New blogs

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

Thursday, January 14, 2010

R1b1 origin: Italy or West Asia?


One thing I have been chewing on
since I read the recent Cruciani paper on R1b1a-V88 is that, at least in that study, the region where R1b1-P25 has highest basal diversity (by haplogroups) are Italy and West Asia. Both regions have R1b1a, R1b1b and some other R1b1* (3/1173=0.25% in Italy and 1/328=0.3% in West Asia).

Another R1b1* was observed in the East Asian sample but this region lacks R1b1a.

The Italian R1b1* seems to belong (by haplotype) to two different subclades, what makes Italy provisionally a good candidate for the origin of R1b1. However notice that the Italian sample is much larger than the West Asian one, almost in direct proportion to the number of R1b1* individuals found.


Italy (incl. Corsica and Sardinia) also has high R1b1a diversity (by subhaplogroups: 2/4 basal sublineages, but lacks R1b1a*). This would suggest that the origin of R1b1a actually lays towards Africa, where basal diversity seems somewhat higher.

Of course it's a very tricky issue. Take with a grain of salt.

30 comments:

waggg said...

Hello.
What east Asian country was it found in (R1b1*)?

As for the R1b1a found in Sardinia and Corsica, isn't it probable that, since it's found in south-east of France and that I2b1 is found both in Sardinia and the Basque country (and a bit in Ibera), they came together in Sardinia?

Couldn't it coincide with the mtDNA H found in the paleolithic Iberian peninsula and Italy? (I see the U subclades being originally mostly linked to y-DNA I, in Europe)?

It could also be in relation (at least largely) with the mtDNA of Taforalt, mostly H and U (BTW, do we know what kind of Y-DNA I* - that was found among the Guanches' remains - was it precisely?).

Maju said...

What east Asian country was it found in (R1b1*)? -

I'll check later if there is any more specific reference (either in some forgotten corner of the main paper or in the supplemental material) but I took it from the main table where it's mentioned as a regional composite of 156 individuals.

For the record, R1b1b1 (the "Uyghur" haplogroup) was found only in the South Asian sample (3.1%, n=288).

As for the R1b1a found in Sardinia and Corsica, isn't it probable that, since it's found in south-east of France and that I2b1 is found both in Sardinia and the Basque country (and a bit in Ibera), they came together in Sardinia? -

It's possible. I think I've read somewhere that I2b1 has highest diversity in SE France.

However I am not sure that I2b1 is found in the Basque Country (I2b is and makes about 50% of Basque and Pyrenean I but I can't recall any reference right now that it is I2b1 specifically). In any case no R1b1a was found in the composite of Western Europe (n=465).

Couldn't it coincide with the mtDNA H found in the paleolithic Iberian peninsula and Italy?.

We are talking of things of totally different sizes here. MtDNA H is a dominant haplogroup, comparable if anything to Y-DNA R1b1b2a. I'd rather compare with something much rarer, like the mtDNA JT(xJ,T) found in Morocco and Italy (and in Paleolithic remains of Taforalt apparently too). Or maybe it could compare with some other rarer mtDNA lineage such as V or U8.

All this with great caution and the knowledge that paternal and maternal lineages do not need to follow the same patterns, specially when they are minor lineages, which are much more sensible to random fluctuations.

However the whole picture of R1b1 and H have some parallels, with middle European (Italy and Germany respectively) and West Asian areas "disputing" the highest diversity position and hence the most likely origin "dignity". But, on the other hand, we don't see much R1b in North Africa (<10% in all cases but Siwa, <5% in all cases west of Egypt, with many cases of just 0% R1b of), when we do see loads of H of apparent SW European origin, concentrated in NW Africa.

So if there is any parallel, it collapses in Africa.

BTW, do we know what kind of Y-DNA I* - that was found among the Guanches' remains - was it precisely?.

No. The paper only mentioned I (they did not test for downstream markers, it seems). Oddly enough, it dropped massively after the Spanish invasion, in spite of Y-DNA I being relatively common among Castilians.

...

If you want a copy of the paper, post with an email.

waggg said...

" We are talking of things of totally different sizes here. MtDNA H is a dominant haplogroup, comparable if anything to Y-DNA R1b1b2a. I'd rather compare with something much rarer, like the mtDNA JT(xJ,T) found in Morocco and Italy (and in Paleolithic remains of Taforalt apparently too). Or maybe it could compare with some other rarer mtDNA lineage such as V or U8.
All this with great caution and the knowledge that paternal and maternal lineages do not need to follow the same patterns, specially when they are minor lineages, which are much more sensible to random fluctuations."


Indeed, and that's the reason of my conjecture. If we try to connect y-DNA and mtDNA it generaly just doesn't work well. When we see some vague pattern we can as well be bold.
After all, we're talking of a time when human populations were much lower and disproportionate evolutions in populations might have been induced by different ways of life, for instance.
Not even mentionning violent events, that could probably have dramatic impact on very localized populations, if we go deep into paleolithic with very low populations, I guess.

The European paleolithic mtDNA might have had a much better success in having offsprings with the neolithic newcomers (presently, it's popular to link R1b1b2 with neolithic in Europe), and R1b1a (if we associate them with some of the paleolithic hunters/gatherers) might have had trouble with survival and was gradually outbred in his own lands, because of farming. Doest it sound plausible "mathematically" speaking given the overwhelming prevalence of R1b1b2 or is it just not thinkable?

I think you still strongly believe R1b1b2 is paleothical in Europe, right? You don't believe much in the datation of the mutations, am I correct?

Do you think it is the most plausible that all the taforalt H, V and U were originally associated with E1b1b1? I'd rather see some west Eurasian Y-DNA hgs with them (even though I can easily believe there were some mtDNA H-related and U-related haplogroup with E1b when it arrived in Morocco).
Do you think the north African E1b arrived from east Africa, north-east Africa (say Egypt) or maybe the near East (Is Natufian an option? - if it is, I guess It could explain better a high level in mtDNA H, at least)?


"If you want a copy of the paper, post with an email."

Yes, thank you. You can send it here, by removing the "anti-spam" parentheses : klg(wrsh(pr222@y)mail.c)om

terryt said...

'The Italian R1b1* seems to belong (by haplotype) to two different subclades, what makes Italy provisionally a good candidate for the origin of R1b1".

It could also mean that R1b1 came into Italy at two different times, by two different routes.

"the knowledge that paternal and maternal lineages do not need to follow the same patterns"

I'm sure you've argued elsewhere that they do necessarily follow the same patterns.

"Is Natufian an option? - if it is, I guess It could explain better a high level in mtDNA H, at least"

It could explain a lot of other things too.

Maju said...

The European paleolithic mtDNA might have had a much better success in having offsprings with the neolithic newcomers...

That's highly conjectural. First we'd have to assume that there was some sort of gender-biased kind of colonization. Not impossible but farmers colonists were not like the pastoralist raiders we know of more advanced times, who did that often, but more like genuine settlers who surely migrated with women, children and goats.

You can look at specific cases anyhow. In (non-Basque) Iberia Eastern Mediterranean Y-DNA is at most 20% in some areas (average could be 10% maybe), mtDNA of likely same origin is not much lower and overall autosomal Eastern Med is also in the 10% range. Similarly North African E1b1b1a(xE-V13) has reasonably similar apportions and distribution as mtDNA U6 in West Iberia.

So in general it looks to me that Neolithic males traveled with their women and children and mixed with natives in a manner that was not gender-biased.

I think you still strongly believe R1b1b2 is paleothical in Europe, right?.

I am as sure as one can be. There is absolutely no way that it could have reached the current density and distribution if it would be otherwise.

You don't believe much in the datation of the mutations, am I correct?.

Yes, I am highly skeptic of molecular clock datations. However most research papers date R1b1b in Europe to the Paleolithic. All that Dienekes/Vizachero speculation about R1b in Europe being Neolithic is a total nonsense and contradicts abundant academic peer-reviewed estimates.

Even if the correct dates would be "recent", Neolithic is only a few thousand years after the Epipaleolithic and Late UP demographic expansions.

However I still have many doubts, very specially about the process of arrival because it is nearly impossible to locate a West or South Asian originated migration into Europe after the original colonization at the MP/UP transition. So I am tempted to attribute it to Aurignacian (and after all Magdalenian is just a refined Aurignacian) and that goes well beyond what any MC estimates reach.

This is a problem that needs a solution. Part of my very tentative approach is to ponder that all age estimates are in fact too recent because of systemic biases in the MC hypothesis, like a too recent estimate of Pan-Homo divergence, under-estimation of what I'd call "conservative drift" (by which the dominant clade almost always drifts, in small population scenarios, out any novel mutated clade or minority arrival), etc.

Do you think it is the most plausible that all the taforalt H, V and U were originally associated with E1b1b1?.

I can't be sure but I'm inclined to think that there was an Epipaleolithic gender-biased sweep with Capsian culture. But this explanation is odd enough for me not to be sure either.

It is also possible that there was such gender-biased sweep in Europe in Epipaleolithic with the Tardenoise-derived (geometric microlithism) cultural pattern. This would explain the apparent higher diversity of R1b in Northern Europe.

All very tentative and cautious, of course. Neither clade looks "Neolithic" in any case.

Do you think the north African E1b arrived from east Africa, north-east Africa (say Egypt) or maybe the near East (Is Natufian an option? -

Upper Egypt (and/or nearby Nubia) looks like the most likely origin for E1b1b. Capsian culture originated over there and the Palestine-Egypt interactions in the Epipaleolithic can perfectly explain the crossed flows of E1b1b to West Asia (and eventually the Balcans as Neolithic founder effect) and of J1 and R1b1 to Africa. However notice that I suspect J1 might be even older in North Africa.

waggg said...

"All that Dienekes/Vizachero speculation about R1b in Europe being Neolithic is a total nonsense and contradicts abundant academic peer-reviewed estimates."

At Dienekes', in the comments of the paper of Underhill about R1a, Anatole Klyosov seemed to agree on the improbable datations. As I think you're a firm believer in the Kurgan hypothesis, how do you reconcile Underhill's statement that there weren't any movements from Europe in Central Asia during bronze age (especially as we know that Kazakhstan and south Siberia were full of Europoid mtDNA during bronze age (or at least we could say "west Eurasian"), associated with R1a1a, at least in south Siberia), with the Kurgan hypothesis?

Don't you agree that R1a1a7 seems to fit a Slavic dispersion rather than the Underhill's explanation?

And the matches between bronze and iron age south Siberia samples and the current individuals related to them, are kind of strangely distributed. Do they really fit so well into the Kurgan hypothesis?

I want to make clear that I myself favor the Kurgan hypothesis at the moment but I still have strong doubts.

Maju said...

Terry (somehow I missed your comment before):

It could also mean that R1b1 came into Italy at two different times, by two different routes.

It does not only have both lineages but also at least two other R1b1* clades (3 individuals: 2 closely related but the other not at all).

An Italy-Dalmatia to Turkey migration in the late UP is not unlikely at all (cave art connections), with Egypt becoming the next destination.

I'm sure you've argued elsewhere that they do necessarily follow the same patterns.

I'm sure the cases are very different.

"Is Natufian an option? - if it is, I guess It could explain better a high level in mtDNA H, at least"

It could explain a lot of other things too
.

Natufian can't explain much re. mtDNA H1, H3, H4 and H7, which are the lineages shared by SW Europe and NW Africa, plus U6 (which surely migrated in the opposite direction). None of these lineages exist or is frequent and old-looking enough in West Asia. Only a direct flow in the West can explain.

Also you have the JT(xJ,T) shared by North Africa and Italy - and nobody else.

Maju said...

As I think you're a firm believer in the Kurgan hypothesis, how do you reconcile Underhill's statement that there weren't any movements from Europe in Central Asia during bronze age.

Well, "believer" is not a word I would use: it has religious undertones.

Anyhow, I'm not sure if I know/recall all the data re this paper of Underhill (date, DOI, link?... some reference please!)

Anyhow, I don't treat Kurgan IEs as specifically European, after all before some Tsar decided otherwise, Europe ended at the Volga, and hence Samara was in Asia. I understand that this Euro-Asian border culture expanded first into Europe and Central Asia (Chalcolithic), from where it would eventually head southwards to the modern Indo-Iranian area. You should not see too much European-specific genetics associated with Kurgan expansion in Asia, unless it's something from the Samara basin as such or some back-flow hard to pinpoint in the archaeological record.

Also the West Eurasian nature of Central Asia is pre-Kurgan, dating at least to Neolithic and probably to the very origins of colonization of West Eurasia c. 50-40,000 BP.

Don't you agree that R1a1a7 seems to fit a Slavic dispersion rather than the Underhill's explanation?.

No. I discussed this at Dienekes (now I recall) and my point is that there is no clear Slavic association. The high presence of the lineage in Greece (but not in Southern Slavs) is a clear signal. The low presence of the lineage in Ukraine and Belarus is also signal of non-Slavic correlation.

Most people just simply do not understand the process of Kurgan spread in Europe and the central importance of Central Poland (and also somewhat East Germany) in the spread of the main branch leading to Corded Ware and modern Western IE languages. The Balcanic spread is even more complicated.

For me that lineage R1a1a7 is precisely a marker of Kurgan spread in Europe, which is not too related to Indo-Iranian spread, which is mostly a distinct phenomenon, nor even to the minor branch that lead to Hittites either.

And the matches between bronze and iron age south Siberia samples and the current individuals related to them, are kind of strangely distributed. Do they really fit so well into the Kurgan hypothesis?.

Not sure what you mean. My impression was that they were very much consistent. Anyhow, for what I know R1a1 could perfectly have a pre-Kurgan distribution, just that it was enhanced by Kurgan migrations. Like in the case of R1b, only a comprehensive detailed analysis can give us all the clues.

terryt said...

"Also you have the JT(xJ,T) shared by North Africa and Italy - and nobody else".

That's interesting. Does it indicate a movement between from Africa to Italy at some stage? What about any connection to Iberia?

Maju said...

It's a paragroup and AFAIK nobody has studied it yet in any depth. It might be a migration in either direction or two different minor branches of JT. Not located in Iberia nor West Asia afaik (this is mentioned by the study on Taforalt aDNA, but not sure its source).

German Dziebel said...

Is there a reasonably common pattern in age and distribution between mtDNA U6 and R1b1?

Maju said...

German: they have no apparent correlation (at least in distribution, that is always more clear). U6 is found in Sudan but not (or not significatively) in Central Africa. U6a probably coalesced at the Nile but overall the highest diversity of U6 is in NW Africa, followed closely, by Iberia. Not sure if there is a local subclade (tiny) in Sardinia too.

I understand that U6 probably represents the early colonization of North Africa from West Asia (Dabban industries, similar to the widespread Aurignacoid complex of that period), H and V the second wave from Iberia (Oranian or Iberomaurusian culture), U6a and various L(xM,N), together with Y-DNA E1b1b, probably the Capsian culture from the Nile area (Afroasiatic language probably) and I'm not sure about mtDNA K and Y-DNA J1, which are also relevant in NW Africa.

Hence the migration of R1b1a into, probably, Sudan and then Central Africa seems to represent a different process and a male-biased one, at least in what regards Central Africa.

One very tentative possibility could be that R1b1/R1b1a migrated indeed with U6 and other Eurasian-originated lineages such as mtDNA X and Y-DNA J1. But there could have been more than one migration (cf. the rock art of Upper Egypt, for instance). The archaeological record is not sufficiently good, I believe, to make a clear judgment.

German Dziebel said...

"One very tentative possibility could be that R1b1/R1b1a migrated indeed with U6 and other Eurasian-originated lineages such as mtDNA X and Y-DNA J1."

European populations pushed down south by the Ice Age: some ended up in southern European refugia, others in Africa?

Maju said...

It's not conclusive that R1b is European. R1b12, which is the main European subclade looks quite more haplotype diverse towards West Asia, and this one also appears so far centered in Africa. So it may be the other way around. It's not too clear cut.

Edfu murals are anyhow dated to c. 15,000 BP (Balanian-Silisian culture but admittedly I know nothing of its toolkit and affinities), what is coincident with post-LGM and with Magdalenian expansion. Upper Egypt anyhow, like Anatolia further north (where possibly related Baldibi is), seems like a crossroads of various regions, with "traffic" now flowing in different directions at different times.

Maju said...

Erratum:

R1b12, obviously means R1b1b2.

German Dziebel said...

I can see the ambiguity. This new study just invoked in my memory an earlier one, which highlighted a surprising connection on the level of a U5 subclade between Saami and Berbers: "Thus, although these previous studies have highlighted the role of the Franco-Cantabrian refuge area as a major source of the hunter-gatherer populations that gradually repopulated much of central and northern Europe when climatic conditions began to improve ∼15 ky ago, the identification of U5b1b now unequivocally links the maternal gene pool of the ancestral Berbers to the same refuge area and indicates that European hunter-gatherers also moved toward the south and, by crossing the Strait of Gibraltar, contributed their U5b1b, H1, H3, and V mtDNAs to modern North Africans." (Achilli et al. Saami and Berbers—An Unexpected Mitochondrial DNA Link, p. 885).

Maju said...

Good finding, German. I was unaware of that paper, so thanks for the reference. It certainly reinforces the idea, specially as the highest basal diversity for U5b1b is in SW Europe (though looks more like Italy than Spain - France is not mentioned, even if it was the main "refuge" area).

German Dziebel said...

More on the same post-LGM migration into North Africa in: Mitochondrial DNA haplogroup H structure in North Africa, by Hajer Ennafaa et al. (2209).

German Dziebel said...

My point is: this migration or migrations may have brought Afroasiatic speakers into Africa, no? I've been a believer in the African origin of this family, but then this article on R1b1 and the origin of Chadic in North Africa made me somewhat doubtful.

Maju said...

I have read Enaffaa's work, thanks. It's commented elsewhere in this blog.

Anyhow, the migration was surely more in the very LGM period (late Middle UP), when Solutrean and Gravettian cultures merge in the Iberian province, creating a unique culture (sometimes included in the general Solutrean maps but in fact very strongly Gravettian) that is probably at the origin of Iberomaurusian/Oranian culture (whose oldest dates are in Morocco, whose tool types seem related to those from SE Spain and Portugal in the same period and whose athropometrical type is considered a Cro-Magnon variant).

When people talk of post-LGM migration, they mean the Magdalenian expansion and maybe parallel (but not too well known) similar processes in Eastern Europe. However that would be c. 15,000 years ago, while the Iberomaurusian genesis is more of c. 22-20,000 years ago, quite older.

As no other process can explain the flow from Iberia to North Africa, this implies that the lineages involved existed already in the Gravettian age. And is one of the very reasons why I dispute the usual MC age estimates. Not just H but H1, H3, H4 and H7 (and also probably other lineages as this U5b branch) must have existed that long ago.

The presence of some (minor but noticeable) links to Italy suggests to me that Gravettian arrived to SW Europe via this country. Alternatively there could have been a separate crossing from Sicily but the evidence in favor of this is rather inconclusive.

My point is: this migration or migrations may have brought Afroasiatic speakers into Africa, no? I've been a believer in the African origin of this family, but then this article on R1b1 and the origin of Chadic in North Africa made me somewhat doubtful.

Afroasiatic is just too recent in time (in spite of being probably the oldest dated language family) to have been implied in this phenomenon. It was surely involved in the Capsian backflow though, which has its origins at the Nile. I'd rather believe that some allegedly Vascoid words that are found in Berber may represent this Oranian substratum (though too unclear - and could also date from the Megalithic period).

German Dziebel said...

Good. Thanks.

waggg said...

Good finding, German. I was unaware of that paper

For the records, I actually had mentionned it (without naming it) and gave you the URL here, towards the end of the comments :)

http://dienekes.blogspot.com/2009/11/sex-biased-admixture-in-americas.html

(the Thursday, December 03, 2009 1:26:00 AM comment)

Maju said...

I see: seems I got the wrong concept because the paper does mention a North African connection and I failed to notice it in that discussion.

Actually Argiedude mentioned it as an isolated Dakar lineage closest to Finnish clades. But that is not what this paper says: within this paper U5b1b is split in four basal sublineages:

1. Italian (22)
2. Spanish-Italian (23, 24)
3. Saami-Yakut (25, 26, 27, 28)
4. Berber-Senegalese (29, 30)

This division makes sense from the viewpoint of a distribution from SW Europe (or rather Italy), with the African and Northern lineages being local offshoots.

I don't understand why it's differently organized in Phylotree, because the reference is this same paper.

O_O

German Dziebel said...

"This division makes sense from the viewpoint of a distribution from SW Europe (or rather Italy), with the African and Northern lineages being local offshoots."

I read it in the same way. A newer version of Uralic dispersals model in linguistics places the homeland of Uralic in East-Central Europe with the Samoyed and Ugric branches being result of a long-distance eastward migration. (The previous model argued for the opposite direction of migration.)

Maju said...

Uralic languages spread in Europe with all likelihood in a Neolithic timeframe, beginning in the Middle Volga and largely paralleling at more nordic latitudes (ecological specialization of the Uralic ethnos and most Y-DNA N in general) the spread of Indoeuropeans, with whom they have at least a sprachbund linguistic relationship (see Indo-Uralic and Comb Ceramic Culture).

Maju said...

Again we see a pattern where the ethno-cultural adscription is largely related to Y-DNA and not mtDNA (patriarchal or at least patrilocal/-lineal). Of course, there's some mtDNA arriving with the Y-DNA (CZ and D) but at much lower levels. Overall modern Finnish look European and their autosomal apportion of Siberian blood is c. 15% only.

This case is probably similar to that of North Africans, mutatis mutandi, and maybe other peoples here and there.

terryt said...

"Again we see a pattern where the ethno-cultural adscription is largely related to Y-DNA and not mtDNA (patriarchal or at least patrilocal/-lineal). Of course, there's some mtDNA arriving with the Y-DNA (CZ and D) but at much lower levels".

And another example of the relative independence of Y and mtDNA haplogroups. You seem quite prepared to accept this as being a common situation in more recent times, but have immense difficulty as seeing the same phenomenon operating in more ancient times. Why?

Maju said...

Because I don't see the mtDNA even at small numbers.

Anyhow Uralics were pastoralists, not hunter-gatherers. Property and in particular pastoralism reinforces the status of males and weakens that of females.

Bradly Jones said...
This comment has been removed by a blog administrator.
Maju said...

^^ Spam: no, thanks.