April 15, 2013

Central European Bell Beaker mtDNA: 88% H

Bell beaker from Hungary (source)
This doctoral thesis by Christina J. Adler is a most important study on the formation of the modern genetic pool (and hence population) in Central Europe. Previously we knew of data from the Paleolithic (U*), Epipaleolithic (U5, U4) and Earliest Neolithic (much more diverse but not yet modern in any sense). Then we had a huge blank until Urnfields (late Bronze), when the genetic pool seemed to be modern already. 

This thesis (found via Eurogenes) fills in the blanks at least to some extent.

Christina Jane Adler, Ancient DNA studies of Human Evolution. University of Adelaide (thesis), 2012. Freely accessibleLINK 

The thesis is, as usual in this kind of studies, extremely long; even the abstract is too long to copy here. Just to mention that the hard data (graphs, tables) is from page 96 on, although there are some other aspects in the text that deserve mention.

Critically Adler could research the ancient mtDNA of Bell Beaker and Únětice culture populations from several German sites, adding important information about the genetic pools of the Late Chalcolithic and Early Bronze Age. She also goes over previous studies on the same area.

The new data (table 2) can be synthesized as follows:
  • Bell Beaker:
    • Quedlinburg XII: 3 H-CRS (H1?), 1 J
    • Rothenschirmbach: 2 H3, 1 H5
    • Alberstedt: 1 H-CRS (H1?)
    • Total (simplified): 7 H, 1 J
  • Únětice:
    • Quedlinburg VIII: 1 U5a1a, 1 U2, 1 U*, 1 H7a, 1 T1
    • Quedlinburg XII: 1 U5a1a
    • Quedlinburg XIV: 1 T2
    • Esperstedt: 2 I*, 1 I1, 1 U5a1, 1 U5b, 1 T2b, 1 T2*, 1 W, 1 X
    • Total (simplified): 4 U5, 3 I, 3 T2, 1 T1, 1 U2, 1 U*, 1 W, 1 X
This extremely high apportion of mtDNA H is almost unprecedented in ancient (and probably also modern) mtDNA samples, in Europe only the Portuguese Neolithic and Epipaleolithic samples by Chandler 2005 seem to be comparable in any way, suggesting that this most important European matrilineage may have expanded from Iberia in the Chalcolithic (aka Late Neolithic in some Anglosaxon literature) with either Megalithism, Bell Beaker or both. 

It also seems to contradict the quite mainstream theory of Central European origins (post Corded Ware) of the Bell Beaker phenomenon and instead support the less popular Iberian origin theory. Until this very day I have been adherent to the Bohemian "Corded" origin theory (with some doubts) but today I have to admit that this genetic data weights heavily for the Iberian origin model, which in turn would fit very nicely with Venneman's Vasconic substrate theory. 

Regardless of what I may think, Adler herself is clearly pushing for the Iberian origin model all along in her thesis, theory which she seems to find the best fit scenario. 

Of course, the genetic landscape was not simply stabilized with Megalithism and Bell Beaker, more waves followed. The Western Indoeuropean Únětice culture seems to fit here as archetypal or potential source of other layers, resulting in modern genetic pools in many places (although as I have mentioned several times the Basque one seems stable since Early Neolithic). These Indoeuropean migrations (Tumuli, Urnfields, Hallstatt, La Tène, etc.) should explain the dilution of the extremely high apportion of H found in these Bell Beaker burials, as well as in Portugal (nowadays H is 40-50% in most of Western and Northern Europe).

The Únětice genetic landscape seems particularly interesting for including which is surely the oldest mtDNA I in Northern Europe (later very common in Viking Era Denmark). The only older case I know is again Early Neolithic Basque (same Paternabidea sample mentioned above) but I don't see any plausible relation. 



PS- I know still have in the "to do" department the paper of Qiaomei Fu, "A Revised Timescale for Human Evolution Based on Ancient Mitochondrial Genomes", which several readers were so kind to send me a copy of weeks ago, encouraging me to write on it. My apologies but I'm on it and I promise to write a review this very week unless the sky falls on my head (or real life equivalent).

38 comments:

  1. May falling pianos avoid you. I speak from experience when I say that piano and falling object related injuries are most unpleasant.

    This piece is yet another very strong piece of evidence for the case that Bell Beaker expansion from Iberia (or perhaps a predecessor megalithic culture expansion from Iberia), in the late Neolithic/Chalcolithic, was the leading demographic source for modern Western Europe's gene pool, regardless of whether this population was indigeneously Iberian, or was merely an intrusive population that expanded from a secondary Iberian staging point. If Bell Beaker has an intrusive to Iberia source, the case that this source was not a population closely genetically related to the Únětice culture people seems pretty solid unless perhaps the intruders were overwhelmingly male and married local women.

    This certainly strengthens the case that there was a homogeneous predominantly R1b male, H female population that was expanding in Western Europe with similar massive demographic impact in both male and female lines (or perhaps greater male than female impact) after the early Neolithic and before the Bronze Age.

    The Únětice data are harder to read. Were T and/or WIX, like U, integrated in by the new culture from pre-existing indigenes (perhaps in a highly male dominated migration), or did it really arrive so late in the game to Western Europe? Conventional wisdom would be to associate T and WIX with first wave Neolithic popluations at the latest, and possibly the immediately preceding populations, not with a Bronze Age or Iron Age source. I don't feel that I know enough to make sense of this data in isolation and would want to know more about physical anthropology of the sources, archaeological context, preceding cultures in the same place and subsequent cultures in the same place, the larger context of WIX and T in Europe, and some informed comment on possibilities.

    ReplyDelete
    Replies
    1. Isnt this old. I guess I will look. It needs more y-dna.

      Delete
    2. "Were T and/or WIX, like U, integrated in by the new culture from pre-existing indigenes (perhaps in a highly male dominated migration), or did it really arrive so late in the game to Western Europe? Conventional wisdom would be to associate T and WIX with first wave Neolithic popluations at the latest, and possibly the immediately preceding populations, not with a Bronze Age or Iron Age source".

      All those lineages are indeed found in Neolithic and "forager Neolithic" contexts (N1a instead of I however, more H than in Unetice also, much less U5). Davidski and the PCA suggest they have Eastern European affinities, although a bit blurry. It could be interpreted in terms of BB pushing back Corded Ware towards Poland or other Eastern regions and these coming back with Unetice. But at this moment I would not make any hard bet for this scenario, just a somewhat plausible explanation.

      "Isnt this old. I guess I will look. It needs more y-dna".

      It's just 2012 and I did not know of it till yesterday. It's only mtDNA - that's what we have.

      Delete
  2. Daniel Kalina said "Isnt this old. I guess I will look. It needs more y-dna".
    Maju said
    "It's just 2012 and I did not know of it till yesterday. It's only mtDNA - that's what we have."

    I think Bell Beaker expansion can explain the mt H expansion for a part only .
    1) Expansion of H from South-West (Franco-Cantabric zone) has been only for some clades H1, H3, and some minor ones Hx, HV4, Vx, U5b, J1c, ... since LGM but other HV and H clades came from Balkans and Western Asia.
    2) If H3 expansion can be explained by Bell Beaker expansion from Iberia, the largest H1, V, .. expansions are more difficult .

    I think the limited mt expansion can be put in parallel with a limited demographic expansion of Y clade .

    As Maju and other, I don’t accept the absolute dates calculated from STR diversity without fudge factor.

    With a fudge factor we can put in parallel the expansion of Bell Beakers, some clades mt H, HV, V and some clades Y derived of the branch R1b-DF27 of R1b-P312.
    There are 3 main branches of P312 : the north-western L21 (maxima > 75% in Ireland, West Scotland, Wales, Cornwall and French Britanny) , the easter U152 (maxima > 50% in North Italy specially in Appennins and Alpes) and the southern DF27 (Maxima in Iberia, and along the French piedmont of Pyrénnees and surprisingly the French coast of Mediterranee (a strong and old BB colony. ). In more, there are the small branches DF19 and the scandinavian L238.

    Since several years, people tried to explain nordist and eastern expansion of the sub-branch of DF23 with SNPs Z196, SR2627, the nordic L165 , the maxima of SR2627 in Val d’Aran (30%) and Catalonia (22%) and Aragon (18%), and they haven’t succeeded with their hypothéses of recent dates and easter origin of R-L51, R_L11, R-P312 and R-U106 and their subclades during chalcolithic times.

    In more, the North-South branch determined by Dr. Kenneth Nordtvedt with the haplotypes and this was confirmed by SNPs foung from the
    project creates the same problem.
    North-South branch of DF23 ---> DF27 –> Z196 –> Z209 –> Z220 –> Z210 –> Z216 –> Z278 –> Z214 –> M153 (mostly basque)
    The North-South clade determined by Nordtvedt includes from Z209 to Z210, but not the most recent Z216, Z218 , Z214 and M153 (too recent to be included in the BB expansion ?).
    Z216, Z218 , Z214 are found in France and Spain centered around the Pyrenees and M153 very mostky in Spanish Basque country.

    Discussion in http://www.worldfamilies.net/forum ahttp://eng.molgen.org/
    “Z220 has a very strong representation in England, Pyrenees, Germany, Eastern Europe, and Netherland. The number of Z220 found in the Netherland is rap idly growing . R1b-L165 under L176.2, sister of the catalan SR2627) is founf in the outer island of Scotland and has been found in Norway. There is an important presence of DF27-derived in the south-wet coast of Norway, terminus of the Bell Beaker expansion to the North.”. Pyrenees and Netherland were secondary expansion origin of Belle Beaker.
    From “Webb” in the DF27 clades of Britain , “Z220 sems predominant in England, SR2627 in the West, L165 in Scotland” .

    ReplyDelete
    Replies
    1. Well, judging on HVS-I haplotype, 50% of the H detected by Chandler, Sykes and Zilhao 2005 in Neolithic Portugal is H1b, while I could not spot any H3 nor any other clear-cut haplogroup, just R0-CRS, which should be other H1. See the ancient DNA maps page.

      Also the spread of H3 (first detected in aDNA in Paternabidea, Basque Country, if my memory is correct) is more limited, not strongly affecting the area primarily under discussion here (Central and Northern Europe). See maps here for reference. However the spread of H1 in the Eastern Baltic and other parts of Eastern Europe may have a different origin.

      What brings me to a key point, which is that nobody is saying that H or H1 as a whole followed that pattern of Chalcolithic expansion as general lineage, just that people with some sublineages did apparently. H and H1 may well be A LOT older and have a different, probably Paleolithic origin. Like the expansion of Y-DNA R1 and other European lineages in America, this has no relation with the formation of the lineage but rather seems a late overflow, which was surely participated by large numbers of people (no marked founder effect therefore).

      On R1b-P312, I don't have a clear opinion because I lack enough detailed data, very especially on France (one of the largest states of Europe and historically and prehistorically even more important in demographic terms) on regional basis. However a tentative hypothesis considering the three main branches may suggest an origin in France, to the South or Southwest. If so, the most plausible mechanism I can think of is Paleolithic (Magdalenian, epi-Magdalenian).

      This does not fit well with the scenario we are discussing here but let's not forget we are discussing only Central Europe, especially Germany, not all Europe. Is perfectly possible for Central Europe to have experimented severe demographic changes in Neolithic and Chalcolithic and even the Bronze Age while other parts of Europe remained relatively stable.

      But we need more data on the genetics of France on regional basis, without that we are half-blind.

      Delete
    2. Expansion of V was almost surely Mesolithic, which can be inferred from its presence in Saami and Berber populations (predominant in the Saami) whose hunter-gatherer culture until recent times (probably maritime hunter-gatherer at first to a great extent) is well attested and whose archaelogical predecessors arrival in Scandinavia can also be pinned down pretty well. Incidentally, V is also high at other Atlantic coast of Europe locations like the Netherlands, which would be en route on the Saami migration. V is very low frequency in most populations with high mtDNA H in many Western European Bell Beaker/megalithic regions, so it was probably not a signficant part of that expansion.

      Delete
    3. I would not bet strongly for that. No V has been reported ever from pre-Neolithic contexts and even proto-Sámi samples lack it altogether.

      Delete
    4. Val d'Aran was estimated 48% SRY2627 albeit it was a sample of some 25 males who claimed ancestry at least 3 generations back. The donors are from the tiny villages of the Aran rather than the cities such as Vielha or Bossost.

      Delete
  3. Maju,
    As I've written elsewhere, I think R1b came to Western Europe:
    Mehrgarh -> Northern Iran -> Armenia/Syria -> NW into Balkans -> South of Alps -> Liguria
    From Liguria forked multiple ways, up the Rhone to Central Europe, down the Italian peninsula, SW into Iberia

    I think based on the Pala paper, that one of the mtDNA lineages that accompanied R1b males was T1a1, and especially T1a1a1.

    Perhaps mtDNA arose in Iberia and spread with Megalithism, and Y-DNA G or I??

    ReplyDelete
    Replies
    1. Not sure why everyone insists on debating R1b. Anyhow Pakistan does not look a likely origin at all, more like West Asia, considering the distribution of its main branches and its almost total absence in the Indian subcontinent. But that was surely deep in the Upper Paleolithic anyhow.

      I see absolutely no parallel between R1b and mtDNA T.

      "Y-DNA G or I??"

      G is West Asian and has been detected in Western Mediterranean Neolithic (G2a, today c. 10% in the area). I is SE European (Ukranian possibly) by origin and spread also in part with Neolithic (detected in Occitania, Sardinian-Pyrenean variant). The third detected Y-DNA Mediterranean Neolithic lineage is E-V13. None of them are too important today in SW Europe, so they must have suffered some sort of setback after the initial expansion.

      No R1b has been detected in these Cardial contexts yet, suggesting that the Neolithic hypothesis is extremely wrong.

      Also R1b in Europe has two main variants: Northern and Southern, each one must have coalesced and expanded from a different origin, possibly Doggerland and the Franco-Cantabrian Region respectively (but in wait for more data to confirm).

      Delete
    2. I should have refined the fact that I'm really talking about R1b-M269 in Western Europe, not all R1b.

      When I say Pakistan, I'm talking about R and R1 - there is some R1b in Northern Iran and very little in Pakistan.

      I think R1b-U106 spread North of the Alps to Central Europe from the Balkans, while R1b-M269 spread South of the Alps, along with R1b-U152.

      In terms of mtDNA, I'm explicitly not talking about T, as T1 and T2 have completely different distributions. Based on the Pala paper T1a1a1 is almost definitely an Indo-European mtDNA marker, and the only one I know of.

      Delete
  4. Correction:

    ...mtDNA H spread from Iberia...

    ReplyDelete
  5. "These Indoeuropean migrations (Tumuli, Urnfields, Hallstatt, La Tène, etc.) should explain the dilution of the extremely high apportion of H found in these Bell Beaker burials"

    However, we should not forget that detailed quantitative studies have shown that Bell Beakers typically only made up 10% to 15% of the population in most of Europe - and that is where they actually were significantly present. So, unless they not only brewed the best beer but also had the most "popular" women, there is no way they should have achieved a high demographic impact, in the first place.

    As to the northern half of the Únětice culture, it is important to point out that this is the core area of Germany (Saale to Eichsfeld to SE Lower Saxony regions) that probably has had the most continuity in all of Central Europe - including the bronze and iron ages, and including Roman times and later attempted invasions (Slavic invasions were expelled swiftly and first, there) - and is also considered the birth area of Germanic languages. Ancient R1b, I, and R1a admixture ratios attest to this continuity.

    Finally, IE entry is widely considered to be very early in the above region - significantly before Italy and Western Europe. So, the direction of spread of language - itself clearly not bound to specific y-DNA haplogroups and so likely also propagated by women - is exactly opposite to that of a presumed, contemporaneous mt H spread from Iberia. This is difficult to reconcile.

    ReplyDelete
    Replies
    1. "However, we should not forget that detailed quantitative studies have shown that Bell Beakers typically only made up 10% to 15% of the population in most of Europe"...

      True and very important potentially. Sometimes I wonder if by looking only at archetypal cultural sites we may be missing lots of population who lived at the margins of those and left less obvious remains. Not just with BB.

      However let's not forget that between early LBK and BB or even Corded Ware, there are 2000 years of what I would describe as quite a bit of agitation in what is now Germany especially:
      (1) LBK already met La Hogette (Cardial-like) culture upon arrival, as well as other non-LBK Neolithic groups in the area of Belgium, North France, Luxemburg and nearby German areas.
      (2) Entering the Chalcolithic we see what looks a tumultuous regionalization process, participated by Nordic Funnelbeaker, Western Megalithism, Eastern Kurgans (Baalberge and successors), which holds many questions open.
      (3) Only then comes to scene the Corded Ware, followed closely by Bell Beaker, which seems a reaction to it (generally inverse funerary customs).

      So between LBK and BB we have a large period of which we know almost nothing genetically speaking. What impact had Michelsberg? Other Funnelbeaker? What about Megalithism? Etc.

      These BB results may well be just the proverbial tip of the iceberg.

      "As to the northern half of the Únětice culture, it is important to point out that this is the core area of Germany (Saale to Eichsfeld to SE Lower Saxony regions) that probably has had the most continuity in all of Central Europe"...

      I'm not at all sure why you claim that. The Upper Elbe was shattered by the irruption of Baalberge and the other Kurgan cultural layers which followed it. Whatever the case, in this paper we have BB and Unetice from the same site (Quedlinburg) with totally different genetic pools.

      ... " the direction of spread of language - itself clearly not bound to specific y-DNA haplogroups and so likely also propagated by women"...

      That language is not linked to Y-DNA doesn't make it linked to mtDNA either. Language is, especially in the long run, a matter of power, ideology-religion-culture and economy. People change language like clothes, they don't change genes.

      Delete
    2. "we should not forget that detailed quantitative studies have shown that Bell Beakers typically only made up 10% to 15% of the population in most of Europe - and that is where they actually were significantly present. So, unless they not only brewed the best beer but also had the most "popular" women, there is no way they should have achieved a high demographic impact, in the first place."

      I agree that there is tension in the data. But, there aren't a lot of other good options besides the Bell Beakers to parsimoniously explain the major population genetic transition that took place after the early Neolithic and before the Bronze Age, so any way plausible to fit the two data points deserves serious consideration.

      One possibility is that the Bell Beaker component was 10%-15% in the first generation that was the source of the estimate, but that their superstrate culture afforded them demographic success in many subsequent generations.

      For example, I recall seeing a study somewhere documenting immense differences in pre-reproductive age mortality for children by social class in late prehistoric Europe.

      Keep in mind that a typical pre-modern fertility pattern is for a woman to have six to eight children in a lifetime, of whom about two or three survive to adulthood (child mortality was high even in high status families in the early 1800s as evidenced by families like those of Euler, George Washington, and Percy Shelley).

      If being part of a better fed and sheltered high status Bell Beaker family rather than a lower status non-Bell Beaker family translated into one additional child surviving to adulthood per generation (i.e. an average of four children who don't reach adulthood rather than five who don't reach adulthood), it would only take about five generations (about 150 years) at that rate for a 10%-15% Bell Beaker admixture proportion to turn into a 50% Bell Beaker admixture proportion if this was a steady advantage that suddenly disappeared after five generations. Bell Beaker derived cultures were dominant in Western Europe for about 1000 years, but one would expect the population genetic distinction between the superstrate class and substrate class (and hence the extent to which there is a population genetic advantage to having Bell Beaker genes because this coincidentally translates into higher socioeconomic status for ancestry informative cultural reasons) to gradually diminish over time. For example, a model with Bell Beaker admixture proportion rising over 10 generations starting with a 50% child survival rate advantage (which is just as 20% child mortality advantage) in generation one and fading at the appropriate proportion with each of the subsequent nine generations until there is no advantage, isn't at all implausible and reconciles the tensions in the data.

      Delete
    3. "As to the northern half of the Únětice culture, it is important to point out that this is the core area of Germany (Saale to Eichsfeld to SE Lower Saxony regions) that probably has had the most continuity in all of Central Europe - including the bronze and iron ages."

      This comment doesn't make much sense in the light of archeological and latest aDNA data.

      Delete
    4. @Andrew: you must mean this entry: http://forwhattheywereweare.blogspot.com.es/2011/03/childhood-and-death-in-neolithic-and.html - notice that obvious class differentiation and child malnutrition was only present in the militarized (feudal?) set of Motilla del Azuer, while in Catalan Neolithic things were more normal with no signs of class nor age differentiation.

      "So, unless they not only brewed the best beer but also had the most "popular" women, there is no way they should have achieved a high demographic impact, in the first place."

      The "popular women" claim sounds like something one would say at a pub as a joke, not in a scientific discussion. Incidentally I just discarded for comment (no time) a paper on sexual selection in mammals (mice under controlled conditions, lost the link sorry) and it seems it's all or nearly all biased to the male side, especially highly successful gigolo (promiscuous) mice. On the female side the only detected difference was minor and said that monogamous females had slight advantage (more litters).

      So when we see such a radical change in the mtDNA we must think in terms of general population replacement, not gender bias. In other words: it must have been coupled with replacement also in the Y-DNA side, even if we can evaluate that part directly.

      "One possibility is that the Bell Beaker component was 10%-15% in the first generation that was the source of the estimate, but that their superstrate culture afforded them demographic success in many subsequent generations".

      That's plausible, especially as BB lasted for many centuries. However, if so, we should also see the telltale signs of cultural replacement (children also inherit parents' culture, more so if socio-economically dominant) and I'm not too sure we can admit to that actually happening on light of the archaeological evidence.

      "Bell Beaker derived cultures were dominant in Western Europe for about 1000 years"...

      Not quite. Cultures with Bell Beaker insertions were but BB as such is a confusing phenomenon. I often compare BB people with Jews in Medieval and Modern Europe: a distinctive, often affluent, minority everywhere. Jews at least did not have any major genetic impact but maybe BB subcultures were more open to admixture (no monotheism, no strict dogma, no obsessive purity rules that we know of before the Christian era). Still it's difficult to see how their genetics could become dominant unless ethnic cleansing or genocide happened (what would have to be documented somehow or remain under strong question mark).

      You present a statistical construct but how well, if at all, is that construct backed by the archaeological data? I'd say that not at all.

      Delete
    5. "You present a statistical construct but how well, if at all, is that construct backed by the archaeological data? I'd say that not at all."

      The genetic data imposes some powerful and pretty much irrefutable boundaries on the range of possibility, however. Something turned a region that was predominantly mtDNA U into a region that was predominantly mtDNA H. It must have happened post-early Neolithic. It must have happened pre-Iron Age. It happened and the question is which is the possibilities, however much they would otherwise be implausible, is the most plausible option.

      All possible explanations involve some combination of differential reproduction rates (a product of total fertility and average generation length) and differential death rates prior to reproduction, which itself can be collapsed into differential net growth rates of gene carriers (which in turn leads to differential percentages of genetic haplogroups).

      Without a decent base number to start from it is basically impossible to get to 50% or so without a really powerful selection agent. There is really only one possibility which can fit what we know for that time period (since H itself is not intrinsically fitness enhancing vis U by itself). This is that initially lactase persistence was present in BB people but not indigeneous people and that the huge growth in H was tied to this correlated fitness enhancing gene. This has the virtue of being possible to persist at a lower level over many more generations since it need not be culturally driven (which tends to change more rapidly within a shared community) since correlation between the fitness enhancing factor (the LP gene rather than cultural status) and the mtDNA H would decay more slowly (i.e. have a statistically significant benefit to the average H carrier) than it would if it were a cultural factor only. If you can spread it out over 1000 years instead of 300 (even beyond BB culture proper) on this decaying H v. LP gene correlation, you can have a much more subtle effect that builds from the 10%-15% base from migration, and hence a less archaeologically noticable one.

      Genocide/ethnic cleansing seeems to be a less solid fit to the archaeology than LP gene correlation, or social class distinction impacts on child survival. We want something subtle and I can't think of anything else that would be more subtle.

      Delete
    6. In the Neolithic it was not anymore "a region that was predominantly mtDNA U", even if you include K ("U" is such a wide term, with authors and bloggers conflating U*, U2, U8, U4 and U5 confusingly as if it was all the same thing). But H was not yet dominant either or in any way the genetic pool was similar to modern one.

      Then in the Late Bronze Age we have the first genetic pool of the region similar to modern in an Urnfields site. But between the Early Neolithic (c. 5000 BCE) and the Late Bronze (c. 1000 BCE) there is a huge blank of 4000 years: twice the length of written History in most of Europe. To fill that huge blank we have now two sets of samples from the Late Chalcolithic (BB) and Early Bronze (Unetice). These can't tell the whole story but they give us some clues.

      You are considering highly conjectural biological adaptive causes, which I deem most unlikely for a number of reasons. Instead the most straightforward explanation is one of demic migrations associated to armed violence. These could well have begun in the time of Early Neolithic, followed by the complex waves related to Funnelbeaker (Michelsberg, Baalberge, Deeply Impressed Pottery...) and Megalithism, then by Corded Ware (surely also violent) and then by BB, Unetice, Tumuli and finally Urnfields. Who brought what exactly is not so clear yet but that these seem to imply to at least some degree peoples in movement (with an economy but also armed force) is an explanation hard to disdain.

      "Genocide/ethnic cleansing seeems to be a less solid fit to the archaeology than LP gene correlation"...

      I disagree: the ability to drink milk is (1) much more widespread than you seem to imagine and (2) not so important for survival (except in purely herding cultures maybe, of which there are none known historically in Europe - but are rather more typical of semi-desertic areas in Africa and Asia).

      Delete
    7. The LP gene has a clear selective advantage and went from essentially 0% as of the dawn of the Neolithic to very large percentages of the population - it is one of the half dozen most strongly selected functional genes in all of the Holocene. It is common now, but that wasn't always so. The discrepencies you note in the linked post are in multiple cases due to the fact that the mix of genes that cause LP in Africa are different than the set that do so in Europe. But, the African LP genes weren't present in Europe (or in all likelihood in Africa in more than trace frequencies).

      I see no basis for your argument that milk drinking isn't powerfully fitness enhancing. Keep in mind that unlike modern societies, through most of farmer history food shortages were pretty much an annual event and that even if one didn't die of starvation, one died from a weakened immune system since underfed. Also, milk specifically includes Vitamin D which is just as immune protective as Vitamin C and is particularly important for darker complexioned people (such as recent migraants from more Southerly latitudes).

      Of course, I am nearly certain that in fact there were multiple causes of the changes in the gene pool. No one seriously believes any more than there was no war or no displacing migration in the time period. No one seriously believes that LP genes dramatic rise in frequency was unrelated to fitness enhancements associated with it. Some superstrate-substrate advantage in mixed communities was probably present too. The exact mix of the different effects is difficult to determine and may be unknowable, but of course, talking about the shift being due to any one cause will always be an oversimplification. There are other possibilities too, of course. For example, an unsuccessful crop year or small pox outbreak wiping out most of the people in a village and leaving a vacuum that BB or some other population filled.

      Delete
    8. I don't think that the Western European LP allele has such a clear selective advantage: milk is just another way to eat fat and proteins, just like meat or beans. And we know that early Central European farmers already made cheese, which is assimilated by lactose intolerant people.

      We also don't know if the early farmers had other alleles which provide also for LP phenotype. You are judging very shallowly here. Italians for example have very low levels of the Western LP allele but their actual LP phenotype is quite high. On the other hand, Eastern Europeans, who probably would benefit at least as much as Western ones from the LP phenotype, have very low levels of LP phenotype in fact, c. 50%, lower than Mediterraneans.

      LP and especially the Western European LP allele seems pretty much accessory and random to me.

      "milk specifically includes Vitamin D"

      It does not. The only edible significant source of vitamin D is fish.

      Milk is one of the few foods that has a positive calcium-phosphate balance however but, except for elderly people at risk of osteoporosis, this does not seem so important.

      Vitamin D metabolism (which generally favors lighter skin shades) might indeed be a biologically important element playing against new immigrants from the South. But that's just a raw untested theory because, for example, Portuguese and Polish have similar skin shades, as we also discussed recently, so I imagine that immigrants from Portugal would have zero vit. D disadvantage vs. natives from Poland, who otherwise seem to be doing quite well.

      The intuitively darker appearance of Portuguese is caused by eye and hair color, which is irrelevant re. vit. D generation. It may also be enhanced by tan but this comes and goes with solar input.

      Delete
  6. Maju,

    I only claimed significant continuity after the Únětice culture - although there certainly also was some, before (but I fully agree with the tumultuous events you mention following LBK etc.; and OT, I find La Hoguette a good candidate for first the northward, and then eastward spread of R1b to the Rhine river and later diffusing beyond).

    Baalberge (~4,300 - 3,100 BCE) is very interesting, because it unifies very early (for the region) megalithic and Funnelbeaker elements. More importantly, several Baalberge sites transition smoothly to Únětice and Corded Ware, respectively - continuing to use the same general grave sites.

    At any rate, the main point I tried to make is that to me, BB in itself seems insufficient to explain the growth of mt H in much of Europe.

    "So when we see such a radical change in the mtDNA we must think in terms of general population replacement, not gender bias."

    Yes, my "most popular women" statement was meant as a contrasting joke - something that normally would exactly *not* explain what happened. Even if BB were the minority underdogs having to give away their daughters to be allowed to stay, those would have hardly achieved a status to successfully found families for generations - especially since unlike a conqueror scenario, there is no reason to believe that the majority had a shortage of women, nor were they growing significantly, at the time. Again, I agree that much more than BB must have taken place, here.

    Finally, unlike male minority domination, with entire population replacement processes you would expect associated a SW European phenotype and autosomal DNA to make huge inroads - which, by all studies I know of, did not take place.

    Andrew,

    The problem is that socioeconomic success of minorities usually leads to males taking on outside females - not the other way around. Outside of selection, disease, or general replacement, it is hard to explain such huge mt DNA replacement. Just look at how much more "cosmopolitan" (diverse) Únětice was in comparison, and how - especially if reflected in autosomal DNA - it should have been better equipped at coping with adversity.

    Davidski,

    As often, I disagree with you completely, but can/will not further comment on your unsubstantiated one-liners.

    ReplyDelete
    Replies
    1. "... unlike male minority domination, with entire population replacement processes you would expect associated a SW European phenotype and autosomal DNA to make huge inroads - which, by all studies I know of, did not take place".

      How consistent are those studies? Old school anthropometry studies often considered the Nordic type a subtype of the Mediterranean one. What about Atlanto-Med and Keltid types? I am generally reluctant to discuss those matters but they are certainly open to interpretation.

      Delete
    2. Old school typology is just pseudo-science. There's no other way to characterize it.

      Western and Northwestern Europeans show significant autosomal influence from the Mediterranean Basin. For instance, here's one of my analyses from many months ago, which just happens to gel very nicely with the latest aDNA data.

      http://eurogenes.blogspot.com.au/2012/03/northwest-eurasians-southwest-eurasians.html

      Delete
    3. Autosomal analysis also has many issues with sampling. For example, I'm planning ahead to study Europeans without dramatically isolated populations like Sardinians or Finns (who distort way too much) and pondering the samples +/- according to real population (why would a Lithuanian sample weight the same as a French one when there are 2 million Lithuanians and 60 million French - and the difference was even more radical in the past?) Said that, your analysis is relatively alright but seems to fall in the usual vices (how do you draw the maps? - I can use ADMIXTURE but I don't seem able to make cool graphs out it). Also it is important to draw a tree of the components based on Fst distances, what should explain better their genesis (always a bit hypothetical, as is their existence in absolute terms).

      Delete
    4. Well Maju, I'm way ahead of you there. In that analysis I only used one Sardinian, one Basque, and very few Finns. See, Finns are OK to use if there aren't too many of them and they don't come from any of the east or north Finnish isolates.

      Now, Lithuanians do behave like an isolate in these sorts of analyses, but if you take them out, then Belorussians will create a very similar cluster. And if you take out the Belorussians, then Russians will take their place...then Ukrainians, then Poles. So you might as well use the Lithuanians. That's been my lesson anyway.

      By the way, the ADMIXTURE manual gives tips on how to create graphics from the output, but not maps. The maps can be done with various freeware.

      Delete
    5. I see: you're doing almost exactly as I would. I agree about Lithuanians also: I did not mean that they are too endogamous to be meaningful, as may be the case with the other mentioned populations, just a matter of demographic weight (i.e. use Russians and Polish instead, for example).

      I will see in due time if I can find the map-making freeware: it seems a very cool feature.

      Delete
    6. @eurologist

      For mtDNA gains, rather than just Y-DNA gains, due to superstrate-substrate distinctions, you would need a gender balanced initial migration and a not terribly long lived but quite strong while it lasted caste system. It wouldn't have to last as long as India's, just 300 years or so (v. 3500 for India's castes) and could have some decay in the strength of the caste system with each subsequent generation (a good comparison case in the historic era might be the duration over which Normans (or pre-Norman Scandinavians) in England maintained themselves as a distinct superstrate caste. There does seem to be some archaeological evidence of endogamy between BB and non-BB, and of BB being superstrate rather than substrate.

      Delete
    7. Andrew,

      Interesting thoughts. I will try to think about it in the proper temporal context.

      Delete
  7. En España donde hay mas H , es donde hay menos R1b .
    Esta claro que no tiene el mismo origen.
    La terminación de los apellidos EZ tiene origen materno.
    Coincide como un guante con las autonomías con más H y menos T.
    Galicia, Asturias y La Rioja. Nativo y no Indo ario.
    Y que en un principio se heredaba de la línea materna como en Portugal.
    Y en algún momento de la edad media se cambio a la línea paterna

    ReplyDelete
    Replies
    1. (Quotes translated)

      "In Spain where there's more H, is where there is less R1b".

      False. Record H and record R1b are both coincident in the Basque Country and nearby areas (it may vary a bit from dataset to dataset, particularly the mtDNA). That's today. In the past probably the Western Third of the Peninsula (at least Southern Portugal) almost certainly had more H, but that changed at some point between the Chalcolithic and present day, probably with the IE expansions, notably the Celtic one. Today Portugal is among the Iberian regions with lowest H but also lowest R1b.

      There are some regions of Europe where mtDNA H is high and R1b low to very low however (Baltic, Italy), so it's not a strict 1:1 relationship but overall there seems to be a serious correlation or at least an appearance of it.

      "The -ez ending of surnames has maternal origin".

      No idea where you got such idea, which is almost certainly wrong. In any case there is a lot of debate about the origin of that patronymic suffix: a traditional belief has been that it may have been Germanic (i.e. similar to the Saxon genitive, as in "'s" of in "Jones") but I know of some linguists who suspect otherwise: that it may be a remnant or even direct loan of Basque -(e)z declension, meaning "made of" (burdin-ez: iron-made, plastikoz: plastic-made, urrez: gold-made, golden, gilded).

      "It fits like a glove with the regions with greater H and less T. Galicia, Asturias and La Rioja. Native and not Indo-Aryan".

      The -ez suffix is clearly a patronym and is also abundant in Portugal, where it's written -es. In the Basque Country there are also lots of Castilian patronyms in -ez, particularly in Araba, where they usually retain a fossilized form followed by "de" placename (i.e. Álvarez de Arcaya) This was the common antique usage of the -ez patronyms, i.e. "son of name from placename". The oldest forms of the -ez patronym I know of are those of the Kings of Pamplona, who were nearly all named García Sánchez and Sancho Garcés (or just García), where the second name is obviously a patronymic, meaning "son of Sancho/García".

      "And that was initially inherited by maternal line as in Portugal. And at some point of the Middle Ages this was changed in favor of the paternal line".

      There is no "Urraquez" or anything like that: that would indeed be a matronymic, but there is López (son of Lope = Wolf), Pérez (son of Pero = Peter), Gómez (son of Gomo), González (son of Gonzalo), Íñiguez (son of Íñigo = Eneko = Angus), Álvarez (son of Álvaro), Velzaquez (son of Velasco = Belasko = little crow), Vélez (son of Vela = Bela = raven or crow), etc. Not a single one comes from a female name.

      As far as I know the custom of maternal inheritance of surnames, which I fully respect, is privative of Portugal in the Iberian Peninsula and has never existed elsewhere.

      Surnames are all in any case a recent phenomenon: there were no surnames in the Middle Ages and they were not fixated until the 19th century, where laws were issued to ensure that. Roman era surnames got lost with the collapse of the Empire and the arrival of the Dark Ages.

      Delete
    2. De eso se trata de ir más allá, actualícese usted por que en el país Vasco lo que hay es 49% de H. Y las que cito tienen más del 55%. Una de ellas esta al lado de el país Vasco. Su conocimiento de España y su historia es de lo más común. Y Cataluña que tiene tanto R como el país Vasco, es donde menos H femenino hay.

      En 3 apellidos con I, y en 4 con E, G, T. en los apellidos mayoritarios.

      Apellido Martínez de La Rioja y Cuenca.
      R = 71 % Indo ario.
      I = 11 % Nativo.
      J = 7 %.
      E,G,T = 10 %.

      Apellido Martín de Ávila.
      R = 68 % Indo ario.
      I = 7 %.
      J = 6 %.
      E,G,T = 18 % Relevante.

      Apellido Pérez de Teruel.
      R = 64 %
      I = 17 % Nativo.
      J = 9 %.
      E,G,T = 9 %.

      Apellido López de Huesca y Lugo.
      R = 63 %.
      I = 12 %. Nativo.
      J = 7 %.
      E,G,T = 17 %. Relevante.

      Apellido Fernández de León, Asturias y Cantabria.
      R = 63 %.
      E,G,T = 23 %. Relevante.

      Apellido Rodríguez de Pontevedra, Orense y Zamora.
      R = 62 %.
      E,G,T = 24 %. Relevante.

      Apellido Sánchez de Salamanca y Cáceres.
      R = 59 %.
      E,G,T = 25 %. Relevante.

      Y el García que es el apellido que mas hay en España tiene.
      R = 70 %. Indo ario.
      I = 9 %.
      J = 8 %.
      E,G,T = 12 %.

      Delete
    3. As you seem that you understand English, why don't you comment in English so everyone can read what you say and discuss it?

      "Update yourself... in the Basque Country there is a 49% of H"

      That would be already high for West European norm, which is around 40% but in fact the most recent studies suggest that the frequency of H is even higher:

      → http://forwhattheywereweare.blogspot.com/2012/02/basque-mtdna.html

      Behar 2012 found 67% in Enkarterriak and 54% in the Basque-speaking areas of the Southern Basque Country. Other grouped regions of the "Greater Basque Area" were in the 38-48% range.

      It's good to check the specific regions because there is some notable variability between them and their groupings may be a bit misleading. For example Araba has 38% of H1, plus 11% of H3, plus 4% H*, plus 2% H58, total: 55%, much higher than the "Spanish-speaking Basque" area it is grouped in.

      In the surrounding Spain the variations are also very dramatic from province to province. Cantabria may have 74% H (although the sample is small), La Rioja 58%, but then Northern Aragon (genetically more akin to Gascony) less than 40%.

      "R = 71 % Indo ario."

      R is not "Indo-Aryan". Actually Indo-Aryan is a linguistic term that only applies to the Indian subcontinent. If you mean Indoeuropean, and you mean by origin in the Samara Valley (and not linguistically, which is mostly unrelated to genetics), only R1a1 may have some relation with that prehistoric flow and that lineage is nimious in Iberia. There is very little Indoeuropean genetic flow in Iberia, really, at least of the kind which can be directly tracked to the steppe.

      R1b and R1a have only a very remote relationship, not much closer than the relation they have with mostly Native American Q, for example or with Papuan M and S. Talking or R in general can only be done in terms of the Early Upper Paleolithic. Later it's two or more different lineages.

      Your argumentation on surnames is anyhow totally inconsistent. Were they inherited matrilineally as you wrongly claimed, there would be no relationship whatsoever with R, mostly R1b-S116. The frequencies you mention of R are absolutely normal for the Iberian Peninsula, where R1b alone is usually above 50-60%. The vast majority of that R has no relation whatsoever with Indoeuropeans but is actually pre-IE.

      You know much less than you think you know, Conchi.

      Delete
    4. Es que para encontrar alguna excepción, abría que ir a las provincias limítrofes con Portugal. El país vasco en proporción no lo es.
      Las R0 en España son el 52%, en el país vasco el 57%.
      En La Rioja, Asturias y Galicia el 62%. Los R1b son hasta un 12% menor a la media.
      Haplogrupo materno H corresponde a cultura campaniforme, pero el R1b en España es posterior, no hay pruebas de lo contrario.

      Delete
    5. MtDNA H has been clearly documented since the Epipaleolithic (in Cantabria, Basque Country and Portugal). Therefore in Iberia mtDNA H as a whole does not "correspond to Bell Beaker culture" (which anyhow seems to originate in Iberia).

      Elsewhere: in Central Europe H appears first in association with first farming but in frequencies much lower than today, in NE Europe may be as old as Gravettian (Sunghir) but without doubt also Epipaleolithic (Karelia). Other areas like Britain may have also hosted H in the Paleolithic but so far undemonstrated beyond doubt.

      "el R1b en España es posterior, no hay pruebas de lo contrario".

      There is indirect evidence via the amazing stability of the Basque genetic pool since Neolithic or soon after it. Basque Y-DNA pool is ~90% R1b, so parsimony obliges...

      see:
      → http://forwhattheywereweare.blogspot.com.es/2013/09/basque-and-other-european-origins.html

      or in Castilian language:
      → http://www.amaata.com/2013/09/el-origen-de-las-vascas-y-otras.html

      Delete
  8. El haplogrup H materno en su conjunto corresponde ha la cultura que motiva este articulo. Como puede haber correspondido a otras anteriores o de otros lugares. Administrador. La población masculina original española oscila entre nada y casi nada. Los romanos la sustituían por esclavos traídos de lo más lejos posible.
    Cataluña era repoblada en masa por occitanos y occitanas. El país vasco estaba despoblado a propósito, era una zona minera y dada en propiedad a mercenarios gascones. Los cantabros se suicidaron, o cantaban muriendo cuando los crucificaban. En Asturias no avía ni un minero nativo, nunca se rindieron. ídem de ídem. La T2 de Aragón es la típica romana. Los pirineos se convirtieron durante siglos, en una zona estratégica por el tema del Hierro para los R1B. Y el 90% del oro Romano lo sacaban de una montaña Asturiana, Trajada por esclavos traídos del otro lado del imperio. Los R1b no dominaron los metales hasta la edad del Hierro. Antes eran Vaqueros. Como los R1b+ V88 africanos. Y aunque los arios son los R1a, los indo europeos R1b, también son de raíz lingüística Indo Aria. Menos los cornudos vascos, que tienen que recurrir a las pruebas genéticas modernas, para enterarse de que padre son. Y esto es lo que hay, le guste o no. Hasta siempre.

    ReplyDelete
  9. The maternal haplogroup H as a whole has corresponds culture that motivates this article. As you may have corresponded to earlier or elsewhere. Administrator. The original Spanish male population ranges from nothing and almost nothing. The Romans replaced by slaves brought from as far away as possible.
    Catalonia was repopulated by mass and Occitan Occitan. The Basque country was depopulated by the way, was a mining area and given ownership to Gascon mercenaries. The Cantabrian committed suicide or dying sang when crucified. In Asturias Avia no mining or a native, they never gave up. ditto. The T2 is the typical Roman Aragon. The Pyrenees became for centuries a strategic location in the topic Iron for R1B. And 90% of gold Romano him out of Asturian mountain, Worked by slaves brought from across the empire. The R1b did not dominate the metals to the Iron Age. Jeans were before. As African R1b + V88. And while the Aryans are R1a, R1b Indo Europeans are also Indo Aryan linguistic root. Less Basque cuckolds, they have to resort to modern genetic testing, to find that father is. And this is what it is, like it or not. Farewell.

    ReplyDelete
    Replies
    1. "The original Spanish male population ranges from nothing and almost nothing. The Romans replaced by slaves brought from as far away as possible".

      I can assure you that doesn't make any sense: it does not fit with the genetic pool data, it does not fit either with the historical data we have re. the Roman period either. Also Roman slaves did not in most cases survive to reproductive age, particularly the males.

      Even in well known plantation slave colonies like Sicily, modern population structure doesn't seem to correlate with any sort of massive slave import, so much less in other larger and less specifically slave plantation provinces. I can't discard some Roman-induced genetic influence in Iberia but it's not easy to discern really: the genetic pool of Iberia seems in essence deeply rooted and that is particularly true for Y-DNA R1b-S116, whose origin seems to be in SW Europe.

      "Catalonia was repopulated by mass and Occitan".

      That's not easy to demonstrate. Personally I would not think so because the Catalan genetic pool is closer to that of Gascony, which back in the day was a Basque-speaking region (not yet Romance), than to that of Provence. However the language is closer to that of Provence and Languedoc.

      You seem to hold a lot of beliefs that seem hard to conciliate with reality. Catalonia was rather conquered by the Frankish than colonized by them. There's no obvious trace of recent colonization associated to Muslim or Christian conquest.

      "The Basque country was depopulated by the way, was a mining area and given ownership to Gascon mercenaries."

      Fantasies of some armchair historian surely. It does not fit with reality either, nor with history.

      You have a lot of absurd fantasies that you treat as if they were "history" and "facts" and, honestly, I'm getting annoyed by them. Moreso as you state them but do not back up your claims.

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