November 1, 2011

Neolithic Iberian DNA

Cardium Pottery example
Dienekes just mentions the publication of a paper on the ancient DNA of some Neolithic people from Catalonia (Avellaner cave, Cogolls, Les Planes, La Garrotxa).

Marie Lacan et al., Ancient DNA suggests the leading role played by men in the Neolithic dissemination. PNAS 2011. Pay per view (depending on world region, free six months after publication). [doi: 10.1073/pnas.1113061108]

From the abstract they have detected two Y-DNA lineages (patrilineages): the pervasive G2a (also in nearby Treilles, Languedoc and in North Italy's Ötzi), which seems to be a trademark of sorts for the Cardium Pottery cultural complex and E1b1b1a1b, which is the Balcanic variant of E1b, often mentioned by its defining mutation V13. 

From the supplemental material, freely accessible, I gather that they have found four HVS-I mtDNA sequences (matrilineages, add 16,000 to all numbers): 
  • CRS (normally H, most commonly H1)
  • 93C, 224C, 311C (must be K1a1a)
  • 126C, 294T, 296T, 304C (must be T2b)
  • 51G, 189C, 270T (possibly U5b1b1a'd, U5b2a1a2 or some other U5b variant - my best quick survey hunch)
Table S2 reminds us of where in Neolithic sites have these specific sequences been found:
  • CRS: Languedoc (6), Spain (3), 'Danubian' Central Europe (4), Sweden pseudo-foragers (1)
  • 93C, 224C, 311C (K1a1a): 'Danubian' Central Europe only (2)
  • 126C, 294T, 296T, 304C (T2b): Languedoc (2), 'Danubian' Central Europe (2)
  • 51G, 189C, 270T (U5b?): nowhere else yet
Table S1 meanwhile reminds us which present day populations share these sequences:
  • CRS: widespread with peaks in Sicily (28%), Wales (25%), etc.
  • 93C, 224C, 311C (K1a1a): Georgians, Cypriots and Austrians are the only ones barely above 3%
  • 126C, 294T, 296T, 304C (T2b): South France (4-5%), Denmark (5%).
  • 51G, 189C, 270T (U5b?): Cornwall (3%), Germany and Estonia are the only other two locations above 1%.
Among modern Catalans (n=133) the frequencies are: 17.29%, 1.50%, 0.00% and 0.00%. Similar apportions (slightly higher for CRS and K1a1a) exist for Balearic islanders (n=67), arguably more likely to have a greater share of Neolithic ancestry but otherwise similar to mainland Catalans. The third lineage (T2b) is found in Iberia only among Galicians and Basques, while the fourth one is totally absent (in this sample at least).

I'm unsure at the moment how many people had each lineage and such.

An interesting note is that we are continuously observing the often thought as 'Neolithic lineages' (Y-DNA G2a, E1b1b1a1b, I2a) in these Cardium Pottery culture sites but we fail to find any of the pre-Neolithic lineages (R1b notably, which now comprises maybe half or the Y-DNA pool) even in a such presumably remote mountain site as this one. Of course the sampling is still thin but the results are kind of surprising. We can, I guess, argue that the relatively rare I2a is pre-Neolithic (???) but that doesn't make up for the absence of R1b, which must have been there somewhere (or expanded massively later on, what would be quite surprising).


  1. I know that you are quite committed to the idea of R1b being pre-Neolithic and I am familiar with the arguments you support this hypothesis with, but, increasingly, the ancient DNA is favoring in interpretation that it is a mid- (what I call "dairy farmer Neolithic" with less than perfect accuracy) to late Neolithic arrival (i.e. Indo-Eurpoean), not a pre-Neolithic or early Neolithic arrival.

  2. Note, to be clear, in the comment above, I used Neolithic in the sence of Cardial Pottery culture or later, in an inapt sense that includes proto-food producing and metal ages as well.

  3. But it's thus impossible to explain the distribution of R1b as most common in the isolated fringes of Europe. Excavations just prove what could be guessed : that at some point in History, parts of Europe were colonized by Neolithic migrants the burial customs differed from the marginalized autochtonous people.

    The only issue here is that it's Catalonia and Catalan people are known to show less West Asian affinity than other Iberian ethnicities in most amateur studies. Either autochtonous people never got to be assimilated by these colonizers or Catalonia saw a later movements of populations coming from the Pyrenees or France.

  4. Excepting one sample from Low Saxony, the oldest known R1b1a2 is Tutankhamun... what offers a lot of weirdo possibilities for those willing to believe in Gengis-Khan-like founder myths.

    Not my inclination.

    Actually no K whatsoever is known before some R1a1 from late Chalcolithic East Germany. Are we going to draw conclusions on that for all R, P, MNOPS and K? I won't.

    Only recently ancient Y-DNA results have begun to come more and more frequently. The samples are still not comprehensive.

    What you say of "dairy farming" lacks any sort of archaeological backing. Sincerely, you should read a book or two on European Prehistory from the viewpoint of archaeology. There's no point on discussing population genetics without knowing the material evidence and how it fits on its own terms.

  5. Heraus wrote:
    "parts of Europe were colonized by Neolithic migrants the burial customs differed from the marginalized autochtonous people."

    From the study's abstract:

    "DNA [was] extracted from human remains excavated in a Spanish funeral cave dating from the beginning of the fifth millennium B.C."

    Anthropologists have known for generations now that there were Neolithic migrations by seafarers from the Middle-East to various parts of the West European seaboard. Why not the Spanish "Levant", too? There is nothing surprising about these results, nor inconsistent with the Paleolithic R1b Theory, it seems to me. If it is confirmed that these are the "Neolithic seafarers" or their descendants, it may even support the Paleolithic R1b Theory.

    This is the same problem various people have pointed out with Out of Africa. Proposition: "Oh, we find the most ancient human remains in the East-African Rift Valley; humans must come from East-Africa." Counter: "Conditions there make it very easy to preserve remains in the Rift Valley. Finding bones in the Rift Valley proves such people were there at a given time, not that they originated there."

    The same logic applies, exactly, in this case. The Neolithic (Semitic-speakers?) who set up colonies in parts of Western-Europe, had burial practices that allow for easier discovery and identification now.

  6. As to G2a, as I have argued before, that could well be of local Balkan/Middle Danube origin, and as such would not carry with it much West Asian autosomal signatures. I think the small portion of ancient E actually tells us not to expect much West Asian autosomal away from Italy and Greece, where it likely arrived in multiple waves since the neolithic.

    I agree with the burial practices ascertainment bias. But I also think that especially LBK was an extremely quickly spreading, thin bubble that would have been an easy target to be overcome by mesolithic DNA from the fringes. So yes, we desperately need ancient DNA data from the fringes - especially y-DNA.

  7. Interesting.

    Dienekes in his update mentions a mtDNA H3 instead of a H1.

    The lack of R1b is kind of surprising but aren't you surprised by the lack of J2 too in the neolithic findings?
    What's your take on it (when and from where (and in which "archeological context") did it arrive in Europe?).

  8. A number of H subclades and other R as well (although it's most rare and has only been reported in Europe in U* and HV(xH)) can be CRS. IDK why D. says H3 but it's possible as well.

    "The lack of R1b is kind of surprising but aren't you surprised by the lack of J2 too in the neolithic findings?"

    Naturally. But we only have five valid examples (4 west of the Alps), because of their consanguinity, and the only thing they say is that G2a appears common and that I2a and E1b-V13 were present as well.

    I remember years ago when the Danubian Neolithic findings produced lots of unexpected mtDNA N1a, which were later both confirmed but also very much relativized (not so many vs. the general sample and concentrated in particular areas like East Germany and Hungary). This is probably the same.

    We need anyhow ancient DNA from Franco-Cantabrian Paleolithic remains. A good number of them should settle all speculations.

  9. Maju wrote:
    "ancient DNA from Franco-Cantabrian Paleolithic remains

    I have heard that Y-DNA decays so fast, that we cannot find it in anything from the Paleolithic. (Yet?)

  10. Epipaleolithic? If they can get from remains from "7000 years ago" (sic), why not 8000? This is the very temporal edge of Neolithic.

  11. What is the oldest Y-DNA ever extracted?

  12. Possibly this one from Cogolls, as well as, of similar age, some individuals from Derensburg (Haak 2010). See:

  13. According to that site, only five excavation sites in Europe from before 2000 BC have yielded any Y-DNA, so far. If that list is comprehensive, the Cogolls cave will be only number six.

    (New) Cogolls cave, Catalonia -- G2a and E1b, circa 5000 BC
    (1) Derenburg Meerenstieg II, Central-Germany -- G2a in 5000 BC.
    (2) Avellaner cave, Catalonia -- Mostly G2a with one E1b in 5000 BC.
    (3) Treilles, Aveyron, France -- Mostly G2a, some I2a in 3000 BC.
    (4) Oetzi the Iceman -- G2a in circa 3200 BC.
    (5) Eulau, Germany -- R1a in 2600 BC.

    I recall that in the amateur paleoanthropological musings of's editor Maciamo (which he presents as fact on his polished website), Y-haplogroup 'I' is implied to be the major haplogroup of Europe's Paleolithic, and R1a and R1b are said to have been totally absent. Does he base this conjecture solely on the finding of two 'I' individuals at the Treilles site?

    He also boldly asserts that R1a migrated to Europe only in the Bronze Age. Yet we have R1a Y-DNA from 2600 BC in Germany, which is before Bronze technology reached the area.

  14. But why is that? Largely because only mtDNA was targeted. Also because technology has been improving. If we can get mtDNA and even decent quality whole nuclear DNA from 35,000 years ago (Neanderthal genome, Denisova genome...), we can get Y-DNA (which is just a part of the nuclear DNA). Of course there will be differences in preservation and on how the technology or the researchers' skills can do this but that it can be done there is no doubt.

    There should not be at least.

    "Does he base this conjecture solely on the finding of two 'I' individuals at the Treilles site?"

    This is an old theory surely based on TMRCA (age estimate) hunches and personal preferences. I'm pretty sure it has nothing to do with Treilles even if this guy may feel some support from that finding.

    In any case the two individuals were very close relatives and should be considered hence as a single effective sample. In practical terms the Neolithic and Chalcolithic of SW Europe has produced four effective samples: two G2a, one I2a and one E1b-V13, in two nearby sites separated by maybe 200 km. That's what we have as of now in the Y-DNA side of the equation.

    "He also boldly asserts that R1a migrated to Europe only in the Bronze Age".

    Whatever. Considering the evidence or rather lack of it that is at the moment wishful thinking. We'll see when the number of samples increases, give this issue five more years: it needs them.

  15. "Anthropologists have known for generations now that there were Neolithic migrations by seafarers from the Middle-East to various parts of the West European seaboard. Why not the Spanish "Levant", too? There is nothing surprising about these results, nor inconsistent with the Paleolithic R1b Theory, it seems to me".

    I agree.

    "I have heard that Y-DNA decays so fast"

    It's not so much that it decays fast, it's just that there is so little of it compared to the amount of mtDNA. Every cell in the body has multiple copies of mtDNA but just one copy of Y-DNA. The last only in men of course.

  16. A discussion on Y-DNA findings for ancient Europe:

    One poster there writes:

    "I think R1b arrived in Europe during the Bronze Age and with it Proto-Indo-European. ... I will be mighty disappointed if some Mesolithic or Paleolithic R1b surfaces in Europe."

    Why are people like this so emotionally-invested in a Bronze-Age R1b? To identify with a mythical race of conquering heroes? An updated version of Aryan Theory? Is that it?

  17. Probably that, Hailtoyou. However there are no meaningful detectable migrations into Europe since Kurgans (origin is borderline Europe, could be considered Asia by classical parameters: East of Volga), unless we consider historical Muslim and Turco-Mongol intrusions, so it's beating a dead horse.

    A different case would be arguing for its scatter in Europe (from some parts of Europe to others) in the Bronze Age. For that there might be some weak and indirect support (it's only since the Bronze Age that we see modern mtDNA pools in Central Europe even if in SW Europe we see them since at least Epipaleolithic) but it's very hard to argue such flows in archaeological terms. Not wholly impossible but quite borderline.

  18. Maju,

    What is your opinion on Indo-European language origin(s)?

  19. Kurgan model; it's very solid and the only one that makes sense at all.


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 (your comment may take some time, maybe days or weeks to appear).