February 4, 2017

The patrilineage R1b-DF27 in North Iberia

Just weeks ago a new study on Northern Iberian Y-DNA, focused specifically on R1b-DF27, was published. It covers Asturias, Cantabria, Basque Country and Aragon, finding greater diversity in the Basque Country and Cantabria and lower in Aragon and Asturias.

Patricia Villaescusa et al., Characterization of the Iberian Y chromosome haplogroup R-DF27 in Northern Spain. FSI-Genetics 2017. Pay per viewLINK [doi:10.1016/j.fsigen.2016.12.013]

Abstract

The European paternal lineage R-DF27 has been proposed as a haplogroup of Iberian origin due to its maximum frequencies in the Iberian Peninsula. In this study, the distribution and structure of DF27 were characterized in 591 unrelated male individuals from four key populations of the north area of the Iberian Peninsula through the analysis of 12 Y-SNPs that define DF27 main sublineages. Additionally, Y-SNP allele frequencies were also gathered from the reference populations in the 1000 Genomes Project to compare and obtain a better landscape of the distribution of DF27. Our results reveal frequencies over 35% of DF27 haplogroup in the four North Iberian populations analyzed and high frequencies for its subhaplogroups. Considering the low frequency of DF27 and its sublineages in most populations outside of the Iberian Peninsula, this haplogroup seems to have geographical significance; thus, indicating a possible Iberian patrilineal origin of vestiges bearing this haplogroup. The dataset presented here contributes with new data to better understand the complex genetic variability of the Y chromosome in the Iberian Peninsula, that can be applied in Forensic Genetics.

The study, quite conveniently, differentiates between "native Basques" (those whose patrilineal ancestors lived in the Basque Country for at least the last three generations) and "resident Basques" (those whose recent patrilineal ancestors immigrated, mostly from NW Iberia).

R1b-DF27 is one of four major R1b sublineages in Western Europe and one of the three "brothers" that can be tracked to an origin somewhere in what is now Southern France, most likely, i.e. together they form part of R1b-S116. The fourth lineage would be, naturally, R1b-U106, "brother" of S116 and found typically around the North Sea. It is the one with the southernmost distribution, being very dominant in Iberia and among Basques. Probably it is also important in all the south of modern France but clear data is missing as of now.

Reconstructed spread of R1b to Western Europe and within it (dates objectively unknown so far, own work)

This is the key data table of the study, showing the frequency of the various sublineages of R1b-DF27 ("*" means "others", so "DF27", without asterisk, means "all DF27" and "DF27*" means instead "remaining DF27 after exclusion of the other mentioned subclades"):

Click to expand (frequencies are absolute, relative to whole sample)
It is also worth sticking this other graph, which shows (top right) the (SNP-based) true phylogeny of the haplogroup R1b-DF27 and, complementarily, the (somewhat messy) haplotype structure based on a limited number of short tandem repeats (STR), in which only Z220 appears clearly defined:

Click to expand

The study is very limited in its scope but it does show that there is a very high diversity for this lineage among Basques. This however does not necessarily indicate that Basques are the direct origin: much more data from the rest of Iberia and very especially from France is required before we can jump to any conclusion. Based on the limited data we have, I am of the opinion that the lineage did not originate in Iberia most likely but rather in what is now Southern France, migrating southwards via the two natural corridors: the Basque Country and Catalonia. 

Sadly enough we just do not have enough modern data, much less ancient one, in order to issue a definitive judgment on the matter. However the overall pattern of distribution of R1b-S116 strongly suggest a "Southern French" origin, not just for "Iberian" DF27 but also for the other two "brother" lineages: "Alpine" U152 and "North Atlantic" M529. 

The big question is how and when did this expansion took place. A "South French" origin was much easier to explain when the Paleolithic continuity model seemed reasonable, however recent ancient DNA findings strongly suggest that the Neolithic and Chalcolithic saw major population changes in much of Europe until stabilization was achieved -- exact patterns vary on specific regions: in some cases this does not happen until the Bronze Age, in others, like the Basque Country and quite possibly the Atlantic parts of France, it may have happened much earlier, even as soon as the early Neolithic. 

So my best recipe for an explanation is that we have to look very carefully at what happened in Western Europe, particularly towards the Atlantic Ocean in that "transitional" period, when not just large cultural phenomena like Dolmenic Megalithism or later also Bell Beaker manifested in quite expansive and dynamic manner but also a dearth of smaller cultures were the actual social or ethnic pieces making them possible. For example it is plausible that Michelsberg culture (originating in Lower Rhineland apparently and swiftly replacing the early Neolithic LBK culture in Germany, North France and nearby areas) could be involved in the expansion southwards of R1b-U106 and other traits of the modern genetic pools we observe. Another culture well worth taking a look at is the Artenac culture, which expanded from Dordogne towards the North up to Belgium soon after the Michelsberg/SOM era. Rather than one single and sudden expansion of a well defined population, it seems to me that we are before a jigsaw puzzle of several cultures and several chronologies, related maybe but not exactly the same.

See also:

Thanks once again to Jean Lohizun.

34 comments:

  1. I think you straight arrows greatly overestimate the navigational skills of Neolithic peoples haha.

    I think M269 is born on the North bank of the Danube in the MN. L23* goes east/south, L11 goes North/west. U106 up the Rhine. All this out of Cucuteni. WHGs all spoke your Basque. EHGs spoke some form of Kipchak or Uralic. Farmers PIE.

    ReplyDelete
    Replies
    1. I know you mean as a joke but probably they do not overestimate anything (even if they weren't indeed meant to argue for such skills either). It's known for instance that HG people in Wight Island "traded" (or robbed) cereals from farmer people in Brittany or Normandy, so either side must have been able to cross the Channel on boat (in summer I presume but still better than you imagine). Notice that those same farmers (generically speaking) settled places like Cyprus, Crete, Sardinia and the Balearic Islands since very early dates, and that there is evidence of deep sea fishing in some of their sites. Cardium Pottery farmers and some of their relatives were excellent sailors, I wouldn't dare to compare with ancient Austronesians and their incredible open ocean journeys but next after them in the scale of skill and daring pretty much yes.

      It's been even argued that the expansion of Dolmenism to Northern Europe might have been caused by cod fishing expeditions to Ireland and such. Never proven but never disproven either. In any case the Megalithic peoples of Brittany and Upper Normandy did not settle Kent but SE England, Wales, Scotland (including Orkney) and surely also Ireland via the Irish Sea. The short crossing through Dover Strait was used by LBK-related peoples of Northwest France, near Belgium, instead, probably less inclined to navigation themselves (LBK is a continental culture).

      Another very "seagoing" population were the Mesolithic peoples of Denmark and surrounding areas (Maglemosean), who did not just live there but also across the North Sea in the Eastern parts of Britain. It's true that their chronology overlaps the end of the last Ice Age, so the now submberged peninsula known as Doggerland, must have acted as intermediate "port of call" (actually it was surely quite densely inhabited, but underwater archaeology is very costly, so only limited direct confirmation exists) but they were still a quite seagoing people on mere logboats.

      Another example are the Pitted Ware subneolithic (farmer influenced hunter-gatherers) culture of the Baltic, which clearly crossed that sea forth and back at a later date.

      So I would not disdain the seagoing capabilities of ancient peoples, particularly where there is evidence, direct or indirect, supporting their naval mobility.

      Delete
    2. "I think M269 is born on the North bank of the Danube in the MN".

      Middle Neolithic? Too late by all means. Such a pattern can only be explained either by Early Neolithic carriers, regardless of which were the cores and times of secondary S116 and U106 expansions (the rest is not really a expansion but rather a trickle) or by older migrations. That would actually put us in the Gravettian period because there's nothing like that (a Balcan or West Asia originated migration to Europe) in between, nor after the early Neolithic either. The only possible frames (on archaeological grounds) for the arrival of new genetics (new peoples) from West Asia (via the Balcans) in significant numbers are c. 49 Ka BP (proto-Aurignacian), c. 32 Ka BP (Gravettian) and c. 9-7 Ka BP (Neolithic). Those are the windows and no known archaeology supports any other window at all.

      "L23* goes east/south, L11 goes North/west."

      Those are just intermediate paragroups between the M269 stage in the Balcans (or West Asia) and the S116 and U106 stages in West Europe. Don't get too worked out on them: they seem to exist (at low frequencies) in certain areas and not others and this geographical scatter could support either a Central European or a Mediterranean route (or both). In any case the phylogeny is M269→L23→M412→L11→S116/U106. Somehow L11 peoples arrived to South France and somewhere at the North Sea maybe and, from those two cores, expanded wildly. At least that's what it seems. But in any case the intermediate phase L23→M412→L11 stage is not expansive but a weak scatter, with very few people carrying those intermediate paragroups today, so they probably migrated inside much more diverse groups before finding their own niches.

      One could argue that, as S116 and U106 both expanded and both are L11 sublineages, that L11 is the lineage to consider but there is nothing obvious about any sort of L11 expansion as such, so a possibility is to consider both S116 and U106 expansions as two distinct phenomena, although it's indeed possible that the carriers were culturally related and benefited from the same kind of historico-political wave. Alternatively U106 can be considered as some sort of peculiar (earlier?) relative of the more numerous S116 expansion.

      If we have to consider both process as one, the raw centroid falls always in what is now France, although hard to say exactly where because it depends on where in Iberia you put the reference for L11* (Iberia is huge!) My personal bet would be Dordogne (core of Artenacian) but Burgundy (Gurgy!!!) could be as well (and also others but always in France).

      Delete
  2. What is your view on the SNP age estimates for R1b M269 on Y Full? I used to like the idea of a Solutrean origin of R1b in Europe but age estimates and the new data make me inclined to believe there is a connection to central European Neolithic cultures.

    ReplyDelete
    Replies
    1. Haven't checked in a long while but last time I did: rubbish. Age estimates have several issues:

      (1) Wrong calibration points (such as a supposed OoA c. 60-70 Ka BP or a Pan-Homo divergence c. 5-6 Ma ago, based only on this you should double all figures to be minimally realistic)

      (2) Belief that modern father-to-son "observed" mutation rates are the standard to follow, when in fact they are totally ignoring the high chances for any mutation, in the Y-DNA or mtDNA too, being deleterious and therefore not surviving beyond the first carrier, the observed mutant.

      (3) There's enough time between mutations (less in the Y-DNA, much more in the mtDNA line) for freaky things to happen, such as simple genetic drift eliminating the mutant strand altogether (even if it is fully functional or even carries a positive advantage of some sort). This is what I call the "cannibal mum" effect, because in a population with stable Ne effective population, chances for a novel mutant to survive are almost zero beginning around Ne=10 and upwards. That's why I suspect that large star-like haplogroups such as M and H (the two largest "stars" in the human phylogeny by far) show reduced mutation rates downstream of those nodes: they may have got lots of "daughters" but most of them had a hard time succeeding on their own because the very "mother" lineage was way too dominant in the population, drifting the others out to the margins or to extinction. Something like that may also happen in the Y-DNA at less dramatic scale. Only if every generation always or at least very very frequently incorporated new mutations, then the "cannibal mum" (or "cannibal dad") effect would not exist at all: every single new line would be unique, but that's not the case.

      I still think that the basics of the molecular clock method may be used but that it requires much more caution, critical experimental approaches and a good calibration point or several, which can only come from updated paleonthology, archaeology or (with due caution) ancient DNA. Assuming all this it would seem easier to do with full Y-DNA than with mtDNA (whose mutations only accumulate extremely slowly, because it's a short chain) and I tried to do something like that here by recalibrating a full Y-DNA tree made by someone else.

      For R1b (and R1a), I applied those results to Underhill's phylogeny (scroll down to last graph), so I get that R1 should be c. 48 Ka old, R1b c. 34 Ka old and R1b-M412 c. 15 Ka old. R1a should instead be 11-10 Ka old. Of course all figures can't be but rough approximations in any case but I still think that the dates make some sense, regardless of possible Neolithic secondary expansions.

      Even if you go by Underhill's estimates (about 1/3 more recent than mine), there's no way that R1b-M412 and it's key descendant R1b-S116 would have expanded primarily in the Neolithic, Epipaleolithic at best.

      In any case we have Villabruna, which is the oldest R1b man known, and is 14 Ka old and a European (paleo-European, probably of Epigravettian culture). So R1b arriving to Europe with Gravettian culture (or earlier) makes all kind of sense, another issue is when it became dominant and how, because it's quite apparent that in Central-Northern Europe at least, it was nowhere to be seen before the Chalcolithic. But in the rest of Europe? We just don't know almost anything (re. ancient DNA) yet.

      Delete
    2. The Underhill link is missing, sorry. Here: http://forwhattheywereweare.blogspot.com/2014/03/y-dna-r1a-spread-from-iran.html

      Delete
  3. Thank you for your comments. For a long time I have been fairly certain about what I don't believe about this study of ancient genetics but conflicted about what I do believe. Your points about known/unknown snps and the molecular clock are especially impactful. Reading your writings you have linked was very enlightening and I look forward to the upcoming plentiful proofs of your statements, in addition to the existing ones of course.

    ReplyDelete
    Replies
    1. Glad to be of help. I don't write (nor analyze even) these days as much as I used to and certainly the "molecular clock" is a quite complex matter, so do not expect too much new from me in the future on this particular subject (sorry but I know that I'm in a downward tendency in regards to productivity on these matters, I must acknowledge my limits), so I can only encourage you and others to take the intergenerational torch and do your own critical analysis, hopefully better than mine. Just keep high the critical thought, in this as in everything. As they say "just because it's published in Nature it doesn't mean it is wrong"... but it doesn't mean it is true either.

      Something I have sadly noticed, very especially on this issue of the "molecular clock", is that academics tend a lot (way too much) to be hyper-conservative and acritical, scholastic in the worst possible sense of the word. This may be because of laziness or other limitations of the people doing the work but quite clearly also because of the hierarchical nature of universities, where a young post-graduate researcher depends way too much on the approval of more senior (and typically obsolete, conservative) personnel, hindering critical research. As consequence they tend to cite all the time the same "old" papers and ignore those that are critical with the usual "molecular clock" interpretations. Scholasticism (appeal to authority, it is "true" because someone prestigious said so in the past) is not science but in fact quite an enemy of science. Of course, we do "walk on the shoulders of giants" sometimes but when those "giants" are rather weak imps we do have a serious problem, so it is crucial to discern which are true "giants" and which are fake ones, and that can only be done through critical analysis, by challenging everything. Even the truest and strongest of giants can be wrong sometimes, but there are also fake giants whose shoulders won't hold you high in the realm of science, who seem to be wrong much more often than should be judging on how much they are cited.

      Delete
  4. Maju, this might be of interest to you, considering this blog entry. It was posted by Nirjhar at Eurogenes:

    "Indeed As per Genetiker , it was M-269 :

    Below are the Y-SNP calls for ATP3, a sample from El Portalón cave in the Atapuerca Mountains of Spain. Positive calls are in bold, and negative calls are in non-bold.

    The calls show that ATP3 belonged to Y haplogroup R1b1a1a2-M269. .
    https://genetiker.wordpress.com/y-snp-calls-for-atp3/"


    Now, if I recall correctly, ATP3 didn't have any (significant?) steppe ancestry - even ATP9 had less than modern Basques - so maybe this hints at a WHG origin of R1b-M269. I'm just a blog lurker who has an interest in archaeology and history (including pre-history) and have by no means any solid knowledge on the topic so bear with me, but what about the following scenario regarding its expansion within western Europe.

    Assuming M269 did exist in Iberia at least in the middle neolithic, that this man was not someone who had come from elsewhere in Europe, and that the lineage had WHG origins: could it be associated with the increased WHG we see in Middle/Late Neolithic as opposed to Early neolithic? Maybe because of climate changes (do you know if any happened during or around this period?) that could have affected EEF a lot more than forager WHG populations. These could be the early Maritime Beaker peoples we see in SW Iberia, who would later trade aong the Atlantic coast, gaining wealth and possibly expanding their genes along Atlantic Europe until they got in contact with steppe-like Central/Eastern Europeans, after which we get the full Beaker package. Around 2500BCE groups of these CEU BBs with steppe ancestry - and possibly IE language - back-migrate into Iberia making a large impact, and turing Iberians autosomaly more similar to modern ones. This could explain why Basques, despite not being IE, have such high levels of r1b - it was not an IE lineage to begin with.
    On the other hand Davidski modelled French_Basque as:
    Basque_French
    Lengyel_LN 0.590±0.027
    Nganasan 0.009±0.016
    Onge 0.015±0.022
    Steppe_EBA 0.285±0.031
    Western_HG 0.096±0.019
    Yoruba 0.006±0.006
    chisq 3.485 tail_prob 0.900346

    which shows quite a lot of Steppe ancestry.



    Just an idea, any constructive criticism or explanations on mistakes would be greatly appreciated.

    Saludos

    ReplyDelete
    Replies
    1. By the way, and just for the record, I still think the "steppe hypothesis" for Western Europe r1b makes more sense, as most WHGs in here appear to be I2. Besides Villabruna, how many WHG r1b do we see here in the West?

      Delete
    2. I find very difficult to give credence to someone whose featured posts are stuff like "More proof of Whites in ancient Peru and Chile", "Statuettes of the White Gods", "The Lady of Cao was White", "Was the Lord of Sipán White?", "Eske Willerslev is an anti-White propagandist" or "The White Gods", really. And that is even if some of his "findings" could be in agreement with my expectations: he just don't seem credible enough to even mention, really, but rather a race-obsessed mind.

      Anyhow, he gets positive calls for R1b-L21, which is a sublineage of S116, one most common in Brittany, Ireland and nearby areas. BUT he also gets a call for R2 (M479), and while it may seem on first sight that R2 is a far fetched lineage, it is not unheard of in West Eurasia and the researchers claimed that his "neighbor" ATP2 was H2, also a South Asian lineage primarily. Those same researchers could not conclude a haplogroup for ATP3 (maybe because it produces both calls for R1b-L21 and for R2, haven't looked it up to such detail). It looks suggestive but I just don't see enough clarity nor I really trust such a race-obsessed guy as Genetiker, really.

      Delete
    3. "how many WHG r1b do we see here in the West?"

      Neither in the West nor in the East, R1b is nowhere to be seen before Villabruna, nor after until the Neolithic, when it begins to be found here and there (at the Volga, but it is a Volga specific subclade, at the Pyrenees but it is an unspecific upstream lineage, and finally with the Bell Beaker individuals, both German and Irish, which are finally S116). No clear pattern in any case yet.

      → http://forwhattheywereweare.blogspot.com/2016/05/large-paleoeuropean-dna-survey.html

      It is not my argument but, if Davidski can arbitrarily claim that, because R1a1 was found in Epipaleolithic Karelia (EHG), then R1a1 must (somehow) have coalesced in Eastern Europe (never mind we have no data on ancient West Asian Y-DNA, where Underhill rationally places its origin), then someone (not me) can equally argue that R1b1 must have coalesced in Western Europe or more specifically in Italy, just because Villabruna is within WHG autosomal clade. The only real thing is that these two are ante-quem dates for the presence of either lineage in Europe but they do not exclude their presence anywhere else. They may be suggestive but not hard evidence of anything on their own: the ancient samples are just too few (except in some specific areas and periods) to be taken as credible unbiased samples.

      My impression is that all R1 sublineages now numerous in Europe, expanded in the Neolithic or Chalcolithic but from various different origins. They should not be conflated together but taken as distinct lineages each on its own right. If anything U106 and S116 may be related in their expansion but they could also be separated process, all other R1 (R1a-Europe, R1a-South/Central Asia, R1b-Volga, R1b-Central Africa, etc.) are clearly distinct processes. It's a bit odd that these are the main expanding lineages but in fact there are others too in that Holocene period: E1b-M81, E1b-M78, J1, J2, G2, etc., often associated or overlapping with R1 subclades. Even I2 variants like the Sardinian-Pyrenean one probably expanded in that period as well, or also Baltic I1, Uralic N1, Sinitic O3, etc. It's a period of widespread change, really.

      Delete
  5. "Anyhow, he gets positive calls for R1b-L21, which is a sublineage of S116, one most common in Brittany, Ireland and nearby areas. BUT he also gets a call for R2 (M479), and while it may seem on first sight that R2 is a far fetched lineage, it is not unheard of in West Eurasia and the researchers claimed that his 'neighbor' ATP2 was H2, also a South Asian lineage primarily."

    IIRC, the H2 referred to here is not a subclade of the typically South Asian H-M69, but rather the typically European & Southwest Asian (Armenian, Assyrian, Arab, etc.) H-P96.

    The Y-DNA haplogroup H tree has been modified a great deal over the past several years. The linkage between European & Southwest Asian H-P96 and South Asian H-M2826 and even more so the linkage between mostly tribal South Asian H-Z5857 (former South Asian F*) and somewhat more mainstream South Asian H-M69 are very weak; in some way, each of these clades may be considered as a basal branch of F-M89, though they have been declared to share a few SNPs in common (34 SNPs for the linkage between H-P96 and the two South Asian clades; two SNPs for the linkage between South Asian H-Z5857 and South Asian H-M69).

    "It is not my argument but, if Davidski can arbitrarily claim that, because R1a1 was found in Epipaleolithic Karelia (EHG), then R1a1 must (somehow) have coalesced in Eastern Europe (never mind we have no data on ancient West Asian Y-DNA, where Underhill rationally places its origin), then someone (not me) can equally argue that R1b1 must have coalesced in Western Europe or more specifically in Italy, just because Villabruna is within WHG autosomal clade. The only real thing is that these two are ante-quem dates for the presence of either lineage in Europe but they do not exclude their presence anywhere else. They may be suggestive but not hard evidence of anything on their own: the ancient samples are just too few (except in some specific areas and periods) to be taken as credible unbiased samples."

    The same concern applies to the finding of Y-DNA haplogroup J (or was it pre-J1?) in a skeleton buried in Mesolithic Karelia. Davidski is apparently eager to consider the Mesolithic Karelian R1a as evidence for an origin of extant R-M17 in Eastern Europe. Is he also so eager to consider the Mesolithic Karelian J as evidence for an origin of extant Caucasian and Semitic J1 in Eastern Europe?

    ReplyDelete
    Replies
    1. There is one J (undefined) in North Russia (Yuzhnyy Oleni Ostrov) but also J(xJ2) in Iran and J2 in Georgia in the Epipaleolithic/Mesolithic. No idea how Davidski wants to interpret that, really, ask him. My point was to recycle his convoluted argument on R1a1, which is one of the theoretical pillars of the "steppe hypothesis" AFAIK, for the context of R1b1, which, on the very same logic, should work against the "steppe hypothesis". A lot of "steppe horde" affiliates just point to the earliest known whatever as "proof" of their dogma but (ignoring the key issue of "absence of evidence is not evidence of absence") they conveniently ignore those "proofs" that are not useful for their speculation, in what is a clear case of cherry-picking the evidence, of building first the "theory" and then selectively using or not the data at whim, much as other pseudoscientific polemists do, be it with creationism or flat earth.

      Instead we, scientific-minded people must accept the data, even if we don't like it. It may be more complicated to provide an answer, it won't always be an easy answer, but, on the good side, whatever we produce will be a pondered answer based on all, and not just some pre-selected, line of evidence.

      "the H2 referred to here is not a subclade of the typically South Asian H-M69, but rather the typically European & Southwest Asian (Armenian, Assyrian, Arab, etc.) H-P96."

      I think you're correct. I just forgot and never really got myself properly updated on the current H phylogeny. It is quite interesting that there is a Western branch of H anyhow.

      "... each of these clades may be considered as a basal branch of F-M89, though they have been declared to share a few SNPs in common (34 SNPs for the linkage between H-P96 and the two South Asian clades; two SNPs for the linkage between South Asian H-Z5857 and South Asian H-M69)."

      Thank you for the explanation. I gather that therefore nowadays all South Asian H is called H1 (earlier it'd be H1 and H2) and that H2 is reserved for the West Eurasian branch.

      Delete
    2. Also this H issue seems to suggest, at least re. Y-DNA, a rather westerly origin of F, because it's not just H2 but also G and IJ, all quite upstream lineages within macro-haplogroup F. Naturally we should also consider all F* lineages in their proper phylogenetic and geographical context, a complicated task, and, if my memory is correct, most of them are in South and SE Asia. But in any case there are a number of important "close to basal" F sublineages in the West (including at least one F-other) and that is intriguing. On the other side CF (and indirectly D) do seem to suggest a more oriental origin for the early macro-Eurasian (or macro-Asian) population after the OoA.

      Any thoughts on this, Ebizur?

      Delete
  6. Hi Maju: Have you seen this? Yamnaya-related admixture in Bronze Age northern Iberia in http://eurogenes.blogspot.cl/
    I think you have a lot to say about this.
    Your commentaries are important.

    ReplyDelete
    Replies
    1. Well, I'm going to stay aside of the discussion because just a few days ago I left another discussion at Davidski's blog in rather bad terms: David was being patronizing while at the same time rejecting to debate the key issues and replacing all debate with a single word: "bullshit", so I accused him of behaving like Trump when tweeting before breakfast and slammed (unsuscribed with due notice). I also suggested him to run for Queen of Australia, as the post will be vacant soon and it seems a trend to have such autocratic characters elected for high office these days.

      A key question is why did Davidski choose to use ATP9 instead of Chalcolithic ATP samples. It is very apparent in the study and my reanalysis that all ATP samples are close to Basques, just that the Chalcolithic ones are also close to Sardinians while ATP9 is rather close to Britons instead. I even argued that ATP9 might well be a mixed Ibero-British individual (emphasis in "individual?, how representative of the actual BA population of the area she is?)

      I just re-read all that entry and many of the comments for the occasion and found this by myself: Update: a Bronze Age ritual burial of weapons not too far from Atapuerca has very clear signature of British origin (one sword is identical to a locally manufactured British one, the ritual also resembles similar British ones). Although the date may be centuries more recent than ATP9, it does reinforce the notion of more or less persistent interactions with the islands in the Bronze Age also in the Upper Ebro area and reinforces the possibility of ATP9 being partly of British ancestry.

      We know that, since Rathlin (late Chalcolithic, Bell Beaker period), people in the islands seem to have some EHG-like genetics (and also R1b-S116 and be very much like modern Brits and Irish, notably Celtic ones). A key issue here is what is that extra-EHG (seemingly) and extra other HG too... really is? The discussion that ended with me slamming was actually about this in the context of Eastern Europe: data strongly suggested SHG (Motala-like) instead of EHG in both pre-Kurgan Lithuania and Ukraine and this issue of SHG (intermediate in "ANE" between WHG and EHG but otherwise WHG-like) has arisen once and again regarding Northern and Western European genetic make up. I don't know for sure if this is "the answer" but it is probably part of the answer. After all we are talking of pre-IE and pre-Kurgan populations of the Western fringes and it seems more likely to me that, say, "refugees" from the Kurgan invasions, maybe from the Rhine basin or Scandinavia were incorporated into the Western genetic pools one way or another than Kurgan people "conquering" without leaving any trace. Or maybe it is a trait that was "all the time" present in Britain or France or wherever and just locally expanded in that late Chalcolithic period within the BB phenomenon.

      Anyhow, the issue is wide open and in wait on more data points from ancient Western Europe and I have no idea (unless I want to think bad of him) why Davidski picks that almost certainly admixed individual, when all ATP samples are similarly close to modern Basques, both ATP9 with her British-like peculiarities and all the older ATP samples with their Sardinian-like ones.

      Feel free to quote me but I'm not personally getting involved in that discussion: too much waste of energies, sorry.

      Delete
  7. " After all we are talking of pre-IE and pre-Kurgan populations of the Western fringes and it seems more likely to me that, say, "refugees" from the Kurgan invasions, maybe from the Rhine basin or Scandinavia were incorporated into the Western genetic pools one way or another than Kurgan people "conquering" without leaving any trace. Or maybe it is a trait that was "all the time" present in Britain or France or wherever and just locally expanded in that late Chalcolithic period within the BB phenomenon. "

    Talking about refugees, what about Doggerland, the piece of land between Great Britain the Netherlands and Germany?
    Studies have dated it around 5800 BC and corresponds to the bridge between North Germany, Great Britain and the Atlantic fringe were proto- proto basques lived.


    http://earth-chronicles.com/histori/doggerland-britains-stone-age-atlantis.html

    ReplyDelete
    Replies
    1. Doggerland is no doubt related to the Maglemosean culture, which is found both in Scandinavia/Low Germany and Eastern/North Britain in the Epipaleolithic. In the old Paleolithic continuity model, it would be ideal to explain R1b-U106 spread but today it seems much less likely because we are realizing that in much of Europe there were important demographic/genetic changes at later periods: first the Neolithic waves, then the not sufficiently clarified HG remix, maybe with waves originating in the West, and finally the Indoeuropean layer (plus maybe some other localized layers in Sicily, Tuscany, etc. from the Eastern Mediterranean).

      Basically Doggerland sinking is too old to matter much. It had disappeared altogether (or almost) by the time the first farmers arrived to continental Western Europe. BTW, in the key period it was much larger than reflected in that article: it was a huge stretch of land joining Denmark, Low Germany, the Low Countries and Britain. This map seems very detailed: http://cdn.zmescience.com/wp-content/uploads/2015/01/31836.jpg

      What you see in the map of your linked article is just Dogger Bank, which is only part of the wider Doggerland and which was an island by the time of Neolithic arrival.

      Delete
  8. I would be very circumspect about taking the Isle of Wight cereal traces as evidence of Mesolithic contact with France. Not only is this a novel methodology, which has been questioned, but the date of the deposit (c. 6000 BC) is too early for either Brittany or Normandy to be the source of the cereal.

    ReplyDelete
    Replies
    1. I'm unaware to that claim being questioned: can you point to a source?

      As for the date it may be wrong, of course. Or it may imply even longer routes. IDK about Central Europe but that date is for the Western Mediterranean roughly the arrival of Cardium Pottery to Liguria or Provence. It would imply a much longer trade route (either by land or sea) and it would also imply that Doggerland was still partly above water, that Britain was a peninsula.

      Interesting in any case because I never noticed before that issue of the dates, which is indeed problematic.

      Delete
    2. Just checked Central Europe's dates and c. 6000 BCE the hypothetical route would have needed to reach to Eastern Hungary or Serbia (Starcevo-Köros culture), much further than Provence. Also Cardium Pottery peoples were clearly very good sailors, so if we are going to give credence to the finding, I'd have to argue for a Cardium Pottery origin, either overland or by boat (although overland seems more reasonable, considering that Southern Iberia was not reached, in any archaeologically documented way, before c. 5500 BCE).

      Delete
  9. Maju sorry for bothering you again, but there's a qpAdm run on Iberian samples at Eurogenes, and some results seem odd.
    https://docs.google.com/spreadsheets/d/1gzxMCTBTJo0NZ5jsTXnM2N6mkbBUUTXHLRgOw2CqQb4/edit

    What is your take on it?

    ReplyDelete
    Replies
    1. No prob. That's what the comment section is for.

      I don't see anything "odd" other than Steppe-EBA (Bronze Age) weighting more than WHG. A problem I see is that those are not comparable stuff: EHG should have been used instead of Steppe-EBA and preferably along SHG, which is suspect of co-responsibility in "strange results" and was important in the past not just in Scandinavia but also in much of Eastern Europe (and maybe also other areas).

      If you compare avocados with apples and oranges, the result will be wrong necessarily, regardless on whether it claims that avocados are more apple-like or more orange-like.

      Delete
    2. Also, how do we know that the labels correspond to what they claim to be. Davidski has often said that such direct analysis based on ancient specimens can't be done, because of imperfect quality of the aDNA? I would like to see the "materials and methods" section because I strongly suspect they are using some sort of "zombified" constructs and not direct comparison with ancient samples.

      Delete
  10. Besides the clearly inflated SSA, I feel there's an issue with the Steppe/WHG ratio, it seems some populations pick more Steppe for whatever reason even if they are relatively close, however the Steppe+WHG amount is very similar amongst all Iberians. Maybe Lengyel isn't the best farmer population either.
    An example, Galicians score 24,1% Steppe and 8,8% WHG. I'd expect the Portuguese to be fairly similar, but instead they get a whopping 30,6% Steppe and a meagre 3,8% WHG - the only Iberian population who scores more Steppe are the Spanish-speakers in Vascongadas, with an extra 0,5%.

    However, as David said, the results are kind of in line with the PCAs

    https://drive.google.com/file/d/0B9o3EYTdM8lQVE5qOVZYRVA2cXc/view
    https://drive.google.com/file/d/0B9o3EYTdM8lQN3hqVy01OFBmVVU/view

    ReplyDelete
    Replies
    1. Not necessarily, because there is no direct reference for the "Basal Eurasian" thingy, which will pull towards Africa (and also exaggeratedly towards CHG maybe, who are heavy on that BA ghostly component).

      "Maybe Lengyel isn't the best farmer population either."

      You're right: they are a late derivative and not the putative source. They may be lowering WHG and increasing other elements relative to LBK/EEF or similar. I'd need to check but that's my intuitive guess.

      "I'd expect the Portuguese to be fairly similar, but instead they get a whopping 30,6% Steppe and a meagre 3,8% WHG"...

      Portuguese could to be more EEF than Galicians (and the change may have happened in the Late Neolithic per a recent study) but that would basically close the bulk of genetic change in the region. Anyhow those fractions of "steppe" and "WHG" are outlandish by comparison to academic studies.

      Have to go without checking the PCAs, anyhow every day less interested on what David has to say, really.

      Delete
    2. Northwest Portuguese seem actually to group very closely to Galicians, genetically speaking.

      Delete
    3. That depends on what marker or what study (what specific sample) you're considering but of course they are neighbors, so they should be great similitudes. However in Adams 2008, North Portugese are unusually low in E1b-M81 relative not just to Galicians but to other Portygese and the whole West Iberian region.

      Anyhow, when I said: "Portuguese could to be more EEF than Galicians", well, that's a typo (a word was deleted and never rewritten) and I seem to have meant "Portuguese could happen to be...", so it seems to me that I was talking in hypothetical case and was not a statement in any case.

      Delete
  11. Hey Maju,

    2 out of 3 males from West German Blätterhöhle have R1: http://biorxiv.org/content/early/2017/03/06/114488

    That's pretty close to the Paris Basin.

    ReplyDelete
    Replies
    1. Very interesting, thank you.

      In the previous entry, on ancient mtDNA, I did mention that this Blätterhöhle site was quite intriguing. I wrote:

      It also caught my eye that a German site (Blätterhöhle, Westfalia, famed because farmer and hunter-gatherers living side by side were located there some years ago), clusters intensely with Iberian hunter-gatherers and related populations [i.e. Neolithic ones which appear more HG-like than usual]. I have to research more on this matter (which I had ignored so far) but I suspect it may be very relevant, because we could get an even longer chain of early "modern" mtDNA pools, adding this site to Paternabidea (Navarre) and Gurgy (Burgundy), spanning a long stretch of Western Europe, an area quite neglected by archaeogenetics so far, it must be said.

      This is the sort of site that could be seeding the later Michelsberg expansion southwards, which wiped out LBK farmers in most of Germany (but not the East) and many nearby areas of the Low Countries, France (where it's known as SOM) and Switzerland. I've been saying for a long while that the answer to the R1b-Western must be in the West, not for any particular obsession but because of internal logic of modern DNA and because every other region studied produced negative results; so there was a big blank area in need of research to the West of Europe, more than suspect of being at the origin of this lineage, and I was 99% sure that once research began in those areas, it would not take long to produce the lineage. It's like that "Sherlock Holmes" catchphrase: once you have eliminated the impossible, what remains, even if unlikely, must be the answer.

      Very interesting indeed.

      Delete
    2. OK, seems I was reading too much and also that you have misreported the findings. ED Table 1 lists only one Blätterhöhle individual carrying R1b1 and another R1 (no downstream affiliation, although I doubt it's "true R1*" but rather that they can't further define the haplogroup), another carried I2a1 and the fourth one is female. The third R1b individual is actually an R1b1a2 from Els Trocs (Aragon, Spain), which I believe was already known to be R1b1 but had not the lineage so finely characterize. This is R1b-V88, which is the African and Mediterranean lineage, so not relevant for the R1b-Western quest (just as Volga R1b isn't either). This Els Trocs population was extreme in their early farmer affiliation via mtDNA, which suggested zero Paleoeuropean admixture (very rare). Indirectly it may tell us about the Sardinian-Pyrenean link but we will still have to wait until R1b-Western is found before the two Late Chalcolithic individuals that are at present the "ante quem" line.

      Delete
    3. Yes, I think I didn't misreport, what I said is quite accurate ;) Read my statement again. That might be just another case of Western European DNA studies being quite unsatisfactory.

      I still think the frequencies alone in this particular location are quite noteworthy. I doubt that these are the direct ancestors of European R1b.

      Delete
    4. Oh, you're right, sorry. I misinterpreted what you said. Mea culpa! :(

      I noticed that the Els Trocs sample had good quality (1.3), while the Blätterhöhle samples have worse quality: 0.8 for the R1b1 man and I think that just 0.1 for the "R1" one. So it's even possible that they are what we are looking for but that it remains unclear because of DNA quality.

      "I doubt that these are the direct ancestors of European R1b."

      It's most unlikely that we find "direct ancestors" anywhere but we should find at some point populations with large frequencies of R1b-S116 and R1b-U106, call them "distant cousins of the direct ancestors". For S116 so far that is some Bell Beaker individuals but BB is clearly an unsatisfactory answer, even if BB would be the vector (which seems not for many areas) there should still be an origin, a source for that peculiar and previously unobserved genetic makeup.

      IMO it should be within or near the triangle formed by Paternabidea, Gurgy and Blätterhole, but we'll see in due time.

      Delete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... OFF (keep it that way, please)