August 13, 2014

Chalcolithic mtDNA from Atapuerca still in the Neolithic range

Bell Beaker Blogger points me to this latest study on ancient mtDNA from the Center-North Iberian Peninsula, including 20 samples from a Chalcolithic site (without Bell Beaker apparently) that clearly shows continuity with mainline Neolithic (Cardium but also similar to Central European Linear Pottery Culture, both sharing the same Thessalian ultimate origins).

Daniel Gómez Sánchez, Iñigo Olalde et al., Mitochondrial DNA from El Mirador Cave (Atapuerca, Spain) Reveals the Heterogeneity of Chalcolithic Populations. PLoS ONE 2014. Open accessLINK [doi:10.1371/journal.pone.0105105]


Previous mitochondrial DNA analyses on ancient European remains have suggested that the current distribution of haplogroup H was modeled by the expansion of the Bell Beaker culture (ca 4,500–4,050 years BP) out of Iberia during the Chalcolithic period. However, little is known on the genetic composition of contemporaneous Iberian populations that do not carry the archaeological tool kit defining this culture. Here we have retrieved mitochondrial DNA (mtDNA) sequences from 19 individuals from a Chalcolithic sample from El Mirador cave in Spain, dated to 4,760–4,200 years BP and we have analyzed the haplogroup composition in the context of modern and ancient populations. Regarding extant African, Asian and European populations, El Mirador shows affinities with Near Eastern groups. In different analyses with other ancient samples, El Mirador clusters with Middle and Late Neolithic populations from Germany, belonging to the Rössen, the Salzmünde and the Baalberge archaeological cultures but not with contemporaneous Bell Beakers. Our analyses support the existence of a common genetic signal between Western and Central Europe during the Middle and Late Neolithic and points to a heterogeneous genetic landscape among Chalcolithic groups.

The results show intense similitude with Catalan Neolithic and Languedoc's Chalcolithic but also with Central European Neolithic. They contrast instead with Portuguese and Basque Neolithic, as well as with Central European Bell Beaker, all them much higher in haplogroup H and, in the Basque case, also in U (being the most modern-like of all ancient mtDNA pools known before the Bronze Age in Europe).

Annotated version of fig. 2
Figure 2. Mitochondrial DNA haplogroup frequency for 21 ancient European samples.
This study: El Mirador (MIR). Published prehistoric cultures [21]: Hunter-gatherer central (HGC), Linear Pottery culture (LBK), Rössen culture (RSC), Schöningen group (SCG), Baalberge culture (BAC), Salzmünde culture (SMC), Bernburg culture (BEC), Corded Ware culture (CWC), Bell Beaker culture (BBC), Unetice culture (UC), Funnel Beaker culture (FBC), Pitted Ware culture (PWC), Hunter-Gatherer south (HGS), (Epi) Cardial (CAR), Neolithic Portugal (NPO), Neolithic Basque Country and Navarre (NBQ), Treilles culture (TRE), Hunter-gatherer east (HGE), Bronze Age Siberia (BAS), Bronze Age Kazakhstan (BAK).

Notice please that the above column for Central European Bell Beaker (BBC) includes the more than dubiously attributed Kromsdorft site, which has a totally different genetic signature. In my previous analysis of European ancient DNA evolution, I treated them separately and I still think that it is much more correct to do it that way.

Notice also that one sample from Mirador was sequenced for autosomal DNA by Evangelia Dasakali, producing an Italian-like sequence, roughly in line with other early European farmers, excepted the Atlantic ones. See here.

So what we see is a "wedge" of Mediterranean or Neolithic ancestry probably penetrating along the Ebro river up to Atapuerca and contrasting with Atlantic Iberian (Basque and Portuguese) ancestry, more modern-like or even "hyper-modern" (by contrast with mainline farmers) in the case of Portugal. This same contrast exists with Central European Bell Beaker sites and also (in the autosomal DNA aspect) with Megalithic farmers from Southern Sweden. 

None of those "modernizing" tendencies can be found instead among Eastern European Neolithic peoples nor among early Indoeuropeans of Central Europe such as those of the Unetice culture. So the overall conclusion can only be that there was in the Chalcolithic (and to some extent Neolithic) a duality of ancestries between the Mainline or Mediterranean (but also Danubian) Neolithic and a more "modern" Atlantic Neolithic, which is related to the Megalithic and Bell Beaker cultural phenomena (and in North-Central Europe also to Funnelbeaker). 

There are still a lot of dark spots in our understanding of how this "modernization" or "Westernization" of the genetic pool happened but, in general terms, it seems to imply a sizable (albeit somewhat irregular) demographic flow from Atlantic Europe into the areas previously occupied by Mainline Neolithic populations of partial West Asian affinity. This is apparent in both mtDNA as in autosomal DNA. 

These sequences from El Mirador (Atapuerca) only underline this phenomenon and the fact that in the Chalcolithic, some 4500 years ago, this process of "modernization" of the European genetic pool was still incomplete.


  1. Maju, these results may also have something to do with the idea that there were two 'poles' of Neolithicozation in Iberia - a northern (Danubian) one derived from LBK and a second , southern (Impressed & Cardial) pole which met in Central Iberia then underwent complex inrermixtures

    1. Simply put: there is no LBK whatsoever in Iberia nor Southern France. What these and other similar results indicate is that the main two Neolithic waves of Europe (Impressed-Cardium and Painted-Linear) were product of a unique meta-ethnicity, with the same genetics, the same origins (Thessaly) and quite probably the same linguistic family (which I believe should be Vasconic). Their genetic heritage is better preserved in Southern Europe though, although here also there was some sort of back-flow from the Atlantic carrying greater frequencies of aboriginal genetics (and later, in the Late Bronze and Iron ages from Central Europe, carrying the Indoeuropean languages).

      I can't imagine where you could get the idea that LBK ever made it past the Loire. It simply didn't - instead Cardium-like La Hoguette culture (and/or other locally specific groups) overlaps with LBK in the areas of Northern France, West Germany and Belgium, making the wider area around the Rhine not so strictly Danubian but a crossroads of both traditions, as well as some innovative local ones, particularly in Belgium.

      Relatively good generic map of early Neolithic in West-Central Europe (not so good re. La Hoguette and other local Neolithic groups though)
      Map showing overlap of LBK and other non-LBK Neolithic groups in the Rhine area
      Decent map of La Hoguette sites (dots) with core, very approximative, Danubian/Cardium areas shaded in grey

      In any event there was absolutely no LBK in Iberia nor Southern France. The explanation to genetic similitudes between LBK and Cardium early farmers must be another: namely a shared origin in Thessaly.

    2. Uh... first map link is broken. It is this one:

      Relatively good generic map of early Neolithic in West-Central Europe (not so good re. La Hoguette and other local Neolithic groups though)

  2. Do you think this "modernization" could have come from NW Africa in the Upper Paleolithic? IE a non-EEF non-WHG population. Would explain a few things I think, especially if it brought R1b with it.

    1. No. For many reasons, both genetic and archaeological.

      The kind of genetic flow is not at all related to NW Africa, certainly not by origin. While mtDNA H is present in NW Africa it clearly derives from SW Europe and not the other way around (almost certainly Solutrean→Oranian flow). Y-DNA R1b is very rare in NW Africa and the immediate origin of the European clades is clearly in Western Europe, while upstream it comes from West Asia, via the Balcans. Autosomally we don't see either NW African components in Iberia or Europe but the opposite is true (much of the NW African genetic pool seems to originate directly in Europe). I can give references but I'm tired now. Please ask later.

      Archaeologically there's also no obvious Africa → Europe influence. Again, if anything, the opposite would be true (Megalithism, some Bell Beaker).

      The source(s) must be in Atlantic Europe.

    2. I thought the phylogeny contradicted an Atlantic origin for even the European R1b subclades. Do you think this modernization was a different wave than whatever spread R1b?

      And any thoughts on how R1b got its present distribution in both Western Europe and in Cameroon?

    3. →

      R1b-S116 quite clearly stems from South France (where it seems most basally diverse), while R1b-U106 may have spread from around the Rhine basin (highest frequency in Netherlands, although unsure about where the greatest basal diversity is).

      Their common precursor (rather rare in "asterisk" variants) R1b-M412 may have spread from Central Europe and/or SW Europe. The ancestor of this one R1b-M269 may have spread from either the Balcans or West Asia.

      Only upstream of this node we enter the wider R1b of extra-European distribution, including the Central Asian/Bashkir and the Afro-Mediterranean clades, as well as some other that are ill studied AFAIK.

      So for me:

      1. R1b (as a whole) originated in West Asia, spreading in various directions.
      2. R1b-M269 (Euro-Asian clade), spread from either West Asia or the Balcans.
      3. R1b-M412 (European clade), spread from either Central or SW Europe.
      4a. R1b-S116 (SW European clade) spread from the Franco-Cantabrian Region.
      4b. R1b-U106 (NW European clade) spread from possibly the Rhine basin or the North Sea area (although an Upper Danube origin is not discarded for all I know).

      Particularly step #4 dissuades me from contemplating a recent origin, as well as some chronological estimates, that would make M412 and S116 Magdalenian, while R1b as a whole would be slightly older than Gravettian.

      As for the Afro-Mediterranean R1b-V88 haplogroup, it has no particular relation with the mainstream European R1b, being another branch of the early R1b spread from West Asia almost certainly. I've discussed it HERE. I would think that V88 as such spread from West Asia, with a branch reaching the Nile (Sudan particularly) and, from there, spreading to Central Africa in the context of the Chadic expansion (Neolithic?) Other branches reached Italy and Sardinia-Corsica, etc. It's not impossible however that it may have spread from Africa, much as E1b did.

  3. LBK was not in southern France but in Northern.
    thirty years ago, the Norman population was supposed to be LBK, essentially, and that from the Neolithic to the Norman period.
    As Norman ( Cotentin peninsula), I am shifted west, I have a lot of Atlantic component and I am a pretty good match for Basques and southwest French( in K15 eurogenes)
    that is difficult to know without ancient samples, but for you Normandy is more Long barrows or more LBK ?

    1. The Neolithic of Northern France, Belgium and much of West Germany presents a good deal of complexity: LBK elements are there for sure but also La Hoguette (Cardium-related) and Limburg (local) cultures, as well as Megalithism, etc.

      Anyhow the LBK genetic pool was clearly not modern: further demographic processes took place that heavily modified it in the Chalcolithic (and maybe also later). With due caution, I would think that the Megalithic phenomenon was largely demographic and "westernized" the rather "oriental" LBK (and Cardium) genetic pool.

      As for Normandy, High Normandy (west) was mostly within the same cultural region as Brittany and nearby West French areas (Armorica in historical terminology). Instead Low Normandy was in the LBK/Limburg/La Hoguette area. But as I just said, the cultural and demographic landscape suffered major ulterior shifts, so these differences were surely blurred later on, particularly as the Artenac culture (Megalithist archers from Dordogne) spread over all West France and Belgium. This may well explain (at least partly) what you say about being "a pretty good match for Basques and southwest French( in K15 eurogenes)".

  4. Thank you for your speed reply

    We read a lot about the long barrows and more about bell beakers, but little about the Artenac culture, I remember you wrote that they stopped the advance LBK, you think they had a significant contribution in terms of demographics in Nothern France, to be honest I had never heard of them before reading your blog.

    1. There's no much info online on Artenacian culture (although there is no doubt more in French, search "culture artenacien"). I know from books however that they were an important late Chalcolithic force which took over the whole West of France (also Belgium), replacing the previous cultures, notably the Seine-Oise-Marne one, which in turn had displaced the ancient Armorican Megalithic culture. They are often understood to be direct precursors of the historical Aquitani (i.e. Basques).

      Its characteristics were AFAIK abundance of arrows (often described as an "archers' culture") and a dolmenic Megalithic funerary fashion much more modest in size than what was styled by Bretons (I would say that a much less hierarchic society therefore, maybe by loss or weakening of the proto-druidism that may have persisted in Great Britain instead). The former core Seine-Oise-Marne area between the Seine and the Rhine (i.e. historical Belgium) had a slightly different facies, maybe reflecting its different, more loosely "Danubian" substrate. In any case that was about the time when the last remains of any Danubian (LBK) legacy were erased in all Europe because Artenacian is contemporary with Corded Ware and Vucedol.

  5. "I would think that V88 as such spread from West Asia, with a branch reaching the Nile (Sudan particularly) and, from there, spreading to Central Africa in the context of the Chadic expansion (Neolithic?) Other branches reached Italy and Sardinia-Corsica, etc. It's not impossible however that it may have spread from Africa, much as E1b did. "

    Many HLA alleles and haplotypes which are shared between Sudan and Italy + Sardinia/Corsica are found in other African populations but not really very much in West Asia if at all. I can provide numerous data to support this.

    1. HLA immunity alleles are not reliable ancestry markers because they are subject to strong selective pressures. Italy for example was one of the areas of Europe most strongly affected by malaria (once known as "Roman fever") and therefore also has unusually high frequencies of thalassemia, a kind of anemia that is an effective anti-malarian protection, which has anyhow closest relation to what is found in Greece, Southern Anatolia, Cyprus and Palestine. It is perfectly possible that the HLA markers were selected for the very same reasons (malaria) or similar ones. Although it suggests some kind of relatedness, being markers subject to strong positive selection subject to environmental pressures, this relatedness may be hair-thin (standard introgression between two otherwise almost totally distinct populations).

      R1b-V88 is found in an arch between Sardinia/Italy and the Eastern half of the Sudanese corridor, with clear presence in West Asia. Considering the overall spread of R1b (centered in West Asia) and the geographic pattern of the subclade, I'd say that a West Asian origin is at the very least quite parsimonious. Direct Sudan-Italy relations are most unlikely in any case, so the Eastern Mediterranean must have acted as corridor at the very least (but I lean towards origin).

  6. The malaria protective HLA allele is B*53. [See: ]
    This is in fact at only a very low frequency in Corsica and Sardinia [both precisely 0.05%].
    The alleles and haplotypes which I am referring to play no protective role with regards to malaria.
    One example:

    B*58:01 in Europe

    Sardinia 6.40% [Peak European frequency - not evidence for selection, probably founder effect]
    Corsica 4.50%
    Switzerland Basel 4.21%
    Italy Rome 3.50%
    Italy North Pop.3 3.30%
    Switzerland Geneva Pop.2 3.26%
    Portugal Castelo Branco 2.80%
    Portugal Leiria 2.60%
    Croatia Pop.2 2.50%
    Portugal Santarem 2.50%
    Portugal Setubal 2.50%
    Portugal South Pop.2 2.50%
    Greece Pop.8 2.41%
    Switzerland Lugano 2.40%
    Portugal Evora 2.40%
    Portugal Portalegre 2.40%
    Portugal Braganca 2.30%
    Portugal Faro 2.30%
    Portugal Vila Real 2.30%
    Spain Ibiza 2.30%
    Switzerland Luzern 2.26%
    etc. etc.

    Lowest Euro frequencies:-
    Belgium 0.50%
    England North West 0.50%
    Serbia Pop.2 0.50%
    Wales 0.50%
    British (a) 0.50%
    British (b) 0.48%
    England Liverpool 0.45%
    Ireland South 0.40%
    Ireland South Pop.2 0.36%
    Ireland Northern 0.30%
    England Newcastle 0.25%
    Serbia 0.20%
    Sweden Uppsala 0.15%
    Scotland Orkney 0.00%
    Spain Gipuzkoa Basque 0.00%
    etc. etc.

    Where might it have entered from?

    Israel Arab Druze 1.40%
    Jordan Amman 1.40%

    Sudan Mixed 4.50%
    Tunisia 4.00%
    Tunisia Ghannouch 3.70%
    Morocco Settat Chaouya 3.40%
    Tunisia Pop.3 3.40%
    Mali Bandiagara 2.20%
    Sudan East Rashaida 1.86%
    Morocco Nador Metalsa Pop.2 1.40%
    [Searching for data from Libya and Egypt]

    Kenya Nandi 10.00%
    Kenya 8.00%
    Guinea Bissau 7.80%
    Kenya Luo 7.00%
    Senegal Niokholo Mandenka 6.90%
    Burkina Faso Fulani 6.10%
    Uganda Kampala Pop.2 6.00%
    Burkina Faso Mossi 5.70%
    Cameroon Yaoundé 5.40%
    Cameroon Bamileke 5.20%
    Sao Tome Island Angolar 4.70%
    Zimbabwe Harare Shona 4.40%
    Burkina Faso Rimaibe 4.30%
    Ghana Ga-Adangbe 4.20%
    Cameroon Bakola Pygmy 4.08%
    South Africa Natal Zulu 4.00%
    Cape Verde SE Islands 4.00%
    Uganda Kampala 4.00%
    Cameroon Beti 3.70%
    Cape Verde NW Islands 3.20%
    etc. etc.


    There are several other HLA alleles and haplotypes in the Sardinia and Corsica + Italy populations which follow a similar pattern, again none of the alleles in question is involved with malaria protection. HLA should not be dismissed out of hand, if one learns which alleles have disease protective or susceptibility associations [they have been studied very extensively over the last 40 years and there is a huge amount of information and data available].

    There seems to be a reluctance on many people's part to consider data which may conflict with the existing paradigms. Many people also seem reluctant to accept the possibility of prehistoric migrations directly across the Mediterranean from Africa into Europe despite data which suggests this. I would be interested to know why. Looking at HLA data [involving alleles with no known disease protective role] I can see evidence of a number of these.

    1. It may have been selected for whatever any other reason (plague epidemics for example) or maybe some aspects of malaria defense are not well known yet. In any case it is clearly an immunity marker subject to strong selective pressures and therefore not really informative re. ancestry. Frequencies do not matter either because in order for an adaptive marker to be selected for in the long run (consider repeated recombination and selection) only one ancestor is needed (regardless of the overall population). A low frequency allele here can become a high frequency one there, after lesser migration from here to there, just because it does work.

      That's why neutral markers (or those believed to be more neutral) are preferred when researching ancestry. Among them haploid lineages are the safest ones because they do not recombine at all (although they can still provide advantages in theory).

    2. In any case your data for West Asia is very limited: you have not provided any data for most of the region and the two samples provided are almost certainly not too representative (particularly not the Druze, who have undergone dramatic endogamy-caused bottlenecks and may be partly of non-local origin).

    3. Also the frequencies you provided for B*53 seem incorrect (per
      → Gaza Palestinians 2.4%
      → Jordan 2.5%
      → Italy: 0.0-2.5%

      Actually per this source, Italians seem to have lower frequencies than Native peoples from the Southern Levant. The largest frequency is in North Pavia pop. 2 (2.5%), which should fall to much lower when averaged with pop. 1 (0.9%). The second largest frequency is from the historically super-cosmopolitan city of Rome (2.0%).

      Nothing to see here...

    4. The All Italy sample (n=159,311) shows a frequency for this allele of just 0.9%!!!


  7. Hi Maju, thanks for your reply.

    Ok, rather than the alleles I'll provide some full haplotypes.

    #1 haplotype: A*30-B*18-C*05 [12.50%]
    [shared with Iberians, Basques, Balearic Islanders, Corsicans, S.French, N.Italians, Swiss, North West Africans, Senegalese, Tuaregs, N.Cameroon, Sudanese, + slightly different variant in Kenya Luo]
    #2 haplotype: A*02-B*58:01-C*07 [4.90%]
    [shared with Iberians, Italians, Swiss, North West Africans, Cameroonians, African Americans, Sudanese, Kenya Nandi, Kenya Luo]

    There are several haplotypes found in Sardinia which are shared with West Asians, however these are not.

    "That's why neutral markers (or those believed to be more neutral) are preferred when researching ancestry."

    Use of HLA as an ancestry marker is not limited to armchair geneticists or anthropologists like myself. Right now the University of Geneva's Department of Genetics & Evolution's AGP Lab [Laboratory of Anthropology, Genetics and Peopling history] is using them quite extensively in some research projects funded by the Swiss National Science Foundation:-

    "The AGP lab is involved in many research projects on human genetic evolution with the objective to understand better the dispersal of modern humans around the world and the mechanisms of their biological differentiations. Information from other disciplines, namely archaeology, paleoanthropology and linguistics, are also used to investigate the relationships between biological and cultural evolution. The team principally works in silico by developing computer tools and performing data analysis. One of the main polymorphisms used is HLA, but other genetic markers are also analyzed (mtDNA, Y chromosome, nuclear STRs, etc). Since several years, a big effort is devoted to the improvement of the HLA characterization of human populations from a methodological point of view and the maintenance of large databases."

    "Also the frequencies you provided for B*53 seem incorrect"

    I didn't provide any frequencies for B*53! [that's the malaria protective allele].
    The frequencies I provided were only for the B*58:01 allele, I provided no data on B*53 whatsoever.

    To summarise, I can see geneflow into Sardinia and Corsica from three directions looking at HLA: Europe; N.Africa/Sahara; and West Asia. The gene flow from Africa impacts the mainland of Europe to some degree also. My issue is that a lot of people want to bury their head in the sand over this. Eurogenes blog deleted the entire 'Ancient Middle Easterners (AME)' post from 10th August, when this had been discussed in the comments section. I don't know if this is coincidental or not.

    Incidentally I regularly read your blog and would like to say keep up the good work. It is very well researched and interesting and informative. It's also refreshing to have a genetics and anthropology blogger who is from the left of the political spectrum rather than the right.

    1. I didn't provide any frequencies for B*53! [that's the malaria protective allele].
      The frequencies I provided were only for the B*58:01 allele,

      My bad, sorry.


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).