I have recently discussed the Denisovan admixture in Melanesians discovered by the Neanderthal Genome Project and I discussed back in its day the Neanderthal admixture in all non-Africans (see here and here).
While the Neanderthal admixture episode may be easy to explain and was thus explained by Green et al. as happening early in the migration out of Africa, probably before arrival to South Asia. The Denisovan admixture in islands far away from Altai is not so easy to understand and has not been satisfactorily explained by anybody I know so far.
What were the Denisovans?
 |
| Denisova cave |
First of all we have not a very clear idea of what kind of hominin were the Denisovans. Well, actually we know that their tooth clusters with Indonesian H. erectus, H. habilis and australopithecines (but also with the H. sapiens of Pestera cu Oase, quite divergent from the rest in this aspect).
We also know that the Denisovan mtDNA belongs to a branch older than that of H. ergaster and descendants, because it is almost twice older in its divergence from that of Neanderthal and Sapiens mtDNA (both derived from H. ergaster c. one million years ago by all accounts that make any sense). What diverges in the common tree of Humankind (senso lato) almost twice that time? Asian H. erectus, believed to derive from a population represented by H. georgicus.
Nothing else does. Hence the Denisovan mtDNA, found in two different individuals (a finger and a tooth actually) must be that of Asian H. erectus.
However the Denisovan nuclear DNA is not so distant from Neanderthals. What does it mean? Most probably that they were a hybrid Erectus-Neanderthal population, what fits well with their presence in Altai (at the crossroads of known homelands of both species), their use of Mousterian technology (typical of Neanderthals) and the presence of Neanderthals in similar dates at nearby sites.
So my theory about Denisovan identity is this one: they were a hybrid population of Neanderthals and H. erectus, with maternal lineages of the latter species and technology of the former.
Melanesians in Siberia? No way!
 |
| blond Melanesians |
Quite obviously Melanesian ancestors were never in Siberia. This is not just a matter of the coastal migration model, that also, but specially a matter of pigmentation. The name Melanesia means Islands of the Blacks in modified Greek and, if the ancestors of these peoples would have been in Siberia for any extended period, they would have lost their tropical pigmentation for sure because otherwise they would not be getting enough vitamin D and their children would be extremely unfit for that reason (retarded, schizophrenic, rickety, etc.) And, as the case of Native Americans clearly illustrates, re-evolving black pigmentation, once it is lost, is no easy matter. In maybe 15,000 years tropical native Americans have only got a tan.
So the ancestors of Melanesians and other very dark tropical Asians have definitively not lived in Siberia at any time. Besides, it is totally non-parsimonious in what regards to modern human mtDNA and Y-DNA spread, the tropical route is much more logical and natural.
So they must have admixed with some relative of Denisovans elsewhere, for example in Sundaland, where some Homo erectus are known to have lived in dates that are perfectly compatible with this scenario. An encounter of the first of our species arriving to that area and Homo erectus soloensis is almost sure to have happened.
So we do have a plausible and even likely scenario for this admixture event in the ancestors of Melanesians, not in Altai but in SE Asia.
Admixture detection by proxy... interesting.
Certainly that we can detect admixture happening in Java by studying distant relatives in Altai is interesting. And it makes sense. If you compare a modern French-Vietnamese with French and Altaians it's likely that he will appear as a mixture of French and Altaians, even if the proportions may not be exactly correct.
I'll get to this matter of proportions later on because it is relevant too.
What happens if we get the son of an Punjabi and Vietnamese and compare with French and Altaians? He will surely still show up as admixed. A simplistic conclusion might be that he is descendant from French and Altaians. This conclusion would be wrong, even if the confusion is understandable.
The Narmada hominin and "Neanderthal" admixture
 |
| Narmada skull |
Thinking about this brought me (with some important help from readers - feedback is crucial) to the mysterious Narmada or Hathnora hominin (see here for an open access reference), the oldest of really big-brained humans and possibly a relative of Neanderthals (but not a true Neanderthal, among other reasons because they did not use Mousterian technology but Acheulean). The skeletal record of South Asia is quite scarce but this big-headed hominin is the last people we know about before African-like Middle Paleolithic technology appears c. 120,000 years ago (see here), probably with the first members of our species.
Yes, you read right: 120,000 years ago (more or less), the idea of a much more recent Out of Africa is almost certainly wrong, even if you will surely read such nonsenses for a while: the molecular clock pseudo-science cannot overrule archaeology.
It is at this moment uncertain whether the Narmada specimen and the probably much larger population it belonged to was a descendant of H. erectus or a descendant from H. heidelbergensis (and hence closely related to Neanderthals). Depending on which of these two options is correct, the scenario presented below will make sense or need to be revised.
I will consider, as suggested here by Michael Petraglia, that the Narmada specimen and related Indian population of the Early Paleolithic (which lasted until c. 100,000 years ago) were descendant of H. heidelbergensis, and hence cousins of Neanderthals. Why? Because they had Acheulean technology, which is associated with at least the late H. ergaster.
If this is correct, when we talk (after Green 2010) of Neanderthal admixture at low levels in non-African modern humans we may well be talking of admixture with anything within the broader Neanderthal family, in other words, with its ancestor H. heidelbergensis (cousin of our most direct ancestor H. rhodesiensis) and their descendants (Neanderthals and others, including probably the Narmada hominin and broader Acheulean-using population of South Asia.
A hypothesis strongly consistent with the coastal (or tropical) migration model
And I finally reach here to my hypothesis, to my explanation of the admixture episodes revealed by the Neanderthal Genome Project this eventful year of 2010. And I will do it with few words:
 |
| click to expand |
The first admixture refers to the general non-African admixture with "Neanderthals", which would actually have happened with their Indian cousins instead upon arrival to South Asia. This admixture would have affected all non-Africans, but as the case of the Karitiana (who only show some 0.9% of such admixture, much less than the rest) evidence, maybe not all populations exactly in the same amounts.
The second admixture refers to the specifically Melanesian hybridization with "Denisovans", which would actually have been with their Indonesian pureblood relatives, H. erectus soloensis.
Makes sense? I think so. Of course, it is not set on stone but it seems a good hypothesis and should at least get some people chewing on this.
Special thanks for some key references to Terry T. and Manju (but in general to all readers who take part in the discussions at the comments sections: keep the flow of ideas vibrant, please). I suspect that Terry will not like my conclusions because they end up in the coastal migration model that he hates so much. But well...
Appendix 1: the real apportion of Melanesian admixture with archaic hominins may be lower than suggested by Reich 2010.
They suggest (supp. info 8) that Melanesians would have as much as 7.4% of admixture with archaic species: 4.8% Denisovan plus 2.5% Neanderthal. But, if Denisovans are hybrids of H. erectus and H. neanderthalensis (as seems most likely, see above), then the real admixture with H. erectus would be an undetermined percentage but always less than 4.8%. As we know that the Neanderthal (or Heidelbergensis) component is 2.5%, it is most likely that the actual Erectus admixture in Melanesians is of only 2.3% or 2.4%, totaling 4.8%.
Appendix 2: very serious inconsistencies in the age estimates derived from nuclear DNA in Reich 2010.
In supp. info. 6, the authors provide data of genetic divergence between various modern populations, Neanderthals and Denisovans, expressed as fractions of the Homo-Pan divergence. They use the wrong time frame for this event (6.5 Ma) but even then the results make no sense.
Using a much more correct (according to modern best scientific understanding) of 8 million years, I get that the age of the migration Out of Africa would have happened between 650,000 and 500,000 years ago. The first figure is the distance between Yorubas and non-Africans and the latter the one between non-Africans.
This is a total nonsense (120,000 years makes sense, add some tens of thousands more if you wish but half a million years is simply not possible) and there must be a critical error somewhere. However I have not been able yet to discover what exactly is wrong. In any case, word of warning about accepting molecular clock age estimates in general and in particular when using nuclear (autosomal) DNA for this purpose.
Greetings Maju,
ReplyDeleteIn relation to your hypothesis, which I think makes sense, one thing that does not fit me finish.
Correct me if I'm wrong, but I think the archaeological materials of Denisova level 11 can be ascribed not so clearly a Mousterian context.
The material culture of this level 11 can be ascribed to an Upper Paleolithic culture, closer to Chatelperronian that Mousterian.
While applying logic, your hypothesis erectus and Neanderthal hybrid makes sense, would not be supported by a context typically associated Mousterian Neanderthals.
The thing is complicated by the recent doubts about the authorship of culture Chatelperronian by Neanderthals.
Then we have to assume that the mix-erectus Neanderthal is able to develop an advanced culture or that Neanderthals have capabilities similar to the material culture sapiens displayed on Level 11.
It is unclear why the culture that can aportal erectus is Acheulean.
A third hypothesis is that the man of Denisova is actually a distinct species with no mixture of Asians, who share a common ancestor with Neanderthals.
But again, correct me if I'm wrong, because I have not read anything about level 11 since last May, and perhaps new data have come about.
"Makes sense? I think so".
ReplyDeleteNot quite, although we are definitely moving much closer together.
"I suspect that Terry will not like my conclusions because they end up in the coastal migration model that he hates so much. But well..."
And that brings us to the problem:
"The Denisovan admixture in islands far away from Altai is not so easy to understand and has not been satisfactorily explained by anybody I know so far".
And your explanation is not really adequate, although I agree with much of what you wrote.
"Quite obviously Melanesian ancestors were never in Siberia".
But some 2-4% of their ancestors may have been. And even if the region of their hybridisation is SE Asia that just moves the problem back a step.
"So we do have a plausible and even likely scenario for this admixture event in the ancestors of Melanesians, not in Altai but in SE Asia".
Yes, but... What is the connection between the Denisovans and SE Asian Homo erectus? They must both have formed part of a single widespread population at some time. The connection is almost certainly not through India where we find a different species, the Narmada. And in your previous post on the subject you mentioned Mongolian 4% and Han Chinese 3.2% Denisovan. What level in India? And I seem to remember someone saying the Denisova level in the Japanese was comparable to the Melanesian and Han Chinese. So the Denisova/SE Asia H. erectus species cannot have spread via the southern coastal route.
@David: that's probably the level of H. sapiens colonization, c. 30 Ka ago, right? I'd have to check.
ReplyDeleteRight now I remember that Reich et al. do mention (in one of the sections of the S.I. PDF) that they took radiocarbon dates from several objects from the site. Excepting a hyena bone, all were either "infinite" (>50 Ka, which C-14 cannot measure) or from 30 Ka till present. They said that they believed (cautiously) that the infinite dates corresponded to 'Denisovan' hominins' occupation and that the more recent ones to occupation by H. sapiens.
This makes sense and is consistent with the overall chronology generally accepted for the whole area. Denisova is rather to the North of the region but other caves further South have older dates (40 Ka or older) with "Aurignacoid" industries, on top of Mousterian levels (Neanderthal).
All the research in the region is relatively new, so there are some uncertainties but this seems to be the basic frame: pre-40 or rather pre-50 Ka Mousterian (Neanderthal and 'Denisovans', till this year believed to be Neanderthals too) and post-50/40 Ka "Aurignacoid" industries eventually also yielding H. sapiens skeletal remains.
But I have to check specifically for Denisova cave's "Chatelperronian", which may also be work of H. sapiens (initially it was believed to be the case in Europe too, and Chatelperronian and Gravettian were both known as a single culture: Perigordian, a label now deprecated, because continuity through time could not be demonstrated and Chatelperronian came to be thought eventually as a late Neanderthal industry).
"It is unclear why the culture that can aportal erectus is Acheulean".
I do not understand well this sentence ("aportal"?) but AFAIK Acheulean never reached Eastern Asia, where the last Erectus are known to have lived.
"A third hypothesis is that the man of Denisova is actually a distinct species with no mixture of Asians, who share a common ancestor with Neanderthals".
That would not fit the baseline theory by which proto-Neanderthal would have diverged from proto-Sapiens (at the Ergaster stage) with the spread of Acheulean c. 900,000 years ago (or maybe even more). I understand that this one is the correct baseline (though there are other hypothesis around, they lack any clear evidence and are based either on feeble molecular clock hunches or on equally feeble anthropometric speculations).
And by the way Denisova is a woman (at least one of the two is, they checked that too).
@Terry:
ReplyDelete"But some 2-4% of their ancestors may have been. And even if the region of their hybridisation is SE Asia that just moves the problem back a step".
Re-read please the section in which I ponder about detecting admixture "by proxy".
I believe that the problem is easily solved once we understand this. Of course this is a theory that makes a testable prediction: that, when H. erectus in other latitudes has its DNA successfully sequenced, it will be evident, with due analysis, that it is the same source of Melanesian admixture. Only that can prove my theory right or wrong in this aspect.
"What is the connection between the Denisovans and SE Asian Homo erectus? They must both have formed part of a single widespread population at some time".
Agreed. I am thinking all the time in Asian H. erectus as a species spread (at least by the time of the OoA) only through East Asia. Altai would be then their westernmost (and maybe their northernmost) habitat. And Indonesia would be their southernmost extent.
"And in your previous post on the subject you mentioned Mongolian 4% and Han Chinese 3.2% Denisovan".
Those are the percentages of Neanderthal admixture (or Heidelbergensis/Narmada).
I did mention in small type that it looked like Cambodian, Chinese and French might have also some very minimal admixture with Denisovans/Erectus but it's a doubt I have from the data in SI-8, nothing certain. Probably it is a misunderstanding on my side.
"And I seem to remember someone saying the Denisova level in the Japanese was comparable to the Melanesian and Han Chinese".
Nobody said that in this blog.
"So the Denisova/SE Asia H. erectus species cannot have spread via the southern coastal route".
We have very little idea on how H. erectus spread through Asia, at least I do not know. It doesn't matter because we know that before the OoA of H. sapiens they are found in all East Asia (China and Indonesia at least).
What I did say of the Japanese (only studied in the previous paper) is that they have somewhat lower Neanderthal (or Heidelbergensis) admixture than Chinese and Europeans. Actually only one Japanese was compared, so this only applies for sure to this individual.
ReplyDeleteI mentioned this fact because of the Karitiana very low levels of Neanderthal admixture (this paper). It may be related (i.e. an unadmixed or low admixture population may have followed the Eastern Asian coasts northwards and later spread to America; the Japanese could have got their "dose" of Neanderthal later, maybe as late as in the Neolithic or Iron Age).
"It is unclear why the culture that can APORTAL - provide - erectus is Acheulean"
ReplyDeleteThis sentence is badly expressed, sorry. Sometimes I use the google translator.
I wanted to refer to a possible admixture between Neanderthals and erectus, modern culture that erectus can provide for their tradition is the Acheulean culture, compared to a culture Mousterian or Neanderthal Chatelperronian that would bring.
We found evidence of Acheulean 800,000 years ago in the Boise Valley, China, East Asia, but then there remains a empty of Acheulean until about 300,000 years, when it already appears more frequently in Asian deposits.
There are several theories about this, but in summary are being told that this Acheulean empty exists because erectus used perishable materials like wood.
"Agreed. I am thinking all the time in Asian H. erectus as a species spread (at least by the time of the OoA) only through East Asia. Altai would be then their westernmost (and maybe their northernmost) habitat. And Indonesia would be their southernmost extent".
ReplyDeleteAgreed. This is my take on the situation although I realise you'd rather I started my own blog. But between us I think we are making huge strides.
It looks very much as though at the time of the OoA the 'modern' humans encountered two separtae populations of what we might call 'archaic' humans. We have Neanderthal/Narmada people through Europe, SW Asia and in India (at least in some of it), and we have Denisova/SE Asian H. erectus stretching from the Altai through Mongolia and China and thence down to Java. The two types may be associated with some sort of erectus/ergaster ancient separation, with an overlay of H. heidelbergensis through some of the total geographic distribution. Anyway the split fits the old Movius Line reasonably well. As you said, 'AFAIK Acheulean never reached Eastern Asia'; the defining element of the Movius Line.
But to me it's reasonable to assume gene flow between the two archaic types, so we probably have a series of clines with the four geographic extremities Dienekes has demonstrated exist in modern humans being the most different: Africa, Europe, East Asia and SE Asia.
Into this soup mixture came the African haplogroups and their associated aDNA. They mixed, but the African-derived haplogroups eventually replaced all the older ones in Eurasia, but at least some archaic aDNA remains. And of course eventually humans were able to move well beyond the archaic human geographic range.
As the haplogroups moved through the population they diversified into regional varieties. Many of these have since expanded in their own right. But at the margins of haplogroup expansion drift or selection (call it what you will) frequently greatly reduced haplogroup diversity.
@David:
ReplyDelete"Boise valley"? It may be a coincidence of names but when I make a search I always end up in Idaho, USA.
It must be Bose basin (found this paper). Bose is in Guangxi-Zhuang, what is just North of Vietnam.
This would actually be the evidence that Terry asked for in relation of a possible coastal (or inland but tropical anyhow) migration, if not of H. erectus at least of whoever carried the Acheulean, because it's the oldest and only known site claiming such technology in East Asia.
I cannot find any reference to the Acheulean presence in East Asia since c. 300 Ka ago. Can you post a link to your references - either just copy-paste or you can use HTML as follows:
Preliminary note: I have to write <...> as [...] because otherwise a code error will happen, but you have to replace the square brackets: [...] by greater-than/lesser-than signs: <...> when you post a real embedded link. Ok?
HTML code is:
[a href=LINK]TEXT[/a]
This way the LINK shows up as clickable blue TEXT.
Btw, if you have problems expressing yourself in English, I understand Spanish perfectly. However many other readers probably do not. I have no particular policy on comments' language, though English is generally preferable.
@Terry: I'm sorry but I do not understand this time what you mean or where you want to arrive to.
ReplyDeleteThe first admixture refers to the general non-African admixture with "Neanderthals", which would actually have happened with their Indian cousins instead upon arrival to South Asia. This admixture would have affected all non-Africans, but as the case of the Karitiana (who only show some 0.9% of such admixture, much less than the rest) evidence, maybe not all populations exactly in the same amounts.
ReplyDeleteThe second admixture refers to the specifically Melanesian hybridization with "Denisovans", which would actually have been with their Indonesian pureblood relatives, H. erectus soloensis.
Makes sense?
Yes, it does. That certainly suggests the potential proof for the coastal migration. But as the things stand today, Neanderthal admixture in Middle East and Denisovan admixture in Siberia, the northern route is the most plausible explanation. Since Melanesians are not light skinned we can certainly infer that they didn't stay there long and moved South. Remember, we are talking about the rapid migration of the early Sapiens.
It is the explanation that requires less thought but not the most plausible one, IMO.
ReplyDeleteFirst of all we have not any evidence of MSA in Altai or anywhere along that route and instead we have it in South Asia, very clearly since c. 120 Ka ago.
Second, there is the key issue of pigmentation, which makes a northern route extremely unlikely, specially for Melanesians.
So the northern route may be ok for a "first thought" but it does not stand a critical "second thought". Q.E.D.
"Since Melanesians are not light skinned we can certainly infer that they didn't stay there long and moved South".
"Not long" may well mean just a few millennia. Rapid is not instantaneous. And for admixture to happen some time should have passed with such introgressing neighbors.
In those millennia, barring preliminary pigmentary (and other, technological, such as needles) adaptation, the proto-Melanesians should have died off because of extreme lack of fitness.
First of all we have not any evidence of MSA in Altai or anywhere along that route and instead we have it in South Asia, very clearly since c. 120 Ka ago.
ReplyDeleteCreated by H. Heidelbergensis. Homo Sapiens MSA is around 30000 years ago in Karnataka.
The relevant link arose in the other discussion: there is MSA-like MP in South Asia since c. 120 Ka ago (Petraglia 2010), in Didwana (Pakistan), Patpara (Son valley) and Bhimbetka (Madhya Pradesh), overlapping briefly in time with the last Early Paleolithic of Acheulean nature, which would correspond to the Hathnora hominin (whose age has been pushed forward to c. 160-185 Ka or at most 230 Ka).
ReplyDeleteThis is also, roughly, the date of earliest known stone blades, which would eventually become a key feature in some (but not all) of H. sapiens techno-cultures (mode 4).
Btw, I thanked (in particular) Terry for input but now I realize that one of the two key links/info, the one explaining that Hathnora is probably derived from H. heidelbergensis, was provided by you. I have expanded my hat-tip paragraph accordingly.
But H. heidelbergensis is not associated (nor should be) with MSA. Theirs is the Acheulean culture (in part), at least in principle (and the Indian fossil record agrees here).
Hola Maju.
ReplyDeleteTe escribo en español, sólo por esta vez,para aclarar lo del valle del Bose. A pesar de mis dificultades con el inglés me intento integrar en las noticias y debates de tu blog, realmente interesantes, y por respeto a quienes no entienden el castellano.
Evidentemente, me refería al link que muestras. La cuenca del Bose, al norte de Vietnam.
Lo de los 300.000 años tengo que reconocer que ha sido un error de interpretación mío de dos libros diferentes: el primero es Homínidos: Las Primeras ocupaciones de los continentes, página 343.
http://books.google.es/books?id=zXbT4aDj-LoC&pg=PA341&lpg=PA341&dq=800ka&source=bl&ots=GqSNRu_2bk&sig=QH2nbtdSc--fDXKNXJKVJF9czRw&hl=es&ei=EksXTbvzD8HLswar0qj6DQ&sa=X&oi=book_result&ct=result&resnum=2&ved=0CCwQ6AEwAQ#v=onepage&q&f=false
El segundo es Nociones de Prehistoria General, página 176:
http://books.google.es/books?id=P1M-bMtQWEkC&printsec=frontcover&dq=nociones+de+prehistoria+general&hl=es&ei=NVEXTfnhOYiPswb985X6DQ&sa=X&oi=book_result&ct=result&resnum=1&ved=0CCkQ6AEwAA#v=onepage&q&f=false
Espero salgan bien los enlaces :)
Estaba intentando dar una hipótesis, basada en la teoría de Hartung&Koeberl,1994, según la cual no aparece tecnología del modo 2 o achelense debido a la escasez de materia prima, y por lo tanto a la utilización de materiales perecederos (resumiendo la teoría completa del impacto de los meteoritos). La hipótesis que te planteba era que si habían aparecido cantos trabajados de cuarcita en la cuenca el río Amur, en los yacimientos de Ulalinka y Filimoshki, con dataciones entre 700.000 y 130.000, tal vez Erectus habitaba esta zona de Siberia en época muy temprana y en base a la anterior teoría no aparecían industrias del Modo 2 por haber utilizado materiales perecederos.
No obstante quizás me he enredado yo bastante con fechas, zonas geográficas y conceptos. Espero no te hayas molestado, ya que en ningún caso pretendía cuestionar tu hipótesis, la cual me parece muy interesante.
Saludos!!:)
One possibility is that the tempo of the OoA is a bit later. In fig. 2 it is evident that South Asian MP sites begin c. 120 Ka. but in fig. 4, it is made evident that there is a void between earliest MP and what can be safely considered as MP made by H. sapiens. The void is between c. 100 Ka and c. 80 Ka. and the corresponding Arabian sites only appear to begin c. 90 Ka.
ReplyDeleteSo guess it can be argued that successful colonization by H. sapiens only happened since c. 80 Ka. But the fact that there are H. sapiens sites (albeit with Mousterian) in Palestine c. 130 Ka, still allows for a less clear early OoA, which could be supported by East Asian fossils and technologies, dating to more than 100 Ka.
But I must declare myself agnostic on which of the two scenarios is the most valid one, because the evidence in favor of a c. 120 Ka OoA is admittedly weak at this stage (but not totally lacking).
Vale, David, entendido.
ReplyDelete"Estaba intentando dar una hipótesis, basada en la teoría de Hartung&Koeberl,1994, según la cual no aparece tecnología del modo 2 o achelense debido a la escasez de materia prima, y por lo tanto a la utilización de materiales perecederos"...
No muy probable, puesto que sí que hay modo 2 en Bose Valley (aunque sea excepcional).
"Espero no te hayas molestado, ya que en ningún caso pretendía cuestionar tu hipótesis, la cual me parece muy interesante".
No, en absoluto. Me parece bien que me cuestionen, faltaría más. Se propone una hipótesis o teoría, se cuestiona, se saca una más refinada... mi propio pensamiento no está para nada cerrado y si tengo preferencias es porque entiendo que los datos y el sentido común las favorecen (puedo equivocarme en esto también).
Synthesis of the Spanish-language discussion:
David says that he got confused and that he tried to propose a theory by which mode 2 (Acheulean) did not exist in East Asia for lack of appropriate raw materials.
I say that this sounds most unlikely. In any case there seems not to be Acheulean in East Asia excepted the Bose Valley brief incursion.
Very interesting, Maju!
ReplyDeleteI don't have many time now to discuss, but I'm wondering whether the different % of neandertal admixture is due to different admixture rates betweem different groups or it's just only due to bottlenecks and/or founder effects.
The Chinese/Cambodians show the highest %, while the Karitiana shows the lowest, but I highly doubt that there were more than just one interbreeding with neanderthals, if in Europe they were clearly replaced. In figures of the study, it doesn't seem that the Han are more closely related to neanderthals than de French. The same isn't true for Melanesians vs other Non-africans and Denisovans: we can clearly see more affinities with the formers than with the second ones.
But H. heidelbergensis is not associated (nor should be) with MSA. Theirs is the Acheulean culture (in part), at least in principle (and the Indian fossil record agrees here).
ReplyDeleteI keep my fingers crossed. There are speculations that Aurignacian culture could be Neanderthal too. I guess if we go with that open mind Heidelbergensis and MSA shouldn't surprise us.
@ML:
ReplyDeleteThat's not easy or maybe even possible to know. Random drift, along with founder effects, such as the ones affecting Native Americans before and after Beringia, can have caused the differences we see.
"it doesn't seem that the Han are more closely related to neanderthals than de French".
It does seem so: all East Asian groups show levels that are markedly greater (45-70% greater) than those of Europeans or Melanesians. Exceptions: Japanese (previous study) and Karitiana. I already raised an eyebrow on this matter with Green's paper, and Reich's one seems to confirm it.
Of course it may be a matter of random events but there it is anyhow.
"The same isn't true for Melanesians vs other Non-africans and Denisovans"...
This is a yes/no case it seems. It's not a matter of degree but of absolute yes/no.
@Manju:
"There are speculations that Aurignacian culture could be Neanderthal too".
Well there are also recent speculations that Chatelperronian could have been made by H. sapiens.
Sure: one should always keep the mind open but also try to follow the threads of patterns.
As for Aurignacian being made by Neanderthals, it is most unlikely because Aurignacoid industries are found nearly everywhere in association with H. sapiens fossils. So for me, unless demonstrated otherwise, Aurignacoid industries mean H. sapiens.
"It does seem so: all East Asian groups show levels that are markedly greater (45-70% greater) than those of Europeans or Melanesians. Exceptions: Japanese (previous study) and Karitiana. I already raised an eyebrow on this matter with Green's paper, and Reich's one seems to confirm it. "
ReplyDeleteBut it's a matter of degrees (it seems so) I don't know if I'm right, but what I understood is that the % are only averages, it's possible to find a French person with more neanderthal DNA than a Han person, but that not true for Denisovans, when you compare a French/Han and a Melanesian person.
I don't know whether these studies can detect more interbreeding events or not, but the general line is, as for neanderthal DNA, that all non-Africans have the same amount, at least this was what they said the first time, that there were no significant differences.
Anyway it's quite rare, because at the same time Melanesians are more admixed with archaics such as Denisova than East Asians. Maybe they came from different migrations? If Denisovans could intebreed with modern humans, does that mean maybe other hominids such as H. floresiensis could as well?
Many time ago I read that near the site where the fossils of that hominid were found, there's a village where people usually are less than 1,40 m tall. Of course, that doesn't mean anything at all, but I'm curious.
What I found to be completely unintelligible is that aparently humans mixed with these hominids only in "two" occasions. What does this mean? Maybe they meet each other many times but nothing happened?
The Jeffrey Long's study dated these "two" occasions: 60.000 in the case of neandertals and 45.000 in the case of Denisovans (or any other relative). How can this be possible, if modern humans left Africa more than 100.000 years ago, and interacted with neanderthals for a long time? What about the modern human jaw found in China and dated more than 100K? Maybe these dates are wrong?
"If Denisovans could intebreed with modern humans, does that mean maybe other hominids such as H. floresiensis could as well?"
ReplyDeleteIt's difficult to say because there's no agreement on what exactly is H. floresiensis. Probably they were just a pygmy H. erectus variant but some even say they were Australopithecus. I'd say that Australopithecus are too diverged for a realistic chance of hybridization... but who knows?
"What I found to be completely unintelligible is that aparently humans mixed with these hominids only in "two" occasions. What does this mean? Maybe they meet each other many times but nothing happened?"
I do not know. The low levels of admixture suggest a gradual dilution of the archaic blood, which must have begun with either individuals or populations who were 50/50.
I already mentioned with the previous paper that it is likely that there was initially a population or group (or even individual) more intensely admixed, say 50/50, and that these must have been absorbed into a larger "pureblood" population.
If there was one initial 50/50 hybrid individual in a population otherwise 100% H. sapiens, this population would have been of c. 20 people (effectively reproducing population). If the initial hybrid population were 20, then the overall population would be 400... and so on.
It's only logical that people almost only reproduced with other people like themselves and not with strange hominins. I really see no issue with that, otherwise there would be no H. sapiens in Eurasia - because all would have been absorbed into the archaic populations and multirregionalism would be correct (but Humankind would be very different too).
"The Jeffrey Long's study dated these "two" occasions: 60.000 in the case of neandertals and 45.000 in the case of Denisovans (or any other relative)".
Sorry but I do not know which is this paper.
"How can this be possible, if modern humans left Africa more than 100.000 years ago, and interacted with neanderthals for a long time? What about the modern human jaw found in China and dated more than 100K? Maybe these dates are wrong?"
I was just discussing this matter a few comments earlier, if you read Petraglia 2010, there are two possible dates for the OoA (on archaeological grounds): one before 100 Ka, whose evidence is tenuous and controversial and another one that would be more clear since c. 90 Ka in Arabia and c. 80 Ka in South Asia. The fossil evidence so far for SE Asia (other references) could point to c. 70 Ka.
This is just another case where archaeology is so far limited in providing clear-cut evidence. One possibility could be that there was a first weak migration that got mixed with the archaics and then another one, much stronger, that did not but absorbed some of those first migrants. Just guessing.
"In any case there seems not to be Acheulean in East Asia excepted the Bose Valley brief incursion".
ReplyDeleteOr SE Asia. The Acheulean didn't make it east or north beyond the Movius Line. The point I was making is that the line holds for more separation than just Acheulean/Oldowan though. In fact it seems to separate the Denisova/Melanesian type from Neanderthal/Narmada and the erectus from ergaster. So it has probably been significant through our whole evolution.
"If Denisovans could interbreed with modern humans, does that mean maybe other hominids such as H. floresiensis could as well?"
I very strongly suspect that time will tell us that the differences between the various human 'species' has been greatly exaggerated.
"The low levels of admixture suggest a gradual dilution of the archaic blood, which must have begun with either individuals or populations who were 50/50".
A dilution through more effective incoming culture rather than any inability to form fertile hybrids would be my guess.
greetings maju,
ReplyDeletei try again in english,
I've been thinking about your admixture hypothesis between neanderthals and erectus communities.
I would like to know your opinion on a hypothesis I want to make, because your knowledge of genetics and prehistory in general.
At this point I would like to extrapolate these admixture theories to the emergence of the Neanderthals, perhaps much theorizing starting with my little knowledge about genetics, but at least I can base myself to throw this hypothesis in the cultural behavior of Neanderthals.
Today we discuss the phylogeny of Homo antecessor, and it seems that its discoverers are in the line of the view that it is descended from an Asian population a type homo georgicus and then its storyline seems it disappears, or at least they doubt seriously that their lineage comes to the Neanderthals.
Both homo antecessor and georgicus had developed olduvai technology or modo 1; in the case of Antecessor presents a behavioral pattern associated with cannibalism, and after European archaeological record, particularly in Atapuerca, seems to reflect a cultural break with the arrival of Homo heidelbergensis, which brings the Acheulean technology or mode 2 to Europe, and it looks like some kind of special treatment to the deceased.
Neandertals present cultural patterns common to these two species, and therefore could be the result of a hybridization of both.
Regards.
I just read that idea at your blog precisely. It's an interesting idea (and I think Martinón and Gómez know well what they are talking about, even if, so far, they are less famous than Trinkaus) but I do not think it would change much in regard to the origin of Neanderthals in H. ergaster, via H. heidelbergensis, right?
ReplyDeleteYet the possibility of hybridization of H. ergaster/heidelbergensis (OoA-2) with H. erectus/georgicus/antecessor (OoA-1) is very intriguing. Obviously this would add another element of uncertainty in regards to our own (OoA-3) minor admixture with Neanderthals and "Denisovans".
However, based on mtDNA, I still think that the lineages of Neanderthals belong to OoA-2 (H. ergaster, Acheulean tech) and those of "Denisovans" to OoA-1 (H. georgicus/erectus, Oldowayan tech). So the general outline remains.
Have you seen this?
ReplyDeleteNot exactly that article but, yes, I've been reading variants of that story for the last 24 hrs. or more.
ReplyDeleteWonder why I am not commenting? Because a bunch of ill classified teeth with a nationalist/neo-biblical mediatic hype do not mean much, really.
I trust much more mtDNA (and Y-DNA, and a solid fossil record not going beyond 200 Ka, all in Africa).
"Sorry but I do not know which is this paper. "
ReplyDeletehttp://www.dailymail.co.uk/sciencetech/article-1268003/Neanderthals-humans-interbred-twice-scientists.html
"This is just another case where archaeology is so far limited in providing clear-cut evidence. One possibility could be that there was a first weak migration that got mixed with the archaics and then another one, much stronger, that did not but absorbed some of those first migrants. Just guessing."
Sounds quite well, but how can we detect (genetically) the first migration? Apparently, all humans are descended from one migration, right? This doesn't change things a lot if we're talking about getting along with very different people. Were neandertals viewed equally human by the first immigrants, 100.000 years ago, despite being separated by more than 500.000/1.000.000, and perceived very different by the second ones, 70-80.000 years ago?
That's in the case of this model being right. Which haplogroups were carrying the first immigrants? Apparently not N/M, maybe L? How could we distinguish them?
I've been searching and could find no paper for either Jeffrey Long or Sarah Joyce (apparently the lead researcher) with the Neanderthal tag.
ReplyDeleteWhat I have found is more science-specific publications that mention their research (in Spring this year, before the Green paper) and their intention of publishing something later on. These are an article in Nature and an interview with Long at Science Interviews. I use Eccellio Science to search for specialized materials, for the record.
It is interesting that they propose two admixture events, one early on in the migration OoA and another later on in East Asia, before the colonization of Near Oceania.
Notice that this is different than what the NGP say, the second admixture event would have affected all East Asians and Oceanians. However, while they say that they could not think of other source for this intrusive component than Neanderthals, it is almost impossible to conceive Neanderthal admixture in East Asia, so maybe is the same as found for Melanesians (remember I mentioned that there was also some much weaker, uncertain, appearance of admixture with "Denisovans" in some East Asians... and the French).
Interestingly, unlike the NGP, they did compare "genetic data from 1,983 individuals from 99 populations in Africa, Europe, Asia, Oceania and the Americas", so even if they lacked the raw material of ancient DNA from Neanderthals and "Denisovans", they did make a much more complete analysis of modern humans.
"Sounds quite well, but how can we detect (genetically) the first migration?"
I do not know. I was not stating a falsifiable hypothesis, just speculating on hard-to-explain archaeological data.
"Apparently, all humans are descended from one migration, right?"
Via mtDNA yes, I'd say so. Not so clearly from Y-DNA and there's no easy way to know from autosomal DNA.
"Were neandertals viewed equally human by the first immigrants, 100.000 years ago, despite being separated by more than 500.000/1.000.000, and perceived very different by the second ones, 70-80.000 years ago?"
I cannot say. Such perceptions may have been shaped by other factors such as culture (xenophobia and racism are mostly cultural) and the real numbers involved. If the second migration was better armed (for instance) and more numerous, they may have feel less inclined to collaborate with the "natives", so to say. Or inversely: a less well equipped and less numerous population (as would have been the hypothetical first wave) may have been more inclined to deal with the "natives" in equal terms. But just speculating here anyhow.
"Which haplogroups were carrying the first immigrants? Apparently not N/M, maybe L? How could we distinguish them?"
Totally unsure. We can wonder if Y-DNA CF belongs to the hypothetical first wave but most probably not. Probably no lineages of the first wave have been preserved, if it happened at all and assuming the second wave was much more resolute and expansive.
"What I have found is more science-specific publications that mention their research (in Spring this year, before the Green paper) and their intention of publishing something later on. These are an article in Nature and an interview with Long at Science Interviews. I use Eccellio Science to search for specialized materials, for the record."
ReplyDeleteThanks for the Eccellio, I didn't know it. I use mainly google scholar and google news to find new articles. But you're right, this work hasn't been published yet. Dienekes commented on it (me too) because it appeared on ASHG 2010 abstracts, but nothing more.
"Notice that this is different than what the NGP say, the second admixture event would have affected all East Asians and Oceanians. However, while they say that they could not think of other source for this intrusive component than Neanderthals, it is almost impossible to conceive Neanderthal admixture in East Asia, so maybe is the same as found for Melanesians (remember I mentioned that there was also some much weaker, uncertain, appearance of admixture with "Denisovans" in some East Asians... and the French)."
No, they never said that these admixture events involved neandertals, just archaic hominids living in Eurasia, which of course, include neandertals. I found it very interesting because independent studies found (nearly) the same they predicted.
I can't tell you anything more about this, because I don't know. I hope we'll hear more news in the next year about this. This work is based on microsatellite data (...) and I don't know if they can detect all interbreeding events or just only the most important ones.
"Totally unsure. We can wonder if Y-DNA CF belongs to the hypothetical first wave but most probably not. Probably no lineages of the first wave have been preserved, if it happened at all and assuming the second wave was much more resolute and expansive."
mtDNA/Y-chromosome lineages could be lost, much like what happened with the neanderthal ones, but in autosomal DNA, if your model is correct, we'd expect to find at least 2,5% of autosomal DNA from the first wave, but if human DNA is so similar it might be very difficult to detect.
"What I have found is more science-specific publications that mention their research (in Spring this year, before the Green paper) and their intention of publishing something later on. These are an article in Nature and an interview with Long at Science Interviews. I use Eccellio Science to search for specialized materials, for the record."
ReplyDeleteThanks for the Eccellio, I didn't know it. I use mainly google scholar and google news to find new articles. But you're right, this work hasn't been published yet. Dienekes commented on it (me too) because it appeared on ASHG 2010 abstracts, but nothing more.
"Notice that this is different than what the NGP say, the second admixture event would have affected all East Asians and Oceanians. However, while they say that they could not think of other source for this intrusive component than Neanderthals, it is almost impossible to conceive Neanderthal admixture in East Asia, so maybe is the same as found for Melanesians (remember I mentioned that there was also some much weaker, uncertain, appearance of admixture with "Denisovans" in some East Asians... and the French)."
No, they never said that these admixture events involved neandertals, just archaic hominids living in Eurasia, which of course, include neandertals. I found it very interesting because independent studies found (nearly) the same they predicted.
I can't tell you anything more about this, because I don't know. I hope we'll hear more news in the next year about this. This work is based on microsatellite data (...) and I don't know if they can detect all interbreeding events or just only the most important ones.
"Totally unsure. We can wonder if Y-DNA CF belongs to the hypothetical first wave but most probably not. Probably no lineages of the first wave have been preserved, if it happened at all and assuming the second wave was much more resolute and expansive."
mtDNA/Y-chromosome lineages could be lost, much like what happened with the neanderthal ones, but in autosomal DNA, if your model is correct, we'd expect to find at least 2,5% of autosomal DNA from the first wave, but if human DNA is so similar it might be very difficult to detect.
"I use Eccellio Science to search for specialized materials, for the record."
ReplyDeleteThanks for the Eccellio, I didn't know it. I use mainly google scholar and google news to find new articles. But you're right, this work hasn't been published yet. Dienekes commented on it (me too) because it appeared on ASHG 2010 abstracts, but nothing more.
"Notice that this is different than what the NGP say, the second admixture event would have affected all East Asians and Oceanians. However, while they say that they could not think of other source for this intrusive component than Neanderthals, it is almost impossible to conceive Neanderthal admixture in East Asia, so maybe is the same as found for Melanesians"
No, they never said that these admixture events involved neandertals, just archaic hominids living in Eurasia, which of course, include neandertals. I found it very interesting because independent studies found (nearly) the same they predicted.
I can't tell you anything more about this, because I don't know. I hope we'll hear more news in the next year about this. This work is based on microsatellite data (...) and I don't know if they can detect all interbreeding events or just only the most important ones.
"Totally unsure. We can wonder if Y-DNA CF belongs to the hypothetical first wave but most probably not. Probably no lineages of the first wave have been preserved, if it happened at all and assuming the second wave was much more resolute and expansive."
mtDNA/Y-chromosome lineages could be lost, much like what happened with the neanderthal ones, but in autosomal DNA, if your model is correct, we'd expect to find at least 2,5% of autosomal DNA from the first wave, but if human DNA is so similar it might be very difficult to detect.
In your link, Sarah Joyce is mentioned as follows:
ReplyDelete"She found that the best explanation was if there were two periods of interbreeding between modern humans and another human species - and that Neanderthals were the only likely candidate.
One period of interbreeding occurred around 60,000 years ago in the eastern Mediterranean, the other around 45,000 years ago in eastern Asia.
However in the interview Long gives a different version, acknowledging that there are other candidates.
"... in autosomal DNA, if your model is correct, we'd expect to find at least 2,5% of autosomal DNA from the first wave, but if human DNA is so similar it might be very difficult to detect".
Exactly. It would be almost impossible, if not plainly impossible, to detect the genetic "cargo" of two waves spawning from the same population with "so little" difference of time (just a few dozen thousand years".
That is unless... there was a preserved lineage. And while it seems that Y-DNA and mtDNA do not preserve anything from that hypothetical event, I recall research in X-DNA (from 2006 or 2007) which detected a haplotype (B006 if I recall correctly, most common among Basques and Native Americans, but most diverse around Mongolia) that was only seldom found in Africa and looked older than the main Eurasian branch.
It could be also a borrowing from archaic Eurasian humans or it could be just a randomly "amplified" African lineage (Africa was poorly sampled and South Asia was not really surveyed, as often happens). But it's something intriguing that was spotted in two consecutive researches. Then the matter was not further researched, sadly.
"She found that the best explanation was if there were two periods of interbreeding between modern humans and another human species - and that Neanderthals were the only likely candidate."
ReplyDeleteAs you said, Long gives a different explanation. It was completely impossible to discover if the archaics with whom modern humans admixed were neandertals or not, because when they published? the study, the neanderthal genome wasn't sequenced.
"That is unless... there was a preserved lineage. And while it seems that Y-DNA and mtDNA do not preserve anything from that hypothetical event, I recall research in X-DNA (from 2006 or 2007) which detected a haplotype (B006 if I recall correctly, most common among Basques and Native Americans, but most diverse around Mongolia) that was only seldom found in Africa and looked older than the main Eurasian branch.
How interesting! Many studies have analyzed modern human DNA and have found indeed signals for admixture with archaic humans (like the one from Wall & Plagnol, in 2006). It's a very interesting issue, I hope more studies will do more research.
Another interesting haplotype was TAU H2, very european-specific, but apparently it has nothing to do with neandertals.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2820164/
"So they must have admixed with some relative of Denisovans elsewhere, for example in Sundaland, where some Homo erectus are known to have lived in dates that are perfectly compatible with this scenario".
ReplyDeleteYour theory on how Denisovan aDNA managed to appear in New Guinea is pure speculation. There's no evidence at all that SE Asian H. soloensis possessed those genes.
Onthe other hand a migration of mtDNA N from somewhere near the region of Denisova across Asia towards Japan and then down the east Eurasian coast to New Guinea/Australia fits perfectly. Of course under that scenarion the Australian Aborigines would have more Denisovan aDNA than would Papuans but, as far as I know, no Aborigines were tested.
"So the ancestors of Melanesians and other very dark tropical Asians have definitively not lived in Siberia at any time".
They could easily have done so. For a start it only involved a small percentage of their ancestry. And Y-hap R, for example, has its deep origin in SE Asia yet most men who have it do not look at all like SE Asians, past or present.
"Your theory on how Denisovan aDNA managed to appear in New Guinea is pure speculation. There's no evidence at all that SE Asian H. soloensis possessed those genes".
ReplyDeleteBy elimination: Denisovans are not by mtDNA Sapiens nor Neanderthal (hence not even Ergaster). So logically it should be Erectus, what else? Homo floresiensis? Australopithecus robustus?
So if the Denisovans are at least partly H. erectus, as their mtDNA quite clearly indicates to anyone familiar with the relevant archaeology and palaeoanthropology, the we can conclude easily that "their" autosomal DNA in Melanesians represents autosomal DNA from H. erectus, maybe from both H. erectus and Neanderthal (overlapping with the general Neanderthal admixture).
That's my theory and I am making a falsifiable prediction in it: go extract and sequence Erectus DNA and prove me wrong (or right). That's how science works.
"Of course under that scenarion the Australian Aborigines would have more Denisovan aDNA than would Papuans"... and Europeans too, which are 99% mtDNA N. There's no population in the World more N than West Eurasians, not even Australian Aborigines.
Get real.
"For a start it only involved a small percentage of their ancestry".
So are you telling me that some random humans went to Altai, admixed with Denisovans (but not anymore with the Neanderthals who were also there), left no archaeological record we can see anywhere, left no genetic record either in modern populations in the area (or anywhere else) and then went to New Guinea and got extremely diluted among Melanesians, who, I imagine were already there waiting...
I think it's easier to get all billiard balls in their holes in the first shoot, so convoluted your claim is.
I tell you: go to Java or whatever museum Solo man is, extract the DNA and prove me wrong. Guess that any other Asian Erectus would do (Pekin man, Bodo, Dali...)
"So logically it should be Erectus, what else?"
ReplyDeleteHow about Denisovans? Were they definitely erectus? And even if they are were can you assume they were genetically the same as the population that existed in SE Asia?
"go extract and sequence Erectus DNA and prove me wrong (or right). That's how science works".
Yes. And I'm betting that SE Asia H. erectus is different genetically from the Denisovans. Prove me wrong.
"and Europeans too, which are 99% mtDNA N".
You know as well as I do the European N is almost completely R. Are you really claiming that a single small population can migrate a massive distance and maintain its genetic purity? Get real. Isn't that taking that 'exclusive mania' I mentioned you suffer from a little far? Talk about racist.
"So are you telling me that some random humans went to Altai, admixed with Denisovans"
The evidence points strongly in that direction.
"(but not anymore with the Neanderthals who were also there)"
I'm sure you were claiming that they show Neanderthal admixture too.
"left no archaeological record we can see anywhere"
There's plenty of evidence for human presence across Central Asia. It's just that, once more as a product of your 'exclusive mania', you claim the fact they didn't possess an 'Upper Paleolithic' culture proves they're not 'Modern Humans'.
"left no genetic record either in modern populations in the area"
Basal haplogroups of both mtDNA N and Y-hap C are widespread across the region.
"then went to New Guinea and got extremely diluted among Melanesians"
Were already diluted obviously, and became more so as time passed, especially once members of mtDNA M and Y-hap KMNOPS entered the region.
"Guess that any other Asian Erectus would do (Pekin man, Bodo, Dali...)"
I would guess that H. erectus varied regionally as much as, or perhaps more than, we do today. So we'd have to sample a wide regional variety of them to trace exactly what genes entered the modern human genome, and where from.
"How about Denisovans? Were they definitely erectus?"
ReplyDeleteThat's what I'm saying: half Erectus half Neanderthal. Their mtDNA is Erectus and their autosomal DNA is half Neanderthal.
"And I'm betting that SE Asia H. erectus is different genetically from the Denisovans".
Everybody is different genetically to some extent. Said that way it's trivial.
What I'm saying is that there was a population spanning from Indonesia to North (and NE) China that we call "Asian H. erectus" and which is descended from the migrations c. 1.8 Ma or maybe a little later with Olduwayan technologies. This species is attested archaeologically in many sites (key supporting evidence already existant) and diverged from us and Neanderthals maybe 1.8 times or 1.5 earlier than proto-Neanderthals did (arguably).
Compared to H. ergaster (Neanderthals and us) they were a clearly distinct population (species, superspecies, whatever) with distinct genetics branching out earlier, confirmed in Denisovan mtDNA.
Can you understand all this?
"Prove me wrong".
I was first and I do have a consistent theory: you have nothing. In any case this will be settled when H. erectus DNA is researched and I bet Chinese geneticists are already considering and even pulling threads to do such a research, before Westerners do.
Let's wait a few years and we'll know for sure.
"European N is almost completely R".
R is N. And there's also several N(xR) clades.
"Are you really claiming that a single small population can migrate a massive distance and maintain its genetic purity?"
If they come from Altai as you claim, Europe and Arabia are closer than Australia.
"Isn't that taking that 'exclusive mania' I mentioned you suffer from a little far? Talk about racist".
I do not understand this and if you're going to make such grave accusations (racism) you better be more clear about what the heck you're talking about.
...
"There's plenty of evidence for human presence across Central Asia".
ReplyDeleteHuman in what sense and in what chronological frame. Human can mean Homo sp. in general or Homo sapiens. The chronology seems to support a layered presence of the different species, whit modern humans only present after the others (true Neanderthals and Neanderthal-Erectus hybrids of Denisova and Okladnikov caves) were gone. In the case of Denisova there seems to be a gap of more than 10 Ka.
"Basal haplogroups of both mtDNA N and Y-hap C are widespread across the region".
Really? I do not think so. Maybe there is C5 (not a Melanesian clade) and maybe some mtDNA A or X, derived from East Asia and West Asia respectively. You are trying to force-bend the evidence to your preconceptions (again). It gets boring, honestly and the bent evidence is already broken of so much bending.
Besides Papuans are not even strongly mtDNA N(xR) nor Y-DNA C, so what the heck are you ranting about? Papuans are essentially mtDNA R and M and Y-DNA MNOPS. They are very much like Europeans in all this, but with some "other" stuff.
I'd make more sense of the theories of some Iranian-Swedish guy argued that West Eurasians are some sort of depigmented Papuans. I'm not subscribing to that but he made better sense than you do anyhow.
"I would guess that H. erectus varied regionally as much as, or perhaps more than, we do today".
Of course. But we have no more DNA evidence of the whole species than these sequences from Denisova.
"So we'd have to sample a wide regional variety of them to trace exactly what genes entered the modern human genome, and where from".
Not really because, as I said in my little essay, you can find admixture by proxy. We know of Neanderthal admixture in Asia using European Neanderthal sequences for example. With all likelihood, neither the Vindija specimens nor any other European Neanderthal directly took part in the admixture process - but we can still detect admixture, with great precision, comparing with them.
"Of course. But we have no more DNA evidence of the whole species than these sequences from Denisova".
ReplyDeleteSpecifically Denisova, note. Not from SE Asian H. erectus. And it's very likely that SE Asian H. erectus shared far fewer aDNA with Denisovans than with Neanderthals.
"Not really because, as I said in my little essay, you can find admixture by proxy".
And that essay is full of 'possiblys' and 'maybes', and is unlikely to account for the evidence in this case.
"Besides Papuans are not even strongly mtDNA N(xR) nor Y-DNA C, so what the heck are you ranting about? Papuans are essentially mtDNA R and M and Y-DNA MNOPS".
Exactly. Hence my comment about Aborigines probaly having more of the Denisova aDNA, and my comment, 'Were already diluted obviously, and became more so as time passed, especially once members of mtDNA M and Y-hap KMNOPS entered the region'.
"I'd make more sense of the theories of some Iranian-Swedish guy argued that West Eurasians are some sort of depigmented Papuans. I'm not subscribing to that but he made better sense than you do anyhow".
That seems to be exactly the parallel you're drawing when you deny any possibility of Papuans having passed through Denisova territory yet conceding European haplogroups come from SE Asia.
"Human in what sense and in what chronological frame. Human can mean Homo sp. in general or Homo sapiens".
Doesn't the present evidence support the concept that these different human 'species' interbred? In which case it's irrelevant which Homo sp. lived across Central Asia.
"Really? I do not think so. Maybe there is C5 (not a Melanesian clade)"
Why do you ignore C3 and C1?
"maybe some mtDNA A or X, derived from East Asia and West Asia respectively".
So you're claiming different rules for haplogroup M and haplogroup N. The fact that M haplogroups are spread in a continuity across southern Asia is proof they expanded together but the fact that N haplogroups are spread in a continuity across central Asia is proof they entered from two separate regions.
"Everybody is different genetically to some extent. Said that way it's trivial".
ReplyDeleteTry not to be stupid Maju. You are very well aware that humans vary regionally as well as individually.
"What I'm saying is that there was a population spanning from Indonesia to North (and NE) China that we call 'Asian H. erectus' and which is descended from the migrations c. 1.8 Ma or maybe a little later with Olduwayan technologies".
Agreed, although the Chinese and SE Asians eventually diversified to quite an extent.
"and diverged from us and Neanderthals maybe 1.8 times or 1.5 earlier than proto-Neanderthals did (arguably)".
And certainly from the Denisovans.
"I was first and I do have a consistent theory"
But it doesn't make sense. You're postulating a common gene pool all the way from the Altai into China and down to SE Asia. Extremely unlikely.
"R is N. And there's also several N(xR) clades".
Yes, but N(xR) clades are a very small minority and R definitely emerged from SE Asia. R's expansion is very likely to have been independent of basal N's expansion despite your beliefs to the contrary.
"If they come from Altai as you claim, Europe and Arabia are closer than Australia".
It's you who is claiming the maintenance of 'genetic purity' over huge migrations. Part of your 'exclusivity mania'?
"I do not understand this and if you're going to make such grave accusations (racism) you better be more clear about what the heck you're talking about".
You see absolute separation between species and cultures etc., denying any mixing whatsoever. You see thing as absolutely black and white. As I said, we cannot view the world in any way other than what we are. And what we are is a product of our upbringing, especially our very early expariences.
The mtDNA of Denisovans, which is consistent across individuals, is pre-Ergaster, hence Denisovans are not (most probably) an unknown species akin to Neanderthals, as the nuclear DNA could suggest, but a hybrid of a (known) species of Homo which diverged earlier. This one can only be H. erectus, which (we know from the fossil record) lived in East Asia at least until 200 Ka ago and in Indonesia at least until 50 Ka ago.
ReplyDeleteMost probably is not "possibly" or "maybe".
"my comment about Aborigines probaly having more of the Denisova aDNA"...
This is a "possibly" which we do not know.
"Doesn't the present evidence support the concept that these different human 'species' interbred? In which case it's irrelevant which Homo sp. lived across Central Asia".
It is most relevant. It is not the same Homo erectus than Homo neanderthalensis or Homo sapiens. They are clearly distinct populations, regardless if you wish to call them subspecies or keep the species concept. At the current mainstream standards occasional inter-fertility does not make two species one - but that is not the issue anyhow.
"Why do you ignore C3 and C1?"
Because they are from NE Asia, towards the coast. Whatever there is in Central Asia arrived from further East.
I'll slightly correct your tonight's first post's last sentence (corrections in bold type) so you get what I claim in fact (and that you do not understand it yet makes me doubt about your intelligence or honesty):
"The fact that M basal haplogroups are spread in a continuity across southern Asia is proof they expanded together but the fact that N highly derived haplogroups are spread in a continuity across central Asia is proof they entered from two separate regions".
Basal sublineages are not the same as highly derived sublineages with clear origins in regions that are not Central Asia. You do not want to make a fool of yourself trying to prove that X or A coalesced in Altai, do you? Because I challenge you to try it.
"And certainly from the Denisovans".
ReplyDeleteYou have no fucking idea what the Denisovans are (were). Why do you claim a totally different and new species when there is absolutely no evidence for such thing?
"You're postulating a common gene pool all the way from the Altai into China and down to SE Asia. Extremely unlikely".
I'm postulating that, yes. Because they were descendants from the first OoA, that of H. erectus with chopper industries. There is a quite large consensus about this.
This gene pool, originating maybe 800 or 1000 Ka before the proto-Neanderthal migration with Acheulean, must have created (and retained in Eastern Asia) a gene pool very distinct from that of the populations derived from African H. ergaster (i.e. Neanderthals and us).
That is what the Denisova mtDNA confirms.
"Yes, but N(xR) clades are a very small minority"...
Also among Papuans, AFAIK.
"R's expansion is very likely to have been independent of basal N's expansion despite your beliefs to the contrary".
Since you demonstrated that R must have coalesced in SEA (novel data from this year), it is clear that R must have migrated with some minor N westward. More clear than before.
"It's you who is claiming the maintenance of 'genetic purity' over huge migrations".
WTF?!
Which genetic purity and which huge migrations? Generalizations are pointless rhetorical acrobatics, we need to be specific. This "huge migration" may be totally different from that other "huge migration" and this genetic pool may be purer than that other genetic pool. Let's be specific.
Spanish saying for someone who speaks clearly: he/she "calls the bread bread and the wine wine". You don't: you say "bread" when you mean wine just to keep the discussion going and going in circles. I understand that is abusive, insincere, mischievous, malevolent. That's not the kind of non-debate I want to have.
Alternatively it can be idiotic but I do not want to think you are a fool. So please do not try to fool me with silly rhetoric exercises, science is not the same as politics.
"You see absolute separation between species and cultures etc., denying any mixing whatsoever".
Not sure. Which cases? Farmers and foragers maybe? Erectus and Neanderthals maybe? I'm willing to go through each case but I cannot argue on generalities.
"You see thing as absolutely black and white".
No.
"... we cannot view the world in any way other than what we are".
That is your belief. Instead I think we can see the world with many eyes, it just requires some open-mindedness (some drugs may help, in some cases, I guess to open the mind). Of course, I will never have the life experiences of a Hadza, for example, but I can try to understand their point of view.
And as we all are essentially foragers in modern clothing, this is easier than seeing the point of view of a farmer, even if I have some more experience with gardening and working the land than with hunting. At least for me: one of the reasons I am interested in anthropology is because I do feel particularly identified with huntergatherer cultures (it's just too easy to understand their way of life, easier than modern life often).
"And what we are is a product of our upbringing, especially our very early expariences".
Yes but not so much (here for instance you see in B&W and I am seeing in technicolor and even 3D resolution with odorama). We make choices and we can sabotage our learning. I do not know how much effort have you put on that but you can believe me when I say that I have. And for a reason: I want to be free, not a slave of the chains of my parents. And to a large extent I think I have achieved it, keeping the best and throwing away the worst, the useless, the parasitic policeman inside.
"That is what the Denisova mtDNA confirms".
ReplyDeleteIt confirms it for the Denisovans but says nothing about SE Asian H. erectus.
"It is most relevant. It is not the same Homo erectus than Homo neanderthalensis or Homo sapiens. They are clearly distinct populations"
As, presumably, were the SE Asian H. erectus. Are you denying genetic drift in the H. erectus population? That's a stupid stance.
"You have no fucking idea what the Denisovans are (were). Why do you claim a totally different and new species when there is absolutely no evidence for such thing?"
And you're claiming complete genetic continuity from Denisovans to SE Asian H. erectus over a period of close to a million years. Even modern humans have diversified regionally over just 100,000 years or so. Obviously you have no fucking idea of genetic drift.
"Which genetic purity and which huge migrations?"
You appear to be claiming that H. erectus changed not one little bit after it left Africa. That's 'genetic purity' surely.
"Basal sublineages are not the same as highly derived sublineages"
N lineages are as basal as the M ones. The difference is that M diversifies after three mutations and N diversifies after five. So why do you claim one set as 'basal' and the other as 'highly derived'?
"Since you demonstrated that R must have coalesced in SEA (novel data from this year), it is clear that R must have migrated with some minor N westward. More clear than before".
Why with any N, let alone 'minor? And I'd remind you adopted the same arrogant refusla to consider the evidence whenever I suggested mtDNA R had expanded from SE Asia. And whenever I suggested the ancestor of Y-hap R had done so too. How about pulling your arrogant head in occasionally.
I am not talking specifically of the H. erectus of SE Asia but of all the archaic humankind that left Africa c. 1,8 Ma ago. That's what the Denisova mtDNA is.
ReplyDeleteIf the non-Neanderthal part of Denisovans and the other Asian H. erectus derive from that migration, as seems to be correct (and if you think otherwise, please explain why), their DNA overall must be related, just like ours and Neanderthals' is.
"You appear to be claiming that H. erectus changed not one little bit after it left Africa".
Not at all. But the genetic pool from which it drew remained the same one, more or less. If there were other instances of Heidlebergensis or Neanderthal admixture other than the ones obvious in Altai, they should be demonstrated. My general impression from the fossil record is that the second OoA wave (to which Heidelberg/Neanderthal belong) never made many inroads East of Bengal. So the East Asian genetic pool remained essentially that of the first wave.
"N lineages are as basal as the M ones".
Are you pretending to be dumb? X2 is not a basal lineage of N. It is a basal lineage of X, which in turn is a basal lineage of N. We know anyhow that both X and X2 coalesced in West Asia, not in Altai. The only thing that the presence of X2-derived clades in Northern Asia and America show is some sort of migration from West Asia (which is confirmed by Y-DNA Q, for example).
Same for A-derived lineages but in regards to East Asia. Altai is a place of encounter between West and East Eurasia, a "silk road" of the Paleolithic.
"Why with any N, let alone 'minor?"
Because if both N and R (separated by a single control region mutation) coalesced in East Asia and are found in West Asia, they must have migrated westwards roughly in the same dates. Besides most of these N basal sublineages show presence in South Asia (X being the only exception), what confirms their southern route also in Westward direction. In this sense also, all the archaeology leading to West Eurasian Upper Paleolithic (mode 4), directly related to the colonization by our species, is in South Asia (earliest use of stone blades), not in East Asia, where such technology only arrives later.
"... you adopted the same arrogant refusla to consider the evidence whenever I suggested mtDNA R had expanded from SE Asia".
You did not provide any evidence. You just made a claim without any support. If I recall correctly, I had to look for the evidence myself. If you had bothered providing evidence (new data bout Eastern R basal sublineages), I would have accepted it. Even if I am politely attributing you some help in finding out this, actually you were not too helpful, because you did not provide clear evidence at any moment and I had to find it on my own. You were making such claims even before any evidence was available at all, so it's more of a case of the donkey who played the flute by chance.
Instead of arguing and arguing, you should bother taking your time to ponder the evidence, sharing it if you find something but specially pondering it dispassionately, without preconceptions, open to all options.
"Even if I am politely attributing you some help in finding out this, actually you were not too helpful, because you did not provide clear evidence at any moment"
ReplyDeleteI kept providing explanations and evidence as to why I believed it to be so. However you were blinded by your absolutely unquestioning acceptance that all Eurasian haplogroups coalesced in India.
"you should bother taking your time to ponder the evidence"
And that's exactly how I came to the conclusion that both mtDNA R and the ancestor of Y-hap R (which at the time was called 'K' but is now MNOPS) had originated somewhere in SE Asia. I was able to do this because I looked at the evidence (the distribution of haplogroups) and then looked for an explanation. I didn't work in reverse, as you do: adopt a theory and then look at the evidence from the perspective of supporting that theory.
"pondering it dispassionately, without preconceptions, open to all options".
That's exactly what I did while you seemed unable to do so.
"You were making such claims even before any evidence was available at all, so it's more of a case of the donkey who played the flute by chance".
There was any amount of evidence. It's just that you were blinded by your theory.
"I am not talking specifically of the H. erectus of SE Asia but of all the archaic humankind that left Africa c. 1,8 Ma ago. That's what the Denisova mtDNA is".
And there's another thing you were arrogantly dismissive of. You scoffed whenever I claimed that 'modern humans' were able to breed with any earlier ones. I'm pleased to see you now accept my position.
"their DNA overall must be related, just like ours and Neanderthals' is".
And the Denisove DNA is almost certainly much more closely related to Neanderthal than it is to any SE Asian conemporary population. Because:
"If there were other instances of Heidlebergensis or Neanderthal admixture other than the ones obvious in Altai, they should be demonstrated".
Such movements would have served to increase the genetic distance between SE Asian populations and all the other Eurasian populations, including the Denisovans.
"Because if both N and R (separated by a single control region mutation) coalesced in East Asia and are found in West Asia"
That's a huge IF as far as N is concerned. In fact you make that assumption to enable you to make the evidence fit your theory.
"X2 is not a basal lineage of N. It is a basal lineage of X, which in turn is a basal lineage of N. We know anyhow that both X and X2 coalesced in West Asia"
And neither X nor its ancestors were ever anywhere near SE Asia.
"Besides most of these N basal sublineages show presence in South Asia"
None of the 'basal' lineages actually appear in India, only downstream ones. That makes it difficult to argue a route through India for any N haplogroups other than R.
"all the archaeology leading to West Eurasian Upper Paleolithic (mode 4), directly related to the colonization by our species, is in South Asia (earliest use of stone blades)"
I agree 100%. If you're going to confine your perspective to Europe you will inevitably assume an India origin, which is exactly the mistake you made originally.
"not in East Asia, where such technology only arrives later".
Exactly. Long after mtDNA R and Y-hap MNOPS had formed there, and even after they had left. As I keep trying to get through to you, the 'Upper Paleolithic' is not the defining characteristic of the first 'modern' humans. Presumably it was develeoped by them, and possibly in India, but they didn't emerge from Africa with a fully developed Upper Paleolithic.
Terry: we can keep discussing in circles, as long as you do not present evidence and keep throwing empty accusations.
ReplyDeleteFor example, you say:
"That's a huge IF as far as N is concerned. In fact you make that assumption to enable you to make the evidence fit your theory".
No evidence and an accusation of partiality, bias and prejudice.
Yet, I have made as clear as I can why I do "believe", think that N coalesced in SE Asia. As with every other clade, I have mapped the basal sublineages and estimated a center of gravity. This center of gravity falls in SE Asia. 6 out of 12 (50%) basal sublineages are exclusive of SE Eurasia (SEA+Near Oceania), the rest are R (agreed to coalesced in SEA), two East Asian clades and three West/South Asian ones.
The conclusion is self-evident for anyone who looks at it dispassionately. Not you, of course.
Yet you accuse me of bias and prejudice but provide not a single piece of evidence against this model.
And it's almost always the same. So it's only logical that I'm annoyed.
"None of the 'basal' lineages actually appear in India, only downstream ones. That makes it difficult to argue a route through India for any N haplogroups other than R".
N2a, W and N5 appear in South Asia, as well as a host of R sublineages. Still you might have a point that the SA route westward is not clearly proven but that says nothing against the origin of N in SEA. If anything it may suggest a possible alternative route (equally unproven or rather less logical by all measures) through Siberia/Central Asia, but in westward direction if anything.
"If you're going to confine your perspective to Europe you will inevitably assume an India origin"...
First, I do not restrict myself to Europe, though the archaeological record over here is pretty good and therefore an unavoidable reference.
But most importantly I do not consider Europe as distinct from West Asia and other parts of Western Eurasia. It is all West Eurasia which displays a quite clear South Asian origin and connection. Well, actually all Eurasia and annex areas do but this is even more clear and recent in the case of West Eurasia.
"... the 'Upper Paleolithic' is not the defining characteristic of the first 'modern' humans".
I know that perfectly well, please. In West Eurasia however there is a quite lineal connection.
"N2a, W and N5 appear in South Asia, as well as a host of R sublineages".
ReplyDeleteWe've agreed that R originated in SE Asia, so let's look at the other two (it's really N2a/W).
N5 is no more basal than is N1, found in Central Eurasia, so we have no way of knowing whether N1 moved out of India or N5 moved into it from the west. In fact it's more likely to be the latter because N5 is far less diverse than is N1. It is also difficult to mount a convincing argument for N2/W having either moved out of India or into it from the west, although W's highest concentration is evidently in Northern Pakistan, sort of borderland between SW Asia and South Asia.
"Yet, I have made as clear as I can why I do 'believe', think that N coalesced in SE Asia. As with every other clade, I have mapped the basal sublineages and estimated a center of gravity. This center of gravity falls in SE Asia. 6 out of 12 (50%) basal sublineages are exclusive of SE Eurasia (SEA+Near Oceania)"
OK. Let's accept that N moved from Africa to SE Asia leaving no descendants along the way. That leaves open the option of considering mtDNA M in a similar manner.
M spread from India, correct? But India is a big place. But when we look at the haplogroup we see that something like half of all the basal M haplogroups are found in the region stretching from Bangla Desh, the Khasi Hills, Assam and into Yunan. Many other Indian M haplogroups are found down the east coast, in Anhra Pradesh. I think we've agreed that D and M8/CZ moved from somewhere round the North Burma/Yunan region. So M probably spread from that region. This would certainly explain easily the basal Ms in the Andamans.
So that leaves M too spreading from SE Asia, slightly to the west of where N spread from: the region between the malay Peninsular, Vietnam and Kwangsi. But it says nothing about either haplogroups' route there.
Another thing: Your idea of an nancient connection between the Denisovans and SE Asian H. erectus leads inexorably to consideration of the Movius Line. I'm not sure if you're familiar with the concept:
ReplyDeletehttp://en.wikipedia.org/wiki/Movius_Line
"Movius had noticed that assemblages of palaeolithic stone tools from sites east of northern India never contained handaxes and tended to be characterised by less formal implements known as chopping tools. These were sometimes as extensively worked as the Acheulean tools from further west but could not be described as true handaxes. Movius then drew a line on a map of India to show where the difference occurred, dividing the tools of Africa, Europe and Western and Southern Asia from those of Eastern and Southeastern Asia".
You can see that the line follows the Himalayas and the SE Asian mountains, or the Ganges/Brahmaputra delta depending on where you care to draw the line at the eatern end.
"Fossil evidence also suggests a difference in the evolutionary development of the people who made the two different tool types across the Movius Line and it has remained in use as a convenient distinction between the two traditions".
Suggests somewhat the distinction between H. erectus and H. eragster. But you can see how the line does suggest a connection between the region of the Denisova fossils and SE Asia, in contrast to Africa, South Asia and Europe. However the line speaks against the connection being via India, or any 'southern route'.
A reasonable recent analysis of the Movius Line:
http://replicatedtypo.wordpress.com/2010/05/29/the-movius-line-represents-the-crossing-of-a-demographic-threshold/
"N5 is no more basal than is N1, found in Central Eurasia, so we have no way of knowing whether N1 moved out of India or N5 moved into it from the west".
ReplyDeleteTechnically this is true.
"In fact it's more likely to be the latter because N5 is far less diverse than is N1".
This is wrong instead: it is trivial which sublineage is more diverse. Obviously N1 had much better chances of expansion in the West than N5 in already populated South Asia.
"It is also difficult to mount a convincing argument for N2/W having either moved out of India or into it from the west"...
Agreed.
"... although W's highest concentration is evidently in Northern Pakistan, sort of borderland between SW Asia and South Asia".
It is a hint in favor of my thesis but we'd need to look at the structure and local diversity of W.
My point is however that:
1. There's nothing of this in Central Asia/Siberia. Weighting a lot against this conjectural route.
2. Besides the presence of these two haplogroups in South Asia, we also have clear evidence that their more clear "sister" R has greater basal diversity in the subcontinent. This is oblique support, but still some support, for a route via South Asia.
So I have to choose between a southern route with some support and a northern route with no support at all. So I cannot but choose the southern route.
"OK. Let's accept that N moved from Africa to SE Asia leaving no descendants along the way".
I can only wish this acceptance is stable.
"That leaves open the option of considering mtDNA M in a similar manner".
Absolutely. The difference is that in this case we are before greatest basal diversity in South Asia (by quite a bit).
"M spread from India, correct? But India is a big place".
Yes but most M sublineages are widespread through the subcontinent (or vast stretches of it) and often look like most diverse towards the NW if anything.
I have tried to discern within India but the information available is limited and the subcontinent looks largely unstructured. With enough data, we should be able to discern something but I have not been able to gather such resolution so far.
"I think we've agreed that D and M8/CZ moved from somewhere round the North Burma/Yunan region".
I do not have this too clear but maybe it is just my ignorance of the fine details. I do suspect that the Burma-Yunnan corridor played a role in the early colonization but other routes such as the coast and the other major rivers or the area cannot be discarded and were probably complementary. With enough resolution of data we should be able to best-guess the most likely routes but at the moment I cannot take a strong stand.
"So M probably spread from that region".
From Burma? I do not think so. On of my first guesses was the Bengal area (in order to accommodate the many Eastern lineages) but, as I say, most SA M sublineages appear to have a NW origin, say the Sindh-Gujarat-Mumbay arch, or nearby zones like the Narmada basin, the Punjabs or the West Indian coast.
I guess that a correction towards the ancestor's origin (after finding out the geometrical/geographical centroid of the basal sublineages) is appropriate. In the cases of M and N it'd be towards L3. So my best hunch for N is Burma and my best hunch for M is NW South Asia, near the Arabian Sea coasts probably.
But I reckon this is arguable and we'd also need much more clear samples to be able to discern with clarity, South Asia for M and SEA for N are good enough for me as of now.
"... consideration of the Movius Line"...
ReplyDeleteI was not familiar with the idea until recently but it makes some good sense. Just that Central Asia (Altai specially) must fall to the west of the line (unlike what the Wikipedia map suggests).
Notice that the concept is from 1948 and since then there have been some major advances in the knowledge of the area. I would also say that the line keeps some meaning after the Eurasian expansion of H. sapiens, with blade-based industries to the West and flake-based industries to the East (at least for some time). But again Altai and Central Asia fall in the Western half.
"I would also say that the line keeps some meaning after the Eurasian expansion of H. sapiens"
ReplyDeleteexactly the reason I brought it to your attention.
"But again Altai and Central Asia fall in the Western half".
Not so. The Movius line is south of that region, thus connecting the Altai/Central Asian region with SE Asia.
"we also have clear evidence that their more clear 'sister' R has greater basal diversity in the subcontinent. This is oblique support, but still some support, for a route via South Asia".
Support for R having migrated through India but it says nothing about N.
"Yes but most M sublineages are widespread through the subcontinent (or vast stretches of it) and often look like most diverse towards the NW if anything".
Not 'most of them' at all.
"I have tried to discern within India but the information available is limited and the subcontinent looks largely unstructured. With enough data, we should be able to discern something but I have not been able to gather such resolution so far".
I've spent most of the morning hunting down as much information as possible. I'll make another post.
"most SA M sublineages appear to have a NW origin, say the Sindh-Gujarat-Mumbay arch, or nearby zones like the Narmada basin, the Punjabs or the West Indian coast".
ReplyDeleteI've had no success at locating M25, M50, M60 and M73 (although I suspect the last is SE Asian) but if we accept the region between Bangla Desh'/Yunan as a separate region ('Borderland'?) to the rest of India here's what we get:
Oz/New Guinea, M14, M15, M27, M28, M29'Q and M42'74, (6 haps).
East Asia, D, M7, M8/CZ, M10 and M12'G, (5 haps).
Southeast Asia, M9/E, M17, M21, M22, M23'77 and M72, (6 haps).
Borderland, M11, M13'46'61, M21, M31, M32'56, M44, M47 and M49, (9 haps).
Andhra Pradesh, M35, M36, M39, M40'62 and M41, (5 haps).
India, M2, M3, M4''64, M5, M6, M33, M34'57, M53 and M52'58, (9 haps, the same as the borderland).
But if we examine the Indian haplogroups further we find (according mostly to Wiki) that M2's highest concentration is in SE India and Bangla Desh, M5's is South India and Orissa, M34'57's is South India, and both M4''64 and M53 are Central Indian. M52'58's earliest branches are in Northeast India. That leaves just M3 in West and Northwest India, M6 in kashmir and kerala and M33 in Northwest India. Just three of the more than forty M haplogroups.
Terryt : "I've had no success at locating M25, M50, M60 and M73"
ReplyDeleteThere's also the Tibetan M16 * (not to be confounded with the M found in Madagascar that was first named M16 then renamed M62b).
* http://www.nature.com/nchina/2009/091216/full/nchina.2009.236.html
"The Movius line is south of that region, thus connecting the Altai/Central Asian region with SE Asia".
ReplyDeleteOnly according to the map in Wikipedia, not reality. Maybe Movius described it that way, though I imagine he was not that precise, but anyhow it was in 1948, long ago.
"Support for R having migrated through India but it says nothing about N".
On absence of any other clear evidence, R, as major N subclade, specially in South and West Eurasia, is best placed to have left a more clear track than the other smaller clades found west of the Movius line: N1'5, N2 and X.
It's not "N" what we are concerned about but some small N sublineages, those mentioned just above. You can choose the (a) UFO transport route, the (b) Atai route (equally unproven and unlikely) and (c) the South Asian route, which at least has some supporting evidence: strong for R, weaker but some for N1'5 and N2 and none for X.
"Not 'most of them' at all".
Care to discuss the details behind this protest? I'm interested.
"There's also the Tibetan M16"
ReplyDeleteIs that a basal haplogroup?
"Only according to the map in Wikipedia, not reality".
You're quite free to put the line where you like in order to fit your theory, but you won't convince anybody else.
"weaker but some for N1'5 and N2 and none for X".
None for the other two either.
"Care to discuss the details behind this protest? I'm interested".
wiki could be the starting point:
http://en.wikipedia.org/wiki/Haplogroup_M_(mtDNA)
"Haplogroup M2 [2] - found in South Asia, with highest concentrations in SE India and Bangladesh;[9] oldest haplogroup M lineage on the Indian sub-continent.[1]
M2a - most common in Bangladesh
M2b - most common in SE India
Haplogroup M3 [3] - found mainly in South Asia, with highest concentrations in west and NW India[9]
M4"45
Haplogroup M4 [4] - found mainly in South Asia but some sequences in Eastern Saudi Arabia
Haplogroup M4a - found in Gujarat, India[14]
Haplogroup M30 - mainly in India, found in Middle East and North Africa.
Haplogroup M37
Haplogroup M37a - found in Gujarat, India[14]
Haplogroup M5 [5] - found in South Asia
Haplogroup M5a - found in Orissa, India[14]
Haplogroup M6 [6] - found mainly in South Asia, with highest concentrations in mid-eastern India and Kashmir[9]
Haplogroup M6b - found in Kerala, India[14]
Haplogroup M7 [7] - found in East Asia, especially in Japan
Haplogroup M8
Haplogroup M8a: [8] - found in East Asia, especially in Japan
Haplogroup CZ
Haplogroup C [9] - found especially in Northeast Asia
Haplogroup C1 [10] - found in Asia and America (Native Americans and Hispanics in particular)
Haplogroup C4
Haplogroup Z [11] - found among diverse Eurasian populations: Hazara, Finns, Japanese, Sami and some Russians.
Haplogroup M9 [12] - found in East and Southeast Asia
Haplogroup E - a subclade of M9 - found especially in SE Asia
Haplogroup M10 [13] - small clade found in East Asia, Southeast Asia, Bangladesh, Central Asia, and southern Siberia
Haplogroup M11 [14] - small clade found especially among the Chinese and some Bangladeshis
Haplogroup M12'G
Haplogroup M12 [15] - small clade found in Japan
Haplogroup G [16] - found especially in Japan with some isolated instances in diverse places of Asia
Haplogroup M21 [17] - small clade found in SE Asia and Bangladesh
Haplogroup M27 [18] - found in Melanesia
Haplogroup M28 [19] - found in Melanesia
Haplogroup M29'Q
Haplogroup M29 [20] - found in Melanesia
Haplogroup Q [21] - found in Melanesia and Australia (Aborigines)
Haplogroup M31 [22] - found among the Onge, in the Andaman Islands[14]
Haplogroup M32 [23] - found in Andaman Islands
Haplogroup M33 [24] - small clade found in South Asia and Belarus
Haplogroup M33a - found in Gujarat, India[14]
Haplogroup M34 [25] - small clade found in South Asia
Haplogroup M34a - found in Karnataka, India[14]
Haplogroup M35 [26] - small clade found in South Asia and Slovakia
Haplogroup M39 [27] - found in South Asia[14]
Haplogroup M40 [28] - found in South Asia[14]
Haplogroup M41 - found in South Asia
Haplogroup M41b - found in Andhra Pradesh, India[14]
Haplogroup M41c - found in Andrah Pradesh, India[14]
Haplogroup M42 [29] - found among Australian Abrorigines
Haplogroup M48 [30] - rare clade found at least in Saudi Arabia"
The Philippines paper provides evidence for several SE Asian ones. And a paper on the Khasi hills also provided evidence some time back, but I can't find it now. Other papers are also available.
"Atai route (equally unproven and unlikely)"
Well, we do find A reasonably close to there, and stretching away to the east. And although W is concentrated in Northern Pakistan (not Southern or India note) but is found through much of SW Asia and even in Europe. So we have a trail across Asia from X in the west, N1'5, N2/W, A, N9/Y, R, N21, N22, O, S, N13, N14.
I missed Wagg's comment early.
ReplyDeleteI am thinking that it is the "Tibetan M16" which has been re-described as M62b. Because the Malagasy haplogroup was described by Ricaut 2009 and the M62b of PhyloTree is mentioned as described by Zhao 2009 as M16 and by Qin 2010 as M62 (without the "b" yet), also by Chandrasekar 2009 (now as M62a).
Terry:
ReplyDelete"You're quite free to put the line where you like in order to fit your theory, but you won't convince anybody else".
You mean you (not "anybody else" but "somebody else" specifically) will remain refractory to reasoning. Your problem.
"wiki could be the starting point"...
Sure but what's your point? It's most imprecise, and even if you take your time pondering each of the references you can seldom be certain of where a clade is present or where is most internally diverse, etc. Subcontinental region precision is the best I can hope for in most cases at this stage of knowledge (except maybe the largest or most localized clades).
"Well, we do find A reasonably close to there"...
A is one single haplogroup under N, it is original from further east as far as I can tell and it is not more related to N1'5, N2 and X than R is. At this moment I am not even sure if A is found in Altai at all.
All I can find right now on A makes its spread westwards (Central Asia, SW Siberia) very thin and imprecise. Mongols have 8% but Kazakhs are reported as having 0% mtDNA A in this paper, for example.
A most likely coalesced in NE Asia or even further south in East Asia and only arrived to such locations probably with the Turco-Mongol migrations in historical times. It doesn't even seem related to the Uralic flows, which are quite older and carried other oriental haplogroups to as far west as Europe (CZ, D).
"And although W is concentrated in Northern Pakistan (not Southern or India note)"...
It's still South Asia and it has been found in India anyhow, where N2a is also found.
"... but is found through much of SW Asia and even in Europe".
In Europe it looks, like X (and surely K and JT), totally a Neolithic input. N1 is probably older in Europe but it is probably also older in its West Asian spread, comparable to R0 or U.
"So we have a trail across Asia from X in the west, N1'5, N2/W, A, N9/Y, R, N21, N22, O, S, N13, N14".
Stubborn like a mule!
Across what part of Asia? South Asia! Only some derived X (and maybe A) is found in the Siberian corridor at all. It is irrelevant for our analysis because these "erratics" come from other places: West Asia (X2) and NE Asia (A, if any).
What mtDNA lineages do Kazakhs have? 20% G, 20% HV (H essentially), 10% D, 10% U5, 10% JT, 10% CZ, etc. No A anywhere.
Should we conclude that U coalesced in Central Asia because there is some U5 among Kazakhs? No way! Even G, which is maybe the most "typically Central Asian" lineage of all, has another origin: East Asia. Believe it or not it is the case for all Central Asian lineages: they arrived from elsewhere: West Asia, East Asia, Europe and even South Asia in some cases. It is a crossroads getting inputs from all around.
Seriously... make your homework, try to think about all this with an open mind. If you have a hypothesis and you want to defend it, you need evidence, but all you have is vague concepts, which you try to bend to your pre-existent model. That's not science: it's mere pseudo-science. And it's tiresome.
I'll give up on this subject for now until more evidence come to light, which I'm sure will turn you towards my idea. But a few other comments are in order.
ReplyDelete"you have is vague concepts, which you try to bend to your pre-existent model".
You're obviously talking about yourself here. After all it is you who wishes to shift the Movius Line to fit your model.
"Subcontinental region precision is the best I can hope for in most cases at this stage of knowledge (except maybe the largest or most localized clades)".
we can get a good idea of haplogroup distribution by looking at the nomenclature, a product of the history of their discovery. As an example we can see the order mtDNA haplogroups as a whole was discovered.
The first work was done by Americans on Native Americans. Hence haplogropups A, B, C and D. Then their East Asian origins: E, F and G; and Americans of European ancestry: H, I, J and K. Then the deeper caldes: L, M, N and so on.
So to haplogroup M. M1 closes to Europe. Then the common Indian ones: M2, M3, M4, M5 and M6. Then a series of East Asian ones: M7, M8, M9, M10, M11, M12 and M13. The Australian M14 and M15 and SE Asian M17, M21, M22 and M23. Melanesian M27, M28, M29. Andaman M31 and M32. Then, as the focus became more defined haplogroups M33 and M34 appeared. And with more research in even smaller regions regional haplogroups were discovered, many of them in single studies: M35, M36, M39, M40, M41, and then M44, M47, M49, M52, M71 and so on. Most of these last haplogroups are confined to the Northeast tribals. And M35-M412 are Andhra Pradesh. So not too many are northwest Indan.
"You're obviously talking about yourself here".
ReplyDeleteObviously I am not.
"After all it is you who wishes to shift the Movius Line to fit your model".
I "wish" to modify a Wikipedia map on light of modern (and not 1948) knowledge. It is something only logical to do (Kepler would have never made his famous breakthroughts if he had remained stuck into Platonian thought, right?)
I am not sure your logic on M subclades' nomenclature is justified. I think it is not in fact (even if there is some lesser truth to the overall "history of nomenclature" thing).
@ Maju & terryt about M16/M62 :
ReplyDelete"I am thinking that it is the "Tibetan M16" which has been re-described as M62b"
Ooops. Somehow I misread it, you're right.
confirmation : http://www.pnas.org/content/107/11/E38.full
"Third, the M16 haplogroup that is referred by Zhao et al. was recently also found by others who labeled it M62 (5), and it shares a basal mutation with haplogroup M40. Acknowledging first publication date, I propose to follow the M62 label. Consequently, Zhao et al.’s M16, M16a, and M16b are now relabeled M62b, M62b1, and M62b2, respectively. "
As for the discrepancy between M16 and M23 in Ricaut in its Madagascar paper :
http://www.biomedcentral.com/1471-2164/10/605/comments/comments
Terryt : " "There's also the Tibetan M16" - Is that a basal haplogroup? "
Given the words of the article I had assumed it was but obviously it's not (It's M62b which is not), sorry for the misfire.
"I 'wish' to modify a Wikipedia map on light of modern (and not 1948) knowledge".
ReplyDeleteWhere do you wish to change it, and what is your justification? Just because it's 'old' doesn't mean it's 'incorrect'.
"I am not sure your logic on M subclades' nomenclature is justified. I think it is not in fact"
OK. Let's compare M with N.
"Obviously N1 had much better chances of expansion in the West than N5 in already populated South Asia".
So applying the same criteria to M, look at M4''64: one mutation and then 8 haplogroups (M4, M18'38, M30, M37, M43, M45, M63, M64). And M3: no mutations, but instantly splitting into 3 (M3a, M3b, M3c). Turning to N we find that N1'5 splits into 2 after just one mutation but N1 merely splits into 2 after a further 2 mutations. Doesn't that suggest that M4''64 and M3 are far more likely to owe their respective expansions to moving into a previously unoccupied region than does N1? That 'unoccupied region' being greater India.
And your arguments about 'basal diversity' indicating origin would certainly point to Assam/Yunan being the region of origin for haplogroup M.
Again I'm sure you will not be convinced until you work it out for yourself.
"what is your justification?"
ReplyDeletePrior to H. sapiens colonization, the area of Altai is only documented to have Mousterian tech and Neanderthal remains AFAIK. The presence of the distinct Denisovan DNA suggests a different East Asian (H. erectus) presence but that does not change the fact that it was also Neanderthal territory with mode 3. It is the absence of not mode 3 but even mode 2, what justifies the distinction Movius made some 60 years ago. If Altai has mode 3 then it is clearly west of the line.
The same happens after H. sapiens colonization, when this area is one of the first ones with mode 4 (blades), after India and contemporaneously with Western Eurasia.
You know all that, so why all the fuzz?
...
I do not understand well what you mean with the M and N clades elaboration. M4"64 expanded right after M and is to M like R is to N: the main heir, so to say. Like R expanded largely in the same areas N did, M4"64 expanded largely in the same areas M did: South Asia. This suggests a partly joint expansion, when mother and daughter clades existed side by side, in the same tribes and clans.
"And your arguments about 'basal diversity' indicating origin would certainly point to Assam/Yunan being the region of origin for haplogroup M".
I do not think so because AFAIK there is much less diversity of M East of Bengal, even if it is still quite impressive. Instead of imagining where the haplogroups exist, in imaginary corners of Assam not documented, why don't you make some secretary work and list where each M subclade is reported to exist? With all that data clarified it should be a lot easier to confirm or, most likely, to reject your conjecture.
I have done that in the past and South Asia is always way ahead of East Asia (incl. Assam) in the basal sublineages count under M.
So I challenge you to prove me wrong. You would shatter to pieces all the coastal migration model if you can do that.
Until then, let's stop the wild speculation, please.
"Prior to H. sapiens colonization, the area of Altai is only documented to have Mousterian tech and Neanderthal remains AFAIK".
ReplyDeleteHang on. Haven't you consistently been trying to link the pre-sapiens population in that region to that of SE Asia? If anything the Movius line offers some support for your connection.
"The presence of the distinct Denisovan DNA suggests a different East Asian (H. erectus) presence but that does not change the fact that it was also Neanderthal territory with mode 3".
And so presumably the Denisova population would have been much closer to the Neanderthals than to any SE Asian population.
"I do not think so because AFAIK there is much less diversity of M East of Bengal"
That statement appears now to be incorrect. It may have held true before so many new haplogroups were discovered in the region east of Bengal.
"why don't you make some secretary work and list where each M subclade is reported to exist?"
I'll try to get some information together with links when I have time.
"Haven't you consistently been trying to link the pre-sapiens population in that region to that of SE Asia?"
ReplyDeleteNot quite. One thing is the Neanderthals in most of Altai and another thing is the "Denisovans" (likely a Neanderthal-Erectus hybrid) in Denisova and (probably) Okladnikov caves, which are a peripheral area to the North.
I already said that Altai is a crossroads but I also think it is West of the said Movius line, because even "Denisovans" used Mousterian tech, which is west of it (by definition).
"And so presumably the Denisova population would have been much closer to the Neanderthals than to any SE Asian population".
Not per the mtDNA, by which Neanderthals are much closer to us than to "Denisovans".
"That statement appears now to be incorrect. It may have held true before so many new haplogroups were discovered in the region east of Bengal".
Can you document this claim. So far I have only read what looks nothing but fallacious speculations.
"Not per the mtDNA, by which Neanderthals are much closer to us than to 'Denisovans'".
ReplyDeleteAnd I strongly suspect that the SE Asian H. erectus mtDNA will have branched off even earlier than the Denisova/Neanderthal split.
"Can you document this claim. So far I have only read what looks nothing but fallacious speculations".
Try this:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0007447
A whole heap of tribal Indian populations. I've had a quick look through it and basically all I've had to change from my above list is the Andhra Pradesh haplogroups. They are actually spread across South/Central India, not just confined to the east. I'll summarise in a separate post.
Most eastern haplogroups are not included, for obvious reasons. They include: D, M7, M14, M15, M17, M21, M22, M23'77, M27, M28, M29'Q, M42'75, M44, M47, M50, M71, M72 and M73. Many of these were what I called 'borderland' haplogroups, but it seems the borderland they were part of excludes India. They may have been missed i n the study of course, but that seems unlikely. Eighteen haplogroups east of India.
ReplyDeleteMembers of seven other haplogroups are included in the study although it is obvious their origin is further east: M8, M9, M10, M11, M12'G, M13'46'61 and M52'58. These haplogroups are all concentrated in the Northeast Indian section of the 'Borderland'. A further haplogroup, M4''64 seems to fit the same region. Eight haplogroups. That makes 26 of the 45 M haplogroups being east of India.
Turning to India a further six haplogroups are confined to the borderland within India or to the east: M6, M31, M32'56, M40'62, M49, M60. Two of these are found in the Andamans, also suggesting a borderland origin for the haplogroups. A running total of 32 of the 45 M haplogroups centred in northeast India or further east.
So, to the other Indian haplogroups. Eight are basically spread across Central/South India: M25, M33, M34'57, M35, M36, M39, M41, M53. The other three are widespread through India: M2, M3, M5 (which I had as East Indian, seems it's actually more widespread).
Two western M haplogroups are obviously not included: M1 and M48. Two haplogroups. Makes 45.
Several haplogroups appear in the article that I can't locate: M54 (south and central), M55 (west), M58 (border), M59 (east). These don't actually alter the ratios.
I invite you to check it all.
I was going to reply in more detail but my computer crashed and my autosave notepad failed in the process, so I lost all notes and with them most of hours of work. I will work on it again at a later moment.
ReplyDeleteThere are anyhow at least 14 South Asian and West Eurasian basal M sublineages (excluding NE India, Andamanese not shared with mainland India and Bangladeshi lineages shared with SEA). After I excluded 9 "unknown" lineages, they amounted to 49% of all remaining M sublineages (34 out of 43).
I was working on clarifying these unknown lineages when my PC crashed.
It is relatively easy (with some work to locate all the haplogroups) to discern this matter so I'll come back on this later on, when I recover from my frustration. Your link should be particularly useful to clarify some otherwise unreported clades. But as you have to admit many of these small clades are not from NE India or not exclusively so.
"There are anyhow at least 14 South Asian and West Eurasian basal M sublineages"
ReplyDeleteFourteen out of forty-five is not a huge proportion.
"After I excluded 9 'unknown' lineages, they amounted to 49% of all remaining M sublineages (34 out of 43)".
Don't you mean 14 out of 36? The only 'unknown' lineage I have is M75 which, although I can't access it, appears in a paper on the Philippines. So what are your unknown lineages?
A few comments on some of the haplogroups:
I've placed M13'46'61 in the borderland of India because of the three branches just M61 is found in India. I put M52'58 in the borderland because I recently read a paper (can't remember which one) that claimed the earliest branches of M58 are in the northeast of India. I've put M4''64 in the borderland because even though it is found right across India its members are spread as far east as Bihar (M18 and M45) so it could be eastern in origin.
Unsure. Maybe you can list them all?
ReplyDeleteThere were 9 unknown lineages for me at the time of the computer crash. Since then I have decided to start a Wiki to "cloud" and share these kinds of notes. However it's getting a bit more complicated that I expected at first. Maybe I should accept adverts and try Wikia instead?
There were only 43 basal lineages under M in my last count.
Perhaps it would be productive for us to turn out attention to the eastern end of the Movius Line. It seems that the stretch of hill country making up the border region between India, China and Borneo has usually been difficult for humans to pass through. However if we consider the possibility that M coalesced ther a lot of difficulties I've had with haplogroup M dissappear. The mountains and valleys in what could be called Indo-China would each have produces their own haplogroups, explaining the star-like diversity.
ReplyDeleteIt then becomes very easy to see how the Andamans were populated by basal haplogrous (although perhaps not so basal it seems). And it explains the sudden appearance of basal haplogroups in New Guinea, Melanesia and Australia.
But then we have M entering an empty India, from the east. That does make it difficult to see where the great southern coastal migration fits into the picture. Unless M's migration is a back-migration preceded by a massive extinction event (Toba?).
And N's distribution still basically forms an even more obvious arc right around M's. It's difficult to believe that if the two haplogroups expanded together then just the N haplogroups would be the ones that reached the extremities of that expansion. Perhaps M's expansion is later than N's, and failed to move as far as N had. But that raises problems concerning the mutation tails. M with 3 and N with 5. Of course there's no reason why daughter mutations would replace mother haplogroups at a constant rate.
So what about R's expansion? At the same time also? R does seem to have moved from SE Asia through India and China. But it managed to travel much further west than did M. To me R has the appearance of having expanded through an already occupied India, making its expansion later than that of M.
This is what I came up with yesterday:
ReplyDeleteNew Guinea:
M27
M28
M29'Q
Australia:
M14
M15
M42'75
East Asia:
D
M7
M8/CZ
M10
M12'G
Southeast Asia:
M9/E
M22
M23'77
M17
M72
Borderland (xIndia):
M21
M47
M44
M71
M50?
Borderland (incl. India):
M52'58
M31
M32'56
M49
M13'46'61
M11
M6
M60
M40'62
South and East India:
M36
M39
M41
Central India:
M33
M34'57
M35
M53
M25
M3
Western India:
M2
M4''64
M5
Southwest Asia:
M1'51
M48
Unknown (probably SE Asia):
M73
I refered to M75 in my previous post, so Australian M42 and Philippines M75 share an ancestor, which makes sense.
I've just come across the link I couldn't be bothered following up regarding Northeast Indian Khasi haplogroups:
ReplyDeletehttp://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0001141
You'll find it useful when you do another blog on haplogroup M. I look forward to it.
At least some of which you have declared "borderland incl. India" are clearly Indian/South Asian, notably M6. But otherwise your categorization seems to resemble mine (from memory) quite a bit.
ReplyDeleteAnyhow, in the most optimistic case for my position (assuming all your clades in that category are "Indian" by origin, which I know is at least arguable), then SA+WEA has 23/43, what would still be a majority but a quite smaller one than I used to consider.
So I have to admit that you have at least some reason, because at least 20/23 basal lineages seem exclusive from East of the Ganges Delta.
Of course some of the newly discovered lineages may end up forming larger subhaplogroups (that is normal) but we do not have that information as of yet.
"But then we have M entering an empty India, from the east. That does make it difficult to see where the great southern coastal migration fits into the picture. Unless M's migration is a back-migration preceded by a massive extinction event (Toba?)".
ReplyDeleteLet's imagine for the moment that your SEA origin hypothesis is right.
It'd fit with what we know for mtDNA N and R, as well as Y-DNA C, D and MNOPS. However it does not fit at all with Y-DNA F and most of its sublineages, excepted MNOPS.
What I do not gather is how to fit the lack of a South Asian distinct mtDNA lineage in that case. If it all happened as in your hypothesis, with people making the arduous route through Altai, probably along many many millennia (but leaving no trace of their pass in the archaeological record, nor even losing the tropical tan), and then arriving to SEA and booming, some other people should still have arrived to SA by the West, either the coast or by land. These people no doubt would have left their distinctive lineages there, however if M is from SEA, there is no single clade that could have done that and survived.
On the other hand we have clear evidence of Y-DNA dong precisely what we could expect: populating SA by the West and arriving (in derived form) to SEA to join C and D in the colonization of the area (or even displace them, whatever interpretation you prefer).
I think that a key issue is that, after removing "the borderlands", with their mysterious apparent diversity, there is still less diversity in East Eurasia (EA+OCE), with 16 basal lineages altogether. While SEA+WEA have more (I have to check this because I'm quite sure you have cheated with more than just M6 but I need time to discern the details).
In general, I think that M must still be of SA origin, though you do raise issues of apparent relevance, as a lot of new sublineages are towards the East.
"At least some of which you have declared "borderland incl. India" are clearly Indian/South Asian, notably M6".
ReplyDeleteYes. I had it in 'Borderland, including India'.
"So I have to admit that you have at least some reason, because at least 20/23 basal lineages seem exclusive from East of the Ganges Delta".
Basically half the haplogroups are east of the Ganges Delta and half are west if it. I've gone through the references in Phylotree and placed all the haplogroups. I'll email you with the result.
"Of course some of the newly discovered lineages may end up forming larger subhaplogroups (that is normal)"
To me that was the logical explanation for the apparently basal haplogroups in Australia, Melanesia and New Guinea. But if the origin is further east this problem dissappears.
"However it does not fit at all with Y-DNA F and most of its sublineages, excepted MNOPS".
It's probably possible to come up with a scenario that does fit F.
"What I do not gather is how to fit the lack of a South Asian distinct mtDNA lineage in that case".
Why do you find that a problem? That is exactly the problem I had with haplogroup N.
"some other people should still have arrived to SA by the West, either the coast or by land".
Perhaps not. Most of the papers on South Asia emphasise the lack of movement across the northwest boundary of the subcontinent. There appears to have been more interchange at the eastern end.
"we have clear evidence of Y-DNA dong precisely what we could expect: populating SA by the West and arriving (in derived form) to SEA to join C and D in the colonization of the area"
I don't believe we do have 'clear evidence' of that at all. It just fits the current theory.
"I think that a key issue is that, after removing "the borderlands", with their mysterious apparent diversity"
Why would you remove them?
"Yes. I had it in 'Borderland, including India'".
ReplyDeleteSo all "borderland, including India" are likely to be genuinely SA and "borderland excluding India" are awaiting further inquiry? Where did you put the two lineages reported in Bengal (Bangladesh) and also SE Asia?
"I'll email you with the result".
Great, thanks. :)
"It's probably possible to come up with a scenario that does fit F".
It seems most hard, impossible, to me.
"Why do you find that a problem? That is exactly the problem I had with haplogroup N".
Not me. I tend to see N as a group of lineages that scatter here and there in more or less loose or stricter connection with M or R. They do not look like a clearly distinct ancestral population, though maybe there was something of that at some point. R instead looks more compact, as does M.
"Most of the papers on South Asia emphasise the lack of movement across the northwest boundary of the subcontinent".
I disagree. A lot of lineages scatter between NW SA and Central Asia, as well West Asia in most other cases. The overall connection with West and Central Asia is quite stronger than with East Asia, with some exceptions restricted to the NE corridor and the Asutroasiatic tribes mostly.
Additionally mode 4 appears first in SA and spreads to Central and West Asia (Aurignacoid and other blade industries). So the Western connection is very clear and strong.
"Why would you remove them?"
Because they are geographically transitional and not well defined. So in order to compare "pure" SA(+WEA) and "pure" Eastern Eurasia, they are like something in between that is not clear where it belongs to.
But well... it's an open matter.
"clearly Indian/South Asian, notably M6".
ReplyDeleteM is certainly fairly widespread, but in the link I provided the other day M6 is shown as being most common in the Sonowai Kachari of the far northeast, and in the Pauri Bhuiya of Orissa. At half that frequency it is also present in the Munda of Bihar, the Betta Kuruba of the south and the Hill Kolam of Central India. It is next most frequent in the Korku of Madya Pradesh and the Malpaharia of West Bengal. Much lesser frequency in the Andh of central India and the Gallong of the far northeast. If you wish you could sort out where the basal M6s are found.
"Where did you put the two lineages reported in Bengal (Bangladesh) and also SE Asia?"
Presumably 'borderland', but which haplogroups are you refering to?
"Because they are geographically transitional and not well defined".
Most of them are well-defined, and perhaps they're not 'transitional'.
"So in order to compare 'pure' SA(+WEA) and 'pure' Eastern Eurasia, they are like something in between that is not clear where it belongs to".
Surely they belong where they're found, in Northeast India. Why would there be only two groups, pure East Eurasian and pure South Eurasia?
By the way. Have you received my email? I had problems making an attachment.
"If you wish you could sort out where the basal M6s are found".
ReplyDeleteSilly me. There's only one lineage. That indicates it has spread recently, or it surely would have diverged being so widely spread.
You cannot judge only on that paper, Terry. M6 has been reported across India, specially in Bengal and the South in much greater numbers than add all the Tribals of the NE together. That is clear case of South Asian lineage with penetration in the NE and that specific tribe is probably for that reason to be considered of SA origin overall, as Assam and surroundings is a transitional area in which flows from both directions are to be found.
ReplyDelete"Presumably 'borderland', but which haplogroups are you refering to?"
M10, M11 and M21. Just found they are three.
"Surely they belong where they're found, in Northeast India. Why would there be only two groups, pure East Eurasian and pure South Eurasia?"
True if true, but most lineages are not so local as far as I know. It is different a clade found ONLY in a tribe of Arunachal than a clade found in a tribe of Arunachal and a zillion other places.
"By the way. Have you received my email?"
No, sorry. Have you removed the anti-spam protection "DELETETHIS" if you got my email from my profile?
There used to be at least two sublineages under M6: M6a and M6b (Metspalu 2004), however for some reason PhyloTree has removed this distinction. I am sure that M6 must have a marked downstream differentiation, if only because it's a large clade.
ReplyDeleteIt seems PhyloTree is ignoring Metspalu 2004 (not found), while using other papers from this author instead.
Anyhow, M6a is found in the SE (and Oman), while M6b has a more northernly distribution (Bengal, Kashmir, Gujarat). It is a very clearly SA clade, even if a small one (frequency seldom raises above 5%).
"Have you removed the anti-spam protection 'DELETETHIS' if you got my email from my profile?"
ReplyDeleteNo, I hadn't. I'll try again.
"M6 has been reported across India, specially in Bengal and the South in much greater numbers than add all the Tribals of the NE together".
But the tribals should give us the most useful information. Once a haplogroup moves out of the tribals it would most likely spread very widely, and so provide very little information as to its original distribution.
"as Assam and surroundings is a transitional area in which flows from both directions are to be found".
But we're examining the possibility that Assam may be the place of origin for the whole haplogroup, so Assam may not be merely 'a transitional area' for many haplogroups.
"M10, M11 and M21. Just found they are three".
I've put M10 in 'East Asia' as I gather it's widespread there. Its presence in Bangladesh could be the result of movement from the east. M21 seems to be a Southeast Asian haplogropup and again its presence in Bangladesh result from back movement. M11 I've placed it as indigenous NE Indian so its presence in Bangladesh is not surprising.
"True if true, but most lineages are not so local as far as I know".
But they presumably each separately coalesced in a single region. It's the single region for each haplogroup that I am searching for, and I hope you will help. Or perhaps we're helping each other.
Thanks for the link. This is an interesting comment in the light of where I might be going:
ReplyDelete"The variation [of M] among caste populations climbs from approximately 40% in Gujarat and (Indian-) Punjab to 65% in the southern states, and peaks at over 70% in West Bengal (Table 2). We observed a similar geographic pattern among tribal populations, where the frequency varied from just over 50% in the northern states of Punjab and Himachal Pradesh, increased to 70%–80% in the southern states, and peaked at 86% in West Bengal (Table 8, see 3)".
And this covering another angle:
"The frequency peak of haplogroup W is 5% in the northwestern states – Gujarat, Punjab and Kashmir. Elsewhere in India its frequency is very low (from 0 to 0.9%) (Table 2) forming a significant spatial cline (Figure 4)".
So W is a northwest haplogroup, could have come in from further west (Iran) rather than being Indian in origin. Especially seeing that:
"Over 90% of the mtDNAs found in Iran belong to haplogroups HV, TJ, U, N1, N2 and X, commonly found in West Eurasia (Table 2)".
"But the tribals should give us the most useful information".
ReplyDeleteI do not see why. Tribals may harbor rare variants and such but are in the end parts of the larger South Asian population. Also many tribals, specially Austroasiatics, are Neolithic arrivals from SE Asia at least in Y-DNA and language. We have to look at the whole population, as we do elsewhere.
"Once a haplogroup moves out of the tribals it would most likely spread very widely"...
I do not understand what you mean here well - looks nonsense on first sight. Do you mean that you believe (wrongly) that lineages shared by tribals and non-tribals are of tribal origin? Always?
If it'd be the case, that is not, a lineage that got out of a tribe should not expand among the common people but rather vanish or remain small at best.
"... and so provide very little information as to its original distribution".
That is true but, as I said before, typical of most Indian lineages. Carefully estudying their diversity and sub-structure may give better results but tribals are not necessarily a better indicator, as they may have concentrated a lineage recently by drift, have migrated along vast areas (specially if non-farmer now or in the past but also in regards to Neolithic flows like the Austroasiatics).
Sedentary mainstream population should provide the best data. Though tribals can be used as a secondary source of info.
"But we're examining the possibility that Assam may be the place of origin for the whole haplogroup, so Assam may not be merely 'a transitional area' for many haplogroups".
Hmm....
It is anyhow transitional between the two subcontinental regions today.
"Its presence in Bangladesh could be the result of movement from the east".
Sure but the opposite is also possible until the underlying structure is discerned. They are "borderland" as it's shared by both regions but could be reclassified if it becomes clear it is ancestral from one or the other.
"But they presumably each separately coalesced in a single region".
Maybe or maybe not, which one in any case? Were not you arguing recently about a huge cloud of mtDNA N covering whole continents at the same time, from Europe to Australia maybe? I think that is very exaggerated but certainly a mtDNA haplogroup can well migrate without mutations happening and "drop daughters" as it does, not as a cloud but as "invisible" pre-sublineages.
The question is from where to where it migrated and the answer is IMO from where the greatest basal diversity is in all directions where it's found.
However it seems in some cases this key parameter is less clear we'd like to.
"I'll try again".
I just got it, thanks.
It seems very difficult to discern some of the key stuff in your list. For example, you mention the Khasi in relation to M49 (which you gave an asterisk, M49*, for some reason) and I cannot find any population with that name in the source.
ReplyDeleteYou also mention M40 "in East Asia also" yet the source is that same Chandrasekar paper, where only Indian tribals are listed.
I fear your list is not reliable and I better stop chatting and get to produce my own.
"Were not you arguing recently about a huge cloud of mtDNA N covering whole continents at the same time, from Europe to Australia maybe?"
ReplyDeleteIf you wish to put it that way. But I have never argued that every single N haplogroup 'formed a huge cloud ... covering whole continents at the same time, from Europe to Australia'. Whereas you seem to be arguing that haplogroup M formed a huge cloud covering the whole of India.
"certainly a mtDNA haplogroup can well migrate without mutations happening and 'drop daughters' as it does, not as a cloud but as 'invisible' pre-sublineages".
Exactly what I have consistently argued is the case with haplogroup N.
"The question is from where to where it migrated and the answer is IMO from where the greatest basal diversity is in all directions where it's found".
And M's greatest basal diversity, and point of dispersal, appears to lie in the Northeast India/South China region.
"They are 'borderland' as it's shared by both regions but could be reclassified if it becomes clear it is ancestral from one or the other".
I've already said several times that I believe there are more shared lineages within M than have been discovered yet, especially in relation to the Australian/Melanesian region. Until such time as these deeper connections are discovered we are reduced to working with the information we currently have.
"Do you mean that you believe (wrongly) that lineages shared by tribals and non-tribals are of tribal origin? Always?"
Obviously tribals can absorb haplogroups from the wider population, but if a haplogroup is found only, or mainly, in tribals within a defined region then it is surely likely the haplogroup has been in that region for some time.
"Tribals may harbor rare variants and such but are in the end parts of the larger South Asian population".
Exactly. That's why their haplogroups can be so revealing.
"you mention the Khasi in relation to M49 (which you gave an asterisk, M49*, for some reason) and I cannot find any population with that name in the source".
The asterisk means it diversified after just one mutation. The Khasi Hills are a region in Northeast India where the tribals with the haplogroup are found.
"You also mention M40 'in East Asia also' yet the source is that same Chandrasekar paper, where only Indian tribals are listed.
Obviously I didn't rely on the Chandraseker paper alone, but if I'd listed all the references it would be a very long list. I think the East Asian reference is from the Sun 2006 or Thangaraj 2006 papers.
"Also many tribals, specially Austroasiatics, are Neolithic arrivals from SE Asia at least in Y-DNA and language".
And probably many mtDNA haplogroups. Indicating even more strongly an eastern location for many Indian haplogoups.
"I fear your list is not reliable and I better stop chatting and get to produce my own".
You're quite free to produce your own, in fact I'm hoping you do. But I'm certain that once you look at the evidence your list will not disagree substantially with mine. It will certainly not have any fewer than the 20 haplogroups found east of India that I have in my list.
"Whereas you seem to be arguing that haplogroup M formed a huge cloud covering the whole of India".
ReplyDeleteAt first of all? No. However soon after that, it did and part of it overflowed to the East. At least that used to be my model.
It is anyhow a lot different to cover a homogeneous smaller region (or large parts of it) than the vastness of Eurasia... plus Oceania. South Asia is much smaller than Asia, right? It's roughly comparable to Europe or SE Asia in extent or people it could harbor.
"And M's greatest basal diversity, and point of dispersal, appears to lie in the Northeast India/South China region".
That's what I want to clarify with patience and in due time. Not now.
"Obviously tribals can absorb haplogroups from the wider population"...
The wider population were also tribals some 6000 years ago... or less.
"if a haplogroup is found only, or mainly, in tribals within a defined region then it is surely likely the haplogroup has been in that region for some time".
Maybe but that is not the case of M6 nor you have bothered checking in other cases.
"Obviously I didn't rely on the Chandraseker paper alone, but if I'd listed all the references it would be a very long list. I think the East Asian reference is from the Sun 2006 or Thangaraj 2006 papers".
That's what I am interesting in discerning clearly. Did Tangaraj ever authored a paper a paper on SEA anyhow?
Thangaraj 2006 is a paper on South Asia, where M40 (and others) is listed as autochtonous Indian haplogroup. These lineages were in that day: M2, M3, M5, M6, M25, M31, M32, M33, M34, M35, M36, M39, M40, M41 and the major sublineage now known as M4"64 but then as M4"30 (I'd say to call this super-haplogroup M0). There were 15 already then, even if two are shared between Andaman and the subcontinent (M35 is partly continental too).
"And probably many mtDNA haplogroups" (of Austroasiatic tribals are immigrant).
I'd suggest you to read the classics on them: they are East Asian by Y-DNA and South Asian by mtDNA, or so was claimed back in the day.
"It will certainly not have any fewer than the 20 haplogroups found east of India that I have in my list".
Of course. That's why I'm puzzled and tempted to go back to my old model in which lineages flowed through all Tropical Asia with ease for a while, or even adopting the Toba catastrophe model.
"Of course. That's why I'm puzzled and tempted to go back to my old model in which lineages flowed through all Tropical Asia with ease for a while"
ReplyDeleteYou will find that a useful excercise.
Of course it's fairly unlikely that M actually originally coalesced in the borderland. Just as it's fairly unlikely that N coalesced in SE Asia, although it's very likely that R did so. Haplogroups M and N are both L3, so presumably they coalesced somewhere near Africa. M1'51 is probably today found in the same region where it always has been. Same with haplogroups X, N1'5 and N2/W.
M's diversity in the 'borderland' is almost certainly a reflection of the formidable geographic boundary that it came up against in the form of the region's mountains and jungle. In other words the expansion east from there was hardly 'rapid'. However that idea eliminates the 'coastal' part of the 'great southern coastal migration'. Yhe only times haplogroup M was involved with the sea was when it crossed Wallacea to reach Australia/Melanesia/New Guinea and when it crossed to the Andamans. The presence of Haplogroup M in the Andamans is not evidence of a remnant of the original rapid movement from Africa to Australia. It's an independent movement from the mainland.
Likewise for haplogroup N. The question is: 'Did N also cross that same formidable barrier in Noretheast India/South China/Mainland SE Asia in its original move east?'
"At first of all? No. However soon after that, it did and part of it overflowed to the East".
ReplyDeleteSee previous post.
"South Asia is much smaller than Asia, right? It's roughly comparable to Europe or SE Asia in extent or people it could harbor".
True. But it's reasonable to assume that the individual haplogroups coalesced in separate regions within South Asia, just as individual haplogroups have coalesced in separate regions within Europe or SE Asia.
"Maybe but that is not the case of M6 nor you have bothered checking in other cases".
We're arguing in circles here. The Chandraseka paper shows M6 as being most common in the Northeast India tribals, the Sonowai Kachari. Take from that whatever you wish.
"Thangaraj 2006 is a paper on South Asia, where M40 (and others) is listed as autochtonous Indian haplogroup".
Have I contradicted that? Have another look at my list.
"These lineages were in that day: M2, M3, M5, M6, M25, M31, M32, M33, M34, M35, M36, M39, M40, M41 and the major sublineage now known as M4"64 but then as M4"30"
If you'd bothered reading what I sent you you would have seen that I placed all those haplogroups in India. So what's your problem?
"I'd suggest you to read the classics on them: they are East Asian by Y-DNA and South Asian by mtDNA, or so was claimed back in the day".
And I agree with that. Again, what's your problem? I have not place any Indian Austro-Asiatic-speaking mtDNA Haplogroups outside India.
"M's diversity in the 'borderland' is almost certainly a reflection"....
ReplyDeleteIt is not demonstrated. See my latest blog entry: SA Austroasiatics are South Asian by mtDNA, while the Khasi (Tibeto-Burman) are instead SEA. Insisting in the "borderlands" without finely discarding all that belongs elsewhere amounts to what we can call intellectual terrorism bordering pseudoscience. Probably we'll be able to clarify these matters in the near future and will not turn out to be anything near what you say.
"that same formidable barrier"
ReplyDeleteThere's no "formidable barrier". Just some hills and jungle (no ice, no deserts, no impassable mountains, no ocean). The real barrier is surely one of lower demographic capacity, what really makes it act as a buffer is the lower population densities than neighbours by either side, helped only up to some point by the geography.
It's really ridiculous how you build up such fantasy products with zero support, just because.
"it's reasonable to assume that the individual haplogroups coalesced in separate regions within South Asia, just as individual haplogroups have coalesced in separate regions within Europe or SE Asia".
It's possible but we cannot often discern such subtleties. Certainly not without the most comprehensive sampling and phylogenetic resolution, still to be done in most cases.
"The Chandraseka paper shows M6 as being most common in the Northeast India tribals, the Sonowai Kachari. Take from that whatever you wish".
M6 is widespread through India, why would a small isolated tribe matter at all? I cannot take you seriously if you do that. Frequency (specially in a drift-prone tribal population) is less relevant (it's almost irrelevant in fact).
By such measures, H would have originated in Araba (where it's more than 80%) but this area was only colonized in the Epipaleolithic and lacks the diversity; V would have coalesced in Lapland and so on. Nonsense!
"Have I contradicted that?"
Yes, you said (in category "NE India", where you arbitrarily placed nothing less than 12 basal lineages:
"M40’62* M40 in East Asia also (1)"
Based on a paper that does not even study SE Asia at all (Chandrasekar'09). You are taking the part for the whole, without any confirmation and clearly brushing towards your side, and that's a very bad dynamics for honest scientific debate.
Other errors are M31 (never located in NE India but in West Bengal), M11 an East Asian clade... I have not much idea yet on the minor clades but I'm sure that your NEI cluster is at least 1/3 smaller if not totally from just external sources.
"Insisting in the 'borderlands' without finely discarding all that belongs elsewhere amounts to what we can call intellectual terrorism bordering pseudoscience".
ReplyDeleteYou have a very strange idea of geography. Do you really believe that there were two completely separate populations of haplogroup M: South and East Asian? And that any haplogroups in any intermediate region must come from one or other of these two regions? Surely we should expect a cline across South Asia, through the borderlands to SE Asia and East Asia.
"There's no 'formidable barrier'. Just some hills and jungle"
Hills and jungle are no barrier? Don't be an idiot. The Movius Line shows that the region has been a barrier for much of human history.
"what really makes it act as a buffer is the lower population densities than neighbours by either side, helped only up to some point by the geography".
Maju. with the development of agriculture Bangla Desh has become one of the most densely populated regions in the world.
"category 'NE India', where you arbitrarily placed nothing less than 12 basal lineages"
It's hardly 'arbitrarily' Maju. Do your own research.
"Based on a paper that does not even study SE Asia at all (Chandrasekar'09)".
So? The fact that M40 is found in East Asia is obviously not from the Chandraseker paper. I can't remember which one now, but I'm sure if you bothered to look you'd soon find that M40 is found in East Asia as well as in Northeast India.
"Other errors are M31 (never located in NE India but in West Bengal)"
M31 is certainly found in NE India, and in the Andamans. It is quite possibly also found in NW India but did it originate there?
"M11 an East Asian clade..."
Found in NE Indian tribals as well. Perhaps it is an East Asian clade, but I was giving the benefit of the doubt to an Indian origin for the bulk of haplogropup M.
"Probably we'll be able to clarify these matters in the near future and will not turn out to be anything near what you say".
It will almost certainly turn out to be exactly as I say.
"Do you really believe that there were two completely separate populations of haplogroup M: South and East Asian?"
ReplyDeleteWhere do I say that. All I say is that we must be careful and strict in our discernment.
"Hills and jungle are no barrier?"
They are not. You admitted so once at least but I have always said so. The Himalaya is a barrier, Wallace line is a barrier, the Sahara is a barrier... but some hills are not at all: they are inhabitable and transitable country for our species.
"The Movius Line shows"...
That "line" is a construct, not any evidence in itself. Please!
It indicates that the West/South and East regions of Eurasia are relatively distinct, what is self-obvious. This is not because of absolute barriers but rather because of choking points at the Ganges Delta and Altai.
"with the development of agriculture Bangla Desh has become one of the most densely populated regions in the world".
It does not mean it was not earlier also. Such a rich mixed alluvial swampland may have hosted lots of people since the beginning. It's much like your dreamland Wallacea: islands and waterbodies of all kinds, but less extremely challenging for crossings, much less.
"It's hardly 'arbitrarily' Maju. Do your own research".
I will, I'm on it. But do not expect results in less than weeks or months.
"M31 is certainly found in NE India, and in the Andamans. It is quite possibly also found in NW India but did it originate there?"
West Bengal, see here.
"Where do I say that. All I say is that we must be careful and strict in our discernment".
ReplyDeleteYou certainly give the impression of believing there are two centres of M haplogroups (South Asian and East Asian) with nothing in between.
"The Himalaya is a barrier, Wallace line is a barrier, the Sahara is a barrier... but some hills are not at all"
Have you seen those 'hills' of the borderland region? 'Mountains' would be a better description. And steep ones at that.
"This is not because of absolute barriers but rather because of choking points at the Ganges Delta and Altai".
And the evidence is fairly overwhelming that the Ganges Delta (and the hills to the east) was a 'choking point' for the expansion of the M clade.
"Such a rich mixed alluvial swampland may have hosted lots of people since the beginning".
I doubt that very much. It's only once the vegetation had been cleared for agriculture that it would have been very useful as human habitat.
"West Bengal, see here".
West Bengal is very close to Bangla Desh, the Ganges/Brahmaputra Delta. And, in the Chandraseker paper, M31 is only listed as being present in the Munda of neighbouring Bihar and in the Pauri Bhuiya of Orissa. So it's eastern presence in India probably indicates an east Indian origin.
"You certainly give the impression of believing there are two centres of M haplogroups (South Asian and East Asian) with nothing in between".
ReplyDeleteYour impression is wrong. I'm just trying to quantify what is SA and what EA and what is truly "intermediate" with some solid criterion and logical methodology.
"Have you seen those 'hills' of the borderland region? 'Mountains' would be a better description. And steep ones at that".
Humans are climbers and certainly hikers, unlike elephants. That's not more complicated that highland New Guine, for where there is evidence of habitation since 50 Ka. ago.
The Hymalaya is a barrier not because of steepness but because of extreme alpine climate.
Anyhow, if the mountains are barriers (what they are not, not in any absolute sense), then only the coastal route would remain. So can you make up your mind.
For your reference: orography of Burma.
"And the evidence is fairly overwhelming that the Ganges Delta (and the hills to the east) was a 'choking point' for the expansion of the M clade".
That I do not see clear. The choking point does not seem to have worked well until population densities grew.
"West Bengal is very close to Bangla Desh".
They are both the same historical country: Bengal and they speak the same language: Bengali. I would not expect major differences between both provinces, they are in what I and everyone calls South Asia. SEA begins East of Bengal.
Shit, I forgot the link!
"And, in the Chandraseker paper, M31 is only listed as being present in the Munda of neighbouring Bihar and in the Pauri Bhuiya of Orissa".
No, because these AA peoples have South Asian mtDNA lineages, we have discussed that above. They are not even in NE India anyhow but West of Bengal.
"So it's eastern presence in India probably indicates an east Indian origin".
East India as in Orissa or Bengal maybe but not as in NE India or Assam, right?
"They are both the same historical country: Bengal and they speak the same language: Bengali. I would not expect major differences between both provinces"
ReplyDeleteExactly. They are both well within the northeastern quarter of the Indian subcontinent.
"they are in what I and everyone calls South Asia. SEA begins East of Bengal".
Of course SE Asia begins east of Bengal. In fact east of Assam. I've certainly never claimed otherwise although you seem unable to discern the difference between NE India and SE Asia.
"No, because these AA peoples have South Asian mtDNA lineages, we have discussed that above. They are not even in NE India anyhow but West of Bengal".
What I'm trying to do, and you seem very afraid of the possible result for some reason, is to narrow down the various regions within India where the individual M haplogroups coalesced. You seem bent on lumping all the Indian haplogroup Ms into one huge region: India. Is that to avoid considering the consequences of regional variation?
"East India as in Orissa or Bengal maybe but not as in NE India or Assam, right?"
OK. India seems to divide easily into the following regions: (1) Northeast India as far west as Bihar and Orissa, (2) Central India including Madhya Pradesh, Maharashtra and Uttar Pradesh, (3) South India including Andhra Pradesh, Karnataka, Kerala and Tamil Nadu, (4) Northwest India and Pakistan including Gujarat, Rajastan, Punjab and Sindh. Try sorting the M haplogroups into those regions. They separate relatively distinctly.
"Shit, I forgot the link!"
And some interesting comments in that link:
"Our recent study on Andaman and Nicobar islanders identified two more mitochondrial DNA (mtDNA) lineage groups, M31 and M32, present in Andaman islanders and absent in 6500 samples from the Indian subcontinent"
Of course connections to the India subcontinent have been since found for both haplogroups.
"This is still consistent with the ancient isolation of these gene pools, albeit not as early as the initial phase of human migration out of Africa".
But they admit that 'the settlement of the Andaman islanders by a population carrying M31 and M32 founders could have happened any time after the out-of-Africa migration of modern humans'. It's just that they seem desperate to hang onto the rapid southern coastal migration theory. There is certainly no reason to believe M reached the Andamans before it had reached Melanesia.
"What I'm trying to do, and you seem very afraid of the possible result for some reason, is to narrow down the various regions within India where the individual M haplogroups coalesced".
ReplyDeleteYou are not using the proper population samples: only tribals.
"You seem bent on lumping all the Indian haplogroup Ms into one huge region: India".
I could never manage to get clear regional differences (I wish it'd be so easy). Most lineages are fairly scattered through the continent. By only looking at one paper on tribals you miss some of this reality of great homogeneity through the subcontinent - with some differences but differences of detail and degree not at all the absolute stuff you want to make up.
It's like if you'd looked at European "tribals" and decided: H is Basque, V is Saami, U5 is (say) Bashkir. Well, you'd just got everything messed up. For example, V is most likely from SW Europe and certainly not original from the Saami. H surely spread from Central Europe instead, as probably did U5...
"(1) Northeast India as far west as Bihar and Orissa"
That's not standard at all. NE India is always said to the states east of Bengal and only them. The Bengal/NEI distinction is very very marked in the ethnological aspect so please keep the usual NEI distinction, thanks.
Bengal and Orissa are within "East India", along with Bihar and Jharkhand.
The other standard regions are North (North and North-Central), West and South.
"I could never manage to get clear regional differences (I wish it'd be so easy). Most lineages are fairly scattered through the continent".
ReplyDeleteThat's why it's necessary to start with the tribals. They are less likely to be heavily influenced by those lineages scattered across the continent. Presumably that's why Chandrasekar et. al. chose tribals for their study.
"you miss some of this reality of great homogeneity through the subcontinent - with some differences but differences of detail and degree not at all the absolute stuff you want to make up".
But surely that homogeneity is a product of later mixing. From the Paper:
"The current Indian mtDNA gene pool was shaped by the initial settlers and was galvanized by minor events of gene flow from the east and west to the restricted zones".
To understand the migration route(s) through India it is necessary to untangle that homogeneity. besides which the authors mention that:
"Northeast Indian mtDNA pool harbors region specific lineages, other Indian lineages and East Asian lineages'.
So it's not as homogenous as you claim.
"It's like if you'd looked at European 'tribals' and decided: H is Basque, V is Saami, U5 is (say) Bashkir".
But even those tribals do help expose the migration patterns into Europe. You wouldn't come to the conclusions you suggest because other evidence argues against those conclusions. The same as the Indian tribals do not provide the sole evidence for my conclusions. It's just that, for some reason, you seem very afraid of what those conclusions about South and SE Asia are saying.
OK. Extract West Bengal, Bihar and Orissa from Northeast India then. The conclusion will be much the same. Tribals in those states are a mixture of the Northeast Indian haplogroups and the Central ones anyway.
"Bengal and Orissa are within 'East India', along with Bihar and Jharkhand".
The solution is simple. I'll make a fifth category. But very few (if any) haplogroups seem indigenous to those states, so it seems a bit pointless to do so.
I disagree that tribals are at all more informative than common people.
ReplyDeleteAnd I disagree that mere frequencies can help untangle that structure: we'd need clear understanding of the phylogenetic structure, as happens in other cases.
Tribals, because of their low numbers, are particularly prone to stochastic drift effects, so their frequencies can well be product of a recent accident.
In the papers where tribals have been compared with other peoples, they often do not share frequencies and geography at all, suggesting either recent migrations by the tribals (often seminomadic) or similarly recent stochastic alterations in their genetic makeup.
"You wouldn't come to the conclusions you suggest because other evidence argues against those conclusions".
Exactly as in India.
"But very few (if any) haplogroups seem indigenous to those states"...
I'd say Bengal and Bihar probably have some diversity of their own. Certainly Bengal looked important in the case of M6 (but along the SE and Kashmir).
"Tribals, because of their low numbers, are particularly prone to stochastic drift effects, so their frequencies can well be product of a recent accident".
ReplyDeleteTrue, but we have to narrow down the search somehow. The Chandrasekar paper is entitled 'Updating Phylogeny of Mitochondrial DNA Macrohaplogroup M in India: Dispersal of Modern Human in South Asian Corridor' after all. So the authors must have considered the excercise useful for elucidating that problem.
"we'd need clear understanding of the phylogenetic structure, as happens in other cases".
Which we largely have for M, although I'm sure that not all basal connections have been uncovered yet. I'd hazard a guess at this stage and say scientists will discover a split between M1'51 and the rest. Then a two-way split between the Central Indian and East Asian/Northeast Indian haplogroups, with the offshore haplogroups being part of the latter grouping.
"In the papers where tribals have been compared with other peoples, they often do not share frequencies and geography at all"
But so to a very limited extent in the Chandrsekar paper. Several haplogroups were present in just one of the tribals. M64 in the Nihal, M63 in the Madia (both tribes from Central India), M10 in the Gallong of NE India although the haplogroup is basically East Asia so that's no surprise. Two other haplogroups that a subclades of wider groups are M62 (part of M40'62) found only in the Dirang of Assam and M56 (part of M32'56) found only in the Korku of Central India. Seven haplogroups were observed in just two, usually neighbouring, tribes and a further nine haplogroups were observed in three tribes, again usually from the same or neighbouring regions.
"I'd say Bengal and Bihar probably have some diversity of their own. Certainly Bengal looked important in the case of M6 (but along the SE and Kashmir)".
I agree that M6 is widespread so it's difficult to come to definite conclusions. However the only haplogroups I can see that may have originated in the Bengal/Bihar/Orissa region are M41and M53, and possibly the M4''64-derived haplogroup M18'38. In all three cases the evidence is weak and they are more likely to have entered the region from outside it. I'll send you my up-dated version of haplogroup distribution if you're interested.
Hi Maju,
ReplyDeleteCongratulations! And excuse me if I didn't read all the comments ( I have problems to move the cursor, it drives me crazy), but I have some questions.
First I'd like to ask you how do you explain that there was an interbreeding in south asia with neanderthal close relatives but there wasn't more in the middle east, in Europe or in the reast of Asia. If I remember well your point was that the cultural barrier was the most important barrier to hybridization, so what made the difference with the indian hominids?
And about the population of Homo erectus soloensis, is it possible that the population was isolated in Java when the first hybridizations happened in south Asia and the land bridge didn't appeared again until the time of the Melanasians ancestors? If the sunda land was emerged for longer how could be different populations between the continent and the Sunda "peninsula"?
About the Karitiana people, I don't understand which is the matter. I thinky is likely a random factor, I red somewhere that probably America was colonized by an small group of people. Such an small group could have had randomly and specially low level of Neanderthal DNA. You speak about the karitiana because is the only native group of america studied or because they have a different frequency in respect to other groups?
I'm not saying that anything must have happened the way I suggest here. It is just an speculation: an alternative model to the simplistic Neanderthal/Denisova inflows taken senso stricto. What I'm saying is that they might be just proxies for relatives, that we need to open our minds in that aspect.
ReplyDelete"... how do you explain that there was an interbreeding in south asia with neanderthal close relatives but there wasn't more in the middle east, in Europe or in the reast of Asia"...
According to the "coastal" (Tropical, not strictly coastal but near the Indian Ocean anyhow) migration model, the first destination of the proto-Eurasian branch of Humankind was Arabia (arid) and then South Asia, where they began expanding. It fits best with the mtDNA, though there's always someone who reads it differently.
This model also fits well with the most recent archaeological findings (Petraglia 2010, etc.), which would place the migration c. 90 Ka (or even as early as c. 125 Ka per Armitage 2011)
So I just tried to fit the new genetic admixture knowledge with the overall expansion paradigm.
West Eurasia excepting the habitable parts of Arabia/Persian Gulf) were only colonized later, c. 48 Ka. onwards, from South Asia (though some lineages may have reached as far as SE Asia and "bounced" back there).
There were no Neanderthals in East Asia nor South Asia, not even West and Central Asia before some date probably - a date that is maybe later than the crossing of proto-Eurasian H. sapiens through the area. We know that there was a big-headed hominin (Hathnora) in South Asia, which IMO was a close relative of Neanderthals but not quite (no Mousterian). We know that there were H. erectus in East Asia and that at least in Java they lasted until our species passed by.
So it's just joining the dots, I'd say, with an open mind. I have proposed other simpler (and more strictly "Neanderthal") models earlier but the Denisova finding forced me to think beyond that simple scheme of admixture in a West Asia that was for our ancestors little more than a corridor they crossed in fast pace (and where there might not even have been Neanderthals in that time yet anyhow).
This is option B. Whatever one may think about Hathnora, I'm 100% sure that proto-Melanesians did not admix with Denisovans in Altai. That's obvious when you look at their skin color (there are other reasons but that one is striking): because living in Altai or similarly high latitudes for many generations demands either vitamin supplements or a biological adaption (light skin) that once happens does not go back (see how Native Americans have never recovered the "black" skin color and remain just more or less "brown" or even white-ish in many cases).
So they must have got that admixture in the Tropics and one possible interpretation is H. erectus soloensis in Sundaland, where they must have been living in their way to Papua and the smaller Melanesian islands.
...
...
ReplyDelete"If I remember well your point was that the cultural barrier was the most important barrier to hybridization, so what made the difference with the indian hominids?"
Not exact, I'd say. It is living together or not. If there are no "multi-species" communities of some sort, hybridization can hardly happen in amounts that we might notice. Unless it happened accidentally (random encounter) when the population was extremely low (80 effective unmixed people: 1 effective 50% hybrid).
I suspect that these tiny numbers are not sustainable, that the Eurasian population must have been at all times at least several hundreds, what means more than just one initial hybrid. So we need a better scenario that accounts for that.
"If the sunda land was emerged for longer how could be different populations between the continent and the Sunda "peninsula"?"
I don't think this is a problem: even today we have different populations: there are no residual Negritos North of the Kraa isthmus, even if the isthmus was a lot thicker in the past. Indochina and Sundaland were surely distinct provinces to some extent, enough to allow for a window of differentiation.
However the admixture may have happened even farther East, in Flores or elsewhere in Wallacea... We do not know enough to say with any certainty.
"About the Karitiana people, I don't understand which is the matter. I thinky is likely a random factor"...
Probably it is but I want to know more anyhow: I want more and more genomes compared with Neanderthals and Denisovans so we can know better if there are regional or individual differences and how do they scatter. Native Americans anyhow have the largest levels of X-DNA haplotype B006 which has very good chances of being Neanderthal. So it must be some other kind of randomness surely.
Ooke! Great! Thanks for all the information! I didn't know about B006! If it was neanderthal it's a paradox that karitiana had such a small amount of archaic DNA!
ReplyDeleteI was thinking today that if "denisovians" ( H. erecrus soloensis) was living in the tropic for a long time sure they had pigmented skin, maybe melanesians could get that fenotyp from the hybridization! In this way the coastal migration wouldn't be necessary!!!
But I guess that if that was the reason we could see it in the skin color haplotyps ( if they were diferent from the african ones). Moreover now I'm thinking that in India also there is very dark people!
Indeed, Adriá. We still have to know with more detail how Neanderthal genes are distributed through humankind. Only a few individuals have been compared in fact.
ReplyDeleteBut I guess that we can think that there was a lot of founder effects and some drift too in all this, so some variance is just naturally random.
...
I think that the persistence of dark ("black") skin colors in several tropical populations is only natural. What it would support is my idea that they did not directly mix with "Denisovans" (living in Siberia, they would have need to achieve whiter skin, and, once achieved, it's not easy to regain black color, as Native Americans demonstrate) but with a tropical relative, possibly H. erectus soloensis.
Nothing else does. Hence the Denisovan mtDNA, found in two different individuals (a finger and a tooth actually) must be that of Asian H. erectus.
ReplyDeleteHowever the Denisovan nuclear DNA is not so distant from Neanderthals. What does it mean?
That H. neanderthalensis is just an European variation of H. erectus.
It's impressive how in these discussion there's always this "essentialist" typology ghost, a somewhat unconscious assumption of a nearly ex-nihilo creation of these types.
if the ancestors of these peoples would have been in Siberia for any extended period, they would have lost their tropical pigmentation for sure because otherwise they would not be getting enough vitamin D and their children would be extremely unfit for that reason (retarded, schizophrenic, rickety, etc.)
Ummm. This fact does not support the hypothetical existence of eskimos. All theories that assume the existence of eskimos must be reviewed. Or perhaps these theories on eskimo existence may be saved by the ad hoc assumption that they would be all retarded or very unfit, if they exist at all.
Anton: that is not what we get from mtDNA nor from archaeology as far as I know (unless you talk of H. erectus senso lato and not senso stricto).
ReplyDeleteArchaeology supports a second Out-of-Africa wave after that of Georgicus/Erectus with Olduwayan: one of H. ergaster with Acheulean. There is some consensus, at least in Spain, around this: Neanderthal appears to have evolved from this second OoA. No other clear OoA happened until that of H. sapiens, which would be the third one of the genus Homo.
Neanderthal autosomal genetics does not seem to support any strong admixture with third species. In fact, depending on whom you read, the divergence between our sibling species would have happened even much later than the c. million years required with the "Atapuerca model" (just made up the name). This other model is either based on some peculiar anthropometric interpretations (Trinkaus) or some slippery DNA 'molecular clock' hunches with a clearly wrong Homo-Pan divergence age (Paabo, Green, etc.) I disregard these models because of obvious inconsistencies but still they evidence that there is NO support for Neanderthal showing any marked hybridization with any third species like H. erectus. This, if anything, would have placed them further from us in the reconstructed (ML, NJ) phylogeny.
The only ex-nihilo ('out of the blue' for those who know no Latin) creation here is "Denisovan". Before these DNA findings there was no any such hominin taxon. So attempting to fit them within the known ones only makes total sense.
...
As for Inuits, they do not have anything like Tropical (black) pigmentation. They have very typical East Asian depigmented skins. East Asian depigmentation patterns are genetically and phenotypically different from those of West Eurasians and are comparatively ill understood... but they are not black in any case.
This loss of pigmentation is even more obvious in Tropical Native Americans, who have not been able in many thousand years to regain anything like the Tropical dark skin we find in the Old World Tropics.
Still the pattern of depigmentation is not just driven by latitude but also by two other quite factors: (1) cloudiness (clearly a major factor in West Europe) and (2) population density (again a major factor in Europe: no other region worldwide is so northernly warm, except a narrow coastal strip in NW America).
"So the ancestors of Melanesians and other very dark tropical Asians have definitively not lived in Siberia at any time. Besides, it is totally non-parsimonious in what regards to modern human mtDNA and Y-DNA spread, the tropical route is much more logical and natural."
ReplyDelete=> How do you explain why majoritary of Ainu, an ethnic group of Siberia, as well as Japanese and Ryukyans, belong to the andamanese haplogroup D ?
Pigmentation is NOT related to patrilineages. People in Tropical Africa and Europe share patrilineages like R1b or E1b and they show no particular resemblance, certainly not in pigmentation.
DeleteOne reason is that DNA overall comes from many ancestors and NOT only the father. In fact, the further back you go in time the less that the patrilineage can explain in terms of personal ancestry. In a couple of centuries or so the proportion of ancestry related to the Y-DNA becomes less than 1%: truly irrelevant.
So maybe I had a blacksmith patrilineal ancestor 180 years ago when the Carlist Wars but of all his DNA I only retain a 2% or so (roughly the same as from Neanderthals). The alleged 18th century corsair that family tradition tells of as an even older patrilineal ancestor only gave me <1% DNA.
Of course there is an statistical "redundancy" effect within populations, so the DNA inside such groups is slightly more similar than when compared to outside groups but, even that it's pretty much irrelevant because most of the variance is at individual level and NOT among groups, not even among "continental" groups or "races".
Even the most endogamous group keeps a lot of internal diversity and even the most distant humans are pretty much the same thing.
These macro-lineages only indicate a common patrilineal ancestor many many millennia ago, maybe 20-30,000 years (or whatever) for the Euro-African example very possibly a much greater figure for your haplogroup D example (not less than 60 Ka IMO, probably more). Notice that Y-DNA D is not an "Ainu" trait but actually an all Japanese trait, and also an all Tibetan one (the lineage seems to come from the near-Tibet area, maybe Yunnan). Half Japanese (who would never consider themselves to be Ainu) carry lineage D, a similar frequency happens with Tibetans (but different subclades).
(continues with the issue of phenotypes)
Probably the most important issue in your quite complex question is understanding the issue of pigmentation in humans. Originally we were all "black" because our origin in Tropical Africa requires that we protect not just our skin but possibly other stuff inside it (keeping folate, vit. B9, in good shape under the UV rays' "aggression" has been proposed as yet another reason for dark skin, which is anyhow common among our closest non-human relatives). But our skin also performs another crucial role: synthesizing vitamin D, which is not just essential for bone development but may most importantly for BRAIN development, especially in childhood.
DeleteThe only alternative source of vitamin D are fish and that was not always available to our ancestors. So, when people began moving outside the tropics, they needed to lighten their skin in order to keep their young healthier (and therefore fitter). We can already see some of that even among Bushmen (San) who have lived for more than 100,000 years in subtropical areas of Africa but the really dramatic change began when people moved into Central and Northern China (what began some 100 Ka ago as well) and later when other group of people began colonizing West and Central Eurasia (since >50 Ka ago). Both groups decreased their levels of pigmentation, sometimes quite dramatically, although they did in a differential manner (each population followed their own adaptive path to "whiteness").
So the ancestors of the Japanese, Ainu, Tibetans (etcetera, it's not just about hg D) were subject to increased adaptive pressure to lose pigmentation. The same happened later (but with an even more dramatic effect at the extreme) in Western Eurasia. Both populations were subject to very intense evolutive pressures to lose or at the very least loosen their originally "black" (brown) pigmentation. The Andamanese, who remained in the Tropics, did not suffer that pressure instead (but the opposite if anything), so they retain the dark color.
We must understand that the color of skin is, in humans, the trait that has suffered the greatest evolutionary pressure by far: survival and health of the new generations (adults are relatively impervious) was at the stake. We must also understand that there are dozens of millennia of evolution behind it and that there are at least two different evolutionary paths to lighter skin colors (oriental and occidental), with different mutations implicated (but producing similar results: convergent evolution).
If you are interesting in more details, I'd suggest to look at the categories "pigmentation" and "vitamin D" in this blog and its predecessor:
→ http://leherensuge.blogspot.com/search/label/human%20pigmentation
→ http://leherensuge.blogspot.com/search/label/vitamin%20D
→ http://forwhattheywereweare.blogspot.com/search/label/pigmentation
"So the ancestors of Melanesians and other very dark tropical Asians have definitively not lived in Siberia at any time. Besides, it is totally non-parsimonious in what regards to modern human mtDNA and Y-DNA spread, the tropical route is much more logical and natural."
ReplyDeleteHow can you explain why majority of Ainu of Siberia and some Japanese and Ryukyuans (due to Siberian Jomon admixture) share Y-DNA haplogroup D (also shared by Andamanese people, who are australoid) on the paternal side, but common siberian haplogroups on the mother side ? Is it possible that partly depigmented australoids reached Japanese archipelago and, furthermore, siberia ?