As I said yesterday I am these days working on updating and re-drawing my ancient mtDNA maps series. But I do not want to be a mere parroting of what authors reported on their papers, often long ago or with questionable criteria, as I did in the past, so I am dedicating a lot of time to double-checking each sequence with the help of PhyloTree. It’d be a lot easier to merely replicate their assumptions but I would feel uneasy about that.
By the moment I have double-checked all Paleolithic and Epipaleolithic (“Mesolithic”) sequences, with the line drawn at the arrival of farmers to Central Europe c. 7500 years ago (exception will be made to accommodate older Euphrates Neolithic data into the Neolithic map rather than in the Late UP/Epipaleolithic one, where it’d be odd). However I have not yet been able to draw aesthetically decent pie charts, so by the moment here you have totally ugly draft maps with mere text indicators (what has one advantage: I can more detailedly name each haplogroup, when known).
Early and Middle Upper Paleolithic (30-17,000 BP):
|click to expand|
Most noticeable is, as I advanced yesterday, that mtDNA H appears to be confirmed already some 25,000 years in Sunghir, Russia. Say whatever you want but H17’27 is the only extant haplogroup that fits with that sequence.
Caution with the Paviland (Wales) site because the DNA was not cloned, what makes it pretty much unreliable.
CRS sequences are still likely to be H or HV but this cannot be confirmed, specially as some aDNA with CRS HVS-I sequences happened to be U* in fact after checking the Coding Region transitions. Technically CRS may mean a wide range of haplogroups downstream of R but most are exotic in relation to Europe and in practical terms it tends to indicate H or HV clades, at least for the vast majority of modern sequences.
A similar case are those labeled R*, which means R* (non-CRS), the mutations detected do not seem to point to any modern haplogroup.
L3* is usually said to be N* but this cannot be correct in the Paglicci case because of the transition at 10873C, so it must be L3(xN). It may still be M but this is uncertain.
Another interesting detail is the finding of the first known U5 and JT (JT*) in Solutrean Iberia.
Late Upper Paleolithic and Epipaleolithic (16-8,000 BP):
|click to expand|
This map has a lot of interesting stuff. Most noticeable maybe is the geographic restriction of R* plus H to the SW parts of Europe and North Morocco, indicating maybe a pruning of everything non-U in the Eastern parts of Europe where the effect of the Last Glacial Maximum was most noticeable (please, read this post).
As I said before we cannot be sure that R* (with or without CRS HVS-I sequence) is H or HV but there is some high likelihood, specially as unmistakable H1b shows up in Portugal (if the case of Sunghir did not persuade you already). Importantly in this sense is that there are no migrations into Europe between Gravettian (Sunghir) and Epipaleolithic (Portugal), unless one wants to argue for a North African origin of mtDNA H, which seems quite hard.
While L3* can be N (or M), there is one Portuguese sequence, originally identified as N*, that quite clearly says L3d2, a haplogroup found (as far as I know) only in Burkina Fasso, however relative L3d1″5* is found in Syria, among other places (Chad, Kenya), so maybe it has a West Asian origin (ref. post), however a North African origin is more likely in my opinion.
Another interesting finding is that U4 was not, it seems, restricted to NE Europe in this period but shows up also in Iberia and Morocco. However the Portuguese case’s identification is not 100% sure, as one of three HVS-I transitions is not found in modern sequences (yet the other two define U4a3, otherwise dump it in the R* category).
All sources and full sequences can be found at Building History’s ancient DNA page (thanks to Jean again for her excellent job). Please feel free to point out any error I may have committed, thanks in advance.