July 09, 2008

mtDNA macro-haplogroup R0

R0 is ancestral to the very widespread HV, V, and H which are frequent in Europe, as well as R0a which is frequent in the Middle East.

BMC Evol Biol. 2008 Jul 4;8(1):191. [Epub ahead of print]

Timing and deciphering mitochondrial DNA macro-haplogroup R0 variability in Central Europe and Middle East.

Brandstaetter A, Zimmermann B, Wagner J, Goebel T, Roeck AW, Salas A, Carracedo A, Parson W.

ABSTRACT: BACKGROUND: Nearly half of the West Eurasian assemblage of human mitochondrial DNA (mtDNA) is fractioned into numerous sub-lineages of the predominant haplogroup (hg) R0. Several hypotheses have been proposed on the origin and the expansion times of some R0 sub-lineages, which were partially inconsistent with each other. Here we describe the phylogenetic structure and genetic variety of hg R0 in five European populations and one population from the Middle East. RESULTS: Our analysis of 1,350 mtDNA haplotypes belonging to R0, including entire control region sequences and 45 single nucleotide polymorphisms from the coding region, revealed significant differences in the distribution of different sub-hgs even between geographically closely located regions. Estimates of coalescence times that were derived using diverse algorithmic approaches consistently affirmed that the major expansions of the different R0 hgs occurred in the terminal Pleistocene and early Holocene. CONCLUSIONS: Given an estimated coalescence time of the distinct lineages of 10 - 18 kya, the differences in the distributions could hint to either limited maternal gene flow after the last glacial maximum due to the alpine nature of the regions involved or to a stochastic loss of diversity due to environmental events and/or disease episodes occurred at different times and in distinctive regions. Our comparison of two different ways of obtaining the timing of the most recent common ancestor confirms that the time of a sudden expansion can be adequately recovered from control region data with valid confidence intervals. For reliable estimates, both procedures should be applied in order to cross-check the results for validity and soundness.



Maju said...

The conclussion is a little misleading because it reads: Given an estimated coalescence time of the distinct lineages of 10 - 18 kya... and then you see in fig. 5 that R0a has a coalescence age estimate of 25-35 KYBP (!!!). The 10-18 KY age is only for HV and H, not for the whole R0 clade.

I also find strange that, being known since the beginning of the century that HV (or at least H and V) probably spread around from Western Europe, this region has not been sampled at all. Similarly I miss a sampling of Central Asia, where some H subclades are most concentrated.

Dienekes said...

I also find strange that, being known since the beginning of the century that HV (or at least H and V) probably spread around from Western Europe, this region has not been sampled at all.

V probably spread from Western Europe, not HV or H.

Maju said...

It's possible. Darn! I have to research that Badgoulian connection mentioned by Pereira 2005 as origin for the Magdalenian.

I was persuaded that Magdalenian was a local Franco-Cantabrian developement (the consensus till some years ago), what basically forbade on archaeo-cultural grounds any LUP arrival from West Asia or even Eastern Europe but maybe it's not the case after all.

Has anybody some further info on this Badagoulian missing-link culture?

Maju said...

Ok. Found something on this Badegoulian transition: it is a French (Franco-Cantabrian) technology that is believed to be influenced (or originated) by very late Aurignacian of Central Europe (Germany, Switzerland, Austrian/Czech border).

It seems it still makes very difficult to think of any West Asia -> Europe migration any time between Gravettian and Neolithic. Aurignacian is after all a European-only culture (even if its origins are maybe in Western and Central Asia) and this is nothing but a derivate, a transitional derivate between Aurigancian proper and the refined version of it that is Magdalenian.

Instead there are connections that can well be interpretated the other way around. Baldibian seems to be influenced by Magdalenian, at least in its artwork and the Epipaleolithic colonists of the Zagros (can't remember the culture's name right now) seem to be epi-Gravettians from Eastern Europe.

It does seem that H* has some of an E->W cline, but the core of the H starlike structure (H*) is anyhow still very very frequent in the West as well, so I have no strong reason to think H coalesced in West Asia, specially as we know that its sister clade V spread from Western Europe too.

If you look at Pereira's graph, it seems evident that while the West Asian (NRE) share of the H* cake is important, that is also true of the NW area (described as "France, Ireland"), followed in importance by the SW one (Iberia). Half of that cake belongs to the Franco-Cantabrian region and its immediate hinterland, while only some 30% is West Asian.

Of course, the case for a western origin is even more clear for the subclades, for all subclades mentioned in that paper: H1, H2, H3, H4, H6, H7 and H13, with only H5a posing the same relative uncertainty as H*. Sure that Pereira seems to have ignored Central Asia, where some H clades seem more relatively important than in Europe. That's my only caveat to a European, a Western European origin for H.

I could say "at least the major fraction of H" but that looks very unlikely considering the starlike structure of the haplorgoup, that strongly suggests a single origin and a rapid expansion, compatible only with the post-LGM circumstance. There is no archaeo-logic that can justify such a rapid expansion from West Asia after Gravettian (assuming it would have a West Asian origin) and before Neolithic.

(And this is the same case for Y-DNA R1b, by the way).

Dienekes said...

so I have no strong reason to think H coalesced in West Asia, specially as we know that its sister clade V spread from Western Europe too.

What its "sister clade" did is irrelevant.

Also, there are very good reasons to think that haplogroup H originated in West Asia and not in Western Europe and there is no reason to think that Western Europe is its point of origin.

(And this is the same case for Y-DNA R1b, by the way).

Similarly there is no reason to think that Western Europe is the point of origin of haplogroup R1b, since Western Europe only has a derivative form of R-M269.

Maju said...

If you are right, Dienekes, then the age guesstimates must be reviewed, most probably to quite older ages. There is just no archaeological logic that could mean a massive migration from West Asia into Central/Western Europe in the LUP, as these age estimates suggest.

It just make no sense, sorry.

Western Europe only has a derivative form of R-M269

You must mean that Western Europe has only the derivative form R-M269 (R1b1b2, formerly R1b1c). This is not actually exact but is close to reality. You cannot forget Sardinian R1b1a (is Sardinia in West Asia? Nope!) nor the fact that the structure of the whole R1b haplogroup is not yet fully resolved.

But, ok, let's talk about R1b1b2, which is actually like 95% of all the R1b in the world anyhow. This clade is clearly West European by origin and I doubt anyone will challenge that, even if the whole R1b may be of Anatolian origin.

In any case, it cannot fit with any LUP migration, at least on light of the archaeological record. As I trust infinitely more the archaeological record than the MCH guessimates, I tend to think that the usual estimates are just necesarily wrong.

Crimson Guard said...

Theres no evidence that living people in those areas are the descended from Cromagnon cultures.

The R1b thing is actually very weak and only put fourth under the bad assumption that Basques are remnants of them and they colonized the North West of Europe--R1b is actually much younger than previously thought.

And the only known genetic testing on Cromagnon remains showed them to have Near Eastern Affiliations. R1b besides came from the Middle East as well, it wasnt like god or primordial ooze sprinkled R1b on the soil and sprung up skeleton's outve the ground in Western Europe miraculously kinda like something Ray Harryhausen's "Jason and the Argonauts" fame.

Dienekes said...

But, ok, let's talk about R1b1b2, which is actually like 95% of all the R1b in the world anyhow. This clade is clearly West European by origin and I doubt anyone will challenge that, even if the whole R1b may be of Anatolian origin.

Incorrect, R-M269 may be frequent in West Europe but it is much more diverse in East Europe and West Asia. There is no reason to think that R-M269 originated in Western Europe. Its high frequency there is probably the result of the fact that Western Europe has been less affected by subsequent migration movements.

Maju said...

Crimson: I'm not sure what you mean by "Cro-Magnon", but whatever the case all archaeological evidence points to a large degree of continuity at least in Atlantic and Western Mediterranean Europe in the Neolithic transition.

If you bother in researching for instance Cardium Pottery culture a little bit, you'll find that there was only occasional colonization and much more commonly partial aculturation (agriculture and pottery imported along with local Epipapelolithic tool sets). This is not replacement, clearly not. The case is even stronger for Atlantic peoples, totally marginal for the Neolithic advance (maybe for climatic reasons, maybe because the Andalusian and Portuguese Neolithic acted as barrier). It is less clear in the case of Central Europe (Danubian Neolithic) though, where cultural transformation is nearly total - but in any case starting off not from Asia but Hungary (and ultimately derived from northern Greece).

Obviously there was some populaton flow from West Asia into Europe, and from the Balcans into the rest of the subcontinent (and even from Northern Africa via Spain) in the Neolithic but neither archaeology nor the current genetic distribution suggest any sort of large replacement from West Asia. There were migrations and probably even localized replacements but these had a more limited geographic scope. Neolithic is a very bad suggestion for demic changes other than localized ones, like the arrival of Y-DNA E to Greece and Iberia, the spread of J in some Mediterranean areas or that of I2 possibly.

If Neolithic is not, the only options I see are:

1. The early colonization c. 40,000 BP (Bakokirian and Aurignacian)

2. The second possible wave c. 30,000 BP (Gravettian, to what the Cromagnoid type is most directly linked)

The LUP shows no waves from West Asia but may mean some difussion to Anatolia from Europe (Baldibian), as well as the famous post-LGM demic expansion from the FCR, nor does the Epipaleolithic, that can at most justify migrations from the Rhin/Seine to the SW (Tardenoisian-derived influences), as well as the epi-Magdalenian colonization of Northern Europe. Again the Epipaleolithic shows instead a migration from epi-Gravettian Eastern Europe into the Zagros area.

That is what we have. We also have very modern-looking human remains in the Magdalenian, including the first known impacted wisdom tooth (due to major reduction of jaw size, typical of this period).

To make a case for a Neolithic or post-Neolithic replacement, you'd have to work really hard, as the pretty complete archaeological record rather suggests non-replacement, at least in the West and SW. And I won't accept anything going against the grain of archaeology, specially if it's founded in mere MCH guesstimates and nothing more.

So the only realistic options are Gravettian and Aurignacian. If that means that the whole set of haploid date estimates must be pushed back a lot, I really have no big problem with that, as the South Asian archaeological record can well support H. Sapiens presence since c. 150,000 BP.

Maju said...

Oops! I meant c. 105,000 BP. Sorry.

Ebizur said...

maju said,

"You must mean that Western Europe has only the derivative form R-M269 (R1b1b2, formerly R1b1c). This is not actually exact but is close to reality. You cannot forget Sardinian R1b1a (is Sardinia in West Asia? Nope!) nor the fact that the structure of the whole R1b haplogroup is not yet fully resolved."

Haplogroup R1b1a-M18 has been found among Druze and general Muslim Lebanese as well as in Sardinia. It's silly to claim that the presence of haplogroup R1b1a in Sardinia somehow makes it a strong candidate for the origin of haplogroup R1b; it's a single island in the middle of the Mediterranean Sea, and not even a particularly fertile one. Sardinia is not the sort of place one should expect to have served as a source of population dispersal.

If anything, Lebanon (or ancient Phoenicia) would be a much better candidate for the origin of haplogroup R1b if we were going to place so much emphasis on the tiny R1b1a-M18 subclade.

Maju said...

Haplogroup R1b1a-M18 has been found among Druze and general Muslim Lebanese...

That I didn't know. It's an interesting piece of info certainly, even if the Druze don't seem to be representative of anything (they are too much of an outlier group in the area), other Lebanese may be.

In any case, Sardinia, like most Med islands was colonized in the Neolithic, not before. I didn't mean that it was the center of anything just that it's closer to Western Europe than to anything else (except Italy, of course).

Other Sardinian clades (I2) are related to the Western Balcans and SE France, so I did not think of any extra-European origin and I still think that R1b is not sufficiently well mapped. For instance Ysearch yeilds almost 200 R1a* (not even looking at subclades) for their "Western Europe" region (didn't look anywhere else), most of them in Central Europe in fact (and Greece too). It is not like the other R1b subclades do not exist in Europe but that they are tiny in comparison with massively widespread R1b1b2.

I can anyhow tentatively agree with an Anatolian origin for H and R1b but I do have a real problem with a post-Gravettian timing for the expansion to Central/Western Europe. That is a serious problem.

pconroy said...


What I think you're missing here is that farming is something like 20 times more productive than simple hunting/gathering - so that it doesn't take a massive migration of people from Anatolia or Lebanon/Phoenicia to produce major demographic change. It just takes a few thousand years for farmers to completely dominate hunter/gathers numerically. If these farmers also brought novel diseases/pathogens with them, which is likely, then only a few hundred years - like the conquest of North America.

I should mention again, that I have a very close match - GD=1 at basic 12 markers - with a Syrian, who describes himself as of Arab descent. I am R1b1b2e, M222+ (formerly R1b1c7) and this guy is R1b1b2. On another forum I found that a number of other guys, from the Ireland or Britain, also had similar matches as I did. Now of course this could be due to crusader DNA ending up in that area, but it is also as likely that this might be the homeland of much of Western Europeans?!

Maju said...

I am not missing that. I realize that the model could hypothetically work (more or less). But I contrast the model with the archeological evidence and this tells us that when Neolithics (Cardium Pottery) arrived to Mediterranean Western Europe, they colonized some areas/spots but in most cases, the natives were assimilated. We can't know the details but we know that while they adopted the pottery, the farming, the sheperding and probably the sailing innovations... they also kept their Epipaleolithic stone tool traditions in most places.

So if anything this is a story of admixture, of cultural assimlation maybe but not of population replacement. Because once the natives had acquired the Neolithic techs they were able to produce at the same rates. And in this case they adopted them very quickly.

Also archaeology doesn't clearly speak of a direct West Asian origin. The core are of Cardium Pottery were the Western Balcans (Bosnia, Dalmatia, coastal Albania and Montenegro). It had a precendent in early Neolithic Greece (called in some books proto-Sesklo, different from pre-Sesklo but in the same early layers) and an unclear connection in Lebanon (Amuq-Biblos culture, but only the Biblos facies), where Cardium Pottery also became common for some time.

Obviously there have been many other connections between West Asia and Mediterranean, even Atlantic Europe, not just the crusaders. Muslims, Romans, Phoenicians, Greeks (both Classical and Mycenaean), Jewish diaspora, and many ill-known prehistoric cultures (Cypriot influence, Megalithism, Sea Peoples...) that have scrambled peoples all around. Syria to Britain is possible but Britain to Syria is not impossible either.

Actually, West Asians, Arabs, see the Syrians as a very "admixed" people. I can't know if that's true but I've already read several times that Syrians (and Lebanese) are the most European-like Arabs. Why exactly, I can't say. It's not like the Crusaders ever conquered Damascus, so it must be older connections.

I strongly think that West Asians and Europeans are tightly related. But I don't think that can be oversimplified to only E>W migrations nor to a single one in the Neolithic. That's not what we can see, neither in archaeology nor, IMO, in genetics. I think we were already first cousins before Neolithic began and that phenomenon only intensified our ties, it did not create them.

Maju said...

Btw, former R1b1c7 is a typical, almost exclussive, Irish-Scottish clade, probably a local founder effect. I think it's extremely unlikely it originated in Syria, sincerely.

Maybe it's different in other cases but in this case, I think it's the Syrian who has a British/Irish ancestor and not the other way around. Crusader? Roman legionaire or slave? Someone who joined the Phoenicians? Impossible to tell for sure.

pconroy said...


Well I think Iberia can be compared to Madagascar, in that in the mountainous uplands of the latter, we find people who are genetically similar to Malayo-Polynesians - providing the earliest evidence of occupation - even though the coastal areas are occupied by peoples of African and Arab descent.

While in Iberia we have people on the central plateau who are 33% Y-DNA I, and this percentage rapidly declines on all sides - so perhaps these are descendant of the original population.

Maju said...

Hmmm... rather not. I am not sure where you get those figures of 33% I but in any case the Central Plateau is not like the that has been inhabited for longer nor that has been immune to demic penetration. Not at all.

In the Paleolithic it was a mostly desertic area, with the most densely populated areas in the northern Atlantic ("Cantabrian") strip and in the East/SE Mediterranean coasts. There are some indications of human presence but they are rare. The area was mostly colonized in the Epipaleolithic and Neolithic. Consider pleas that the most fertile part of the plateau is 700 m above sea level. It was not frozen in the Ice Age but it was very very cold in any case. In general the interior areas, even today, tend to be dry and move between the extremes of temperature.

The semi-deserted Plateau only shows a distinct local culture, of seminomadic pastoralist character, in the Bronze Age (Cogotas I), that was succeeded in the Iron Age by Cogotas II of mixed but mostly Celtic character. Not only the Celts found it easier to conquer the plateau than the coasts but later also the Goths, who mostly settled in that area. Haplogroup I may partly have travelled with Indoeuropean migrations of that kind, though it may also be partly of Neolithic (Balcanic) origin.

Overall, even today, Iberia has alway been more densely populated in the coasts. True that the coasts are in principle more exposed to foreign arrivals, specially those of the Mediterranean, but tribal Indoeuropeans at least seem to have prefered the interior, maybe easier for their horse-based combat tactics.

You also find a lot of I in France for instance, and it's a much more exposed region. Yet this frequence decays cleary to the SW. Are you telling me that Gauls were older in the area than Basques/Aquitanians? Or than Ligurians and Iberians? That is not realistic nor makes any sense in relation with all we know. Probably Ligurians and Iberians had an important Eastern Mediterranean component (Neolithic and of later times maybe too) and you can see it in genetic elements such as Y-DNA J (or E in the case of southern Iberia). But these components are relatively small. And you see the same picture in mtDNA, except that then there is no plateau specifity (AFAIK).

I'd rather think that most of that I component of the Plateau is Celtic (or otherwise Central European) from the Hallstatt period. Though Goths may have also left some mark.

Unlike Madagascar, Iberia was never desert but, specially, it was never tropical. So the plateau is climatically rather disadvantageous: a semi-arid region of lesser interest. Only with the consolidated Neolithic it could be exploited in some meaningful way. But even today it's only relevant because it's between the coasts, not because of its own merits. Excepting Madrid (that only exists as big city because it's a political capital), the whole plateau has now like 4 million people. Not much more than Lithuania or Switzerland, in spite of being much larger, and less than 10% of all the peninsular population. It's little more than a passage, pastures and cereal fields. And it was that even much more in the time before Castile existed.