November 20, 2012

Ancient DNA from Eastern Europe and Sardinia

A very interesting doctoral thesis has been known these days (h/t Jean). The thesis by Clio S. I. Dersarkissian (directed by A. Cooper and W. Haak) includes novel ancient mtDNA from North Eastern Europe (Karelia and surroundings) specially and also some Scythian and Sardinian burials from the Metal Ages.

Clio Simone Irmgard Dersarkissian, Mitochondrial DNA in ancient human populations of Europe. University of Adelaide, 2011 (thesis). Freely accessible ··> LINK [identifier: http://hdl.handle.net/2440/74221]

The most interesting findings may be those from Karelia:
  • First pre-Neolithic mtDNA H in Northern and Eastern Europe and one of the few findings strongly confirmed in such haplogroup before Neolithic. It clearly reinforces the already well established notion that mtDNA H existed in Europe before the Neolithic.
  • U2e - which might well be descendant or otherwise related to the U2 of Kostenki.
  • C1 - suggesting pre-Neolithic Siberian influences in Northern and Eastern Europe. The specific sublineage (named as "C1f") has not yet been sequenced elsewhere.
There are some more interesting data regarding ancient NE Europeans, Scythians and Sardinians but let's see that by parts.


Epipaleolithic peoples from Karelia and Northern Russia


Possibly the most impacting findings of this paper are those regarding two Epipaleolithic sites in Karelia (Uznyi Oleni Ostrov) and nearby parts of Northern Russia (Popovo, in Russia proper but not far from the Karelian border), as well as one more recent site from Sápmi (Lapland).



As I mentioned above, the U2e and C1 ("C1f") findings are unusual and suggestive of ancestral connections with Kostenki (Early Upper Paleolithic site from Southern Russia with U2 mtDNA) and Central Asia and Siberia. In fact an overall comparison with modern populations, shows strong affinities with West Siberians and Uyghurs for these Epipaleolithic Karelians.

Instead the Bronze Age Sami site shows more generic or distributed Siberian affinities, although there are populations in West Siberia (Nenets?) that also fit well with that mtDNA genetic pool. Bashkirs show similar affinity to both ancient populations (see ch. 1, fig. 3 - p. 103).

Not shown here are the results for the 18th century Sami site of Chalmny-Varre, which look a very modern Sami mtDNA pool, dominated by V7e and complemented by U5b1b1 and U5a1. 

Confirming the existence of mtDNA H in pre-Neolithic Europe


I really want to underline this, because certain influential people have been dead set into denying the existence of mtDNA haplogroup H in Europe altogether before the Neolithic. Why? Because they have a theory (a hypothesis more properly speaking) and they can't accept to be wrong about it.

That hypothesis (very popular in some circles) states that European aboriginal hunter-gatherers were very radically annihilated by Neolithic invaders from West Asia (never mind that archaeology alone is much more complicated than that, they don't seem to like thinking too much, much less looking at the matter from all the angles).

And a central battle they have fought is denying the possibility that mtDNA H (he most common haplogroup today in Western Europe) existed in the continent before Neolithic. The whole haplogroup, in their imaginary reality, could only have arrived with the industrious (and seemingly quite genocidal) farmers from West Asia (who almost never even mixed with anyone aborigine, how odd).

Reality began questioning their findings since 2005 but back in the day only HVS-I or at best HVS-II (control regions of the mtDNA chain) were used, leading to inconclusive results, specially in regards to short-stemmed haplogroup H. So they could still deny and deny...

But, recently, two different new studies have found unmistakable mtDNA H in Magdalenian people from Cantabria and Epipaleolithic people from the Basque Country. The reaction of some such knowledgeable aficionados has been simply unbelievable: they have flatly rejected the results without any reason; these findings are simply too inconvenient truths for their conjectures to be accepted. They are so obsessed with their fantasies that they can't even accept mounting evidence against them: they have stopped being scientific and begun being fanatics.

Very sad, really.

This finding in Karelia adds to the mounting unquestionable evidence on the matter: mtDNA haplogroup H not only existed in pre-Neolithic Eruope but it was quite extended, roughly through the areas in which is today abundant (and not just SW Europe as I came to suspect for some time). However in most regions was still far less common than it is today (or even totally missing, as seems to be the case in Central Europe).

Said that, it is not too clear yet where does all the improved knowledge of ancient genetics lead us to but what is clear is that mtDNA H is older and specifically older-in-Europe than some (too many) people have been insisting on.

Also it seems more and more obvious that the popular Neolithic farmers did not define the modern genetic landscape of Europe at all. They certainly introduced lineages that surely did not exist before but their overall influence seems limited and it does look like, after an initial burst, they declined also quite abruptly.

This is something that has been in the news these days (but no paper yet) and that I observed also in 2009 in relation to some similar studies (see: here and here). The age that we begin seeing modern-like mtDNA pools actually varies a lot, for example:
  • SW Europe: Basque Country: Neolithic (at least) ··> Hervella 2009 (discussed here).
  • Central Europe: Elbe Basin: Bronze Age or Chalcolithic ··> Schilz 2006[de], Schweitzer 2008.
  • Far North Europe: Sápmi: some time after the Bronze Age and before the 18th century (this study).
  • Central Asia: Iron Age (see below).
I conjecture here that (before the Medieval agricultural revolution) Northern latitudes could in general support lower population densities, being also more susceptible to the effect of climatic fluctuations. But more data is needed before we can have some consolidated certainty.

In any case, I took some time to make a couple of updated maps of the European and North African (1) Late Upper Paleolithic (Magdalenian and Oranian cultures) known ancient mtDNA and (2) Epipaleolithic. With this last one I found some conceptual difficulties so I had to take decisions, which were:
  • A most recent date boundary of 4000 BCE (which already overlaps with Neolithic in most regions since 1500 or more years before). Actually the most recent sites are c. 4200 BCE from Lithuania and c. 4600 BCE from Navarre.
  • No inclusion of any Neolithic data even if contemporary. The only possible exception was Franchti Cave (Greece), which has a sequence beginning in the Epipaleolithic (or Mesolithic) but is largely Neolithic. The exact adscription of the sequenced individual is not known.

The results are:

Late Upper Paleolithic mtDNA from Europe and North Africa
R* and specially R*-CRS can well be H and have often been reported as such but we do not know for sure

Epipaleolithic mtDNA from Europe (until 4000 BCE)
R* and specially R*-CRS can well be H and have often been reported as such but we do not know for sure

Some of these data (and others from more recent periods) can be seen in the dedicated Ancient mtDNA maps page at this blog. It needs some updating however: not much time has passed since I created those maps but new findings do pile up quickly these days. 


Ancient Scythian mtDNA


Another point of interest of the thesis is the ancient Scythian tombs from the Don basin (Iron Age, proto-historical). The results show some greater Eastern genetic influence than modern peoples (Russians) do.


The results, which place ancient Scythians closer to modern Central Asians than to Eastern Europeans are consistent with other recent studies that show an inflow of Eastern Asian mtDNA lineages into Central Asia even before the Turkic invasions of the Roman period and early Middle Ages.


Bronze Age Sardinian mtDNA


Finally the thesis deals with Sardinians from the Bronze Age (Nuraghic period). The sites are both from the most central parts of Sardinia, so they may be more representative of an early refuge population than to the overall Bronze Age of the island but still they are curious and interesting:


Dersarkissian argues that this suggests continuity but with many doubts, partly because the source of the genetic data (isolated teeth) did not allow for any certain identification of individuals. Still the resulting mtDNA pool (no matter how you look at it) is not really modern but rather reminds of Central European and Mediterranean Neolithic sites. 

The may well be some of the last Neolithic immigrants, who, instead of replacing the hunter-gatherer aborigines all around (as some imagined too dearly) were the ones taking refuge in this turbulent period in the highlands of Sardinia.

Who knows?!

9 comments:

  1. So, H is present in pre-Neolithic contexts in Europe at low frequencies for an mtDNA haplogroup that is now modal in Europe in Iberia, Britain and Karelia suggests the possibility of late Paleolithic/Epipaleolithic maritime migration (perhaps by peoples for whom ocean fishing is an important source of food; perhaps in the same wave that brings mtDNA hg V to the Saami).

    Is your late Paleolithic map pre-LGM or post-LGM? The notion that H may have arrived in Europe as part of a partially SW Asian sourced repopulation of Europe post-LGM still seems quite plausible, particularly in light of the increased frequency of H in the Neolithic which likely had a meaningful demic migration component.

    ReplyDelete
    Replies
    1. Well, the frequencies are not low in Iberia nor Morocco but rather similar to modern ones, especially if we assume that much of all that (HVS) R* is actually H, as is probably the case. I mean: today H is the largest haplogroup in Western Europe but also almost everywhere under 50% of the sample (c. 40-50% is the most common figure). However the quality and density of the research is so variable that not much more can be said by the moment, IMO.

      I did speculate recently with Megalithism (spread from Portugal and then Brittany and Western France, i.e. Armorica) being a possible vector of genetic scatter in the Chalcolithic (known as "Neolithic" in Britain) and I keep open that door but I am by no means certain that it is the case. There may have been several vectors in fact: all tending overall to create a certain homogeneity but of course it seems necessary in Central Europe (and maybe other parts of Northern and Western Europe) that there was some expansion(s) that would have consolidated mtDNA at c. 40% frequencies.

      "Is your late Paleolithic map pre-LGM or post-LGM?"

      Post-LGM. It is "Late Upper Paleolithic", what always means the post-LGM period, and in text I say "Magdalenian and Oranian cultures".

      "The notion that H may have arrived in Europe as part of a partially SW Asian sourced repopulation of Europe post-LGM still seems quite plausible"...

      It makes no sense archaeologically speaking: Magdalenian is quite clearly a European culture by origin (probably coalesced in Dordogne although influences from the Low Countries late Aurignacian remnants are probable).

      I don't understand well why that obsession for imagining everything having late dates and original from West Asia (West Asia is not larger than Europe and if Europe had problems of cold, West Asia had issues of aridity and in some areas both aridity and cold). Whatever the case there is one case in Britain of R*-CRS (H?) and another (two relatives) in Central Russia (Sunghir) with an unmistakable H17'27 in the earlier phases of Upper Paleolithic. A lot of Europe, notably the whole Franco-Cantabrian region, remains unsampled for pre-Magdalenian dates.

      ... "in light of the increased frequency of H in the Neolithic"...

      That only happens in Central Europe. You can't say that for Iberia for example. And I think it is critical to discern that because we simply can't extrapolate what happened in Central Europe to all the continent (as too many people have carelessly done).

      Also a lot of data re. mtDNA H is blurry: is R*-CRS H? And R* other? No way to know but in many cases it's probably the case.

      Delete
  2. "Why? Because they have a theory (a hypothesis more properly speaking) and they can't accept to be wrong about it".

    You are so consistently guilty of exactly that.

    "they have flatly rejected the results without any reason; these findings are simply too inconvenient truths for their conjectures to be accepted. They are so obsessed with their fantasies that they can't even accept mounting evidence against them: they have stopped being scientific and begun being fanatics"

    Again, that whole comment seems directed at yourself. Takes one to know one?

    Having said that, I have no problem accepting pre-Neolithic H in Europe.

    ReplyDelete
    Replies
    1. Stop spamming, Terry, please. You're getting trollingly abusive and I won't tolerate that.

      Delete
  3. Awesome synthesis maju! I learned a lot very quickly.

    For now, I'm very interested u2e as I want to make it a part of a migration theory I'm working on.

    ReplyDelete
  4. "As I mentioned above, the U2e and C1 ("C1f") findings are unusual..."

    Why do you think the u2e finding was unusual?

    ReplyDelete
    Replies
    1. U2 in general is very rare in Europe. It's usually considered a South Asian or South and Central Asian clade. It has been found however in some individuals (<1%) in Western Europe (France, Spain) but you would not usually expect it to show up so insistently in Eastern Europe (or anywhere in Europe) and yet it has appeared first in Kostenki (Early UP) and now in Karelia (Epipaleolithic), well above the frequency we could expect.

      My question is why wouldn't you think U2 is unusual or otherwise strange?

      Delete
  5. Maju: "U2e - which might well be descendant or otherwise related to the U2 of Kostenki"

    I dare say it's almost certain.

    "Central Russia (Sunghir) with an unmistakable H17'27 in the earlier phases of Upper Paleolithic"

    About that, this H in the Dersarkissian thesis, what can be said about it? Couldn't it be related to the 20,000+ yrs old (possibly) H17'27 lineage of Sungir?

    "My question is why wouldn't you think U2 is unusual or otherwise strange?"

    Because U2e (the "European"/north eurasian clade) has been consistently found in aDNA of Ural/Siberia and central Asia (even as far as the Udeges of south-eastern Siberia (it was there w/ other "west eurasian" lineages in a study)). it seems omnipresent (while not very frequent), in this large region, so why not in the adjacent region - the ancient north Russia - north of where it was more than 30,000 yrs ago (near Voronezh). It's not mind-bending.

    ReplyDelete
    Replies
    1. Re. H: probably not related. There is a very clear HVS-I haplotype that describes H17'21 and that's why we could identify it previously. I have not checked the haplotypes but I trust the authors did.

      Re. relation with Kostenki: I prefer to remain on the side of caution. Both suggest Central Asian affinities of the relevant peoples but not clearly relation between each other (it remains a possibility however).

      Re. U2e being found in later specimens, you seem to be right re. aDNA but most of it is Asian or borderline (Andronovo). Only a couple of individuals from Germany (Chalcolithic) and Denmark (antiquity) show that signature in ancient DNA that I can locate. It's more frequent than I'd have expected but nothing that makes me think that I should not be surprised at finding it in Karelia so early in time.

      Delete

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 (sorry, too many trolls).