detail. Even if it's wrong, it may be interesting.

terryt said...

"At certain, older, levels they are "randomly distributed".

We would expect that. However even these broken distributions are most easily explained as having been part of an earlier continuous distribution broken up, and replaced in several regions, by later expansions.

"On the C/N pairing I'm willing to concede just out of researching laziness".

I've had a quick look and found Y-hap C with the following early N derived mt-haps: Nivkhs - Y; Buryats - N, A and Y; Altaians - N, I and X; Uzbeks - N, W and A; Northern Han - N, A and Y. Interestingly Mongols have just a small amount of A. The other mt-haps are R derived or M. And early N derived haplogroups are not at all common, even with Y-hap C, north of the Central Asian mountains.

Ebizur wrote, "Haplogroup B2 is the only Native American mtDNA haplogroup that is derived from macrohaplogroup R". And further "is also odd for being more commonly found among East/Southeast Asians and Oceanians (e.g. Japanese, Polynesians, Thais, Indonesians) than among North Asians". To which Maju wrote, "I can only see a more or less coastal migration along the Indo-Pacific arch in all that".

I realise you were refering to all the haplogroups with this comment but I agree totally it's true for B. Doen't that mean there's a good chance B originated in SE Asia? And what about its parent haplogroup?

"the Australian founder combo ... appears to connect best with Eastern Asians by the paternal side and with West Eurasians ... by the maternal one".

Most reseachers believe the first Australians looked more like contemporary populations in East Asia rather than contemporary populations in SE Asia. Therefore they probably moved down the East coast of Eurasia before they were able to move across Wallace's Line.

"There at least Q appears to have arrived to Berigina via the mainland, while mtDNA D and C, as well as A and B, appear to be original from East or NE Asia".

I mentioned somewhere (possibly here) that Q appears to have picked up several mt-haps as it moved east towards Alaska. It doesn't mean, "that all these clades were just scrambled together in Beringia". However it does indicate they probably became scrambled in Beringia before moving further south through America.

You wrote, "but the coincidence of Y-DNA K (Q) in America with mtDNA M (D and C) can hardly be the product of a long shared origin". Obviously it's not. By the way, it's also possible mt-hap C's origin is fairly western. It tends to be concentrated north of the Central Asian mountains. And its sister clade is a sort of Hazara/Pakistan/Iran haplogroup.

"Btw, why don't you start writing your own blog or website where you explain your hypothesis in detail".

It's pretty much all up at remotecentral. It's possible for you to read the argument in order of increasing complexity if you start with the contents page. You'd get an idea of the thread just by looking at those contents.

Maju said...

We would expect that. However even these broken distributions are most easily explained as having been part of an earlier continuous distribution broken up, and replaced in several regions, by later expansions.

Earlier continuous distribution? I read in that "rapid coastal migration", ahem.

Please provide some sort of archaeologically consistent support for your hypothetical "later expansions", specially when these are claimed to have continental dimensions.

Otherwise I just cannot believe in it: the archaeology I know of rather suggests no continent-wide migrations after the first expansion, so all the clades we see now must have been there when this early expansion happened (though there could have been even more, now extinct by mere drift).

Most reseachers believe the first Australians looked more like contemporary populations in East Asia rather than contemporary populations in SE Asia.

Reference?

AFAIK, the term "australoid" has been used rather arbitrarily to describe very different sets of populations that have nothing in common save an archaic-looking phenotype not easy to describe using the other commonly acknowledged racial descriptions.

By the way, it's also possible mt-hap C's origin is fairly western. It tends to be concentrated north of the Central Asian mountains. And its sister clade is a sort of Hazara/Pakistan/Iran haplogroup.

Nah. Z is actually a subclade of C (also called CZ). It is nevertheless the clade with a more clearly western distribution within CZ, being found among Uralic Europeans (who have other genetic links to Siberia/East Asia) as well as the Hazara (who still largely retain the "Mongoloid" phenotype and are generally considered to be related by origin to the Turco-Mongol peoples, even if they now speak Persian). The remnants found among Tajik (eastern Persians) and Pakistanis must be original from the Hazara community or some related Turco-Mongol peoples).

It's pretty much all up at remotecentral. It's possible for you to read the argument in order of increasing complexity if you start with the contents page. You'd get an idea of the thread just by looking at those contents.

Your essays are basically on evolution as a whole and have the particular intent of persuading religious fanatics of the truth in evolution and science. A laudable attempt but not something I need to read.

There is only one essay among them dealing with human genetic genealogy, AFAIK, and I can't agree with you for the reasons already mentioned a thousand times:

1. You're twisting genetic data to fit your preconcieved idea (probably on good faith but it doesn't really matter)
2. You are obsessed with boating being only invented in SE Asia and not before, what is ilogical
3. You seem to need a secondary expansion to explain the distribution of haplogroups like Y-DNA K and mtDNA R (possibly even others)
4. You seem to believe that these clades represent somehow a "superior" subset of humankind, something that is hilarious to me
but anyhow sounds to Eurocentric racism
5. You insist on arguing against all evidence for an early continental migration through Central Asia
6. You lack of any evidence for all those claims other than patterns created in your own mind, other than your own subjectivity

Sorry to be so demolishing critical but if there is something I'm good at is at making criticisms, normally good ones. I know that they often hurt but can't help it.

terryt said...

"I read in that 'rapid coastal migration', ahem".

It's absolutely impossible to reconcile the distribution of mt-haps A, X and Y as being the result of a great Exodus from India, coastal or otherwise. Unless you're "twisting genetic data to fit your preconcieved idea (probably on good faith but it doesn't really matter)". Haplogroups A, X and Y obviously derive from N but, because their distribution is basically regional, they almost certainly evolved in those paricular regions after an expansion of N. There's no way they left any region, let alone India, together already differentiated.

"AFAIK, the term 'australoid' has been used rather arbitrarily to describe very different sets of populations that have nothing in common".

It should have been obvious I don't have such a view of Australoid. I pointed out to you long ago exactly what you've just claimed for yourself. As for references, any half-pie decent account of Australian prehistory covers the connections. And even the more general Stringer and McKie 1996 book "African Exodus" covers it. The first Australians are not actually "Australoid".

"Z is ... the clade with a more clearly western distribution within CZ, being found among Uralic Europeans".

There's no earthly reason why it must have come in with people from "Siberia/East Asia". There are any number of other genetic connections to account for the phenotypic similarity. Anyway it seems you accept that C and Z were basically simply separated by geography. Why are you so vehemently opposed to the same idea when applied to the distribution of deeper haplogroup divisions?

"A laudable attempt but not something I need to read".

But many of your comments reveal you have a very poor understanding of the mechanism of evolution. Sure, my essays use the perspective of explaining evolution to non-believers but by the time I get to the essay you mention I have provided a huge range of evidence supporting the position I take.

"You seem to need a secondary expansion to explain the distribution of haplogroups like Y-DNA K and mtDNA R (possibly even others)".

Do you really believe that there have been no other haplogroup expansions since some mythical one out of India? I'm sure I've read comments of yours where you appear to accept secondary expansions.

"You seem to believe that these clades represent somehow a 'superior' subset of humankind".

How on earth do you come to that conclusion? I'd bet a huge amount of money that you are simply interpreting what I say through the prism of your own prejudices. Surely possession of different technology doesn't at all indicate genetic superiority.

"if there is something I'm good at is at making criticisms, normally good ones".

And I greatly appreciate your input. Very useful. But everything you've said concerning an India Exodus lacks "any evidence for all those claims other than patterns created in your own mind, other than your own subjectivity".

Unknown said...

Terry:

we are watching your postings for the last 18 months or so.
You wrote being not familiar with Y haplo F also on the same Dienekes forum.
So far not a single writing made sense.
You keep on writing against Southern dispersal theory. You seems like India hater guy or some thing. I am not India lover either.

I believe southern dispersal theory has scientific proof on its side in the studies "sofar". If somebody finds proof of Continental migration than I agree with you.

Just show a single Y haplo or Mt Haplo of Melanesia or Papua NG or Aborigine Australia in East Asia with same diversity. At least show the groups.

Just give some examples.

Maju said...

It's absolutely impossible to reconcile the distribution of mt-haps A, X and Y as being the result of a great Exodus from India, coastal or otherwise.

Why? They are local/regional derivates of N.

Haplogroups A, X and Y obviously derive from N but, because their distribution is basically regional, they almost certainly evolved in those paricular regions after an expansion of N. There's no way they left any region, let alone India, together already differentiated.

So? They are local/regional derivates of N: that's obvious. Undifferentiated N and already derived R migrated together (and with mostly undifferentiated M as well). That's the idea.

There's no earthly reason why it must have come in with people from "Siberia/East Asia".

If the fact that it's most common among Finnic and Hazaras (both with evident partial ancestry from North/East Asia) is not sufficient to you...

Anyway it seems you accept that C and Z were basically simply separated by geography. Why are you so vehemently opposed to the same idea when applied to the distribution of deeper haplogroup divisions?

Neither I do believe specifically that C and Z got separated by anything in particular (Z is just one among several subclades of C and anyhow I have not researched this haplogroup in any particular depth), nor I oppose that geography can be a involved in haplogroup diversification - not at all.

But many of your comments reveal you have a very poor understanding of the mechanism of evolution. Sure, my essays use the perspective of explaining evolution to non-believers but by the time I get to the essay you mention I have provided a huge range of evidence supporting the position I take.

Neither I accept that I could have a poorer understanding of evolution than you nor I'm willing to read all that. I like things synthetic and clear.

Do you really believe that there have been no other haplogroup expansions since some mythical one out of India? I'm sure I've read comments of yours where you appear to accept secondary expansions.

I don't believe anything. I think, I understand, I suspect, I gather..

Your question anyhow is too generical to be relevant. Ask me about particular haplogroups and not "haplogroups" in general. As I have said before it's not the same Y-DNA CT than F or P or R1b1b2a1b4c1. They are all related but they are not the same at all.

"You seem to believe that these clades represent somehow a 'superior' subset of humankind".

How on earth do you come to that conclusion?


Youa argued quite clearly before that the expansion of K must have been related to some superior tehnology, society and maybe even human type.

I say: BS.

Surely possession of different technology doesn't at all indicate genetic superiority.


What "superior" technology?! If even the technological differences between AMHs and Neanderthals are extremely unclear, what the heck of "superior" technology are you talking about? Toothpicks maybe?

Explain yourself. Point in the archaeological record what evidence exists of that "superior" technology?

...

And I totally agree with South Central Haplo that you have provided no viable or logical alternative model. You should do your homework first of all.

Maju said...

Btw, Z is not just found among Central/West Eurasians. I suspect that you are confused about that: mtDNA Z is also found among East Asians, like Japanese, Chinese and Koryaks (ref).

terryt said...

"I believe southern dispersal theory has scientific proof on its side in the studies 'sofar'. If somebody finds proof of Continental migration than I agree with you".

I guess we'll just have to wait, won't we?

"Undifferentiated N and already derived R migrated together (and with mostly undifferentiated M as well)".

Very unlikely. N's diversification is earlier than R's. R is a subset of N and its expansions looks unrelated to N's.

"I like things synthetic and clear".

Doesn't really happen like that. It's never simple.

"As I have said before it's not the same Y-DNA CT than F or P or R1b1b2a1b4c1. They are all related but they are not the same at all".

And their individual distributions are the product of a series of movements. Not just a single one.

"What 'superior' technology?!"

I'd guess open water navigation but you obviously believe humans had dugout canoes more than 50,000 years ago.

"I suspect that you are confused about that: mtDNA Z is also found among East Asians, like Japanese, Chinese and Koryaks".

Yes. A bit confused anyway. The information I have is that it's most common in Hazaras and Northern Han Chinese with lesser proportions in Altaians, Kets and Buryats. No mention of Japanese but thanks for the reference.

Maju said...

"Undifferentiated N and already derived R migrated together (and with mostly undifferentiated M as well)".

Very unlikely. N's diversification is earlier than R's. R is a subset of N and its expansions looks unrelated to N's.


Let's go to the fine detail, check the various philogenetic trees in Ian Logan's mtDNA site for reference.

Branchings of N (ad-hoc nomenclature in [square brackets] for mere reference):

1. (1 snp), [pre-RPYN9]includes:
1.1. (1 snp), [pre-YN9]includes:
1.1.1. (8 snps): Y
1.1.2. (1 snp): N9 (divided in N9a and N9b after 2 shared snps)
1.2. (1 snp), [pre-RP] (known simply as R in other materials) includes:
1.2.1. (2 snps) R
1.2.2. (2 snps) N* "Asian sequences" (10 main sub-branches, none of them named, each defined by long series of snps - almost all in South Asia - only one case in China)
1.2.3. (1 snp) P
2. (8 snps): A
3. (1 snp), [pre-IX] includes:
3.1. (7 snps): X
3.2. (4 snps): N1 (including I)
3.3. (18 snps): certain Indian N*
4. (16 snps): some Ket and European N* [Ket N*]
5. (1 snp), [pre-W] includes:
5.1. (10 snps): W
5.2. (14 snps): some Australian N* sequences
6. (1 snp): S

So R (as described by Logan) is 4 SNPs away from N, while pre-RP (or just R in other nomenclature, like the one used in Wikipedia) is only 2 SNPs away from the N node. Furthermore this pre-RP or R macro-haplogroup belongs to even a larger one with Y and N9, that I called pre-RPYN9, which is only 1 SNP away from the N node.

You have five other branches but only 3 of them are comparable in genetic distance measured in SNPs: pre-IX, pre-W and S. Of them, pre-W shows some of the ample geographic spread of N, pre-RPYN9 and R itself (or pre-RP), connceting West Eurasia and Australia, while the other two are consolidated in each of these extremes (West Eurasia and Australia). A and "Ket N" appear to have experienced major epysodes of drift (small populations, no expansion) but A is now relatively common in spite of that.

Furthermore, pre-RPYN9 is as widespread as N and R. So the main expansion at least in East Asia must have happened within the process of evolution from N to R, via pre-RPYN9. The other clades seem more localized but must have been involved in some specific epysodes anyhow, either still undifferentiated or already defined.

Overall N appears to have spread along the coasts of the Indian Ocean first and foremost. Maybe from South Asia to east (Australia but also East Asia) and west (West Asia).

The following genealogical phase would be pre-RPYN9, which has two branches: R (pre-RP) and pre-YN9. The latter is limited to East Asia, while the other is widespread. This may suggest a rather eastern area of coalescence (SE Asia or Eastern South Asia?). If you have any reason in your idea of forth and back movements, this may be your key moment.

Then it comes R (pre-RP), which is generally acknowledged as South Asian by origin (highest diversity there both at pre-RP and "R proper" stages).

In general a South Asian origin is a safe conclusion. Pre-YN9 could well have migrated to East Asia (along maybe with A and "Ket N", as well as so many M subclades) in an epysode that went only in that direction. It could well have been part of a large migration that included undiferentiated M, M33, pre-W, S, proto-A and proto-"Ket N" (well, "Ket N" can also have some other history, I really don't know). That's like 6 founding mothers (surely more but some lineages went extinct, while others became regional, like S), not very different of what we find in Beringia/America at a later moment.

Beyond the safe conclusion of South Asia as likely urheimat, I would like to know more of the role played maybe by areas like the Gulf of Bengal (which seems pivotal for M and maybe for pre-RPYN9) and also about the mystery of the West Asia-Sahul connection, apparent in pre-W. But still all appears to be pivoting around India. Talk maybe of a "Greater Southern Asian" urheimat (with some unclear extensions beyond the borders of modern South Asia) but the Indo-Pacific arch seems very clear in any case also for N and R as migration route.

"What 'superior' technology?!"

I'd guess open water navigation but you obviously believe humans had dugout canoes more than 50,000 years ago.


Dugout canoes and open water navigation do not mix, at least normally. Canoes are for coastal, lake, swamp and river navigation.

Also what has to do navigation with your alleged route through the steppes?

Unknown said...

Terry wrote:

I guess we'll just have to wait, won't we?

SO Terry as usual no hard data. No logical agreement and throwing some ideas and arguing. I was polite in saying that. as usual that is your only chance. .

You are waiting for some one like god comes and prove that your argument wins and keep dreaming about that .

Even if the neighborhood of wallace lane genetics is identical to some other geenetics in coastal areas of some other countries you dont agree with that.

You just want your dream come true. No proof and data.

terryt said...

Thanks for that updated link regarding mt-hap N Maju. This diagram from it is very interesting:

http://www.ianlogan.co.uk/discussion/gifs/N_star_gif.htm

A chart of the divergences. It divides into six branches. The three simplest branches lead to A, Ket N* and S. A fourth leads to W and some Australian N*. Indian N is also found on two different branches. One of these became divided into three: X, IN and Indian N*. The sixth branch includes Y and Far East N*. This latter branch divides further into N9, R and P, some of which are also found in India. But as you point out this final branch diverged from the basal haplogroup quite early and is basically a separate haplogroup. It's very widely spread through the Far East and Australia as well as India.

"R (pre-RP) and pre-YN9. The latter is limited to East Asia, while the other is widespread".

Became widespread later? Of course I'll allow you to interpret this distribution in any manner you wish.

"A and "Ket N" appear to have experienced major epysodes of drift".

Which is not surprising considering where they are found.

"Also what has to do navigation with your alleged route through the steppes?"

The route through the steppes has nothing to do with mt-hap R or Y-hap K's expansion. Itwas some time before that.

"It could well have been part of a large migration that included undiferentiated M, M33, pre-W, S, proto-A and proto-'Ket N'".

You can regard this as just an aside but do you believe Y-hap E also originated in India? If not how did D get so far to the East? And, finally, what Y-haplogroup/s do you "think, ... understand, ... suspect, ... gather.." came out of Africa originally?

terryt said...

"Even if the neighborhood of wallace lane genetics is identical to some other geenetics in coastal areas of some other countries you dont agree with that".

Except for what can be interpreted as relatively recent movements through the region the genetics either side of Wallace's line is not identical. Of course immediately either side we have the same ones but the line fairly effectively divides the haplogroups.

Unknown said...

again care to write
which one?
some data about humans please.

Maju said...

The route through the steppes has nothing to do with mt-hap R or Y-hap K's expansion. Itwas some time before that.

Evidence. Just repeating your pre-concieved idea one and a zillion times wil not give you the reason.

You can regard this as just an aside but do you believe Y-hap E also originated in India? If not how did D get so far to the East?

IMO, D (or proto-D) travelled to South Asia and beyond with the other Eurasian Y-DNA clades. It was surely never very numerous anyhow because it's only found in some groups, mostly in East Asia, and specially in random peripheric groups (Andamanese, Tibetan, Ainu/Japanese), where it must have become fixated due to random events (localized founder effects).

DE is a sybling clade of the major Eurasian haplogroup, CF. It's no wonder that they traveled together. I'd be more surprised if it was A or B or some other distantly related clade. But, as it is, it fits pretty well with the parallel mtDNA tree and the existence of a, likely single, expansion OOA at the moment of divergence of Y-DNA CT and mtDNA L3.

terryt said...

Did either of you people even bother to look at the diagram inthe Link Maju provided? It seems to be you two who "just want your dream come true. No proof and data".

Ebizur said...

terryt said...
"Did either of you people even bother to look at the diagram inthe Link Maju provided? It seems to be you two who "just want your dream come true. No proof and data"."

I don't know about Maju or SCH, but I did bother to look at the diagram, and something about it is making me want to look carefully at people in Japan and vicinity.

Maju said...

Did either of you people even bother to look at the diagram inthe Link Maju provided? It seems to be you two who "just want your dream come true. No proof and data".


What's the issue now?

The diagram shows 6 branches of N, of which:

- 2 (the only ones with an specifically Northern/Eastern distribution) appear to have gone through long and important periods of small population sizes (many many succesive SNPs)
- 3 are most important east and west of South Asia: in West Eurasia and Australia, with one of them being present in both regions.
- And then there is pre-R (including N9 and Y), which is widespread and appears to connect specially East and South Asia (main cores for N9/Y and R respectively).

It could be used to argue for more important roles in Eurasian genesis for West and East (SE?) Asia, I reckon, but hardly for any "northern corridor".

Maju said...

...but I did bother to look at the diagram, and something about it is making me want to look carefully at people in Japan and vicinity.

You may be on something. I have been today reviewing the genealogy of mtDNA M (as it's obvious that we can't look only at N to understand early Eurasian population dynamics) and I've found that, after South Asia, the area of highest top-level dversity for M seems to be East Asia (8 top-level M subclades, South Asia has 11), with many of those clades being most important in Japan (not sure how much of this impression may be due to comparative oversampling though).

Whatever the case, if we consider M and pre-R (R plus M9/Y), I do find some parallel in both macrogroups being most diverse in these two regions: South and East Asia (but not SE Asia it seems). N(x pre-R) is different though and does appear more clearly concentrated in the Indian Ocean arch (and I would not totally exclude a West Asian origin, though a South Asian one is also possible).

What has me intrigued is the not very apparent role of SE Asia in all these migrations (after reviewing the matter, SE Asia appears to have only 2 high level M subclades, one, M9/E, shared with East Asia and the other, M21, typical of Negritos). Even if we consider later repopulations of the region, we should find higher diversity among Negritos and admixed groups. Maybe it's an issue of undersampling, of lack of sufficient research? Am I missing something important?

While it's possible that people may have been migrating between South and East Asia sidelining SE Asia (though I can't imagine why), it's impossible that Paleolithic peoples could have migrated to Sahul without transiting by SE Asia. And yet, this realtive lack of SE Asian mtDNA diversity is puzzling. At the moment I can only think of incomplete data for the region as possible explanation.

Maju said...

PS - Another possible explanation could be the Toba supervolcano but that would mean an even earlier human expansion in Eurasia, i.e. before 74,000 BP, something that cannot be confirmed by any available archaeological data, at least East of India.

terryt said...

"something about it is making me want to look carefully at people in Japan and vicinity".

Exactly. You both appear to be unbelievably Eurocentric. So much so that, just because virtually all mt-haps in Europe (HV, JT and UK) originated in India (or nearby), you assume most of the others did too. And they all emerged from there at the same time.

We should start with what we all agree on: the above haplogroups originate in or near Northern India. But there is a long history written in the haplogroups. Possibly back to the origin of modern humanity, but not necessarily so.

Let's start with the common ancestor of the above haplogroups: R. Its closest relation is Indian N9. So far so good for the expansion from India theory.

Its next closest relation, P, is found only in Australia and New Guinea. Strange that no pockets of P managed to survive in India, or further north in Asia, after the migration. Perhaps some official at the start of the mass migration out of India ordered all those with mt-hap P to turn right, jump into dugout canoes and paddle to New Guinea. Of course there's no possibility at all that P and RN9 simply originated on either side of Wallace's Line, is there?

So let's look at PRN9's origin. We now arrive at far Eastern N* and Y. It's obvious to us all that they must have come from India, because they are extremely rare in the subcontinent itself. It must have come from India after R had formed? Correct? Perhaps. We know it's impossible that Far Eastern N*Y came across Central Asia. And of course it's completely impossible that the ancestor of PRN9 had actually moved south from the Far East.

Surprisingly this Far Eastern N*Y sits in the middle of an arc of related, reasonably regionally discrete mt haplogroups. In the west we find a complex group with a single origin: IXIndian N. Perhaps it is complex mainly because it has been well researched. After all some of the branches are found in Europe and North America.

To the east of this IXIndian N haplogroup we find Ket N* and A. Ket N*? What's it doing there? The same official mentioned earlier must have ordered all those with Ket N* to keep moving through 20 metre snow drifts and expansive deserts until they finally reached Central Asia. I mean it's impossible that any early humans could have lived in Central Asia, isn't it? But it certainly looks as though either Ket N* or Far East N*Y must have moved through Central Asia.

At the other end of this arc of related, reasonably regionally discrete mt haplogroups we have S, in Australia.

Oh. Sorry. we also have the enigmatic Oz N*W. So let's build a whole case on just this single haplogroup and not look for alternative explanations for its distribution. We might just get away with it. We'll have to use huge amounts of random drift, founder effect and bottlenecks to explain away any discrepancies though. We'll say that Oz N* and W both came from India and went in opposite directions. But hang on. They separated after the basal N haplogroup had already split into six. Ah. They must have left India at the same time as HV, JT and UK along with the other, more than twenty, haplogroups required to fit the theory.

And I'm prepared to bet that just three haplogroups had come out of Africa in the first place: mt-haps M and N and Y-hap CT. I'd guess that E moved back into Africa from somewhere in Asia (India? but more likely the Arabian Peninsula) after the eventual diversification.

Maju said...

Of course there's no possibility at all that P and RN9 simply originated on either side of Wallace's Line, is there?

"Originated" or rather "evolved"? IMO P as such coalesced into its present form in New Guinea but its immdeiate ancestor (R) did not. As is not found elsewhere on Earth, that's about all.

We now arrive at far Eastern N* and Y.

I've found places where this clade is just called N9 and therefore I'll use this term hereafter.

It's obvious to us all that they must have come from India, because they are extremely rare in the subcontinent itself. It must have come from India after R had formed?

Not necesarily. It may have arrived before R was formed or to be more malabaristic with the words, while R was forming in India and just before or simultaneously with the arrival of the very few R subclades that are common in East Asia, namely B and F.

How can this happen "before or simultaneously" or even "before and simultaneously and after too" all at the same time? Because migrations are not just picking a couple from place A and moving them to place B on a map, they are many people moving into new lands, coming maybe back to visit their relatives before a new generation advances even farther (or maybe moves back), etc. There may be a few cases where a migration is a total break in one day or year of all what was behind (relatives, known landscapes from infancy, etc.) but most normally it is not - but rather a gradual process taking many generations. Even the crossing into Sahul has been claimed to have happened that way (after the supposed accidental discovery of Australia by a lost canoe) and not as the product of a mere shipwreck or the accumulated effect of many of them.

So when the genetic distance between various clades is small, it is perfectly possible that they all participated in the same migratory phenomenon, either in a simple (mother and daughter) or complex (distant cousins, like members of the same tribe) manner.

Surprisingly this Far Eastern N*Y sits in the middle of an arc of related, reasonably regionally discrete mt haplogroups. In the west we find a complex group with a single origin: IXIndian N.

Only in your mind. That arc is a pattern you are drawing and that I do not see (check my later meditation on macro-haplogroup N). For me "macro-X" (including X, N1 and related Indian N*) is the only clear West Eurasian subclade of N (after I reconsidered "macro-W" as likely South Asian). They have absolutely no relation except by their branches in South Asia.

Why all N clades that are not exclusively local (i.e. macro-R, macro-X and macro-W) have one or more South Asian subclades?

To the east of this IXIndian N haplogroup we find Ket N* and A. Ket N*? What's it doing there?

This is interesting, yes. But my greatest interest is not because of their geography but because they remain undivided for a long span (till the present in the case of Ket N*). They must be (in their present definition) rather young haplogroups but that have gone, since the N node (i.e. since long ago), by a long intimate (and marginal) history.

I do wonder if they migrated by the steppary corridor instead of doing it by southern Asia and that would be why they show such a reduced population for so long. But this is just an speculation so far and a concession I do to your reasoning.

It cannot apply to the other four clades that show a rapid expansion (branching) right after the N node split. Hence there could be a "happy trail" along Tropical Asia and a "rough trail" through North Asia for N subclades. But the latter is so far just a mere hypothesis: they could also have migrated via East Asia and the long coalescence be caused by other reasons (there are many other marginal niches besides Ice Age Siberia).

We'll say that Oz N* and W both came from India and went in opposite directions.

Most likely correct, even if I have some doubts. W is found in highest percentages in Pakistan and has also sublineages as far east as Thailand.

But hang on. They separated after the basal N haplogroup had already split into six. Ah. They must have left India at the same time as HV, JT and UK along with the other, more than twenty, haplogroups required to fit the theory.

Hold your horses, man! HV is about 5 SNPs apart from the N node (maybe older than "evolved" A but hardly comparable with the 4 fast-spreading N subclades), while macro-W is only one.

The W-Australian N* node is comparable to the R-N9 one and must represent roughly simultaneous events (rings any bell?).

In any case, the impresion is that the expansion between the stages of N node to the best known derived clades of today (such as HV), passing by the R node (and parallel but less influential ones) happened probably in a quite short time span.

So you (or rather my own review of the data after you sparked this controversy) is persuading me more and more that the "rapid coastal migration" model (or something very similar) is correct in the end.

And I'm prepared to bet that just three haplogroups had come out of Africa in the first place: mt-haps M and N and Y-hap CT.

Why not just L3 and CT? Or why not M and N (mtDNA) and CF and D (Y-DNA)? Or even C, F and D on the male side?

I'd guess that E moved back into Africa from somewhere in Asia (India? but more likely the Arabian Peninsula) after the eventual diversification.

Why? Why not D being the one moving into Asia after split from E? E certainly appears totally centered in East Africa and by no means in Asia. In fact E could be just the product of CT in East Africa, after drift wiped out other variants. The date suggested for the expansion of E is certainly much more recent than that for the expansion of the rest of CT, what would suggest a long coalescence in East Africa before it found the occasion to expand.

terryt said...

Thanks for that effort Maju. We seem to be moving closer together.

"I do wonder if they migrated by the steppary corridor instead of doing it by southern Asia and that would be why they show such a reduced population for so long".

At least you now accept the possibility.

"they remain undivided for a long span (till the present in the case of Ket N*). They must be (in their present definition) rather young haplogroups".

Ah. Now we can agree on something else. This apparent young age, or at least diversity, is probably a product of earlir drift and bottlenecks. Ket N* certainly derives from basal N so it must have separated from other N haplogroups long ago and just hung around in a small population.

"It cannot apply to the other four clades that show a rapid expansion (branching) right after the N node split".

If it's only the lack of rapid expansion you're worried about see above.

From your own blog, "macro-R: including huge and widespread haplogroup R (South Asian by origin without any serious doubt), some Indian N* and East Asian N9 (that includes Y)".

This first haplogroup in your list doesn't just 'include' Y. Y is an early branch. You haven't taken this into consideration. In the diagram on your blog you have R leading to N9 whereas it's much more likely to be the other way round. This N9 haplogroup has a connection to KetN* otherwise we would most likely find the N haplogroup first splits into two: a western and an eastern version. That's not what we find.

"Overall I'd say that N is most diverse in South Asia".

Possibly because it came in from two directions, east and west.

"migrations are not just picking a couple from place A and moving them to place B on a map, they are many people moving into new lands".

Very true. But the haplogroup pattern doesn't allow us to unravel that detail. What it does tell us is the pattern of movement of various single clades long after the event.

"The W-Australian N* node is comparable to the R-N9 one and must represent roughly simultaneous events (rings any bell?)".

Certainly. Brings to mind my idea of an expansion from Wallacea involving R haplogroups.

"Why not just L3 and CT? Or why not M and N (mtDNA) and CF and D (Y-DNA)? Or even C, F and D on the male side?"

We still have far fewer haplogroups involved than your postulated exodus from India. M and N are distinct branches of L3 so their split is probably early in the diversification of that haplogroup. For that reason I suspect the two branches of L3 came out of Africa. For this reason I doubt the comment on your blog, "possibly indicating that the N-M split (from L3) happened in relation to events and geography of South Asia".

"In fact E could be just the product of CT in East Africa, after drift wiped out other variants".

Could be. But we require a huge level of drift to achieve that. Anyway, I'm just thinking about it at this stage.

Maju said...

Ah. Now we can agree on something else. This apparent young age, or at least diversity, is probably a product of earlir drift and bottlenecks. Ket N* certainly derives from basal N so it must have separated from other N haplogroups long ago and just hung around in a small population.


Sure. That seems to be the case. And to a large extent also applies to A (and surely other derived clades of whatever super-lineages here and there).

If it's only the lack of rapid expansion you're worried about see above.

I am worried at your lack of acknowledgment of such rapid expansion and its relevance.

From your own blog, "macro-R: including huge and widespread haplogroup R (South Asian by origin without any serious doubt), some Indian N* and East Asian N9 (that includes Y)".

In fact I commited an error there, because of the somewhat sloppery nomenclature and confusing graphics. There's probably no Indian N* in that "transitional" macro-haplogroup but actually Logan seems to be refering to Indian R sequences.

So R-N9 (or "macro-R" as I called it before) only has two sublineages: R and N9, the first one centered in South Asia and the latter in East Asia.

I guess you'd like this better, as it offers more opporunities for speculation about the pre-R period of N.

This first haplogroup in your list doesn't just 'include' Y. Y is an early branch. You haven't taken this into consideration.

Y is an early branch within N9 (or "N9-Y" if you wish) but it is not directly related to R. They both share the mutation at 5417 and therefore must be considered a single haplogroup for our purpose.

"Overall I'd say that N is most diverse in South Asia".

Possibly because it came in from two directions, east and west.


That doesn't make any sense.

N has four rapidly expanding subclades, right? One (macro-X-N1) went to West Asia, another (S) to Australia, yet another (macro-W) to all Southern Asia (West, South and SE Asia) and Australia too and then another is found in both East and South Asia.

Two directions only? Why not 10?! And from outer space as well, why not?

The original geography of N is not totally clear maybe, not like M and R (both neatly South Asian it seems) but you cannot just hang desperately to this ambiguity to defend your pre-concieved theory. You are building the house from the roof down and that is not the proper way to do things, much less in science.

"The W-Australian N* node is comparable to the R-N9 one and must represent roughly simultaneous events (rings any bell?)".

Certainly. Brings to mind my idea of an expansion from Wallacea involving R haplogroups.


And what then about another comparable node: X-N1-Indian N*? This has absolutely no relationship with "Wallacea" (what a name! I still don't know what it may mean: Indonesia or Sahul or maybe even the Marianas...).

And anyhow there is not enough diversity of R (or M for the case) around Wallace line (guess it has something to do with your "Wallacea") to justify that R (or M) sprung from there. And most Eurasian lineages are derived from R or M, not N(xR).

We still have far fewer haplogroups involved than your postulated exodus from India. M and N are distinct branches of L3 so their split is probably early in the diversification of that haplogroup. For that reason I suspect the two branches of L3 came out of Africa. For this reason I doubt the comment on your blog, "possibly indicating that the N-M split (from L3) happened in relation to events and geography of South Asia".

M and N are not directly related but L3 has many other subclades (7), all them virtually exclusive of Africa. M and N probably coalesced for a period of intermediate length (4 and 5 SNPs at root respectively) but we do not know where exactly. South Asia is the best candidate and could have harbored both perfectly (it's big enough). They could also have lived together all the time: drift does not necesarily drives to fixation in a single clade, in fact often it is several clades, specially re. mtDNA, which are found in heavily drifted peoples; even Beringians had at least 5 different mtDNA clades (and that was the end of the world, so to say), even Asutralians or Papuans show a an array of different mtDNA clades in spite of heir nealry total isolation, even Bushmen have more than one single mtDNA lineage. I an't in fact think of a single population of the world that has one single mtDNA clade, no matter how isolated it has been through history and prehistory.

So M and N could perfectly have been shared by a single population between the OOA and the Eurasian expansion. But M appears to have been dominant and, would not it be for R, it would still be that way.

"In fact E could be just the product of CT in East Africa, after drift wiped out other variants".

Could be. But we require a huge level of drift to achieve that. Anyway, I'm just thinking about it at this stage.


Y-DNA drifts much more extremely than mtDNA.

...

Anyhow, back to mtDNA, check my latest post on the issue: it does explore how the three Eurasian macro-haplogroups may have expanded. I have dedicated like three days to analyze the matter and I believe it is pretty transparent:

1. First expansion pulse (maybe c. 70,000 BP?): expands M and N. Covers all the Indo-Pacific "T" and also an N "avantguard" to West Asia (macro-X-N1).

2. Second expansion pulse (maybe c. 55,000 BP?) expands M and R. It also reaches all the Indo-pacific "T" and (maybe in the aftermath) we can see the bulk of Western clades (R-derived) arriving to West Asia.

M6, A and M1 are the main top-level clades that see expansion after these two succesive (maybe even continuous) epysodes.

You could argue maybe that "R people" were in posesion of some advantage (unclear, doesn't seem to be boating in any case) and drove the second expansive pulse because of that reason. But this expansion also includes many M subclades, so it would be more like the "R and M people" if anything. And this pulse, maybe even more clearly than the first one, seems centered in South Asia.

terryt said...

That's a pretty good post you've made at your blog. I agree with most of it. And I'm impressed that you accept at least two migrations into Australia. But where's Y in your N haplogroup diagram? You wrote in this blog that, "They both share the mutation at 5417 and therefore must be considered a single haplogroup for our purpose".

"So R-N9 (or 'macro-R' as I called it before) only has two sublineages: R and N9, the first one centered in South Asia and the latter in East Asia".

What about P? And Y? Both part of it if you accept six N haplogroups.

"That doesn't make any sense. N has four rapidly expanding subclades, right?"

Correct. But two others that sat around for a while before they were able to expand. You cannot ignore them just because it suits your theory. Once you include them you have two ends to the arc. One reaching Northwest India the other reaching East India. Make sense now? You're obsessed with the idea that everything originated in India.

"And what then about another comparable node: X-N1-Indian N*? This has absolutely no relationship with 'Wallacea'".

Of course not. It's at the Western end. By the way, Wallacea is the generally accepted term for the groups of islands that don't fit with either Sahul or Sunda. They've never been connected to any continent. Such groups as Suluwesi, Philippines, Halmahera etc. The Marianas are not part of Wallacea.

"And anyhow there is not enough diversity of R (or M for the case) around Wallace line".

No. But the haplogroups are sure split by it. Several are found only East of Wallace's Line and others only immediately west of it.

"M and N probably coalesced for a period of intermediate length (4 and 5 SNPs at root respectively) but we do not know where exactly. South Asia is the best candidate".

Not necessarily the best candidate. After all they didn't simply jump from Africa to India. I'd agree that M reached India reasonably quickly though, even if its immediately ancestral haplogroups are not found there (see Mathilda's Blog).

"drift does not necesarily drives to fixation in a single clade, in fact often it is several clades".

True. But, apart possibly from America, for most of the examples you give the variety seems to be a product of mixing rather than multiple haplogroups at origin.

"So M and N could perfectly have been shared by a single population between the OOA and the Eurasian expansion".

Almost certainly were. But it may not have been India.

"You could argue maybe that 'R people' were in posesion of some advantage".

I hope I've never led you to consider that I believe technology is never shared. The R mt-hap, along with K Y-hap do seem to have led some sort of expansion but the expansion certainly picked up other people (for example mt-hap W perhaps?). Besides which it seems that Y-hap K was originally associated with mt-hap M but must have picked up mt-hap R somewhere.

Maju said...

That's a pretty good post you've made at your blog. I agree with most of it.

Read (or rather watch) also my last one: a visual essay on a plausible sequence of Eurasian mtDNA diversification/expansion. It has absolutely no conclusions, it's mostly intended for contemplation and meditation.

Certainly the ultimate origin of M and N is not well determined and SE Asia could, in principle be an alternative to South Asia because of its central position but South Asia still holds the record of diversity and is a most likely origin for M and R (and therefore probably N too).

And I'm impressed that you accept at least two migrations into Australia.

It's a less rotund concept but certainly it does look like South Asians have been pouring to Australia at several times (the opposite does not seem likely).

What I think I've found is a rather continuous process of expansion in all the Southern-Eastern Eurasian (and continental Oceanian) "T" but maybe paused at an specific rather short period, pause that is coincident with the formation of R.

What about P? And Y? Both part of it if you accept six N haplogroups.

P is a derivate from R and Y from N9 (or N9-Y, or call it "macro-Y" if you hate numbers, though N9 is in fact a more important, more extended and older, clade). They are all part of R-N9 and of its two branches (R and N9).

But two others that sat around for a while before they were able to expand.

And maybe a thousand others that never made it to our time or even to birth.

You cannot ignore them just because it suits your theory.

I am not ignoring them. I was the first one in this dscussion who ever paid any attention to them. What I do is to place them in their right place: minor erratic clades that took very long in finding an opportunity for expansion, opprotunity that never really came in the case of Ket N* and that arrived pertty late in the case of A. A in practical terms (i.e. finished A and no coalescing proto-A) is more or less contemporary of the expansion of C and Z and the consolidation of Y, and is best called upon when dealing with the colonization of North Asia (and later America too), and not the main "explosion" of early Eurasians.

In the case of Ket N* you can surely ignore it alltogether and won't make any difference at all.

Once you include them you have two ends to the arc.

No, because we really have not the slightest idea where was proto-A at the time of the main Eurasian demic explosion. It could have been in North Asia but probably not. There were surely many other N* distinct lineages then that never made it. A was "lucky" but that doesn't make it more important before it began expanding.

Of course not. It's at the Western end. By the way, Wallacea is the generally accepted term for the groups of islands that don't fit with either Sahul or Sunda. They've never been connected to any continent. Such groups as Suluwesi, Philippines, Halmahera etc.

Thanks for the explanation. Those islands have no (AFAIK) any particularly ancestral haplogroup of any sort. Forget it.

Would you be talking about a really relevant place of the kind of Sunda... but still nope.

"And anyhow there is not enough diversity of R (or M for the case) around Wallace line".

No. But the haplogroups are sure split by it. Several are found only East of Wallace's Line and others only immediately west of it.


Obviously: it is a major barrier. I can't think of any other barrier in Eurasia as important as that one, except for the Himalayas and, arguably, Beringia.

When people are separated by a barrier they tend to evolve separately, even if this evolution is mere drift.

The surprising thing here is the many clades shared at some level between South Asia and Sahul, not the barrier at Wallace line.

Not necessarily the best candidate.

South Asia has several clear things in favor to be considered the urheimat of Eurasian humankind:

1. It has the highest M and R diversity (and the N diversity is at least comparable to any other region)

2. It is geographically closer to Africa, from where the ancestors had necesarily to come. This advantage is only surpassed by West Asia but there the diversity is much lower.

3. It does have archaeological sequences that appear to suggest pre and post Toba continuity, as well as migration routes from West to East.

4. It has a mostly tropical climate and c. 60,000 years ago was basically Savannah (the ideal climate for expatriated East Africans)

5. It is reasonably well communicated to all other Eurasian regions.

The ony serious contender would be SE Asia (maybe expanded to include a big chunk of East Asia) but the only advantage it has is a more central position re. the early expansion. The main disadvantage is the clearly lower diversity, something that kind of yells: "not here!"

So it does look like South Asia provided the "manpower", so to say, and SE Asia the "roads".

After all they didn't simply jump from Africa to India. I'd agree that M reached India reasonably quickly though, even if its immediately ancestral haplogroups are not found there (see Mathilda's Blog).

What are you talking about? The only ancestral haplogroup to M is L3.

True. But, apart possibly from America, for most of the examples you give the variety seems to be a product of mixing rather than multiple haplogroups at origin.

Ambiguous pretext. I don't care here if S and Q arrived (long ago) by different ways to Australia, what matters is that after many tens of thousands of years, they still keep those clades and many others. Hence drift does not drive to single-clade fixation necesarily. In fact it seldom does in all I know of human mtDNA.

What is surprising is not that there were two L3 subclades in the early Eurasian population but that there only two mtDNA haplogroups (or so it seems). They must have gone by a quite severe bottleneck and/or a most narrow founder effect. Otherwise I make little sense of it.

But, well, the bottleneck/founder effect seems to be something that pops up every time Eurasians are considered in comparison to Africans.

I hope I've never led you to consider that I believe technology is never shared.

Uh? Technology is always shared or copied. Pygmies use (European-inspired) crossbows nowadays, Talibans make their own Kalashnikov and Astra guns, gunpowder's, silk's, paper's and porcelain's secrets were all taken out from China. Technology, if useful, is always subject to be copied.

I don't know what you 'believe' but I am a quite radical difussionist - not of mere faith but analytical conviction.

The R mt-hap, along with K Y-hap do seem to have led some sort of expansion but the expansion certainly picked up other people (for example mt-hap W perhaps?). Besides which it seems that Y-hap K was originally associated with mt-hap M but must have picked up mt-hap R somewhere.

Psah. Haplogroups are not tribes. Tribes (and sometimes nations, which are a derivate of the tribe) have mythical ancestors, haplogroups have real ones. Sometimes they may be coincident but most often they are not. Adoption and alliance pull people together, strife pulls them apart. Clans and tribes have to be reinvented every other day, and meanwhile the real biological lineages sneak their way into all them as long as they are at reach. This is probably even more true for maternal lineages (at least in patrilocal societies).

Your prejudices, preconceptions, blind your judgement.

By this I don't mean that real biological ancestry may not go in parallel to real cultural advantages, transmitted from parent to child (roughly). But the link is limited and the more that time passes the weaker it should be: cultural difussion should dilute the advantage and biological exchange should make cultural and biological ancestry less and less connected.

terryt said...

"Your prejudices, preconceptions, blind your judgement".

That's, as we say, an example of the pot calling the kettle black.

"What are you talking about? The only ancestral haplogroup to M is L3".

So what is Mathilda talking about here?

http://mathildasanthropologyblog.wordpress.com/2009/01/28/mtdna-variation-in-the-south-african-kung-and-khwe%e2%80%94and-their-genetic-relationships-to-other-african-populations/#comments

"Certainly the ultimate origin of M and N is not well determined and SE Asia could, in principle be an alternative to South Asia".

I see now that you've been labouring under a huge misunderstanding. I have never claimed a Southeast Asian origin for M and N. I have merely claimed the evidence of N's distribution does not support a movement to the east through India.

"P is a derivate from R and Y from N9".

According to Ian Logan's diagram you so kindly supplied P diverged from the N9 line (or whatever you like to call it) before R had separated from N9. Likewise Y, and what he calls Far Eastern N, had separated even earlier. And that's just a single one of the six haplogroups N had split into even before then. How do you explain all that?

"I am not ignoring them".

You're certainly giving every indication of doing so.

"What I do is to place them in their right place".

Why don't you go back to your diagram of N's distribution and draw a line connecting all six early N haplogroup branches. By the way, I'd actually put KetN* a little further south than you've placed it but it still works where it is. I agree that it couldn't have lived that far north in very ancient times.

"minor erratic clades that took very long in finding an opportunity for expansion".

How many times does Dienekes have to point out that the time of a haplogroup's expansion can be long after that haplogroup has formed? Haplogroups A and KetN* must have eventually expanded from somewhere near where they formed and survived. You can propose any theory you like for where that was but I'd guess it was somewhere near where they're still found.

"In the case of Ket N* you can surely ignore it alltogether and won't make any difference at all".

You certainly 'believe' that. Others might disagree. Of course it's possible to prove anything if you're prepared to ignore enough evidence.

"Those islands have no (AFAIK) any particularly ancestral haplogroup of any sort".

It's generally accepted that, even though some show signs of early human presence, humans had died out on them by the time of the Austronesian expansion. Their haplogroups are releatively recent arrivals. So we have to look further afield to understand SE Asian haplogroup distribution. This explains what you have called "the many clades shared at some level between South Asia and Sahul". But most of the main ones are not shared.

"Haplogroups are not tribes".

Exactly. But we can use the distribution of haplogroups to determine the broad sweep of human migration. After all tribes contain haplogroups.

I feel we've spent enough time on this subject so I'll let you have the last word. Thanks for the stimulating discussion.

terryt said...

Sorry. One final comment. Mutation 16223 forms the root of the Y complex, or whatever you wish to call it.

Then 5417 forms the root of Y Far Eastern N* (including N9), and 12705 forms the root of P, R* and what he calls Asian sequences.

Then 15607 separates P from the other two, R* and Asian sequences.

Interesting, isn't it.

Maju said...

So what is Mathilda talking about here?

No idea: she says that there is no L3 in India, when in fact it's like 100%. She says that there is no M in India, when in fact it's like 80%.

Reminds me of the blind guy and the elephant.

I see now that you've been labouring under a huge misunderstanding. I have never claimed a Southeast Asian origin for M and N. I have merely claimed the evidence of N's distribution does not support a movement to the east through India.

You have been claiming that peoples got out of Ethiopia, went to Siberia and then to Australia and that from that area (aka "Wallacea") they expanded to India, more or less. So basically India was the last part of Eurasia to be colonized of all (no matter that everybody understands otherwise)

Stop playing with words: that's your model: defend it or shut up. But don't try to confuse me with "I said this and now I say that".

According to Ian Logan's diagram you so kindly supplied P diverged from the N9 line (or whatever you like to call it) before R had separated from N9.

The difference between you and me is that I have been painfully reviewing ALL the trees and discussions in that site (relevant to Eurasia), not just that one.

So I know what I'm talking about finally and you are still stuck in that confusing anotation.

Why don't you go back to your diagram of N's distribution and draw a line connecting all six early N haplogroup branches.

Because that is not the right thing to do, actually it's misleading: showing data about events that happened in totally separated times. I did something much better instead.

Haplogroups A and KetN* must have eventually expanded from somewhere near where they formed and survived. You can propose any theory you like for where that was but I'd guess it was somewhere near where they're still found.

A theory: they were in Mars. And I challenge you to disprove it. You can't: there is absolutely no evidence of where these minor lines were expressed between the N node and, in the case of A, the A node and, in the case of Ket N*, the present time.

And what "somewhere near where they're still found"?

In the case of Ket N* it could be Europe, where a related sequence has been found, or Central Asia, or East Asia or... whatever.

In the case of A, I'm betting for the Bahamas in fact. I think it is near to where they are found at the present, right?

Of course I'm not being really serious: Mars and the Bahamas are not realistic posibilities and neither is Wallacea nor your early Siberian route.

Stick to the facts. Join the dots, ok, but please do not join the Shanghai Tower with the Pyramids and these with Lascaux because then you are mixing apples and oranges.

Ket N* is like the Shanghai Tower: it belongs to the present. It may have been there for a while now but we really have no damn idea of where it was in the year 60,000 BCE. "Somewhere in Eurasia" is as close as we can get.

And we do not know how many Ket N*-like subclades have existed or where. Probably many but they were drifted out. This also applies to some Australian and Papuan mtDNA clades with very looong roots: they may have been coalescing for long in Sahul... or in Indonesia, or in India...

You say it's all the same. I say it's not: that each one belongs to its time and that very long chains of SNPs assembled in a unique but extremely rare lineage do not say anything about the history of that clade, nor give much info either on the ones related to it (in the Oceanian cases, other macro-F and W repsectively).

They are just odd survivors.

You certainly 'believe' that. Others might disagree. Of course it's possible to prove anything if you're prepared to ignore enough evidence.

You will disagree, I'm sure. Because such rare clades appear to be important in your improbable assmbling of the pieces of the puzzle.

Well, get it straight: "Ket N*" is as European as Siberian, so maybe it represents an early remnant of a very early colonization of Europe by H. sapiens, long before Aurignacian. Why not?

I chose to call it Ket N* capriciously but Euro-Siberian would be more precise, as two sequences are listed: one Ket and the other European.

But it doesn't really matter: it's obviously just an erratic.

Sorry. One final comment. Mutation 16223 forms the root of the Y complex, or whatever you wish to call it.

Then 5417 forms the root of Y Far Eastern N* (including N9), and 12705 forms the root of P, R* and what he calls Asian sequences.

Then 15607 separates P from the other two, R* and Asian sequences.

Interesting, isn't it.


In other words: R-N9 splits in:

- R (including P and Asian R lineages, as well as European R, as he calls it it that graph)
- N9 (including Y)

That is precisely what I'm saying all the time - at least since I realized that the "Asian sequences" are nothing but Asian-specific R lineages (most of them South Asian). That graph is confuse and confusing and you should look at other graphs in the same site (or elsewhere) and take due notes until you get your facts right.