April 22, 2011

Haplogroup O3 downstream structure refined

A new paper has revised some details of the phylogeny of Y-DNA haplogroup O, specially of O3:

Shi Yan et al.,  An updated tree of Y-chromosome Haplogroup O and revised phylogenetic positions of mutations P164 and PK4. European Journal of Human Genetics, 2011. Pay per view. [doi: 10.1038/ejhg.2011.64]

A copy of the paper can be read (for a year) here.

Most importantly, three subclades (O3a1, O3a2 and O3a4 prior to this review) have been joined in a single subclade of O3a that gets the name O3a1, defined by newly described mutations L127, KL1 and KL2.

Former O3a3, the most common lineage in China, is now relabeled O3a2 (per the paper's proposal) and subclade O3a3b2 (P164) goes up in the hierarchy as O3a2c, absorbing its "uncle" (O3a3c - M134) as a subclade (now O3a2c1) and becoming independent from its former "father" lineage O3a3b (M7), which retains the name.

Another subclade moved up in the phylogenetic hierarchy is former O2a1a (PK4), which is now revealed to be ancestral (and not descendant) to former O2a (M95). 

Figure 1 should help understanding these changes:

click to expand

The three regions (East, North, South) refer to regions of China (rough references: Shanghai, Beijing and Guangzhou respectively, I guess). Most haplogroups are quite evenly distributed but some are only (or almost only) found in the South. These are (by the new nomenclature):  O1a2, O2a*, O2b and O3a2a. Inversely O3a1a is not found in the South (dominant in the East instead). Notice that because of the samples being taken from students at a Shanghai university, East China is oversampled.

A discussion in Chinese can be found also in this Forum of Molecular Anthropology. Thanks to Natsuya for the info.

59 comments:

  1. From the paper:

    ‘East’ refers to the samples whose origins are from the provinces of Jiangsu, Anhui, Zhejiang, and Shanghai, whereas ‘North’ and ‘South’ refers to the other provinces of which the capitals locate northern or southern to the Line of Qinling Mountains-Huai River, respectively.

    ReplyDelete
  2. Hey Maju, I think you may have mistaken P164 for M162? M162 should be O3a2c1a1 now. (BTW, I'm Natsuya.)

    ReplyDelete
  3. Thanks extremely for that Maju, especially the diagram. I've been hoping for some time that someone would sort out Y-hap O as it's been obvious that it hasn't made sense. I've got some work to do matching up the distributions of the old names with the new. But meanwhile:

    "some are only (or almost only) found in the South. These are (by the new nomenclature): O1a2, O2a*, O2b and O3a2a".

    O1a2 is associated with the Austonesians and Taiwanese people so it's not surprising it's present to some extent in Southern China. Its brother haplogroup, O1a1, appears to be an east coast Chinese haplogroup, which again makes sense.

    O2a has always been considered a southern branch of O2, but I can't see where the Japanese O2b fits in here, although we have always known some O2b reached SE Asia at some time. O2b seems quite uncommon in China from this diagram.

    O3a2a is very much a downstream mutation and it's only just present in South China, 1.5%. The general distribution of O3 tends to argue against a Southeast Asian origin for O3 as a whole.

    "Inversely O3a1a is not found in the South (dominant in the East instead"

    which means we still have work to do to find where O3a separated into O3a1'2'4 and O3a3.

    ReplyDelete
  4. "I think you may have mistaken P164 for M162? M162 should be O3a2c1a1 now".

    Right. On second thought maybe I should have been more succinct in commenting the changes, because it's a bit of a labyrinth.

    I corrected that. Thanks.

    ReplyDelete
  5. "‘East’ refers to the samples whose origins are from the provinces of Jiangsu, Anhui, Zhejiang, and Shanghai".

    That's a smaller area of what Wikipedia calls East China, but I imagine it is a matter of who choses the definition. In any case, it's the area around Shanghai.

    How can you know if one or another definition applies?

    ReplyDelete
  6. @Terry:

    I remind you that the (so far unchallenged) paper proposing a southernly origin of O3 is Hong Shi 2005.

    In any case there is no sublineage that I can spot showing any sort of preference for the North. The Eastern region, specially by the Wikipedia definition, is rather southernly and anyhow should be related to coastal migration patterns (which you also stubbornly reject, AFAIK).

    It's not just O3 and O in general, all Y-DNA haplogroups in Eastern Asia show at some level that same structure indicating quite strongly a southernly origin, often via the coast. There may be some exception but I cannot think of a single one.

    "but I can't see where the Japanese O2b fits in here"...

    Sorpresas te da la vida... (life surprises you).

    I can see how it fits indeed.

    ReplyDelete
  7. "some are only (or almost only) found in the South. These are (by the new nomenclature): O1a2, O2a*, O2b and O3a2a".

    The subsamples for these hgs are so small that a Southern trend for them with this data alone is probably not statistically significant in these cases compared to a null hypothesis of homogeneous hg distribution by region with all of the deviation from a homogeneous hg distribution being due to sampling error. The n's for the hgs in question are:

    O1a2 = 3 (1 East, 2 South)
    O2a* = 5 (1 East, 1 North, 3 South)
    O2b = 1 (1 South)
    O3a2a = 1 (1 South)

    The total sample size is 361, with the sample in the East having n=167, the North having n=129, and the South having n=65.

    Four of the described hgs aren't even present in the sample at all.

    In the case of O1a2, the data from outside the scope of the paper re Austronesians from terryt would seem to support a Southern affiliation. But, the fact that a single example of a rare hg ends up in one region rather than another in the sample isn't very meaningful. The odds that at least one of those is due simply to random chance is very high.

    ReplyDelete
  8. Muy interesante tu trabajo, me gustaría saber tu opinión sobre la población balear, no sé si has publicado algo antes. Gracias por anticipado.

    ReplyDelete
  9. @Andrew: true that the sample is small and not systematic. We cannot conclude much from this one paper. But in any case notice that the most undersampled group is the Southern one.

    As I said above, the support for a southern origin of Y-DNA O3 is not to be found here but in Hong Shi 2005.

    ReplyDelete
  10. @Santiago:

    [It's an off-topic reply on Baleric islands]

    No creo que hay mucho material re. las Baleares específicamente. Mira en los artículos bajo etiqueta Iberia en este blog y en mi viejo blog Leherensuge.

    En general, yo diría que los Baleares, especialmente los menorquines, son próximos a los Catalanes pero quizá algo más "Neolíticos" (o "Mediterráneos"). Los ibicencos, que tienen una prehistoria muy diferente son un grupo aparte, con gran aportación "Mediterránea" y marcados efectos fundacionales.

    En particular yo recomendaría estas entradas (aunque son en general sobre Iberia):

    - Guest post by Argiedude: A West-East Y-DNA gradient in Iberia?
    - Neolithic spread in Iberia, the Y-DNA perspective.
    - Iberian Y-DNA.

    Por desgracia ninguno de los estudios sobre los que versan estas entradas tratan el Mediterráneo Occidental en una perspectiva más amplia. Cuando lo hacen, Baleares no es normalmente su foco de interés. Lo menciono porque no me extrañaría que las Baleares tuvieran conexiones con Occitania, Córcega y Cerdeña, además de la Península.

    ReplyDelete
  11. "there is no sublineage that I can spot showing any sort of preference for the North".

    And you wrote:

    "I can see how it fits indeed".

    So can I, now that I've printed your diagram off. It's O2* that is northern. Presumably O2b proper is confined to regions north of China, including Japan and Korea. O2a moved south.

    "The Eastern region, specially by the Wikipedia definition, is rather southernly and anyhow should be related to coastal migration patterns (which you also stubbornly reject, AFAIK)".

    I certainly don't reject it for movement along the east coast of Eurasia. In fact I'm sure that Y-hap NO originally moved north from SE Asia along that coast. As probably did C originally, but in the other direction. What I have difficulty accepting is a movement along the coast all the way from the Horn of Africa to SE Asia.

    "I remind you that the (so far unchallenged) paper proposing a southernly origin of O3 is Hong Shi 2005".

    But that doesn't fit the evidence of the various clades of O3. As far as I'm aware O3a1 (now O3a1a) is still Northern China, although just 3 examples of O3a1 are shown in this study. And a recent paper claimed one C3a clade was associated with Sino-Tibetan-speaking people (perhaps O3a2c1* in the new nomenclature) and a closely related clade was associated with Hmong-Mien (possibly O3a2b in the new nomenclature). Anyway, as Andrew said, too few examples are represented in this study to really draw any conclusions, however the paper will surely help untangle the expansions of various O haplogroups.

    ReplyDelete
  12. "As I said above, the support for a southern origin of Y-DNA O3 is not to be found here but in Hong Shi 2005".

    And I've argued with you elsewhere why that paper doesn't make sense. For a start the authors are mostly from Kunming and, like Dienekes with his recdent post on the origin of Indo-European, they are keen to show that their region is the origin for the most common Chinese Y-hap. But having looked at the paper again, here we go:

    "It was estimated that the early northward migration of the O3-M122 lineages in East Asia occurred ~25,000–30,000 years ago, consistent with the fossil records of modern humans in East Asia".

    If the expansion of O3 was really that ancient we would not find the identical clades so widely spread. O3 would have broken into regional clades by now. The current paper certainly argues against such ancient regional diversification. The earlier paper even says, later:

    "In general, the distribution of the O3-M122 haplotypes did not show distinctive divergence between southern and northern populations, with all the major subhaplogroups shared between them—except for O3-M7, which was observed only in the southern populations and therefore indicates a recent common ancestry of the O3-M122 lineage in East Asia".

    Their own comment suggests a relatively recent expansion surely. And:

    "there was a lot of similar STR evolution after the emergence of O3-M122, and many shared STR haplotypes were observed between northern and southern populations, again confirmation of the recent common ancestry of the M122 lineage in East Asia".

    So they again contradict their earlier statement.

    "Previous studies have shown that O2-M95 and O1-M119 are prevalent in SEAS and probably originated in the south"

    O2-M95? No-one argues anything else for M95, but that conveniently ignores O2b-M176, which is basically northern. O2 is O2-P31.

    "Most of the populations sampled were from southern and southwestern China, where ~80% of the Chinese ethnic populations live"

    Hmmm. Even I can see that would skew their results.

    ReplyDelete
  13. "The Yangtze River was used as the geographic border to separate the SEAS and NEAS".

    Precisely the region where I would guess the homeland of Y-hap O lies. And they continue:

    "In the SEAS, there are 14 Tibeto-Burman–speaking populations with a recorded history of migration from northern China ~3,000 years ago"

    I know you completely disagree with that statement. And this one:

    "In the MDS map, the Hmong-Mien populations were clustered closely with Han populations, which reflects the recorded history of admixture"

    And this:

    "It has been well documented that the Tibeto-Burman populations living in southwestern China were originally, during the late Neolithic period, from the north, but they have been under extensive influence from the southern ethnic groups, including Daic- and Austro-Asiatic–speaking populations"

    I think the evidence is pretty strong that the Daic- and Austro-Asiatic–speaking populations have also been influenced from the north. So, to then ensure that they get the result they desire:

    "To remove the influence of relatively recent population admixture, we constructed the STR network excluding the Tibeto-Burman, Altaic, Hmong-Mien, and southern Han populations".

    Then they use a convoluted argument to support their idea of a southern origin:

    "It should be noted that the lack of genetic divergence, in view of the gene diversity between the southern and northern O3-M122 lineages, indicates that the O3-M122 lineages were probably dominant in the population involved in the initial northward migration; therefore, no obvious bottleneck occurred for the O3-M122 lineage, in contrast with the skewed distribution of the O2-M95 and O1-M119 lineages (Su et al. 1999; Wen et al. 2004b). However, recent gene flows due to the expansion of Han culture could also have contributed partly to the homogeneity of the O3-M122 lineage"

    In fact I guess that the homogeneity of the O3-M122 lineage is a product of an expansion going back well beyond the Han expansion.

    ReplyDelete
  14. O2b is only found in the South in this sample (1/65). No idea what you're talking about. O2* has higher frequency in the north but that does not mean anything without further info (we do not know the actual diversity under this paragroup label and all three regions have quite a decent frequency of it anyhow).

    "In fact I'm sure that Y-hap NO originally moved north from SE Asia along that coast. As probably did C originally, but in the other direction".

    I have no idea how can you reach such conclusions. I understand that NO(xN,O) is very rare and restricted to South and East China (and some in SW Japan too), so it never reached the North in its undivided form. The rest (and at least part of the previous) path must have been done as separate N and O (or O-number) haplogroups.

    As for C, it is quite clear it must have migrated from South to North somehow. There's no other way it could have reached Australasia.

    Even East-Asia-specific haplogroup D has been determined to have a southern origin.

    "What I have difficulty accepting is a movement along the coast all the way from the Horn of Africa to SE Asia".

    I do not. It's just a prejudice you have. If H. ergaster could cross Gibraltar Strait and various hominins are found in islands from Crete to Flores and Luzon, I see absolutely no reason for H. sapiens, which was then already the same as we are, not to be able to navigate along the coasts of Asia from Africa.

    It's absurd to think that they could cross tens of thousands of kilometers to Australia and could not navigate along the coast. You are obsessed with that prejudice, as if our ancestors needed to learn the art of making rafts from H. floresiensis or something. Nonsense!

    "As far as I'm aware O3a1 (now O3a1a) is still Northern China"...

    In Hong Shi 2005 this clade is only mentioned in one Cambodian and one "Han Zibo" (H4, Shandong). But the latter is 1/98 (1%) and the former is 1/14 (7%). So not sure what you are talking about.

    "And I've argued with you elsewhere why that paper doesn't make sense".

    And I've explained you once and again why it does, no matter how you look at it. Should we re-start it all over again?

    "For a start the authors are mostly from Kunming and (...) they are keen to show that their region is the origin for the most common Chinese Y-hap".

    Some of the authors work at Kunming University but that probably only directed their interest and helped them in availability of sampling Yunnan, which is indeed widely surveyed in that paper, what allows for a finer characterization of the diversity in the area that otherwise may have gone unnoticed.

    ...

    ReplyDelete
  15. ...

    "If the expansion of O3 was really that ancient"...

    25-30,000 is not "ancient". In fact I think it's an understimate by maybe as much as half. 50 Ka is well within the realistic possibilities IMO (for a c. 70 Ka arrival of H. sapiens to the region, even older dates are possible too).

    "... we would not find the identical clades so widely spread".

    It is very possible that the population of East Asia (specially China), and South Asia too, was much much larger in the Paleolithic than that of Europe or even all West Eurasia together.

    Without small population sizes, there is very little drift and the genetic composition could well be to a large extent "fossilized".

    Complementarily, there is a lot of substructure to be unraveled, much like talking of R1b (or even R1b1b2) and E1b (or even E1b1b1) hides a lot of the important substructure under these categories and leads people to wrong conclusions.

    "... conveniently ignores O2b-M176, which is basically northern".

    Maybe but I'm quite intrigued at the instances of O2b in southern populations, which, if studied, may end up indicating a southern origin for this clade as well.

    Not all or even anything at all is frequency. Again the example of R1b illustrates how a region of quite low frequency such as West Asia is the likely origin of the haplogroup, while a region of extremely high frequency as West Europe can only be considered originary for much more derived sublineages. Frequency may be totally misleading.

    "So, to then ensure that they get the result they desire:

    "To remove the influence of relatively recent population admixture, we constructed the STR network excluding the Tibeto-Burman, Altaic, Hmong-Mien, and southern Han populations"".

    That's just a precaution and you get the results with both sets in fig. 6: (a) with all populations and (b) with the cleared set.

    So why if you already know that, you insist in denouncing what is so transparently offered to you in both versions. In the uncropped fig. 6a, you still get southern ancestral nodes for all the clades, even if the structure is more complex (logically) and in some cases these root nodes belong to Southern Han/Sino-Tibetan (in green).

    You're just being grumpy without any justification: you just want your preliminary idea (of Northern origins) to succeed against all evidence and for that purpose you will charge once and again no matter what. You will repeat failed argumentations like the one I just refuted for n-th time. You will just entrench yourself and pay no attention to reason.

    Tiresome.

    ReplyDelete
  16. Maju said,

    "I have no idea how can you reach such conclusions. I understand that NO(xN,O) is very rare and restricted to South and East China (and some in SW Japan too), so it never reached the North in its undivided form. The rest (and at least part of the previous) path must have been done as separate N and O (or O-number) haplogroups."

    As far as I know, no confirmed example of haplogroup NO-M214(xN-M231, O-M175) has ever been reported in the literature. Haplogroup N*-M231(xN1-LLY22g) has been reported as singletons in a few Han Chinese
    samples from various regions of China, and it seems likely that the known Japanese cases of NO-M214(xN1-LLY22g, O-M175) also should belong to N*-M231(xN1-LLY22g).

    ReplyDelete
  17. I was surely mistaking one and the other. I stand corrected then, thanks.

    ReplyDelete
  18. "'North' and 'South' refers to the other provinces of which the capitals locate northern or southern to the Line of Qinling Mountains-Huai River, respectively".

    So 'South China' in this paper does not mean just Yunan, Guizhou and Guangxi. It includes the whole Yangtze River catchment, a major region of the Chinese Neolithic.

    "O2b is only found in the South in this sample (1/65). No idea what you're talking about".

    Of course you wouldn't. I wasn't talking about this paper. I was taking all previous studies into account.

    "O2* has higher frequency in the north but that does not mean anything without further info"

    Here is further info:

    http://en.wikipedia.org/wiki/Haplogroup_O2_(Y-DNA)

    Quote:

    "Haplogroup O2 is also notable for the fact that it can be divided into two major subclades ... One of these subclades, Haplogroup O2a (M95), is found among some (mostly tribal) populations of South and Southeast Asia, as well as among the Khmers of Cambodia and the Balinese of Indonesia ... The other major subclade, Haplogroup O2b (SRY465, M176), is found almost exclusively among the Korean, Japanese, Thai, Vietnamese and Indonesian".

    "Maybe but I'm quite intrigued at the instances of O2b in southern populations, which, if studied, may end up indicating a southern origin for this clade as well"

    More info:

    http://en.wikipedia.org/wiki/Haplogroup_O2b_(Y-DNA)

    Quote:

    "Haplogroup O2b is found mainly in the northernmost parts of East Asia, from the Uriankhai and Zakhchin peoples of western Mongolia[1] to the Japanese of Japan, though it also has been detected sporadically in the Buryats[3] and Udegeys[12] of southern Siberia, very rarely among populations of Southeast Asia (including Indonesia,[8][3] the Philippines,[3] Thailand,[3] and Vietnam[8][3]), and Micronesians".

    Note: 'very rarely among populations of Southeast Asia'.

    (continued)

    ReplyDelete
  19. (continued)

    "And I've explained you once and again why it does, no matter how you look at it. Should we re-start it all over again?"

    No. Because you are obviously not interesting in discovering what actually happened.

    "25-30,000 is not 'ancient'. In fact I think it's an understimate by maybe as much as half. 50 Ka is well within the realistic possibilities IMO (for a c. 70 Ka arrival of H. sapiens to the region, even older dates are possible too)".

    There is no way that O's main expansion is anywhere near even 25-30,000 years old.

    "It is very possible that the population of East Asia (specially China), and South Asia too, was much much larger in the Paleolithic than that of Europe or even all West Eurasia together".

    Extremely unlikely. Especially considering that any evidence for the region is so sparse.

    "Complementarily, there is a lot of substructure to be unraveled, much like talking of R1b (or even R1b1b2) and E1b (or even E1b1b1) hides a lot of the important substructure under these categories and leads people to wrong conclusions".

    Some of that substructure is easily visible in your diagram. I know you are very opposed to the idea that the Han expansion was anything more than just cultural, but two haplogroups have all the characteristics of rapid, recent, widespread expansion: O3a1c and o3a2c (as named in the recent paper). Both are fairly evenly spread through all three regions. For example O3a2c as a whole makes up 32.4% of 'East China', 34.1% of 'North China' and 21.5% of 'South China'. Together the two haplogroups O3a1c and O3a2c make up 44.6% of the Chinese Y-haps in the study, both evenly distributed through the three regions.

    If Both O3a1c and O3a2c are Han Chinese in origin, and it's difficult to see them as anything else, they moved south, not north. Whether they are the northern branches of O3a1 and O3a2 respectively remains to be discovered, but to me it looks as though at least O3a1 is northern. And although O3a2b is present especially in the Daxi and Hmong-Mien it is quite likely these groups are members of a pre-Han southerly movement.

    ReplyDelete
  20. "So 'South China' in this paper does not mean just Yunan, Guizhou and Guangxi. It includes the whole Yangtze River catchment, a major region of the Chinese Neolithic".

    In this paper and in nearly all contexts, yes. All that southern half of China where rice is the main crop as enjoys subtropical climate. All that southern half of China that was only recently (in historical times) recycled into Han identity.

    As for O2b and O2* you don't clarify anything at all.

    "There is no way that O's main expansion is anywhere near even 25-30,000 years old".

    Why not? From mtDNA I estimate that the pan-Eurasian flows (excepting some lesser ones via Siberia) were cut almost completely after the expansion of mtDNA R, which (I believe you agree) should be coincident with the expansion of Y-DNA MNOPS, both from SE Asia (between Shanghai and Bali). The expansion of MNOPS in East Asia is the same as the expansion of NO, which in China and SE Asia is the same as the expansion of O: there may be some time between these various steps but should not be much in any case or otherwise we would not see such hegemony of O and such lack of MNOPS or NO basal diversity (we would have instead lots of NO-number groups here and there - but, as Ebizur explained, there's no NO* anywhere).

    The expansion of mtDNA R cannot be more recent than c. 50 Ka ago, because R is super-important in the colonization of West Eurasia since that date. So the expansion of MNOPS cannot be more recent either, with Y-DNA R or R1 (or R1b maybe) having an age of c. 50-40 Ka necessarily. So that is roughly the age of NO and O. The exact age of the lineages under O I cannot estimate because I lack clear archaeological references but should not be much more recent than this date of c. 50-40 Ka estimated for their West Eurasian cousins.

    It is in any case a very reasonable estimate based on what we know combinedly of archaeology and genetics.

    As for the estimate of East Asia and South Asia having large populations in the Paleolithic this has been inferred from modern genetic diversity (I can't find the paper right now but I'm sure you know which I'm talking about). This is very logical anyhow: we would just not find so many different (and high in the phylogenetic tree) mtDNA haplogroups in these regions would not be because of early "superpopulation" (for Paleolithic standards). Where West Eurasia has some five basal M or N sub lineages, America has five too, Australia has also five or a bit more (not counting Melanesia, which also has a few of its own)... East Asia and South Asia are each by the many dozens, almost ten times these other "peripheral" regions. So these two regions sustained large (and therefore genetically diverse) populations since very early on.

    ...

    ReplyDelete
  21. ...

    "I know you are very opposed to the idea that the Han expansion was anything more than just cultural"...

    I would not put it that way. Localized colonizations surely happened, specially partial colonizations and sustained flows (not necessarily from the Han core area though). But the bulk of the sinization process was no doubt political (China as state) and cultural (even today it is language and some cultural elements what make you Han or not: China is an inclusive assimilative nation much like France is - for good or bad). This is specially true as we are talking of a process that happened long after Neolithic was established in all the region (so populations were dense, consolidated and deeply rooted in their respective areas).

    "... two haplogroups have all the characteristics of rapid, recent, widespread expansion: O3a1c and O3a2c (as named in the recent paper)".

    I was going to say that maybe you were onto something in the O3a2c case (but only in this case, O3a1c is more frequent in the South if anything) but when we look at M134 and M117 in Hong Shi 2005, they look as spawned from the South, so in fact O3a2c seems to represent a northwards colonization and not a southwards one.

    That is what the data says. If you looked at the haplotype structure trees instead than to mere frequency data you'd know.

    Hong Shi did not yet know of the defining mutation of O3a1c, so this clade is not pondered separately in that paper but I imagine it is just like the rest: they represent a northward flow.

    ReplyDelete
  22. "O3a1c is more frequent in the South if anything"

    In your diagram you have O3a1c 16.2%, 17.1% and 18.5% of the haplogroups in East, North and South China respectively. So it is only just 'more frequent in the South'. Near enough to evenly spread, I'd say.

    "This is specially true as we are talking of a process that happened long after Neolithic was established in all the region (so populations were dense, consolidated and deeply rooted in their respective areas)".

    But the two haplogroups' spread presumably became established during the Neolithic, long before the Han political hegemony was begun. And the evidence shows that O3a1c and O3a2c are by no means 'deeply rooted in their respective areas'. They are widespread.

    "when we look at M134 and M117 in Hong Shi 2005, they look as spawned from the South, so in fact O3a2c seems to represent a northwards colonization and not a southwards one".

    But there is no evidence for any such expansion that could explain the homogeneous spread of either haplogroup. Whereas there is overwhelming evidence for at least cultural expansion southwards, starting more recently than 10,000 years ago, which would easily explain the homogeneous spread of both haplogroups.

    "That is what the data says".

    Not so. Data 'says' nothing. It has to be interpreted. You choose to interpret the haplogroup distribution as representing an ancient northwards colonization that suffered no drift, bottlenecks or founder effects. To me that is interpreting the data from a pre-determined viewpoint. As is Hong Shi in the 2005 paper.

    "If you looked at the haplotype structure trees instead than to mere frequency data you'd know".

    So where else in the world is an apparently anciently diversified clade of Y-haps remain spread so evenly through a particular region? Various Y-hap Rs certainly spread fairly evenly through many parts of the world today, but no-one is going to argue an anccient expansion to explain that fact. At no earlier period of its existence has Y-hap R ever had two of its clades spread evenly through a single region as do the two O haplogroups O3a2c and O2a3c today. They have all the characteristics of a relatively recent expansion, certainly the most recent widespread expansion within China.

    ReplyDelete
  23. Sorry. I missed your earlier comment.

    "the expansion of mtDNA R, which (I believe you agree) should be coincident with the expansion of Y-DNA MNOPS, both from SE Asia (between Shanghai and Bali). The expansion of MNOPS in East Asia is the same as the expansion of NO, which in China and SE Asia is the same as the expansion of O"

    Pretty much agree, but your use of 'coincident with' is a bit indiscriminate. The breakup goes MNOPS to M, S, NO and P in SE Asia. Then P breaks into R and Q, probably somewhere in India. NO breaks into N and O somewhere in East Asia. So the expansion of MNOPS in East Asia is NOT the same as the expansion of NO, and NO's expansion is NOT the same as the expansion of O.

    "there may be some time between these various steps"

    Quite some time, possibly.

    "but should not be much in any case or otherwise we would not see such hegemony of O and such lack of MNOPS or NO basal diversity"

    The fact we see such lack of MNOPS or NO basal diversity suggests they were drifted out. Which in turn suggests there was quite some time between the breakup of MNOPS and the spread of O. Time enough for descendant haplogroups to replace their ancestral haplogroups, whether stationary or on the move.

    "As for the estimate of East Asia and South Asia having large populations in the Paleolithic this has been inferred from modern genetic diversity"

    But the present paper shows that that genetic diversity is overlain at some time by an expansion of much lesser genetic diversity.

    ReplyDelete
  24. "In your diagram you have"...

    What an obsession with the undue use of the second person form. It's not "my" diagram but that of Shi Yan and other authors of the paper. I just reproduced it.

    "So it is only just 'more frequent in the South'. Near enough to evenly spread, I'd say".

    Exactly what I said: "O3a1c is more frequent in the South if anything".

    "But the two haplogroups' spread presumably became established during the Neolithic, long before the Han political hegemony was begun".

    If that would be true it'd be impossible, I understand, to attribute their spread to any archaeologically factual culture: the north-to-south Neolithic flow in China (of any sort) is only in your imagination as far as I can tell.

    But I'm not sure what you have in mind, as you seem to think that Guandong or Hunan are "North". As we have clarified above: North China is essentially the Yellow River basin.

    "And the evidence shows that O3a1c and O3a2c are by no means 'deeply rooted in their respective areas'. They are widespread".

    The evidence shows nothing because there is not enough evidence (as I discussed above).

    When I said 'deeply rooted in their respective areas' I meant populations, specifically post-Neolithic populations. It is you who make undue associations with haplogroups.

    If we were to judge by your method, at a phylogenetic depth only some 6 steps under MNOPS, we'd have to agree with Balaresque (and others) when she argues (wrongly) that R1b1b expanded within Neolithic from Turkey to Europe. This is a shallow understanding: what we see in most cases at such phylogenetic levels is always (or almost always) deeply Paleolithic.

    Of course each case should be judged by its own merits but, in any case, the only expansion we can detect in Y-DNA O3 so far is from South to North. Naturally this is just one rather shallow layer of the onion but I cannot adventure what further peeling will reveal exactly (though I imagine that some regionally specific structure will be revealed for your tranquility - and mine).

    "But there is no evidence for any such expansion that could explain the homogeneous spread of either haplogroup".

    Fair enough if you mean "archaeological evidence". There is a great lack of archaeological evidence in most of Asia in general, so we can't really say much about that and we have to rely, sadly enough, on archaeo-genetics, so to say.

    "Whereas there is overwhelming evidence for at least cultural expansion southwards, starting more recently than 10,000 years ago, which would easily explain the homogeneous spread of both haplogroups".

    Not within China - and, as far as I can discern, not either from China southwards into Indochina and the Malay Archipielago.

    In China the only meaningful "north-to-south" process that can be discerned archaeologically is that of the expansion of the Chinese state since the Bronze and specially the Iron Age.

    Further south I fail to see any clear pre-Austronesian flow either. Bellwood himself, in the paper you linked for me recently, clearly argues for demographic continuity in at least parts of Indonesia and I could clearly spot a sickle among his Toalean implements (in Timor), almost direct evidence of cereal collection (and hence probably of rice farming).

    "Data 'says' nothing. It has to be interpreted".

    The data of Hong Shin 2005 has only one possible interpretation as far as I can tell: south-to-north migration of all discernible O3 subhaplogroups.

    ...

    ReplyDelete
  25. ...

    "You choose to interpret the haplogroup distribution"...

    I do not interpret haplogroup distribution. It is only you who seems obsessed with remaining at that shallow level.

    Hong Shi and colleagues interpret haplotype structure and distribution. I agree with method and conclusions, even if the paper is getting a bit old - I hope something better comes soon, now that some better understanding of O phylogeny and context has been achieved.

    "... representing an ancient northwards colonization that suffered no drift, bottlenecks or founder effects".

    The very fact that we do not see O1, or MNOPS(xNO) or almost F(xMNOPS) or C(xC1,C3), that we see some regionally unique clades as NO, C3, C1, D... are founder effects.

    But you do have a point and this seems in fact to address my doubts when faced on how to trace (lacking archaeological evidence or obvious geographic barriers) the "migration routes" north of the mountain zone (Zomia): there are no major barriers between Hanoi and Beijing, hence it makes some sense that the population is so homogeneous in all this area.

    There were probably few or no bottlenecks in this flow: you made an interesting discovery. We can call this therefore the Tothill model of colonization of China - I know you'll hate it, that's why I am suggesting :p

    However there must have been some localization by drift, even if minimal because the population was always large. This part we do not know about because there is not enough resolution. It's like watching one of those low resolution videos, you know: you lose a lot.

    Partly this difficulty may be circumvented by studying the haplotype structures. And that is what Hong Shi et al. did in 2005.

    "So where else in the world is an apparently anciently diversified clade of Y-haps remain spread so evenly through a particular region?"

    I already mentioned haplogroup R1b before recent improvements to resolution. R1a is an even better example because, as happens with O3 subclades, we still have to crack the code. What about Q1a3 in Native America? What about haplogroup E1b1b1? E1b1a? H? When you have low resolution everything looks oddly amorphous.

    "Various Y-hap Rs certainly spread fairly evenly through many parts of the world today, but no-one is going to argue an anccient expansion to explain that fact".

    I have been arguing for that for long. And the more I look into the matter, the more clear it seems to me that it is an ancient expansion.

    "At no earlier period of its existence has Y-hap R ever had two of its clades spread evenly through a single region as do the two O haplogroups O3a2c and O2a3c today".

    But that is because China did not probably experience an LGM or even an original colonization bottleneck. The internal geography of China, between the deserts and steppes of the North and the Zomia mountains, is pretty much boundless.

    This is the Toothill model. Thank yourself for the finding. :)

    ReplyDelete
  26. "So the expansion of MNOPS in East Asia is NOT the same as the expansion of NO"...

    M, S and P all expanded outside of SE Asia and are not found in it (excepting maybe Wallacea). So the expansion of MNOPS in East Asia is that of NO because MNOPS in East Asia is NO and, as far as I know, only that.

    "... and NO's expansion is NOT the same as the expansion of O".

    Not North of Mongolia-Manchuria, where N is actually the main and almost only NO representative. But otherwise it's almost the same, unless you are going to claim something unexpected, like a once numerous N-dominated population somewhere that was replaced (this is the kind of stuff that fills your mind: replacement and more replacement, even when there is no need for it nor evidence of it).

    Only under O we see some geographical structure, which is interesting on its own and IMO can be considered to parallel the NO split, rather than considering the two successive phenomenons (as a simplistic read of the phylogenetic tree would suggest).

    "Quite some time, possibly".

    Or very little. This is hard to discern and, considering the abundance of time in the Paleolithic, not that extremely important anyhow: we have at least 60,000 years to make it fit before Neolithic arrives with its boring stability.

    "The fact we see such lack of MNOPS or NO basal diversity suggests they were drifted out".

    Exactly, what means that this happened before any MNOPS meaningful expansion in East Asia. Once that population began expanding, generalized drift was not anymore an option because they were all the way around.

    So we have what we can call a proto-Oriental population dominated by NO (both as N and as O but with this last one clearly dominating), which split in four (hypothetical) subpopulations, as can be discerned from Y-DNA:
    · Northern (dominated by N)
    · Central Mainland (dominated by O3)
    · Central Coastal (dominated by O2)
    · Southern (dominated by O1)

    Please discuss for clarity of details. It would seem that there was relatively little drift after that.

    "Which in turn suggests there was quite some time between the breakup of MNOPS and the spread of O".

    What says is that the proto-Oriental population was very small then, so drift was very easy. All this probably happened near the early colonization of Tropical Asia.

    ReplyDelete
  27. "As for O2b and O2* you don't clarify anything at all".

    Not for you, of course.

    "If that would be true it'd be impossible, I understand, to attribute their spread to any archaeologically factual culture: the north-to-south Neolithic flow in China (of any sort) is only in your imagination as far as I can tell".

    It has been accepted for years.

    "As we have clarified above: North China is essentially the Yellow River basin".

    And South China is the Yangtze Basin. Both regions were centres of dispersal of the Chinese Neolithic. I've provided this link before but you obviously haven't read it:

    http://en.wikipedia.org/wiki/List_of_Neolithic_cultures_of_China

    "The evidence shows nothing because there is not enough evidence (as I discussed above)".

    They are the two haplogroups for which we actually have the most evidence in the paper.

    "What about haplogroup E1b1b1? E1b1a?"

    The widespread expansion of such downstream haplogroups is almost certainly relatively recent.

    "What about Q1a3 in Native America?"

    As far as I'm aware the Q haplogroups in America are reasonably discretely distributed, unlike the two Chinese haplogroups we're considering here.

    "There were probably few or no bottlenecks in this flow"

    Then you contradict yourself:

    "However there must have been some localization by drift, even if minimal because the population was always large".

    We have absolutely no evidence that 'the population was always large'. And again:

    "What says is that the proto-Oriental population was very small then, so drift was very easy".

    And it remained small until it developed a Neolithic.

    "And the more I look into the matter, the more clear it seems to me that it is an ancient expansion".

    To New Zealand, Australia and North America? Be realistic. Within Europe the various clades of R1b can hardly be called 'evenly distributed'. The only regions where many R1b haplogroups are 'evenly distribute' are the three countries mentioned above.

    "The internal geography of China, between the deserts and steppes of the North and the Zomia mountains, is pretty much boundless".

    No it's not. Have a look at an atlas.

    "M, S and P all expanded outside of SE Asia and are not found in it (excepting maybe Wallacea). So the expansion of MNOPS in East Asia is that of NO because MNOPS in East Asia is NO and, as far as I know, only that".

    Exactly. Not O.

    "Not North of Mongolia-Manchuria, where N is actually the main and almost only NO representative".

    And presumably N first appeared somewhere near Mongolia-Manchuria. So NO must have made it that far.

    "Only under O we see some geographical structure, which is interesting on its own and IMO can be considered to parallel the NO split, rather than considering the two successive phenomenons"

    I certainly do not consider 'two successive phenomenons'. NO diversified somewhere near where N coalesced.

    "Exactly, what means that this happened before any MNOPS meaningful expansion in East Asia".

    Rubbish. It's extremely doubtful that MNOPS entered China. However it is extremely likely that NO entered China.

    "· Northern (dominated by N)
    · Central Mainland (dominated by O3)
    · Central Coastal (dominated by O2)
    · Southern (dominated by O1)"

    I agree with your first two. But you ignore O2b which to me suggests that O2 first coalesced around the mouth of the Yellow river. And O1 somewhere in the southern Yangtze Basin.

    "It would seem that there was relatively little drift after that".

    Perhaps so. But significant later expansion of all three haplogroups.

    ReplyDelete
  28. Here's apost by Ren which covers the subject of southward-moving Mongoloids entering SE Asia based mainly on a dental study:

    http://s6.zetaboards.com/man/topic/527924/1/

    Some comments from it:

    "The results of both the metric and nonmetric analyses demonstrate close affinities between recent Australo-Melanesian samples and samples representing early Southeast Asia, such as the Early to Middle Holocene series from Vietnam, Malaysia, and Flores. In contrast, the dental characteristics of most modern Southeast Asians exhibit a mixture of traits associated with East Asians and Australo-Melanesians, suggesting that these populations were genetically influenced by immigrants from East Asia. East Asian metric and/or nonmetric traits are also found in some prehistoric samples from Southeast Asia such as Ban Kao (Thailand), implying that immigration probably began in the early Neolithic".

    "The results of metric and nonmetric dental analyses indicate that Hoabinhian Malay series such as those represented by Gua Cha and Guar Kepah, as well as the prehistoric Vietnamese from Bac Son and Da But cultural contexts, are the most differentiated from East Asians. These populations likely represent indigenous people who descended from Late Pleistocene occupants of Sundaland, groups that are closely related to recent Australo-Melanesians and Andaman Islanders. In contrast, our results suggest that the remaining Southeast Asians compared in this study were genetically influenced by East Asians, i.e., primarily Neolithic and later agricultural populations from southern China. Metric traits suggesting gene flow from the north are also found in some prehistoric inhabitants of Southeast Asia such as those buried at Ban Kao in Thailand, implying that the initial wave of immigrants probably occurred in the Middle Holocene. These immigrant populations intermixed with the indigenous Australo-Melanesians as they expanded. Clear evidence for East Asian immigration was found in the early Metal Age Vietnamese and Sulawesi groups".

    "The results of the present study support the two-layer immigration hypothesis for the population history of Southeast Asia. This hypothesis, articulated most clearly in recent years by Bellwood ([1987], [1996], [1997]), proposes that there was a diffusion of agricultural populations from southern East Asia into Southeast Asia beginning in the Neolithic period. Although in the data analyzed here, the genetic influence of East Asian immigrants seems to be stronger in the post-Neolithic rather than the Neolithic period, the present study does not take into account sufficient data from early Neolithic Southeast Asia. Further research on skeletal remains from that period is required to determine the exact timing of prehistoric population dispersals into Southeast Asia".

    ReplyDelete
  29. And this 2007 paper has some interesting diagrams showing the variation in East Asia:

    http://www.jstage.jst.go.jp/article/ase/116/2/135/_pdf

    None of the paper's authors are not Chinese, so they do not have the wish to deny any southerly movement of Mongoloid phenotype.

    ReplyDelete
  30. "It has been accepted for years".

    No. The first widely accepted Neolithic culture of China (and all East Asia) is from the South!!!

    "Both regions were centres of dispersal of the Chinese Neolithic".

    BOTH! So there it goes your silly idea of a N->S flow of any sort.

    "The widespread expansion of [E1b1b1, E1b1a] is almost certainly relatively recent".

    Totally disagree. Go pay conceptual vassalage to Dienekes.

    "As far as I'm aware the Q haplogroups in America are reasonably discretely distributed"...

    Illustrate the matter. Because as far as I am aware that is not the case.

    "We have absolutely no evidence that 'the population was always large'".

    We do, there is strong genetic evidence. "Always" is of course a relative concept but since rather soon after the first colonization in any case.

    "No it's not. Have a look at an atlas".

    Yes it is: look at that atlas yourself: only rivers are "obstacles" between Hanoi and Beijing. But rivers are never true obstacles (rather avenues) so...

    "And presumably N first appeared somewhere near Mongolia-Manchuria".

    Probably not. Probably it coalesced in South China and migrated to Mongolia from there, already formed (but this is not fully discerned).

    "Here's apost by Ren which covers the subject of southward-moving Mongoloids entering SE Asia based mainly on a dental study"...

    That refers to Indonesia and Indochina, not South China. I do not have to agree with it anyhow, but we are here discussing North-South within modern China!!!

    Anyhow, your argumentation is purely morphometric, which is a slippery slope towards nonsense. While anthropometry may at times inform us somewhat, it is anything but conclusive and highly subject to interpretation.

    ReplyDelete
  31. As for the Matsamura paper (which is at least openly accessible), it only says that most Man Bac people were modern-looking (with one individual anomaly, who does not look "Negrito" nor "Papuan" to me but rather "true Australoid" - but whatever because Australian aborigines don't look themselves like Paleolithic skulls of Australia, so it's a total mess).

    The paper anyhow only deals with a single site in a single period in North Vietnam, what is alone of limited informative value, but in the conclusions section makes a review of the literature. After mentioning Bellwood and other apologists of the "two layer theory", they also mention Turner and others who oppose it.

    "Turner (1987, 1989, 1990, 1992) assumed that the array of nonmetric dental traits, so called ‘Sundadont’ traits, possessed by present-day Southeast Asians, are the product of long-standing continuity uninterrupted by significant admixture with Sinodont’ peoples from the north".

    So, erm, modern SE Asians are "Sundadont"? How curious if they have been recently replaced by "Sinodont" Neolithic immigrants from China or even beyond!

    "Multivariate craniometric analyses by Pietrusewsky (1992, 1994, 1999) and Hanihara (1992a, b, 1993a, b, c, 1994) have demonstrated relatively close affinities between prehistoric and modern Southeast Asians, coupled with a distinct dissimilarity to Australo-Melanesians".

    Take note please, Terry. You cannot only look at one side of the picture: there are other interpretations than Bellwood's.

    ReplyDelete
  32. "No. The first widely accepted Neolithic culture of China (and all East Asia) is from the South!!!"

    Did you not read the link?

    "there are other interpretations than Bellwood's".

    And here's one of them:

    http://www.anthropology.hawaii.edu/people/faculty/Bae/pdfs/2002_Gao_and_Norton.pdf

    The authors suggest that just two phases exist for the paleolithic in China. Quote:

    "All archaeological material dating between the late Middle-early Upper Pleistocene (c. 140,000-30,000 years ago) are considered Middle Paleolithic."

    "It is the development of the blade and microblade industries in the late Upper Pleistocene that signals the break from the long and conservative Paleolithic cultural development in China"

    And where does that come from?

    "There is a general consensus that the late Paleolithic small flake toolkits and blade and microblade technocomplexes were a direct outgrowth of the small core-flake industry of the Early Paleolithic in North China based on technological and typological comparisons".

    "For instance, Zhang Senshui (1987; 1990) believes that the blade-microblade assemblages found in North China share many technological and typological features with some assemblages from Mongolia and Russia".

    Back to you:

    "We do, there is strong genetic evidence. 'Always' is of course a relative concept but since rather soon after the first colonization in any case".

    The above link suggests the population was 'always' small.

    "Illustrate the matter. Because as far as I am aware that is not the case".

    http://en.wikipedia.org/wiki/Haplogroup_Q1a3a_(Y-DNA)

    Quote:

    "Q1a3a1 (M3) - Subclade associated with all Indigenous Americas and also found in Evens of Siberia[7]. Origin: Siberia 15,000 years ago[1][8]
    Q1a3a1a (M19) - subclade found among Indigenous South Americas, such as the Ticuna and the Wayuu.[8] Origin: South America approximately 5,000 to 10,000 years ago.[1]
    Q1a3a1b (M194) - It has only been found in South American populations[8]
    Q1a3a1c (M199, P106, P292) - subclades that have only been found in South American populations[1]"

    "Probably not. Probably it coalesced in South China and migrated to Mongolia from there, already formed (but this is not fully discerned)".

    Extremely unlikely. Otherwise we'd surely find N to be reasonably common there. Especially if there has been a minimal amount of drift, bottlenecks and founder effects as you claim.

    "but we are here discussing North-South within modern China!!!"

    We're ultimately discussing the southward movement of the Mongoloid phenotype through China and on into SE Asia.

    ReplyDelete
  33. Another link, although I know I'm wasting my time:

    http://en.wikipedia.org/wiki/Mongoloid_race

    From the link, admittedly Peter Bellwood:

    "Archaeologist Peter Bellwood claims that the 'vast majority' of people in Southeast Asia, the region he calls the 'clinal Mongoloid-Australoid zone', are 'Southern Mongoloids' but have a 'high degree' of Australoid admixture.[17] Ainus are considered Southern Mongoloids even though they live in East Asia".

    So South Chinese lie on a cline between North Chinese and Se Asians. Who in turn lie on a cline between South Chinese and Australo-Melanesians. But here's another view from the same link:

    "A 2006 study of linkage disequilibrium finds that northern populations in East Asia started to expand in number between 34 and 22 thousand years ago, before the last glacial maximum at 21–18 KYA, while southern populations started to expand between 18 and 12 KYA, but then grew faster, and suggests that the northern populations expanded earlier because they could exploit the abundant megafauna of the 'Mammoth Steppe', while the southern populations could increase in number only when a warmer and more stable climate led to more plentiful plant resources such as tubers".

    Puts the earliest Chinese modern human expansion in the north.

    "Geneticist Luigi Luca Cavalli-Sforza claims that there is a genetic division between East and Southeast Asians.[39] In a like manner, Zhou Jixu agrees that there is a physical difference between these two populations".

    ReplyDelete
  34. In truth, Gao & Norton 2002 depicts a very sad reality of archaeology in China and, by extension, all East Asia. This fact, conflated with the fact that Oriental industries were for long flake-based, causes much misunderstanding. We can barely rely on archaeology in East Asia. Hopefully, the same that the situation is changing in South Asia, Arabia, etc. it will also change in East Asia.

    Is the blade technocomplex born locally or evolved from Altai "pseudo-Aurignacian"? Or something in between? I'd be inclined to the third option but it's difficult to judge.

    There is nothing clear on how this blade technocomplex expanded to the South - if it did at all. Or if there was a different evolution in that area or...

    I understand that Hoabinhian senso lato is an industry with blades and is old enough to compete with the North: approximately both "advanced" technocomplexes begin c. the same age: 20 Ka. (or maybe older in the case of Hoabinhian: I've read occasionally more than 30 Ka.) So, at the current stage it is impossible to say much based on archaeology alone.

    ReplyDelete
  35. "Did you not read the link?"

    Very difficult to know what link with such brief one-liners, but I believe I did read all links you posted, at least to some extent. I you at least cared to be more specific...

    "Otherwise we'd surely find N to be reasonably common there".

    It has nothing to do: a haplogroup can coalesce where it is rare: Q looks like coalescing in West Asia and it's rare there, while very common in America. There are many other examples. We must look at diversity, not frequency.

    If you want to insist on arguing based on frequency alone, I'd ask you to do it elsewhere - because it's pointless and annoying.

    "We're ultimately discussing the southward movement of the Mongoloid phenotype through China and on into SE Asia".

    I'm not. I have not until this very moment or the last post when I tried to address some of your issues.

    This paper actually does not mention any other part of SE Asia than what is within the borders of China, so you are diverting the debate to a quite off-topic issue. I have nothing against well placed or otherwise necessary off-topics but when you just claim that the off-topic is the center of the discussion... sorry but nope.

    ""Archaeologist Peter Bellwood claims that the 'vast majority' of people in Southeast Asia, the region he calls the 'clinal Mongoloid-Australoid zone', are 'Southern Mongoloids' but have a 'high degree' of Australoid admixture.[17] Ainus are considered Southern Mongoloids even though they live in East Asia".

    So South Chinese lie on a cline between North Chinese and Se Asians."

    That's what Bellwood and you believe. I do not have to agree at all. I actually think that there is no anything "Australoid" besides the Australian Aborigines and that all those peoples you claim "Australoid" are actually a large number of unique phenotypes and genetic pools on their own right.

    I fail to see what relation can have Ainu with Australian Aborigines, but I also fail to see what relation can have the Onge with the Papuans or... each of those "residual" populations is unique and illustrates a different kind of local isolate. Some may be related with each other in very deep time-frames but often what you see is that they are more related (genetically, mind you) to some "Mongoloids" (or even others) than among each other.

    I would even challenge the idea that there is a unique "Mongoloid" phenotype: Dienekes certainly failed to find it, and instead produced a number of diverse phenotype clusters, while at the same time Thor Heyerdal and Hosni Mubarak insisted in being clustered together. So there is No "Mongoloid" anything in the same sense that there is a "Caucasoid" type, ranging from Norway to Egypt.

    This means probably that the various "Mongoloid" phenotypes, as well as the various "Australoid" (senso lato) phenotypes coalesced probably before the "Caucasoid" type did, c. 40 Ka ago.

    But it's difficult to demonstrate because this is the study of the highly irregular and individualized shapes of nature and we don't even know for sure what part is played by genes and what by environment. Or if some skull type implies "slanted eyes" or not, what in the end is what rings in our minds when we think "Mongoloid", all other traits being secondary.

    "Geneticist Luigi Luca Cavalli-Sforza claims that there is a genetic division between East and Southeast Asians".

    Obsolete unreproducible data, sorry. Most if not all modern comparable researches fail to produce that duality. C-V failed in that (because his range of gene markers was limited).

    ReplyDelete
  36. "There is nothing clear on how this blade technocomplex expanded to the South - if it did at all".

    Most actually involved with studying it agree with a northern origin and southward spread. But of course you have your belief and try to fit the evidence to that belief rather than look at the evidence and then form an explanation. This allows you to ignore any other possibilities.

    "The paper anyhow only deals with a single site in a single period in North Vietnam, what is alone of limited informative value"

    Again everybody involved in actually studying the region agrees that Vietnam is mich more 'Sinicised' than are the other regions of mainland SE Asia. The linked paper explains when this happened.

    "a haplogroup can coalesce where it is rare"

    Yes, but according to your argument N would not have been 'rare' in South China.

    "Q looks like coalescing in West Asia and it's rare there"

    More likely Central Asia I'd say, where it is still common amoung the Ket and Selkup.

    "Yes it is: look at that atlas yourself: only rivers are 'obstacles' between Hanoi and Beijing".

    Rubbish. If that were so why is the population so concentrated in just a few regions? Apart from the Yangtze River Basin and pockets along the coast much of southern China is mountainous and quite sparsely populated.

    "We must look at diversity, not frequency".

    Well, that's the problem here. Up till now I have regarded O2a2c as a single haplogroup. But its downstream mutation O3a2c1 is also evenly spread throughout the three regions of China. Surely that is enough to show that O3a2c did not expand until O3a2c1 had appeared, then they both spread together.

    ReplyDelete
  37. "I actually think that there is no anything 'Australoid' besides the Australian Aborigines and that all those peoples you claim 'Australoid' are actually a large number of unique phenotypes and genetic pools on their own right".

    To some extent there are 'a large number of unique phenotypes and genetic pools'. But those phenotypes do form a cline. From Australia the cline runs north to Papua/New Guinea, then turns east through the Bismark Archipelago, Solomon Islands, Vanuatu/New Caledonia, Fiji, Tonga/Samoa and, finally, Polynesia. It is usually possible to tell reasonably accurately where on that cline most individuals you come across lie. Then the cline swings back west, but there is a break in it. As the Austronesians moved east from Southeast Asia the 'purer' form of the Papua/New Guinea phenotype followed along behind, and that movement has broken the cline. But it has contributed to the steady, gradual west/east change. Various haplogroups support the existence of that cline.

    "what you see is that they are more related (genetically, mind you) to some 'Mongoloids' (or even others) than among each other"

    Exactly. The Polynesians are more related genetically to Mongoloids than are the other members of this part of the cline. And we could regard the Polynesians as being halfway along the cline between the Australians and the northeast Asians 'Mongoloids'.

    As I said, the cline swings back. The other half of the cline jumps back to people who look a lot like Polynesians: filipinos/Indonesians/Malays. Then to Mainland SE Asia, South China, North China/Japan/Korea and finally northeast Asia.

    ReplyDelete
  38. I'm not ignoring any possibilities, just don't see any clear evidence anywhere, rather the opposite - but if you're going to disqualify with a blank slate disdainful sentence... whatever.

    "... according to your argument N would not have been 'rare' in South China".

    According to what I think, N was always rare in South China, it never had room to expand there (because of its dominant "brother" O mostly). However it still coalesced there as a minority lineage (probably, mind you, not "for sure": only further research will clarify this matter).

    "More likely Central Asia I'd say" [locality of Q's coalescence].

    Maybe. It's not like it's too frequent in Central Asia anyhow, besides Alta (where it may be 20-25%). The greatest frequencies area always found in localities that imply a late colonization, such as America or the colder parts of Siberia.

    "Rubbish" [re. the lack of geographical barriers between Tonkin and Manchuria other than rivers]

    Instead of reacting emotionally, could you mention such "barriers"?

    See: satellite image of China. Mountains begin when you approach Tibet or get into Indochina, not in what could well be termed the "great Chinese flatland" (it's actually somewhat hilly, specially in the south, I know, but nothing too marked). No Alps, no Saharas, no Wallace Lines anywhere. Not even a permeable low mountain Zomia-like area.

    "Up till now I have regarded O2a2c [O3a2c] as a single haplogroup. But its downstream mutation O3a2c1 is also evenly spread throughout the three regions of China".

    Both clades were determined by Hong Shi 2005 to have spread from South to North. Judging by frequency alone may be very much misleading.

    So yeah: that's the problem here: that you ignore structure and almost hysterically insist once and again on judgments (or rather wishful thinking) based only on frequency.

    Stop it. You are in denial: look at Hong Shi's structure analysis and you'll realize (as soon as you can put your faith to the side) that these two haplogroups expanded from South to North.

    ReplyDelete
  39. "look at Hong Shi's structure analysis and you'll realize (as soon as you can put your faith to the side) that these two haplogroups expanded from South to North".

    Quote from that paper:

    "In general, the distribution of the O3-M122 haplotypes did not show distinctive divergence between southern and northern populations, with all the major subhaplogroups shared between them—except for O3-M7, which was observed only in the southern populations and therefore indicates a recent common ancestry of the O3-M122 lineage in East Asia".

    Do I read correctly? 'recent common ancestry of the O3-M122 lineage in East Asia'? That's the whole of O3 according to the reconstruction. M7 (now O3a2b) is Hmong-Mien as far as I'm aware, so it is now a 'southern' haplogroup, but may not have first coalesced there anyway. And the paper goes on:

    "there was a lot of similar STR evolution after the emergence of O3-M122, and many shared STR haplotypes were observed between northern and southern populations, again confirmation of the recent common ancestry of the M122 lineage in East Asia".

    'Recent common ancestry', again?

    "Using the STR data, we calculated the gene diversities; no significant differences were observed between SEAS and NEAS or among different language groups (data not shown)".

    Proves that the 'haplogroups expanded from South to North'?

    "However, the MDS analysis showed that the NEAS are closely related by clustering together, whereas the SEAS showed relatively loose connections with larger variance, indicating that SEAS are genetically more polymorphic than are NEAS (fig. 5)".

    As we would expaect if the haploups had expanded from north to south, separating as they did so. Seems the author assumed that the Hmong-Mien, Austro-Asiatic, Tai-Kradai and Austronesian people are aboriginal Southern Chinese, whereas it's more than just possible they are migrants from the north too.

    "However, our previous studies showed that northern Han populations are relatively homogenous, with similar Y-chromosome haplotype distributions"

    The paper you blog here shows that not really to be so. Therefore the Hong She comment, 'It should be noted that the difference in genetic variance between NEAS and SEAS could be due to the sampling-density discrepancy' is correct.

    "it's actually somewhat hilly, specially in the south, I know, but nothing too marked".

    http://www.google.co.nz/imgres?imgurl=http://www.mapcruzin.com/free-maps-thematic/china_population_83.jpg&imgrefurl=http://www.mapcruzin.com/free-world-population-maps.htm&h=1014&w=1056&sz=129&tbnid=miqzJq59bbDrwM:&tbnh=144&tbnw=150&prev=/search%3Fq%3Dchina%2Bpopulation%2Bdistribution%2Bmap%26tbm%3Disch%26tbo%3Du&zoom=1&q=china+population+distribution+map&hl=en&usg=__TyPzf4FucDQvy6exzxFp_4ewWWo=&sa=X&ei=VoO6Td-eFIqssAO5p-TWBQ&sqi=2&ved=0CCQQ9QEwAA

    So how come there is a distinct thinning of the population through that 'somewhat hilly' country? You can see that the population is concentrated along the major river basins and in patches along the coast.

    "According to what I think, N was always rare in South China, it never had room to expand there (because of its dominant "brother" O mostly). However it still coalesced there as a minority lineage"

    There you go, as usual. Come up with a theory and then manipulate the evidence to fit it by invoking unfounded inventions of drift, bottlenecks and founder effects. Why are you not prepared to accept the evidence at face value? Why would it not expand at the same time as its brother O? Especially considering we're dealing with 'what could well be termed the great Chinese flatland'. Your theory doesn't make any sense.

    ReplyDelete
  40. "Why are you not prepared to accept the evidence at face value?"

    What evidence, what face value? Just yelling at me. :(

    "Why would it not expand at the same time as its brother O?"

    Initial drift favored one over the other (without really fixating it totally). It happens a lot.

    "You can see that the population is concentrated along the major river basins and in patches along the coast".

    Somewhat concentrated as you say, indeed. So what? I have always been pretty much conscious of the attractor role that waterways exert over people: that happens everywhere, including the North European Plain (where you must agree that there are no obstacles other than rivers, which actually act as attractors more than true obstacles most of the time).

    This has nothing to do with the existence or lack of obstacles but with the existence of attractor waterways.

    "... it's more than just possible they are migrants from the north too".

    Why? Because someone picks his nose? Or because there is evidence suggesting it? If so, which one? I have yet to see any such evidence: it's all (or most) wild speculation made academic doctrine (what is not the same as science, mind you).

    ...

    ReplyDelete
  41. As for Hong Shi 2005, it seems obvious to me that you insist in misquoting and misinterpreting. A good synthesis is in this paragraph:

    "It should be noted that the lack of genetic divergence, in view of the gene diversity between the southern and northern O3-M122 lineages, indicates that the O3-M122 lineages were probably dominant in the population involved in the initial northward migration; therefore, no obvious bottleneck occurred for the O3-M122 lineage, in contrast with the skewed distribution of the O2-M95 and O1-M119 lineages (Su et al. 1999; Wen et al. 2004b).

    That is what they say.

    "Table 3 lists the age estimations; all the subhaplogroups have a history older than the Neolithic time, with a range of 25,000–30,000 years ago".

    Notice that these are the subhaplogroups, not O3 as such, and that I consider that the Zhivotovski method underestimates age by 15-100% (Pan-Homo divergence systematic underestimation). So we can easily be talking of 60 Ka realistic age for O3 - no kidding.

    Notice also that the "skewed distribution" of O1 and O2 is also paralleled (in its own way) by N, which should be considered for practical purposes of historical reconstruction as a fourth major subhaplogroup of NO rather than as a totally distinct haplogroup with a totally different history.

    All this only the Results section that you misquote so much. The Discussion section is even more clear:

    "As we described above, the distribution of the O3-M122 haplotypes in East Asian populations supports a southern origin".

    And at the very end:

    "In summary, our data about the East Asian–specific haplogroup O3-M122 indicates a southern origin of the O3-M122 lineage, therefore supporting the hypothesized southern origin of modern humans in East Asia. The initial prehistoric northward migration was estimated at 25,000–30,000 years ago".

    Bu why do I have to quote all that when obviously you had read it already? :(

    ReplyDelete
  42. You may find this interesting:

    http://www.investigativegenetics.com/content/2/1/10/abstract

    "Interestingly, a high frequency (31.4%) of haplogroup O2b- SRY465 (and its sublineage) is characteristic of male Koreans, whereas the haplogroup distribution elsewhere in East Asian populations is patchy. The ages of the haplogroup O2b-SRY465 lineages (~9,900 years) and the pattern of variation within the lineages suggested an ancient origin in a nearby part of northeastern Asia, followed by an expansion in the vicinity of the Korean Peninsula. In addition, the coalescence time (~4,400 years) for the age of haplogroup O2b1-47z, and its Y-STR diversity, suggest that this lineage probably originated in Korea".

    And:

    "These findings are consistent with linguistic, archaeological andhistorical evidence, which suggest that the direct ancestors of Koreans were proto-Koreans who inhabited the northeastern region of China and the Korean Peninsula during the Neolithic (8,000-1,000 BC) and Bronze (1,500-400 BC) Ages".

    That takes care of O2b. Now:

    "probably dominant in the population involved in the initial northward migration"

    Several unjustified assumptions straight away.

    "N, which should be considered for practical purposes of historical reconstruction as a fourth major subhaplogroup of NO"

    Agreed. But that becomes a problem for your theory.

    "Initial drift favored one over the other (without really fixating it totally). It happens a lot".

    But you quote from the other paper:

    "therefore, no obvious bottleneck occurred for the O3-M122 lineage"

    So why on earth should just one haplogroup suffer a bottleneck when the others didn't? Especially considering (according to you) the original population entered a region where it was able to expand unhindered? Drift, bottlenecks and founder effects should have had no influence at all.

    "in contrast with the skewed distribution of the O2-M95 and O1-M119 lineages"

    So how do you explain that skewed distribution, if not as indicating independent migrations from different regions? Doesn't make sense, if you're correct, and all four haplogroups expanded from the same region in South China through a region with no obstacles other than rivers.

    "Somewhat concentrated as you say, indeed. So what?"

    The Nanling Mountains are virtually uninhabited. They may not be to high but they sure are steep.

    "all the subhaplogroups have a history older than the Neolithic time, with a range of 25,000–30,000 years ago".

    Maju. You obviously don't realise that the 'age' of a haplogroup is not the same as the time of its 'expansion'. A haplogroup can coalesce thousands of years before it expands.

    "Bu why do I have to quote all that when obviously you had read it already?"

    Because, unlike the case for many links I've provided, you fail to see the weaknesses in the argument.

    ReplyDelete
  43. I'll be back later with the O2b paper. Just remind that this haplogroup is not exclusive of NE Asia but is also found at low frequencies though SE Asia, while AFAIK never in North China, invalidating any possible "Neolithic" flow conjecture. This means that whatever the connection between SE Asian and NE Asian O2b is not through North China but probably by a long erased coastal route of some sort.

    When I say "initial drift", I refer for the initial "NO population" at t=1. When you discuss O3 that is at t=3, some time after t=1, a very different scenario as far as I can infer, when O was expanding all around already.

    Different moments. I hope you can understand this simple idea.

    "how do you explain that skewed distribution, if not as indicating independent migrations from different regions?"

    Actually I think in the following terms:

    1. (t=1) NO proto-Oriental "tribe" (in SE Asia or South China)
    2. (t=2) coalescence of the four major subclades (N, O1, O2 and O3, maybe some of their subclades)
    3. (t=3) expansion of these subclades (and corresponding diversification). Some (N, O2b) reach the far North in localized founder effects), while others (O3 and sublineages) expand in a wide front across China.

    Unsure if the northern C lineages, and maybe even the D ones, migrated also in this process (it's possible indeed).

    "The Nanling Mountains are virtually uninhabited".

    Anecdotal stuff. You are already touching Zomia anyhow.

    "You obviously don't realise that the 'age' of a haplogroup is not the same as the time of its 'expansion'. A haplogroup can coalesce thousands of years before it expands".

    Not really. Not in the case of Y-DNA. Y chromosome mutations are way too frequent for that. That can be true in the much shorter mtDNA ring but not in the comparatively huge Y chromosome. However the time between one phylogenetic stage and the next one (i.e. the process of coalescence as "coalescing" and not yet as "coalesced") can take some good time indeed. That's why t=(1,2,3...)

    In Y-DNA, coalescence as haplogroup (emphasis as always in "group": branching node) and expansion must be the same. Another thing is whether we can estimate their ages with any accuracy however.

    ReplyDelete
  44. "However the time between one phylogenetic stage and the next one (i.e. the process of coalescence as 'coalescing' and not yet as 'coalesced') can take some good time indeed. That's why t=(1,2,3...)"

    That is exactly what I'm getting at. We don't really know what happened to NO, or even MNOPS, during t=1, t=2 etc.

    "When I say 'initial drift', I refer for the initial 'NO population' at t=1. When you discuss O3 that is at t=3, some time after t=1, a very different scenario as far as I can infer".

    I agree with that completely.

    "1. (t=1) NO proto-Oriental 'tribe' (in SE Asia or South China)"

    We know that P moved west rapidly, leaving no remnant behind but, in the form of its descendant haplogroups, became very common through India and out into wider Eurasia. And M and S both spread rapidly to New Guinea/Melanesia. So there's no reason at all why NO was unable to begin moving north immediately.

    "2. (t=2) coalescence of the four major subclades (N, O1, O2 and O3, maybe some of their subclades)"

    Each somewhere within the region that NO had spread through, not necessarily in the region where NO or MNOPS had originally coalesced. I'd guess the region the four haplogroups coalesced encompassed much of what is now China north of Zomia and the Nan Ling, with N in the north, O2 in the northeast, O1 in the east and O3 in western China.

    "3. (t=3) expansion of these subclades (and corresponding diversification)".

    But in that scenario O3's expansion was not subject to a 'corresponding diversification, especially in the case of O3a1c and O3a2c. No R haplogroup (for example) at any level displays such a phenomenon. And as you say:

    "Another thing is whether we can estimate their ages with any accuracy however".

    To me the fact that several O3s are widespread, as opposed to the other haplogroups, suggests their expansion is relatively recent. They have not been overlaid by any more recent expansion. Again, to me that suggests their expansion is more recent than the Paleolithic. Of course regional diversification in those O3 haplogroups may show up at a finer level but we certainly have no hint of such diversification at this stage.

    "This means that whatever the connection between SE Asian and NE Asian O2b is not through North China but probably by a long erased coastal route of some sort".

    I agree with that completely. We actually see regional diversification in O2. But I suspect that the origin of that coastal connection is from somewhere around the Sea Of Japan. The spread probably involved the Ryukyu Islands as O2b is reasonably common there.

    ReplyDelete
  45. There's a reason why NO probably did not move North immediately: we see no traces of such migration such as NO* scattered along the route. You compare with P but I estimate that P coalesced in Bengal, what is almost like Indochina, not really far away in any case from the putative MNOPS urheimat in Sundaland or Indochina. By comparison your obsessive "North" is much farther away, not considering the climate as barrier and obstacle for demographic growth.

    In fact, my null hypothesis right now is that all early Oriental lineages (D, C1'3 and NO migrated more or less jointly nortwards, with N, the several O, D and C subclades having localized founder effects in diverse areas. Notice that I'm not even saying that they were the same tribe or ethnicity, but that they were interacting closely all the time (constituting therefore a single population).

    There are other possibilities, specially earlier D and/or C layers and replacement by an expansive northwards migrating NO but I fail to see the archaeology of such a process.

    "But in that scenario O3's expansion was not subject to a 'corresponding diversification"...

    There is: O3 split in two subclades: O3a1 and O3a2. I think that (thanks to Hong Shi's pioneer work) the case for a southern origin of O3a2 and subclades is quite clear. It may be not so clear for O3a1 (notice comparatively high O3a1* in the north in this paper) but it's not something we may discern right now (because we lack data).

    In any case these two lineages and their subclades were in continuous contact and nothing really opposed their flow and that's why we see them now so interspersed.

    Unless... you can't prove, which you cannot at the present level of knowledge.

    "No R haplogroup (for example) at any level displays such a phenomenon".

    You see that between R1a and R2 in South Asia, M and S in Melanesia. Also several E1b1b1 subclades do clearly overlap and probably E1a1a1 do as well. Most other haplogroups have never been researched so much: what do we know of the structure of H, one of the most important human Y-DNA haplogroups by numbers? Really little.

    ReplyDelete
  46. Erratum:

    "Unless... you can't prove"...

    should read:

    "Unless... you can prove otherwise"...

    ReplyDelete
  47. "I think that (thanks to Hong Shi's pioneer work) the case for a southern origin of O3a2 and subclades is quite clear".

    It's far from 'quite clear'. The phylogeny of O3 is nowhere near finalised so all calculations of its dating and point of origin could be way off.

    "O3 split in two subclades: O3a1 and O3a2".

    We have no way of knowing whether that split occurred during a northward or southward movement.

    "notice comparatively high O3a1* in the north in this paper"

    And O3a*-M324 (in the Korean paper).

    "In any case these two lineages and their subclades were in continuous contact and nothing really opposed their flow and that's why we see them now so interspersed".

    I doubt very much that is the explanation for the current mixed distribution of the two haplogroups.

    "You see that between R1a and R2 in South Asia"

    Those two haplogroups could hardly be called 'evenly distributed' through any part of India. R1a tends to be much more common than is R1b and tends to be northern. R1a has all the characteristics of being indigenous to India while R1b looks to be an immigrant.

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

    Quote:

    "R1a is typical in populations of Eastern Europe, Indian Subcontinent and parts of Central Asia ... Haplogroup R2aHaplogroup R2a ... is mainly found in South Asia and southern Central Asia".

    So R1a tends to be north of R2. Back to you:

    "M and S in Melanesia"

    M tends to be concentrated in the Melanesian islands, S in mainland New Guinea:

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

    Quote:

    "Haplogroup S is commonly found among populations of the highlands of Papua New Guinea.[3] It is also found at lower frequencies in adjacent parts of Indonesia and Melanesia".

    And:

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

    Just M1 is from New Guinea. M1 derivatives M1a and M1b, along with M2 and M3 are offshore.

    "Also several E1b1b1 subclades do clearly overlap"

    They 'overlap' but are not 'evenly distributed'. Each tends to be concentrated in their particular region. Besides which I'd argue for a Neolthic expansion of the E1b1b1 subclades, not Paleolithic.

    "probably E1a1a1 do as well"

    You probably mean E1b1a, which is spread through much of Africa except for the northeast. Again my guess is that the haplogroup is not 'evenly distributed' through the region. And where it is the distribution is a product of relatively recent arrival. E1a1 and E1a2 tend to be west African and E1a is a minor haplogroup overall.

    ReplyDelete
  48. Why has O3a2 been demonstrated to have a Southern origin within China?

    O3a2c and O3a2b, which were demonstrated by Hong Shi 2005 to have a southern origin (with or without considering Southern Han and Sino-Tibetans). He discerned the defining mutations of these lineages and studied their haplotype structure separately, always pointing to a clear Southern origin.

    The only remaining subhaplogroup, O3a2a, was not studied by Hong Shi 2005 but it seems to only exist in the South anyhow, so I understand that it is a moot point. The only doubt could exist re. paragroup O3a2* but eve this one is much more frequent in the South than in the North - for whatever that is worth.

    So I understand that the Southern origin of O3a2 is quite solidly demonstrated, notwithstanding that some new contradictory evidence could arose in the future as can always happen in matters of science.

    I hope that you get that straight because I am not going to repeat this again (you should have got it straight many posts ago and even many years ago, as this issue has been arising once and again in discussions involving you, Terry, and only you).

    ReplyDelete
  49. As for the rest, I do not think that O3 subclades are "evenly distributed": it'd be foolish to think that way. They overlap but they typically have their own "home districts" where they are more dense.

    IDK, for instance (Hong Shi 2005 again) O3a2b is concentrated towards Tonkin and South Fujian, while O3a2c1 extends along the middle and upper Yangtze mostly. Just that no such mapping effort has been made AFAIK.

    It is also impossible that E1b1b1 major subclades spread in the Neolithic, none of them having greatest diversity in the fertile crescent or anything of the like. They spread in most cases before Neolithic, though some late expansion may have happened under Neolithic conditions (this is specially true for E-V13 but only for this one clade).

    ReplyDelete
  50. "O3a2c and O3a2b, which were demonstrated by Hong Shi 2005 to have a southern origin"

    Their work is convincing only if you are desperate to believe. The phylogeny has been altered since their research anyway, so its even less convincing. And we seem to have established that O2 is northern as is N. O1 is indeed fairly southern as in from the Yangtze River mouth. Surely O3 should have coalesce somewhere near where the other three coalesced.

    "IDK, for instance (Hong Shi 2005 again) O3a2b is concentrated towards Tonkin and South Fujian, while O3a2c1 extends along the middle and upper Yangtze mostly".

    But the haplogroups that are evenly distributed (under the new phylogeny) are O3a1c, O3a2c1 and O3a2c1a. And 'the middle and upper Yangtze' certainly encompasses the main Chinese Neolithic.

    "It is also impossible that E1b1b1 major subclades spread in the Neolithic"

    Impossible? Well. This is what wikipedia says:

    http://en.wikipedia.org/wiki/Haplogroup_E1b1b_(Y-DNA)

    Quote:

    'Based on genetic STR variance data, Cruciani et al. (2007) suggests that E1b1b1a1 [E-M78] originated in 'Northeastern Africa', which in their study refers specifically to Egypt and Libya.[Note 4] about 18,600 years ago (17,300 - 20,000 years ago)... based on archaeological data, they propose that the point of origin of E-M78 (as opposed to later dispersals from Egypt) may have been in a refugium which existed on the border of present-day Sudan and Egypt, near Lake Nubia, until the onset of a humid phase around 8500 BC. The northward-moving rainfall belts during this period could have also spurred a rapid migration of Mesolithic foragers northwards in Africa, the Levant and ultimately onwards to Asia Minor and Europe, where they each eventually differentiated into their regionally distinctive branches"

    So members of E1b1b1a1 cannot have reached Europe before 8500 years ago. That's relatively recent as far as I'm concerned. And the downstream clades such as E-V13, E-V65 and E-M521 must be even more recent. The other European E clade is E-V257. Quote from Wiki:

    "Arredi et al. (2004) believe the pattern of distribution and variance to be consistent with the hypothesis of a post Paleolithic 'demic diffusion' from the East. The ancestral lineage of E-M81 in their hypothesis could have been linked with the spread of Neolithic food-producing technologies from the Fertile Crescent via the Nile, although pastoralism rather than agriculture".

    And:

    "In Europe, E-M81 is found everywhere but mostly in the Iberian Peninsula, where unlike in the rest of Europe[Note 6] it is more common than E-M78"

    So the ranges of E1b1b1a and E1b1b1b can hardly be said to 'overlap'. And their arrival in Europe, and probably their spread, is hardly 'ancient'.

    ReplyDelete
  51. "And we seem to have established that O2 is northern as is N".

    No!

    O2b is probably northerner but not all O2. O2a is clearly southerner.

    N is probably of Southern origin (high N(xN1)), even if its greatest expansion in terms of numbers (founder effects) happened in the Far North.

    You know all that already, why are you beating not one but several dead horses?

    "... 'the middle and upper Yangtze' certainly encompasses the main Chinese Neolithic".

    The Middle (but not Upper) Yangtze is the core of the East Asian Neolithic as far as I can tell.

    But it is in the South. If anything it'd would argue for a south to north "Neolithic" flow. Yet the Northern Neolithic is so different and almost immediate in time that there is no archaeological grounds for any meaningful flow in either direction.

    Therefore the concept "Neolithic" is horrible to explain any kind of genetic homogeneity within the borders of modern China, whichever direction of flow you favor.

    Think about this if you can hold still and do it.

    ReplyDelete
  52. Re. E1b1b1, there is another recent thread here, and a bit older but more specific here. I'm not sure if you took part but there is a lot of people who agrees that E1b1b1b1 specifically must have been dominant in NW Africa before c. 20,000 years ago.

    Anyhow what Wikipedia, or rather Cruciani 2007 says is that E1b1b1a1 spread with Capsian (late Paleolithic) and arrived to Europe maybe 8500 BCE, which is a bit earlier for Neolithic (and there is an Epipaleolithic-rooted model of Greek Neolithic anyhow).

    In any case Greek/Albanian centered E1b1b1a1b (V13) is just a second tier subclade and one that I already considered to be of possible Neolithic expansion above.

    Dead horse again, just that this time you don't even seem to realize that I already said, or assumed as generally known, all you would eventually quote from Wikipedia, which, in this particular case only supports almost millimetrically my thesis.

    However I cannot agree with Arredi 2004 for several reasons:

    1. There is no Neolithic cultural flow in North Africa that could explain this E->W alleged migration.

    2. It'd need to be Capsian (which is pre-Neolithic anyhow) and this means origin in Upper (and not Lower) Egypt. But S. Egypt is peripheral in the haplotype structure.

    The real central node of E1b1b1b1 (M81) is in NW Africa, between South Morocco and Tunisia. This seems pretty clear to me. Haplotype 46 (and only that one) could have back-migrated with Capsian.

    There is in any case no particular reason (other than preconceptions and bias) to suggest any Neolithic origin for this haplogroup.

    As I have argued once and again its unusually high presence in Asturias and Cantabria (NW Iberia of deep Paleolithic settlement) strongly argue for a Solutrean founder effect in relation with the Oranian genesis.

    One would need to know quite a bit about Iberian and North African Paleolithic in order to even consider that, I know. But once you know it's quite compelling.

    "In Europe, E-M81 is found everywhere but mostly in the Iberian Peninsula, where unlike in the rest of Europe[Note 6] it is more common than E-M78"

    That only happens in the Western Iberian Peninsula, mind you. But otherwise it is true.

    In Europe this marker (which probably exists at low frequencies and the occasional localized founder effect though the Atlantic) should be considered a quality marker for Neolithic flow from Megalithic Portugal. The low apportion of this marker really indicates that there was just very limited demographic flow through the Atlantic waterways in the Neolithic, Chalcolithic and later on, at least in the South-North direction. But there was some, because the marker is not totally absent (and is at most 9% in West Iberia - so a 1% in say Wales, means a 10% demic flow from West Iberia potentially).

    ReplyDelete
  53. "O2b is probably northerner but not all O2. O2a is clearly southerner".

    They must have a common source, somewhere. Most likely in the north, rather than O2b having moved north.

    "N is probably of Southern origin (high N(xN1)), even if its greatest expansion in terms of numbers (founder effects) happened in the Far North".

    That is an extremely unlikely scenario.

    "The Middle (but not Upper) Yangtze is the core of the East Asian Neolithic as far as I can tell".

    OK. So that's the orgin of O's expansion.

    "If anything it'd would argue for a south to north 'Neolithic' flow".

    Evidence?

    "That only happens in the Western Iberian Peninsula, mind you. But otherwise it is true".

    I haven't really looked at E much before, so thanks for bringing it to my attention. It looks very much as though basically members of two E haplogroups reached Europe: E1b1b1a (V68) and E1b1b1b (V257). But to me it looks quite possible that they are just an eastern (E1b1b1b) and western (E1b1b1a) Mediterranean population. The two may well have entered the Mediterranean together, from the same region of North Africa. One spread east as far as Albania (and beyond) and one spread west to Gibraltar (and beyond). So when did E develop or aquire the ability to travel through the Mediterranean?

    Sardinia looks to hold the key. It has basal members of both haplogroups. So when did people first reacg Sardinia? Species extinction on the island suggests it was during the early Neolithic, or perhaps as recently as 7000 years ago.

    Of course e may not have been the first Y-hap to reach Sardinia. But it does give us the earliest possible date for the expansion of E1b1b1-M35.

    "I'm not sure if you took part but there is a lot of people who agrees that E1b1b1b1 specifically must have been dominant in NW Africa before c. 20,000 years ago".

    No matter what 'lot of people' agree with, E-M35's (E1b1b1) expansion cannot have been much before E1b1b1a and E1b1b1b were able to spread through the Mediterranean and reach Sardinia. So the downstream haplogroup E1b1b1b1 cannot have been 'dominant in NW Africa before c. 20,000 years ago'. Besides which those discussions took place before the revision of E-P2's (E1b1) phylogeny.

    "Cruciani 2007 says is that E1b1b1a1 spread with Capsian (late Paleolithic) and arrived to Europe maybe 8500 BCE, which is a bit earlier for Neolithic"

    But hang on:

    "In any case Greek/Albanian centered E1b1b1a1b (V13) is just a second tier subclade and one that I already considered to be of possible Neolithic expansion above".

    Isn't E1b1b1a1b (V13) a downstream haplogroup from E1b1b1a1? The two statements conflict and don't make sense.

    ReplyDelete
  54. You claim too many things without any support whatsoever, Terry.

    Naturally, the origin of O2, taken alone and on light of the existing data, is inconclusive (not "most likely in the north") but I do apply maximum parsimony (Occam) by looking at the other related clades, what suggest a southern origin, if anything.

    In the case of N, high amounts of N(xN1) may mean highest diversity or at least equally high basal diversity in the South as in the North, so in the worst case we are like with O2 (inconclusive) and in the best case it is a clear case of S->N flow. In no case this one is a "most unlikely scenario" using a rational mind.

    "Evidence?"

    Evidence of an "if anything" case? Please!

    ReplyDelete
  55. As for E1b, I do not think that Sardinia has anything to do: it is a secondary destination and not any origin. Anyhow nowadays it seems Sardinia was colonized as early as the Epipaleolithic (though this is irrelevant because E1b, as well as other clades corresponds to the Neolithic layers in that island AFAIK).

    "The two statements conflict and don't make sense".

    Because you are not putting your finger on the map as you read them:

    1. E1b1b1a1 expands with Capsian (late UP)
    2. E1b1b1a1 arrives to Greece/Albania in the Epipaleolithic
    3. E1b1b1a1 ("Greek" variant V13) expands from the SW Balcans with Neolithic flows into Europe

    ReplyDelete
  56. "Because you are not putting your finger on the map as you read them"

    Yes. I lost count of the 1bs and got confused.

    "As for E1b, I do not think that Sardinia has anything to do: it is a secondary destination and not any origin".

    I never claimed it was the origin. But the Sardinian haplogroups are close to basal, so where-ever E1b1b1-M35 originated it reached Sardinia rapidly.

    "1. E1b1b1a1 expands with Capsian (late UP)
    2. E1b1b1a1 arrives to Greece/Albania in the Epipaleolithic
    3. E1b1b1a1 ('Greek' variant V13) expands from the SW Balcans with Neolithic flows into Europe"

    I don't think so. For a variety of reason E's phylogeny has been particularly well studied. And so we should be able to come to reasonably sound conclusions. You mentioned earlier:

    "In any case Greek/Albanian centered E1b1b1a1b (V13) is just a second tier subclade and one that I already considered to be of possible Neolithic expansion above".

    But it's quite possible that at least the E1b1b1a-M68 and E1b1b1a1-M78 mutations had already formed by the time E1b1b1-M35's expansion began. I'll explain:

    The region of E1b1b1-M35's highest basal diversity readily stands out. From your continued claims that 'basal diversity' indicates 'origin' we can say with some confidence that E1b1b1-M35 originated in the Horn of Africa. Three of its seven basal haplogroups are found only or mainly there: E1b1b1d, E1b1b1f and E1b1b1g. Another of the seven, E1b1b1e, looks to have moved south from there, possibly with the expansion of pastoralism (giving us another dating method for E1b1b's expansion).

    The remaining three basal haplogroups ended up around the Mediterranean. This impies they had boats. So what was their route into the Mediterranean? The Red Sea or the Nile?

    It's possible that E1b1b1c-M123 moved along the Red Sea and diversified through Arabia and the Levant and up into Anatolia. But what about the other two? E1b1b1a-M78 broke into 6 subclades, although the first two in the next list may be combined. One is found through the Horn, Ethiopia and Kenya (E1b1b1a1a2-V32), one in the Sudan and Egypt (E1b1b1a1a-V12), one in the Horn, Sudan, Egypt and the Middle East (E1b1b1a1c-V22), one in the Balkans (E1b1b1a1b-V13), one in Greece (E1b1b1a1e-M521) and one through the Maghreb, Italy and Spain (E1b1b1a1d-V657). So that subclade seems also to have originated in the Horn. Unless E1b1b1a-M78 moved back up the Nile to Sudan and the Horn the most logical explanation is that the haplogroups are spread along what was the route from the Horn to the Mediterranean. In other words we'd have to conclude that the subclades of E1b1b1a-M68 were already part of the original E1b1b1-M35 expansion. E1b1b1a1b-V13's presence in the Balkans is the result of drift and founder effect acting on the populations resulting from that expansion. So the E1b1b1a1b-V13 arrived in the Balkans during E-1b1b1-M35's expansion.

    ReplyDelete
  57. "Sardinian haplogroups are close to basal".

    If you want Sardinia to play any role, the less you can do is document your claims (link preferably). I cannot discuss claims I have never before heard about: for me Sardinian E1b is derived from their surroundings and a "recent" arrival. Never before yesterday I read otherwise.

    "we'd have to conclude that the subclades of E1b1b1a-M68 were already part of the original E1b1b1-M35 expansion".

    No and I don't see how you can reach to any such conclusion.

    It seems the flaw in your quite convoluted and obscure logic is in this:

    "Unless E1b1b1a-M78 moved back up the Nile to Sudan and the Horn"...

    Most likely on light of Battaglia 2008, right? E1b1b1a1 (M78) seems to represent (together with J1 and maybe other more localized lineages like R1b1a) the flow of Afroasiatic languages from the area of Egypt/Sudan. It is true that the languages may have an older cradle further south but a lot on this judgment depends on the belonging or not of Omotic to this phylum.

    It is even possible that there were two successive cradles: one early on in Ethiopia, deep in the Upper Paleolithic and another later on in Nubia in the late Upper Paleolithic. Whatever the case, if we are to make any association between AA languages and E1b, it must be largely pre-Neolithic.

    I'd even argue (weakly but still) that some of the principles of would-be West Asian Neolithic, notably the harvesting of grasses (cereals) to eat their seeds, was first developed in Nubia. But at best we'd be talking of a Mesolithic (proto-Neolithic) and not Neolithic yet.

    Mathilda once also argued for a Mesolithic microlithism flowing from Nubia to Palestine via Egypt. I lack of solid knowledge of this matter to be sure (specially because microlithism is an slippery concept and something invented many times in many places).

    Anyhow, we are getting off-topic. Please continue in the relevant threads if need be.

    ReplyDelete
  58. "Anyhow, we are getting off-topic. Please continue in the relevant threads if need be".

    OK. But I'll make a few last comments here.

    "E1b1b1a1 (M78) seems to represent (together with J1 and maybe other more localized lineages like R1b1a) the flow of Afroasiatic languages from the area of Egypt/Sudan".

    Actually the whole of E1b1b1-M35 could be at the base of Afro-Asiatic language spread. E1b1b1c-M123 is at least as 'Jewish' as E1b1b1a1-M78. Certainly E1b1b1 is very much present in the region where Austro-Asiatic languages are spoken today. The regions where the haplogroup and the languages don't both exist are easily explained way by later language replacement.

    "if we are to make any association between AA languages and E1b, it must be largely pre-Neolithic".

    OK. E1b's expansion is obviously 'largely pre-Neolithic'. But E1b1b1-M35's is later than that of E1b-P2's, and is probably associated with Austro-Asiatic.

    "No and I don't see how you can reach to any such conclusion".

    We need to get this straight then. So you accept that basal haplogroup diversity indicates region of origin, except when that would conflict with what you alreay believe. Is that correct? And E1b1b1-M35's expansion is 'starlike'. When considering other haplogroups you have always claimed that this is an indicator of rapid expansion. Do you always alter your interpretation of the evidence to suit your pre-existing belief? I suspect that a starlike pattern of haplogroup phylogeny can mean a population growth rather than necessarily expansion. But Population expansion is usually preceded by a growth in population, and I'm sure that is what we see in the case of E1b1b1-M35's starlike pattern. The same with O3a_M324.

    "for me Sardinian E1b is derived from their surroundings and a 'recent' arrival".

    I agree with the recent arrival bit. But the only E1b1b-M215 haplogroups in North Africa outside the Northeast are E1b1b1a1 and E1b1b1b. Both haplogroups are also found scattered through the Mediterranean, and neither have close relations in North Africa. So it is most parsimonious to conclude that their presence in North Africa is a product of their presence in the Mediterranean, rather than the other way round.

    "If you want Sardinia to play any role, the less you can do is document your claims (link preferably)".

    Good old Wiki again:

    http://en.wikipedia.org/wiki/Haplogroup_E1b1b_(Y-DNA)

    Quote:

    "E1b1b1a (E-V68), is dominated by its longer-known sub-clade E-M78 (E1b1b1a1). Three 'E-V68*' individuals who are in E-V68 but not E-M78 have been reported in Sardinia, by Trombetta et al. (2011), when announcing its discovery. The authors noted that because E-V68* was not found in the Middle Eastern samples, this appears to be evidence of maritime migration from Africa to southwestern Europe".

    And:

    "E1b1b1b is dominated by its dominant sub-clade E1b1b1b1 (E-M81) ... V257's discovery was announced in Trombetta et al. (2011) ... They found 6 'E-V257*' individuals in their samples who were E-V257, but not E-M81. A Borana from Kenya, a Marrakesh Berber, a Corsican, a Sardinian, a southern Spaniard and a Cantabrian".

    Surely the presence of these two widespread basal haplogroups suggests strongly that E1b1b1-M35's expansion was rapid, and included the Mediterranean islands. And here's the article the haplogroup diagram in Wiki is based on:

    http://www.plosone.org/article/info:doi/10.1371/journal.pone.0016073

    ReplyDelete
  59. No but I'm not going to discuss anymore on E here. Repost where relevant.

    ReplyDelete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).