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.

Sunday, August 31, 2008

Geographic patterns of European Y-DNA (1)

Some like to argue about alleged haplogroup age estimates based on various versions of the molecular clock hypothesis. I just shrug at that. I find geographic and (pre-)historical patterns to be much more substantive. In what regards to me, they should use them to modify their estimates appropiately and maybe that way they would reach to some reasonable MC estimation system.

It's just common sense. For example, Dienekes was recently arguing that haplogroup E-V13 (E1b1a2 by the 2008 nomenclature) had to be of the Bronze Age (he has his own even shorter version of MCH) but I protested that the distribution pattern is only consistent with Neolithic spread (and by no means with Greek historical coastal colonization). A map is worth a thousand words:

The above map's pink shades show the density of E-V13 (the increase towards Finland is just an effect of lack of data for that region though). On it I have added blue lines representing the extent of Balcanic-derived Neolithic waves in Europe. BN stands for Balcanic Neolithic (including Sesklo-derived, Eastern Lineal Pottery and Grey-Black Pottery cultures), and the thick oval in it is the Thessalian region of origin (notice that Sesklo-Dimini, at the origin of most European Neolithic, is as high in E-V13 as Albania, if not even more). DN stands for Danubian Neolithic (aka Western Lineal Pottery) and MN for Mediterranean Neolithic (aka Cardium Pottery), both more or less clearly related to Sesklo and the Balcanic Neolithic.

For me the pattern is very clear: it is a Neolithic clade. And there is no way it can have another explanation.

Another clade often subject to discussion among the MCH fans is the most common Western European haplogroup: R1b1b2 (formerly R1b1c). Some argue it should be Neolithic or even more recent. But its distribution pattern is by no means coincident with Neolithic expansion.

Again a map is worth a thousand words. Even better: two maps:

The above maps show (roughly) the density of R1b1b2 (red-pink shades). Map A shows the origin (purple oval and star) and expansion (blue lines) of Magdalenian culture, as well as of epi-Magdalenian derivates (green). I understand that the Magdalenian-origin interpretation is the only possible one, the patterns being almost exact.

Map B illustrates the only possible Neolithic (or rather Chalcolithic) explanation for R1b1b2 spread: Western Megalithism. The purple oval and star indicate the origin, the blue lines max. expansion and the green lines those areas lost to Indoeuropean expansion c. 2400 BCE. Nevertheless, I understand that Megalithism was not related to any major demic expansion but it was mostly a cultural phenomenon spreading into native groups. We cannot exclude some localized colonizations and it is even possible that some haplotypes might have spread with this cultural/religious phenomenon - but that would be all.

The only intriguing issue is the relatively high R1b1b2 in Italy, that was not part of Magdalenian culture but was of Western Gravettian, where the clade may have its ultimate source. It could be argued that higher density in the subalpine area means Indoeuropean infiltration in proto-historical times from the other side of the Alps. Alternatively it could mean that the origin of R1b1b2 is a Western Gravettian founder effect (with R1b1b1 being maybe part of Eastern Gravettian). Whatever the case, there is no logical way it could be of Neolithic or Chalcolithic origin.

And that's all for today.


terryt said...

"Again a map is worth a thousand words". I've just looked at tthe maps again and something occurred to me. How about the possibility that R1b1b2 actually evolved along the western European shoreline before what I know as 'The English Channel' opened? It was this channel that separated the Spanish and Irish branches.

Maju said...

Terry: there are no "Spanish" and "Irish" branches, except if you wish for some minor subclades. The vast majority of R1b1b2 (or even R1b as a whole) is R1b1b2*, private lineages within the haplogroup and only a handful belong to sublineages within it. What means they have been going on without meaningful events since whenever R1b1b2 expanded.

These minority subclades and their geography, mostly still ill-studied, show only localized distributions (North Sea, Ireland-Scotland, Basque Country, North of the Alps...), what means very limited local founder effects.

No two branches, just one (or millions). If Brits and Iberians are different it is not readily visible in the Y-DNA. But that can also be said of many South Asians and Eastern Europeans, so it's only so good. We must look at other elements to see the differences and, among them, mtDNA shows some differences (Briton/Irish mDNA is closer to that of mainland NW Europe) while nDNA clusters distinctly within Europe forming (at least) three groups in the Magdalenian area: Iberian, Basque and NW European (Bauchet et al, 2007), what again suggests very strongly the three Upper Paleolithic provinces involved in Magdalenian and highlights their small yet visible differences.

We can't judge properly what importance the now submerged shoreline had in the past. Maybe you have some reason after all and it was central. But from the non-submerged sites we can muster that the highest demographic density was surely in the Franco-Cantabrian region, specially in Dordogne (a riverine but inland area). Instead Northern France seems spasely populated, what (together with the few British sites) seems to suggest that the Channel basin was not demographically important at any time.

Instead Central Europe (Rhin-Danub province) shows many more signs of inhabitation, specially after the LGM and the genetic affinities of the Brits do suggest that they were mainly an offshot of the peoples of that area, who surely colonized the islands in the Epipaleolithic period.

This does not exclude further migrations along the coasts, specially in the Chalcolithic period possibly. The R1b haplotype structure would seem to support that up to a point, as well as some details of the nDNA structure.