After many rumors and pointless discussions based on them, finally the Haak et al. study on ancient Kurgan DNA is available only (pre-pub format):
W. Haak et al., Massive migration from the steppe is a source for Indo-European languages in Europe. bioRxiv 2015 (freely available pre-pub) → LINK [doi: http://dx.doi.org/10.1101/013433]
[Update: LINK to the June 2015 final publication, Nature]
The study expands on previous work by Lazaridis 2014 by including a much larger array of ancient DNA from Germany and Hungary, as well as some key ancient DNA from Russia and also some complementary samples from Northern Iberia.
The study expands on previous work by Lazaridis 2014 by including a much larger array of ancient DNA from Germany and Hungary, as well as some key ancient DNA from Russia and also some complementary samples from Northern Iberia.
Autosomal DNA
Even if the authors admit it is difficult to properly quantify, there are clear tendencies that are outlined in fig. 2:
 |
| Figure 2: Population transformations in Europe. (a) PCA analysis, (b) ADMIXTURE analysis. The full ADMIXTURE analysis including present-day humans is shown in Extended Data Fig. 1. Annotations in red by me. |
The main inferred processes of demographic formation of the modern European genetic pool are outlined:
- A baseline of Epipaleolithic hunter-gatherers that pulls in the PC 1 towards the left (geographically would be the Atlantic). I call this layer Paleo-European (PE).
- A first replacement by Neolithic farmers of Thessalian origin (already discussed in depth in Lazaridis 2014), who were similar to modern Sardinians. I call this layer Neo-European (NE).
- A reflux of Paleo-European genetics that alters the Neo-European layer somewhat. This can be associated to Atlantic Neolithic flows (i.e. Megalithism, Funnelbeaker, etc.) We can call this stage Neo-European-2 (NE2).
- A second replacement wave by Steppe tribals, certainly bringing the Indoeuropean languages (IE).
Modern European populations align well along a IE-NE2 axis, whose midpoint seems to fall on North France, depending of course on which references you choose. I drew that axis as dotted red line. It is not substantively different from the Dimension 2 axis, although it is slightly slanted because the IE invaders obviously carried more PE than the NE layer. This new PE is not WHG (Magdalenian) but EHG (Eastern Epi-Gravettian) however - and hence its tendency towards Paleo-Siberian genetics (Ma1 or "ANE").
So Dimension 1 quite apparently contrasts the Paleo-European vs the West Asian components. What does Dimension 2 express? A quite apparent element is the Paleo-Siberian tendency. Alternatively it can also be considered to express the distinction between Lowland and Highland West Asians. Finally it can also be expressed as IE vs NE. All three are surely just variants of the same continental vs peripheral opposition, which is weaker than the PE vs West Asia one.
I must mention that fig. S5.2 offers a slightly different view:
 |
| Figure S5.2: PCA analysis with ancient individuals projected onto the variation of the present-day ones. |
Notable is that the NE-IE axis (not drawn) appears more slanted, with most modern populations showing greater excess of PE tendency and less "obviously" resolved by the late Chalcolithic populations (LNE/EBA in the authors terminology).
As I said above, it seems very difficult to objectively measure the exact fractions of admixture (the tendencies are clear but the quantification not so much) and something that is becoming more and more painfully obvious is that Atlantic European ancient genomes are needed to explain the changes that happened prior to the arrival of Kurgans. Particularly it'd be most interesting to get ancient samples from: Portugal (Neolithic, Megalithic and Bronze Age), Basque Country and Gascony (Neolithic and Megalithic at the very least, preferably from the coastal regions), Brittany and West France (Neolithic, Megalithic 1 and Artenacian), Belgium (non-LBK Neolithic), Britain (several regions preferably, as the British Neolithic seems to have strong regional differences), West Germany (Michelsberg culture). This array or at least a sensible part of it could shed light on key processes taking place before and after the Kurgan migration. Bell Beaker samples from outside Central Europe would also be very interesting.
I would also be interesting to see a PCA without West Asians, whose presence quite apparently does not add much to the analysis. It is known that when the PCA is European-only (or mostly), Basques and Sardinians display clearly different polarities (typically Sardinians vs Russians in PC1 and Basques vs Caucasians in PC2). It would be very interesting to observe how these ancient samples behave in a Europe-only PCA.
Y-DNA
A lot of the upheaval was around the fact of the finding of some R1b in Samara Valley. This is very interesting indeed but it is not the kind R1b that can be considered ancestral to modern European mainline R1b-M412. It is mostly of a different haplogroup whose modern distribution is unknown to me: R1b-Z2103.
*Update: some people have commented that R1b-Z2103 is found in West Asia and some Volga peoples.
*Update: some people have commented that R1b-Z2103 is found in West Asia and some Volga peoples.
Schematically (following YSOGG), R1b-M343 and its sole relevant subclade R1b1-M415 are structured as follows:
- R1b1a (L320)
- R1b1a2a1 (L51/M412)
- R1b1a2a1a (P311)
- R1b1a2a1a1 (U106) → NW Europe
- R1b1a2a1a2 (P312/S116) → SW Europe with scatter elsewhere in the continent, including Ireland, Britain, Italy... Found in Kromsdorf (late Chalcolithic)
- R1b1a2a2 (CTS1078/Z2103) → found in Samara culture
- R1b1b (M335) → minor, West Asia
- R1b1c (V88) → Mediterranean and Africa, particularly important in Sardinia and Central-East Africa.
Otherwise all the spotted R1b in this study is R1b1*: in Samara culture and in Neolithic Aragon (NE Iberia), both of which are hard to relate to anything of modern relevance.
Corded Ware is associated to R1a only at this time. So at least in Europe it makes good sense to associate Kurgan expansion with R1a expansion.
Mitochondrial DNA
There is plenty of mtDNA data but most is recycled from previous studies so not really novel. An interesting detail is that there is no or nearly no mtDNA H within the Kurgan (IE) samples, strongly suggesting that their migration was largely male-biased, at least initially. As happens with Y-DNA R1b, Kurgan immigrants cannot be associated to any increase of mtDNA H, whose origins must therefore be sought in some other origin (namely: Atlantic Neolithic).
Note: my apologies for being so extremely passive in my blogging activity. I don't really know how to explain other than feeling OLD AND TIRED and needing LOTS of "me time". It's time for others to pick up the torch, I guess.
@Maju,
ReplyDelete"After many rumors and pointless discussions based on them, finally the Haak et al. study on ancient Kurgan DNA is available only (pre-pub format):"
I agree it was mostly a waste of time. It did prepare us for this news and most of what we said wasn't far off.
@Maju,
"A lot of the upheaval was around the fact of the finding of some R1b in Samara Valley. This is very interesting indeed but it is not the kind R1b that can be considered ancestral to modern European mainline R1b-M412. It is mostly of a different haplogroup whose modern distribution is unknown to me: R1b-Z2103."
Think about this....
Every pre-historic Y DNA sample from Russia and east of there starting all the way back with MA-1 is R except for a few. MA-1, Mesolithic Samara, Mesolithic Karelia, Samara Yamna, Tarim mummies, Andronovo, later Siberians. Then 6/6 Bell Beaker and CWC are R1 and 50% or more of their ancestry was Yamna-like.
It's more than a coincidence.
About the Yamna R1b.
As you know R1b Z2103 is a brotherclade to European R1b-L51(Which the Bell Beaker sample belonged to), which makes it very connected to Europe. Today Z2103 gets above 30% in some Volga-pops, and is not exclusive to west Asia and the Balkans.
We have to remember R1b didn't just exist in Samara Yamna, every sample had it. The existence of R*, R1a1*, and R1b1* in North Eurasia earlier leaves room for that Z2103 to have evolved in North Eurasia.
Think about it if most R1a-M417 today is Steppe-derived, and R1b1* existed in EHG along with R1a1*, why can't most modern R1b1a-P297 be steppe derived?
R is a very big haplogroup and, together with IJ, the main Y-DNA haplogroup of West Eurasia (or with H of South Asia). The surprising thing is not when R is found, the surprising thing is when it is not found (which in some cases may be a matter of sample size and survey bias anyhow).
DeleteSome people imagine R as something recent, I see it as something as old as the Upper Paleolithic, as the main colonization of West Eurasia (incl. Central Asia and parts of Africa) by H. sapiens. That's R and IJ (and less importantly Q and G). So the only thing strange about R, R1 and R1b is that it's not everywhere... apparently.
Still the Russian data shows that not everything in pre-Neolithic Europe was I2 (as Northern Europe's data might have suggested). That's something and it underlines the need to expand the survey area. We see the same with mtDNA H: nothing in Central-North Europe before Neolithic but something (or even a lot) in East Europe and SW Europe.
So Central-North Europe must be put aside in favor of wider surveys. Otherwise it's like interpreting New York City from the viewpoint of Central Park only. No, NYC is not all trees and lakes.
"As you know R1b Z2103 is a brotherclade to European R1b-L51(Which the Bell Beaker sample belonged to), which makes it very connected to Europe. Today Z2103 gets above 30% in some Volga-pops, and is not exclusive to west Asia and the Balkans".
Precisely: it is NOT L51 nor its two subclades that make up essentially all the West European R1b. It's something else. That it is R-Z2013 or I2a or G or E1b is not different: it is not the lineage we are looking for.
"We have to remember R1b didn't just exist in Samara Yamna, every sample had it."
Yes. And, for what you say, they only left a minor legacy among some Volga peoples. That is interesting and emphasizes that Yamna was not so important in patrilineal legacy as such. If anything, side groups, like the probable one with European type R1a centered around Poland-Ukraine (Corded Ware people probably), were the ones having an impact rather than Yamna senso stricto. And that's exactly what I was saying on Underhill's data and also previous R1a data: not all R1a is probably associated to Kurgan expansion, but one subclade is very apparently in Central-Eastern Europe, via Corded Ware.
... "leaves room for that Z2103 to have evolved in North Eurasia".
North Eurasia is not a valid category: it includes a varied range of regions and paleohistorical populations. That was my contention re. "ANE", that, even if somewhat informative, it was not really representative of anything that could actually matter. Now that we have the Corded Ware data, we can skip not just ANE/Ma1 but even Yamna, because CW is a much better proxy.
....
...
Delete"Think about it if most R1a-M417 today is Steppe-derived"...
No. I very strongly disagree: R1a has two main subclades:
→ Z282 which seems strongly associated to Corded Ware and probably has a core towards Kiev/Smolensk
→ Z93 which is clearly derived from the Iran area
The steppe is a destination in these two expansions, not the origin. Upstream R1a in general also seems to stem from Iran/Kurdistan.
Kurgan expansions, particularly Corded Ware, seem to have been vehicular in some secondary re-expansions but attributing all the R1a expansion (let alone all R) to Kurgan expansion is clearly wrong.
"why can't most modern R1b1a-P297 be steppe derived?"
Because the geostructure of European R1b1a2a1 (L51/M412) and R1b1a2a1a (P311), as well as that of its key subclades (to which essentially every other Western European belongs to): R1b1a2a1a1 (U106) and R1b1a2a1a2 (P312/S116), does not fit at all with any such sort of geography.
It spread from either West Asia or the Balcans to Central/West Europe, still without making much noise and, once in West Europe, it experienced two expansions: S116 from probably Southern France (or Iberia?) and U106 from surely NW Germany or the Netherlands. The pattern does not fit with any sort of expansion from Eastern Europe, where in fact the lineage is nearly non-existent (and where it's found, it's clearly derived, not ancestral).
→ http://leherensuge.blogspot.com/2010/08/r1b1b2a1-is-almost-unique-of-west.html
There's no point in discussing other R1b lineages when trying to understand the origin of Western European R1b-M412. Only when we look for the ultimate origins of R1b it does makes sense but then the arrow points to West Asia and never to Eastern Europe.
Nice response. I see same patterns.
DeleteGood stuff.
ReplyDeleteDo you think the place to look is first Gumelnita-Karanova>Varna>Boleraz/Baden>Michelsburg?
ReplyDeleteThose are a large array of cultures. Why do you put them with ">" signs as if they were derived from each other rather than contemporary?
DeleteAnyhow, they'd be all very interesting to know more but in general terms I expect them to be similiar to other "Neoeuropeans". I'm particularly interested in Michelsberg, since they totally displaced LBK/Rössen before Corded Ware arrived and have Western (Megalithic) links but I'm even more interested in learning about Carnac, Seine-Oise-Marne, Amesbury, Artenac, Zambujal, etc. Basically I'm asking to study France, Britain and Iberia much more seriously than they have done till now.
Okay. I'm not saying one is to the other, but we see similar patterns show up in each. There are parallels within those later groups. Look below for some interesting stats from the paper!!!!
DeleteMaju,
ReplyDeleteI have to say, I always expected a spectrum of miscellaneous, low-frequency haplogroups to appear in Neolithic Western Europe to include a low frequency R1b and other cooky oddballs.
Finding it in the Ebro Valley of all places though, is something I haven't fully been able to unpack just yet. This one clade isn't ancestral to the West, but it certainly wasn't the only one in the Ebro valley, statistically speaking. What are your thoughts on this?
I know previously you had supported a residual population from the LGM then maybe backed off of late. Where do you think this individual's lineage came from? The Northern Cardial trek?
Would a Cardial farmer have an East Danubian lineage?
One sample is inconclusive. It does show that R1b existed in Iberia in the Neolithic but it's not yet the kind of R1b we are looking for. It's not so much R1b what matters but S116 and U106. Of these we still have too few samples, probably because West Europe is undersampled and under-researched.
DeleteSo wait for research to come. When we have a good number of Portuguese and other Western European samples we will be able to say something. So far we know too little.
"Where do you think this individual's lineage came from? The Northern Cardial trek?"
In Aragon? Either a front wave farmer or a pre-Neolithic remnant (it's the mountains!)
"Would a Cardial farmer have an East Danubian lineage?"
In mtDNA we see only very minor differences between Danubian and Mediterranean early farmers. In autosomal DNA they almost totally overlap. So yes, absolutely yes.
Why do you ask that?
@Maju
ReplyDeleteGreat that you are giving some time on this big research.
I will start with these questions, if you please answer:).
1.What is the Difference between the Spanish Basal R1b1 to the Samara basal R1b1?.
2.What is Your Assumption on the Language that the Russian Karelian H-G spoke?.
3.Can R1a and R1b be Branded now as Pure EHG? as David and others considering?.
1. They are both "similar" as far as I can discern, i.e. part of R1b1*. That doesn't mean they are directly related, just that both belong to the same catch-all category (paragroup), which is NOT ancestral but parallel to the main branches that dominate the landscape today.
Delete2. Uralic.
3. No. They are beating the same old dead horse all the time: they put their emotions before the facts, the cart before the horses...
Thank you!! here are the next-
Delete4. The SC Asian type ancestry detected in Post-Neolithic Yamnaya-CWC samples what do we interpret from them?
5.Is Anatolian totally a dead hypothesis now?.
4. Do you mean the "dilution" of the EHG ancestry? I guess that there was some sort of infiltration from West or Central Asia. You have suggested yourself a possible origin in Jarmo→Yangelsk. It'd be interesting to compare Samara with Dniepr-Don in this aspect.
Delete5. For me it always was: ancient languages of West Asia were too diverse and mostly non-IE for Renfrew's hypothesis to be real. It was always a "Santa Claus" hypothesis which aimed quite openly to save Indoeuropeans from their "original sin" of murderous conquest and rape.
IE languages, culture, and violent aristocratic identity must originate in the Western steppe, just like Semitic one originates in the semi-desert and Altaic ones do in the Eastern steppe. They are three variants of the same process of pastoralists turned conquerors.
''For me it always was: ancient languages of West Asia were too diverse and mostly non-IE for Renfrew's hypothesis to be real. It was always a "Santa Claus" hypothesis which aimed quite openly to save Indoeuropeans from their "original sin" of murderous conquest and rape. ''
DeleteI can't even think to reply on that .....
''IE languages, culture, and violent aristocratic identity must originate in the Western steppe, just like Semitic one originates in the semi-desert and Altaic ones do in the Eastern steppe. They are three variants of the same process of pastoralists turned conquerors.''
The Only POINT the Steppe Theory has is the Horse an evidence which also can be discarded-
https://priyadarshi101.wordpress.com/
How come the "only point"? In fact the horse evidence is only of secondary relevance (it can explain fast migration and even military superiority but little more, more so when horses were in those times already domesticated in other areas like Iberia). The big evidence is that of archaeological continuity between the cultures that can be defined as Kurgan or Kurgan-derived: http://forwhattheywereweare.blogspot.com/p/blog-page.html
DeleteIf you don't understand this, you are fighting against windmills that you, and only you, imagine to be giants. By not addressing the archaeological continuity of Kurgan cultures, you are intently ignoring the core of the issue. Horses, chariots, etc. are pretty much secondary.
When the finger points to the moon, the fool looks at the finger. Similarly, when the horse points to the kurgan, the fool looks at the horse? Guess it is convenient not to look at the substance of the matter if all you want is to pretend to debunk it: it's best to look at minor issues that are almost totally irrelevant.
Its completely the Reverse of What you just Said The Asian Archaeology is far Uniform and Unidirectional and also Linguistically FYI they are more IE than you imagine, Asia is indeed and all ways was the best PIE candidate.....
Deleteand of course ''The Kurgan'' uniformity is just a myth now with aDNA confirmation.
No, "Asia" was never the best PIE candidate. Actually the use of the term "Asia" is almost as meaningless as writing "Earth". "Asia" should not be used in opposition to "Europe" but as the superset to which Europe belongs, just as Japan or India do.
DeleteThe aDNA support of "a myth" does not make you wonder about the firmness of your dogma?
''The aDNA support of "a myth" does not make you wonder about the firmness of your dogma?''
DeleteJust Face it this time Maju by getting out of your bubble world of Kurgan fairy land that the current aDNA in simple words speaks of Late-Neolithic/Bronze Age Intrusion to Eastern Europe by populations of Caucasus-N Iran and SC Asian Ancestry most likely bringing IE languages there while mingling with Uralic type populations...
End Of Story.
That Spanish R1b is probably from the Near East, with Cardial. I am thinking that EHG is in EEF. R1b in the Near East, coupled with fstats showing Loschbour as 48%EHG, 52% outgroup, could point to that. If the rate of EHG in the non-Basal part of EEF is lower than Loschbour's rate, then he won't fill the 56% non-basal part, and EHG won't show up in admixture. Loschbour fits in at like 42%, so EEF could be 20%EHG. There is a possibility that some R1b reached Western Europe from Samara, but it won't have the same mutations. There is no migration from Western Europe to the Middle East and Central Asia. He is not ancestral, from Spain, but maybe West Asia.
ReplyDeleteYou have a lot of confusion in your mind. Cardial does NOT originate in the Near East, unless you include Greece in it. Impressed-Cardial, like Painted-Linear originates in Thessaly. In fact Lazaridis already demonstrated a substantial UGH (~WHG) in EEF, which is part of the original alchemy.
DeleteI do not discard at this point that R1b-Western arrived via Cardial but it has to be demonstrated. The data so far is inconclusive. In any case it needs of a secondary expansion process or rather two centered in the Atlantic and North Sea respectively. This can only be explained with the Atlantic Neolithic cum Megalithism (and secondarily Funnelbeaker, Bell Beaker).
Presumably Cardial farmers had ancestors in the Near East/West Asia at some point though.
DeleteIt seems like there are 3 possibilities:
"Old Iberian" R1b arrived from WHG in Iberia.
It arrived from the Near East with early farmers.
It arrived with early farmers, but they acquired it somewhere between Iberia and the Near East.
All 3 seem plausible.
In agreement (I would not use the term "Iberia" but "Western Europe" though). With the current aDNA information we can't decide for either one easily.
DeleteBut the implications of both scenarios are different:
1. If R1b-L51 arrived from West Asia or the Balcans with Neolithic farmers (Cardial for example), then the double expansion of S-116 and U-106 must have happened only very late, c. 4000 BCE, with the Megalithic demographic boom.
2. If R1b-L51 was in West Europe since the late Upper Paleolithic, then S-116 and U-106 could well have coalesced before the Neolithic and they only scattered further but without large obvious founder effects (although some subclades like the "Irish", "Alpine" or "English" ones could be those Neolithic founder effects).
I can't decide myself with the current aDNA data but, based on full chromosome age estimates, I tend to favor the second scenario. The range is ambiguous enough to allow for both scenarios however.
What aDNA evidence do we have and what we do not have? We have two WHG samples (only 2!), none of which is R1b. We have a much larger array of mainline Neolithic samples (I count 41) of which only one is R1b but not L51. None of the Neolithic samples are from the Atlantic facade and only one of them (from Iberia) is associated with H1 (another one from Hungary is associated with H*).
So I'm rather tempted to discard the Neolithic origin of R1b-L51, as well as of the main associated mtDNA lineages, which are H1 and H3 (and possibly other less well defined H). In the case of mtDNA H (and H1), we know it was common in Atlantic Iberia (Portugal, Cantabria, Basque Country) in the Paleolithic already. So my hunch is that we need more aDNA from Paleolithic Western Europe in general and from Neolithic/Megalithic Atlantic Europe. That should solve the riddle.
From yfull tree.. Wait until they sample the western half of Yamnaya.. Pre Yamnaya, sredny stog to Cotafeni. You will see!
ReplyDeleteL150/PF6274/L150.1/PF6274.1/L150.2/PF6274.2 * L23/S141/PF6534 * L49.1/L49.2/PF6276/S349: formed 12200 ybp, TMRCA 6400 ybp
L51/M412/S167/PF6536 * PF6414 * CTS10373/PF6537: formed 6400 ybp, TMRCA 6000 ybp
L151/PF6542 * P310/S129/PF6546 * YSC0000191/PF6543/S1159: formed 6000 ybp, TMRCA 6000 ybp
P312: formed 6000 ybp, TMRCA 5800 ybp
TRMCA is nonsense. You don't even provide sources. You are totally ignoring full chromosome dating, such as the one available with Underhill and others (Mal'ta paper), which necessarily produces much older dates.
DeleteIt's from the Yfull tree... how much more of a source do you need? Your wild hypotheses are going to continue to crumble.. I told you Baalberg wasn't Kurgan influenced. You never listen, or change your ideas. You are already well behind, and falling further back with each paper. Adapt to changes with data or become irrelevant.
DeleteI don't know what is Yfull tree but anyhow each age estimate is an OPINION, an EDUCATED GUESS, and therefore the most logical documentation is to cite author(s).
DeleteIt'd be interesting also if you "TRMCA fans" became used to provide not one but a whole comprehensive array of different age estimates, which do exist, so we could all watch how they become meaningless or at the very least very blurry in their variety. It would be interesting if you would put due emphasis on the rarer but much more credible full chromosome age estimates and also age estimates calibrated to ancient specimens.
By doing just that I come to very different conclusions than you do, Chad.
Y-Full has not yet told us what there dating algorithm is - it's supposed to be released soon - but I expect it *is* based on whole chromosome sequencing, since that's what they do. I don't know if you have read the Hallast et al 2014 paper yet, but that was whole chromosome sequencing too, and is the basis for a lot of the date estimates floating around right now. The central estimate dates given in the actual paper are too young, because they are based on Xue's pedigree rate, and are not compatible with aDNA samples. However, you can recalibrate them with aDNA to your liking. We need way more aDNA though - fully sequenced, because SNP targeting like in the current paper only gives minimum ages - everything has huge error bars at present.
DeleteAn Atlantic Megalithic expansion of L151 would actually fall within the confidence intervals of these dates - Y-Full has 6000 (7900-4900) BP as their estimate. I would make L151 a good deal younger based on the Hallast et al data - only 4700 BP, perhaps Bell Beaker - but that could be way off. aDNA from Atlantic Europe will settle this once we get it.
Interestingly Y-Full puts M269 at 12 200 BP, much older than I have seen elsewhere, though L23 is only 6400 BP. I get 6600 BP for L23 but have no data for older R1b clades.
Underhill puts R1a1a1-M417 at 5800 (4800-6800) BP. I put it at 5500 BP. Y-Full puts it at 5500 (6200-4800) BP. So these estimates are all very close.
Underhill mentions that there is no consensus on the human Ychr mutation rate and then mentions FOUR different scalable references: "Recent estimates include one SNP per: ~100 years,⁵⁸ 122 years,⁴ 151 years⁵ (deep sequencing reanalysis rate), and 162 years.⁵⁹" But then he uses one of the shortest ones (122 years/SNP), even though the longest one is the only one calibrated to aDNA (Anzick) and hence the most realistic.
DeleteSo Underhill's notational estimates must be multiplied x1.5, so if he gets 5500, it'd be 8000, etc.
Furthermore, I personally have a problem even with the Anzick-calibrated Méndez' rate, because this would still make an age(CF), when it must be c. 100 Ka BP judging on archaeology. So I stretch the R1a age significantly more, to c. 10-11 Ka BP, what makes the R1 node 48 Ka old, while with Méndez' date is just 33 Ka old.
I think that's the key test for any mutation rate: what age(CF) produces? If close to 100 Ka BP (no one gets even close!), then it's plausible, else it's trash.
What's going on here? Clap-o-meter! If you go "radical" (even if with a very strict scientific criteria) people who are mostly unable or unwilling to do all the work to test the model, will frown you, because it just doesn't fit with what they have READ before. If you are "moderate" or "conservative" (even if you have no criteria whatsoever), then they smile and pat your back. And it's way too easy to go with the crowd, and this is called scholasticism or scholastic inertia, and it is a most serious vice in science.
So it's almost certain that Pan-Homo split c. 10-13 Ma ago but scholastic inertia keeps pushing for 5-7 Ma ago. It's certain that the Out-of-Africa migration happened c. 125 Ka ago (to Arabia/Palestine) and 100 Ka ago to farther Asia (what is a great calibration point for the age of CF) but scholastic inertia will keep pushing for 60-70 Ka ago because that's what they have been parroting since the 1980s.
So for me it's not important how many authors say this or that, what matters is what is the age of CF and of Pan-Homo split according to that model. Or in other words: if there is any substance whatsoever other than the rancid smell of old ink behind those claims.
Erratum: "this would still make an age(CF), when it must be c. 100 Ka BP" should read "this would still make an age(CF) quite more recent than the correct one (per archaeology): c. 100 Ka BP"
DeleteIf you want to use poorly-understood events 100 thousand years ago speculatively associated with Y-chromosome haplogroups as your fixed and certain calibration point, no one's going to stop you. But I don't think a whole lot of people are going to agree with you either.
DeleteCould you explain your calibration based on Anzick-1? SNP counting from him I get an average age for the Q/R split of 38 kyo (an average of 97 SNPs back, with Anzick-1 missing an average of 32 SNPs compared to modern samples, radiocarbon dated to 12 600 years old). That agrees with the estimates I posted above.
MA-1 gives older dates, using the same method relying on internal branch lengths only: I get 407 years/SNP on that tree, average 109 SNPs to Q/R split, for an age of 44 kyo. (This looks close to the estimated ages from Scozzari et al.)
For Ust'-Ishim Man the authors have put the date of K(xLT) at 50 kyo, with a 95% confidence interval of 47-55 kyo, using a much more sophisticated approach than mine. For comparison, using the MA-1 calibration K is 60 kyo, and using Anzick-1 should be close to 49 kyo (though you can't directly get the age of K from that tree).
So using aDNA (albeit only three samples!) the *highest* date gives a 15-20% increase of the dates I posted above, while Anzick-1 agrees with the dates posted (which were based on the Ust'-Ishim Man calibration).
Yes, that is a wide range, and we have only a few calibration points, none modern humans older than 45 ky.
Pan-Homo split is not IMO a very good calibration point, because not only do we not know for sure when it was, but it involves millions of years of hominid existence very different from modern humans, when we could have experienced quite different mutation rates. I would much rather use modern human aDNA.
I should have added that the Ust'-Ishim man study also gives a mutation rate calibrated by ancient DNA, and it is higher than that of Mendez et al.
DeleteProper calibration is of the essence. When radiocarbon was proposed 60 years ago to measure archaeological time, complementarily to the more complicated stratigraphy, scientists demanded evidence and this was provided by quite accurately measuring organic remains of known age such as Egyptian mummies. Nothing of that is apparently today demanded from "molecular clock" hunches, what underlines the debacle of scientific thought in the West, replaced largely by useless (and even destructive) scholasticism.
DeleteNotice that genetics is not even remotely close to "rocket science". In fact it scores worse than social sciences or ecology in the aspect of scientific rigor, see: http://leherensuge.blogspot.com/2010/04/genetics-and-biolog-rather-soft.html . Genetics is at about the levels of economics and we all know that economists are (generally speaking) rather like priests than like true scientists and that their "predictions" and rationalizations can't be trusted almost at all.
So I don't care if "a whole lot of people are going to agree" with me. What I do care is that a whole lot of people is going to swallow fallacies as "facts", when they are not at all such thing and I would like that they realize the dangerous proximity to pseudoscience that such fallacious logic implies. All I beg those "whole lot of people" is to be less credulous and more rigorous. They don't have to agree with me, but please do not agree with other claims that are likely not any better. Keep the systematic doubt and critical thought at the forefront and, if you ever find a most likely answer, let it be because of well done critical analysis and not just quasi-religious credulity.
"Could you explain your calibration based on Anzick-1?"
It's not my calibration it's from Méndez 2013. Cited by Underhill 2014. Quoted and discussed here: http://forwhattheywereweare.blogspot.com/2014/03/y-dna-r1a-spread-from-iran.html . The links and references will bring you to the source.
"MA-1 gives older dates, using the same method relying on internal branch lengths only"...
Looks odd to me. I was recently looking at the original Ma1 drift parameter tree and Ma1 looks like a good calibration point, which, to my eyes (comparing with other similar full chromosome trees, like the one that appears in Underhill 2014) is more recent than R1b-M269 but older than R1b-L51.
"For Ust'-Ishim Man the authors have put the date of K(xLT) at 50 kyo, with a 95% confidence interval of 47-55 kyo"...
To my eyes that is a bit too recent. K(xLT), now K2, probably has an age of 70-74 Ka BP, coincident with the Toba catastrophe. It should be similar to the age of mtDNA R and maybe slightly more recent than that of N. Why? Because both lineages show associated expansion, we can almost say that mtDNA R and Y-DNA K2 were the same very dynamic people, that spread on the ashes (literally) of the first Asian expansion, attributable especially to mtDNA M and Y-DNA F.
"Pan-Homo split is not IMO a very good calibration point"...
Possibly not, I agree, but it has been used a lot and almost always wrongly (with way too recent dates). This has helped to establish the paradigm of OoA being of c. 60 Ka BP, when it is CERTAINLY of 100 Ka BP (Aterian in India!, H. sapiens in China!, good climatic logic: end of the Abassia Pluvial, causing a pump effect).
One aspect of this which touches on something discussed before
ReplyDeletehttp://bellbeakerblogger.blogspot.co.uk/2014/09/john-koch-presentation-celtic-from-west.html
is the seeming probability now there were (at least) two sources of farmers one of which being the Levant.
If the Levant one is E1b connected and maritime orientated including a fairly rapid spread to Iberia via Sicily, Malta, Sardinia etc this might speak to your thoughts about Vasconic, Atlantic Megalith and Afro-Asiatic.
If - my opinion - BB were IE speaking copper artisans who independently of La Tene later came to Iberia along the maritime trade networks and settled in the Atlantic Megalith coastal settlements and adopted chunks of the culture and the language that might speak to the p/q Celtic thing and the elements in insular celtic which apparently have similarities to afro-asiatic.
As soon as I read Koch I facepalmed. Koch is nonsense: if you read his fancy interpretation of Tartessian as "Celtic" and compare with Untermann's and other serious Iberian/Tartessian transcriptions you will realize why.
DeleteThe "strong arguments" outlined by Bellbeakerblogger are not really arguments but much less they are strong. Both 1 and 2 can perfectly be explained, for example, if Neoeuropeans (NE2 particularly) spoke Vasconic languages. Let's not forget that Celtic was the Western avant-guard, almost "cannon fodder", of Indoeuropean expansion and very probably it coalesced among populations with strong Vasconic substrate of West Germany.
Anyhow: BB is trivial. BB had no significant demographic impact anywhere except in Ireland. The demographic game-changer was its Dolmenic precursor. See: http://forwhattheywereweare.blogspot.com/2014/02/neolithic-and-chalcolithic-demographics.html
If I understood it right the point you were making in the bellbeakerblogger thread was something about finding an afro-asiatic element in Vasconic.
DeleteSo if the Haak paper is suggesting two sets of farmers including a maritime one from the Levant (possibly speaking an afro-asiatic language) that might fit your idea.
My idea that the p and q celtic thing might be related to early (maritime) and late (overland) arrivals is a separate thing.
Afroasiatic is not helpful in order to explain the insular grammatical trait: it's not something AA but something West Asian. I've never read that Berber had that trait, however Turkish and other non-AA West Asian languages do.
DeleteMaju nice wrap up
ReplyDeleteBut, why don't you heed your own advice ?
U say (rightly) "CWC were the ones having an impact rather than Yamna senso stricto. And that's exactly what I was saying on Underhill's data and also previous R1a data: not all R1a is probably associated to Kurgan expansion, but one subclade is very apparently in Central-Eastern Europe, via Corded Ware. "
But then, that's not the Kurgan expansion (!) which stipulated that the kurgans came from yamnaya . To me CWC has little direct relation to yamnaya (R1a vs R1b; admittedly needing more samples ). Moreover ; as you've highlighted ; the distribution of R1a forms an arc in ECE and Central Asia but was missing in the westenr steppe. Either it was never there or R1b was intrusive
I don't interpret that Kurgans come from Yamna but from Khvalynsk culture (and ultimately Samara), as per Gimutas (with minor updates at best): http://forwhattheywereweare.blogspot.com/p/blog-page.html
DeleteYamna would be in this interpretation the root of the Indo-Iranian branch (ethno-linguistically, via Andronovo) but not of the rest. Of course Yamna is directly linked to Khvalynsk and one can well argue that the Indo-Iranian branch is the "purest" of all IE branches but that's about it.
A more critical issue arises from the present genetic data however: Baalberge aDNA from East Germany is not yet drifted towards Yamna/EHG but fully fits the NE2 cluster. How to interpret this? Option A: those who argued that Baalberge was not Kurgan were right; option B Baalberge (at least in Germany, we still don't have data for the Polish branch, which is crucial) only had IE elites and the bulk of the population were pre-IE natives.
If option A is correct, then the Yamna-like elements triggering the dramatic changes associated to Corded Ware were decisive but they were not in any case from Yamna but from Catacombs culture. If option B is correct, the Polish branch of Baalberge (which apparently produced CW via Luboń and Globular Amphorae) should have more CW-like genetics than the German branch. A mixed scenario is also possible.
"the distribution of R1a forms an arc in ECE and Central Asia but was missing in the westenr steppe"...
That's an interesting observation indeed and does seems to match Underhill's data (http://forwhattheywereweare.blogspot.com/2014/03/y-dna-r1a-spread-from-iran.html). For me it matches the model in which R1a-Z282 and R1a-Z93 must be considered as two separate processes: the first one may well be associated to Corded Ware expansion (and judging from leaked Afanasevo aDNA, maybe with other Kurgan expansions) but the latter does not seem to have any relation with Kurgans (whose expansion in Asia, excepting Afanasevo, is of Bronze and Iron Age, quite late) but is probably older and corresponds to a Neolithic scatter instead.
"Either it was never there or R1b was intrusive"...
I don't think we must jump to conclusions. Ancient peoples were not doing genetic paternity tests and could well have a colorful array of lineages, some of which had important founder effects while others didn't. R1b-Z2103 could perfectly have coexisted with R1a-Z282 and R1a-Z93 and even other lineages.
They could also have co-opted other tribes from their area, which then proceeded to make the founder effect(s) instead of the comfortably installed early Kurgan elites, who did not need to emigrate to have a chance at power and wealth. An example could be the secondary Turkic expansions under the Mongols. The Mongols themselves only had a very limited founder effect in Nogai Republic, otherwise it was associated peoples, largely Turkic but in some cases also Iranic, who were doing the imperial job, even if initially the leaders were Mongol.
I caution against overly simplistic explanations, especially when the available data does not justify them.
Maju ; I at least agree with ur genetic observations ; but not your archaic archaeological renditions
DeleteIt's not "archaic": the Balcanic Kurgans, Maykop and Afanasevo at the very least are as old as Early Yamna, and so is Sredny-Stog II. It's a matter of chronology. Khvalynsk is the common root: Yamna is just one branch, the one that stayed put.
DeleteMaju; you have a very primitive understanding of archaeology . Sorry
DeleteWell, you may want to illustrate us on your "advanced" understanding of archaeology. Meanwhile I can only take one-liner comments like the two above one as personal provokation rather than the thought-provoking comments I'd like to read. Don't worry, I have a thick skin but seriously: illustrate us, please, on your "improved" archaeo-vision and stop repeating the same one-liner. Thanks in advance.
DeleteWell for a start; the thread of continuity you claim to construct ; linking yamnaya with khvalynsk is false; apart from being reductionist (as it hangs on one element- the supposed burial rite).
DeleteYamnaya kurhans owe little to the stone cairns at khvalynsk. Rather their origins lie in inspirations from Majkop and are status symbols of new migrants from East Central eruope (Tripolye) and the North Caucasus region (Majkop).
How can Yamna descend from Maikop if both begin at the same date (approx.) and within the territory of Khvalynsk, Maikop in the south and Yamna in the north?
DeleteMajkop is slightly earlier , and far richer
DeleteBut don't just fixate on local forager predecessors - which were unimpressive and otherwise diverse groups. The idea of a kurgan culture is a constructed heuristic concept invented by Gimbutas.
Look at it globally - what really caused the opening of the steppe was the advent of a productive economy from west Asia which allowed development of pastoralism, on the one hand ; and the fall of the balkan copper age centres coincident with the rise of Caucasian ones; in turn catalysed by west and Central Asian trade if not outright colonisation . THe Majkop chiefs were the big dogs ; yamnaya followed and became intermediaries connecting europe with the Caucasus
So your hypothesis is that Maykop originates in some culture South of the Caucasus, right? Well, I have to see any evidence of that: the kurgan burial and other inherited cultural practices suggest otherwise.
Delete"Kurgan culture" could have become a "heuristic concept" for convenience, as happens with all consolidated theories, but I don't think it is in origin nor when looking critically at it with today's eyes: the Kurgan pattern stands for all I know, and this genetic data seems to confirm it, at least for Europe.
"Look at it globally - what really caused the opening of the steppe was the advent of a productive economy from west Asia which allowed development of pastoralism, on the one hand"...
But that had been in the steppe for millennia already. It could not have arrived with Maikop: even horse pastoralism was already established long before Maykop.
"... and the fall of the balkan copper age centres coincident with the rise of Caucasian ones"...
That's precisely what Kurgan invasions caused: maybe you could advocate of a technology transfer Balcan → Caucasus via prisoners of war or whatever. But surely it is much more complex: I think that the Caucasus (i.e. Maykop) centrality re. Metallurgy is quite a bit hyped. They had beautiful designs in the late period but that does not make them central to the development of metallurgy, if anything it gives them a chapter in History of Art.
... "catalysed by west and Central Asian trade if not outright colonisation".
Colonization from where exactly? Mind you?
"THe Majkop chiefs were the big dogs ; yamnaya followed and became intermediaries connecting europe with the Caucasus"...
They did not follow: they are synchronous. A different of a mere century in some date does not justify your "origins" hypothesis: you need evidence of directionality, which is lacking. Probably Yamna evolved in the North of the Khvalynsk culture coincident with the Maykop "secession" or maybe formation as monarchic polity. IMO Maykop was a kingdom, while Yamna was rather a tribal confederation of some sort, maybe adopting the "horde" form at some points but not necessarily including all tribes in the cultural area.
I think that vs. the "heuristic concept" of "Kurgan culture", which is justified by cultural inheritance, you are prisoner of another "heuristic concept": that of West Asian centrality and civilization's "superiority" over pastoralist tribes. However, as we can see for example in Mesopotamia (and also other cases: China, Rome, etc.), it is clear that pastoralist tribes:
1. are nearly impervious to conquest by (pre-industrial) civilizations, for whom the semi-desert or the steppe (or even the mountains often) are "barrens" where they can't extend their control within parameters of reasonable cost/gain (costs are huge, profit near zero).
2. can conquest the civilizations and eventually turn them into the culture of the invaders, with whatever mixture: the example of the Semitic conquest of Mesopotamia is very clear.
"Barren" areas' tribals, usually pastoralists, tended to be armed and dangerous and nearby civilizations acted as a magnet for them: to trade and learn and maybe even sell their services as mercenaries when the civilizations were strong, but to plunder and conquer when they became weak. We see that often in history and we see it also in the prehistoric Metal Ages, just that we don't have the accounts that confirm them.
In the Kurgan case they plundered at the very least the early (and very rich) civilizations of SE Europe (IMO in collaboration with "Pelasgian" Vinca at first, later also against them). They also expanded in other directions anyhow (Altai, Highland West Asia and Central Europe at one point or another), where the plunder was less impressive probably.
Ill try to reply to each point in turn, but Im afraid its difficult to make one see a point, when they don’t possess the theoretical underpinnings for it, especially given the fact that the arguments which you recite are those of yesteryear, despite attempting to make you consider this very aspect before even launching into the details of the matter.
DeleteAnyway, for the final time:
** “So your hypothesis is that Maykop originates in some culture South of the Caucasus, right? Well, I have to see any evidence of that: the kurgan burial and other inherited cultural practices suggest otherwise. “
Well, for a start, Kurgans do not have a single origin. But they do , in fact, appear south of the Caucasus very early. They also appear possibly equally early in certain parts of eastern Europe – west of the steppe.
Whether the Yamnaya culture is equally as early as Majkop or not, is not the main issue. First of all, it still remains that the Majkop culture is earlier. This is certainly the current concensus. Your argument that they’re equally dated rests on the recent carbon dating paper. Nevertheless, the Majkop culture is far richer and more elaborate that the Yamnaya culuture. Clearly, this is where, if anyhere, the major chieftains dwelt. This is not my opinon. This is fact. Accept it.
What catalysed the Majkop culture remains a point of discussion, mostly because the Majkop culture, on which the Yamnaya culture models itself on (more or less), appears as a rather original, new cultural complex which literally just bursts onto the scene. Clearly then, this can not be a simple process of gradual evolution of local foragers-cum-pastoralists on the north Pontic steppe. Something catalyzed it., undoubtedly linked to its role as a major metallurgical centre which supplied vast amounts of Eurasia, from Mesopotamia, to Europe, the Urals, central Asia, and beyond. These local barbarian chiefs are those of the Majkop culture.
The rise of Majkop clearly coincides with the * relative* collapse of the Balkan Cupper Age centres- which were probably due to internal conflicts. The Yamnaya culture arose as intermediaries of contacts developing in the new environment, which connected the Caucasus with the Balkans and Carpathian Basin.
** “that of West Asian centrality and civilization's "superiority" over pastoralist tribes”
I never claimed that. I wasn’t implying any simple diffusionist or centre-periphery perspective. In fact, as I said, it is not known what exactly casued the rise of Majkop. But clearly it was due to its pre-eminence as metallurgical centre. Its rise undoubtedly lies with contacts with Anatolia and northwestern Iran. This does not mean it was simply a “colonization” process by more advances southerners. You’ve misunderstood.
*“ Patoralists “. are nearly impervious to conquest by (pre-industrial) civilizations, for whom the semi-desert or the steppe (or even the mountains often) are "barrens" where they can't extend their control within parameters of reasonable cost/gain”
False. Fact is Pastoralist tribes were defeated and incorportated repeatedly throughout history. ?
** Pastoralists “can conquest the civilizations and eventually turn them into the culture of the invaders, with whatever mixture: the example of the Semitic conquest of Mesopotamia is very clear.”
Sorry, I miss your point. The Semitic tribes have always lived in Mesopotamia, since recorded history.
** “In the Kurgan case they plundered at the very least the early (and very rich) civilizations of SE Europe (IMO in collaboration with "Pelasgian" Vinca at first, later also against them).”
Ah, the Kurgan people first cooperated with the Vinca people. Then the Trolls attacked. But luckily the Elves saved the day
No, the elves didn't save the day, the orcs won. And now we are all bastard orcs and speak orcish. Get real!
Delete"Kurgans do not have a single origin. But they do , in fact, appear south of the Caucasus very early. They also appear possibly equally early in certain parts of eastern Europe – west of the steppe".
DeleteKurgans first appear in Khvalynsk culture, a thousand years before Maykop and the other places you have in mind. Samara culture may be even older. They may be "primitive" but even great Rome once was just a tattered village on a hill. We can follow the development of the kurgan burial in these two cultures from the 6th mil. BCE.
"The rise of Majkop clearly coincides with the * relative* collapse of the Balkan Cupper Age centres- which were probably due to internal conflicts".
Maykop raises some 300 after the collapse of Karanovo-Gumelnita culture (maybe the first centralized monarchy of Europe and the earliest metallurgic center of West Eurasia and possibly the whole World). There is a clear "destruction horizon" coincident in time with the Balcan-Tisza Kurgans. The successor cultures (Cernavoda, Ezero and Cotofeni) are all related to the Russian esteppes. Cernavoda for example is older than Maykop, so claiming an origin of the Kurgan macro-culture at Maykop is clearly wrong. Sredny Stog II is also older than Maykop. So, while the issue surely deserves a careful analysis, it does seem that there was an Indoeuropean or related (steppe) flow with destruction before Maykop and Yamna can be described as such. Hence it's only logical to consider Maykop and Yamna as eastern heirs of the wider and earlier Kurgan/IE conqueror phenomenon and not at the origin.
"... it is not known what exactly casued the rise of Majkop"
Maybe the Khvalynsk first large kurgan at Nalchik can give us a clue?
"But clearly it was due to its pre-eminence as metallurgical centre".
I don't know. What I know is that Maykop looks like evolving into a kingdom and that it had contacts with West Asia since early on, so it's probable that both concepts or at the very least the metallurgic tech they eventually excelled on was imported from West Asia rather than the Balcans (I don't know for sure but sounds plausible particularly considering the relationship it had with Kura-Araxes, which filled a vacuum after the collapse of Ubaid culture and is very roughly coincident with the first Semitic invasion of Mesopotamia).
"Fact is Pastoralist tribes were defeated and incorportated repeatedly throughout history."
DeleteFact is that marginal tribes, often pastoralist, were nearly impossible to conquer (lack of clear centers, which was the Celtic reason of defeat vs Germanics and later also Romans) and actually managed to invade the civilization centers once and again:
(1) The Semites did that at the very least in Mesopotamia (historically documented) and probably also in the Levant. Later we see the Jewish tribes doing the same in Canaan and even later the Arabs conquered half the known world out of a barren land in a matter of decades.
(2) The Turkic peoples first and the Mongols and Manchus later conquered vast lands including China, Iran, the Baghdad Caliphate, the states of India, Kievan Rus and the Eastern Roman Empire.
(3) The Celts historically conquered several civilized lands: Galatia in the remote Asia Minor but also Rome itself, which barely managed to survive as independent entity by paying huge tributes. Their expansion into less civilized lands was much larger, of course.
(4) The Germanics conquered all the Western Roman Empire, in some cases (Britain particularly) totally altering the ethno-linguistic reality, while in others they (as the Akkadians in Sumer or the Mongols and Manchus in China) accepted linguistic (and later ethnic) assimilation by the conquered civilization on practical grounds.
So barbarians invading civilizations is nothing unknown to history at all. Another thing is that literacy may in some cases delay or even put a serious barrier to ethno-linguistic assimilation by the new lords. But there was no (significant?) literacy in the Copper Age.
"Sorry, I miss your point. The Semitic tribes have always lived in Mesopotamia, since recorded history".
Heh! There's pretty much a consensus that Semites are intrusive to Mesopotamia, arriving c. 3900 BCE. Semitic names begin appearing in those dates and new centers arise North of Sumer. The history of "the flood" is often interpreted as meaning a human flood, that of Semites: "amaru" = flood, "a-maru"=Semites (aka Amurru). The genesis of Semitic peoples is normally understood to be relatively marginal (Circum-Arabian Pastoralist Complex) until these invasions of c. 4000 BCE, which would also include the Levant (whose earlier languages we do not know). Naturally some branches also migrated to the South all the way to Ethiopia.
The origin of this CAPC is generally attributed to Harifian, which in turn was probably influenced by Mesolithic migrations from NE Africa (hence Semitic is Afroasiatic), when northerly PPNB conquered all Levant, a distinct population survived in what was then a less-arid semi-desert, eventually becoming a force on their own right.
But of course you can believe in panspermia and let Semites be everywhere since the beginning of times. I do not.
maju
DeleteI can only suggest you pick up books and article written in the last 50 years. There's no point arguing with someone who is of a low aptitude and biased agenda
Out of good faith, ill try help bring you out of 1938, Maju
DeleteRead En"olithic burial mounds in the black sea steppe" by none other Rassamakin. You'll clearly see that the Pit-Grave (yamnaya ) burial mounds form under the influence of the richer elites of Majkop, and not the preceding disparate disconnected steppe forager groups. IN fact, apart from the fact that the earlier stone cairns and mounds are typologically different to later Yamnaya ones, there is a period of 300 - 400 years prior to Yamnaya were mounds etc are no longer used in the steppe.
Now multiple this (very central) misunderstanding with hundreds of others (equally important), you'll see your house of cards collapse with a whimper.
Im not sure why your so blindlingly dullened to this. Im fairly sure its not a lack of aptitude, Must be linked to your ideologies on Basque antiquarianism in western Europe ? ?
"" It has been noted that the earliest “symbolic period” in the development of steppe monumental architecture
Deleteis limited to a steppe zone between the Northern Caucasus and the Balkan-Carpathians (Rassamakin 2002,
p. 60-63). The emergence of these constructions is inluenced by external impulses. The first one is connected to the development of the Balkan-Carpathian metallurgical province (Chernykh et al. 1991) and the shaping of an exchange system of prestige goods. The second one is connected to the new Pre-Maikop cultural system (settlements of Svobodnoe, Meshoko, etc.) of the Northern Caucasus.""
You're welcome
"Out of good faith, ill try help bring you out of 1938, Maju"...
DeleteBeing patronizing is not "good faith" but arrogance. I'd suggest you stick to the facts and opinions. In any case it would not be 1938, but at least 1979 (Gimbutas is not that old).
"Read En"olithic burial mounds in the black sea steppe" by none other Rassamakin."
I'm doing it (link). He says:
"I think that M. Gimbutas was right when she stated that the first burial constructions, such as small stone cairns, can be identified as the first “kurgans”, as a beginning of the so called “kurgan cultures” (Gimbutas 1994)".
However he then goes to claim that they first evolved in the Balcanic-Carpathian area, as well as Maykop, but doesn't explain why he is discarding all the other cultures that Gimbutas considers original, notably Khvalynsk and at least some elements in the Sredny-Stog II stage (which is not of a single cultural affiliation but shows scatter of various affinities: Kurgan, DD and CT). Both would be older than Maykop and Yamna.
Rassamakin then argues for the origin of the kurgan to be in a so-called post-Stog group (Kvitiana culture - I know nothing nor could find anything about it) and the Lower Mykhailivka culture, which is relatively late in chronology (within the 3500-2800 BCE period) to be ancestral to Maykop and Yamna.
Honestly, I don't get a clear picture of what Ramassakin argues for or how do you get the Maykop origins hypothesis from that study, whose focus seems to be Ukraine (and not the wider European steppe area). Maybe you can point me to the relevant passage that illuminates your though - although the paper is not as long that I should have missed anything relevant, I'm far from being perfect.
For what I see in that paper and in the quote you mention from 2002 (almost the same phrasing), it all relies on the stated opinion of Ramassakin about the Carpathian-Balcanic area and Maykop but without providing the evidence that would make that claim meaningful. My impression is that you are being blinded by an unsupported quote that simply fits your "pet theory". But where is the data? The Carpathian-Balcanic cultures did not build kurgans nor anything of the kind, in fact they fell victim to Kurgan tradition invaders, so it's totally far-fetched to imagine them as precursors of the Kurgan macro-culture. Maykop is within it but not at all the oldest in chronology.
All I get from Ramassakin is that Gimbutas was right and hence (per Gimbutas and with whatever question marks needed because of poor old school field work, etc.) Khvalynsk (and possibly Samara before it) stand at the origin of the Kurgan phenomenon. All the rest is blurry and does not point to Maykop in any particular way.
The "post-Stog" group that seems important for Ramassakin (within his Ukrainian focus) is listed as being from ~3900-3400 BCE, again underlining that there are Kurgan elements before Maykop, in this case in Ukraine. But AFAIK some of those elements are also found in Sredny-Stog II.
DeleteAnyways, looking for more online data, I found this recent post by Davidski, where he mentions a study (Morgunova & Khorlovka) that clearly states that: "the Pit-Grave culture [Yamna] developed synchronously with the early stage of the Maikop culture in the northern Caucasus", dating its earliest stage to 4000 BCE (i.e. older than Maykop unless Maykop dates have also been revised).
In general terms we can say that in the 4000-3500 period there was a Kurgan area in all the European steppe (already penetrating into Ukraine) and parts of the Caucasus. Older stages may be not so clear but what is clear is that Khvalynsk and (partly) Sredny-Stog II are key to those origins. The geographic scope of Khvalynsk is perfect to explain the origins of both Maykop and Yamna (and let's not forget the remote Eastern branch of Afanasevo, also contemporary and the first Balcanic Kurgans, which are clearly involved in the plundering and destruction of the Eastern Balcanic civilization of Karanovo-Gumelnita particularly), and Khvalynsk is also typologically related.
DeleteOf course many questions remain and further research would be desirable but, on our current knowledge, that's about it.
Maju: It is known that when the PCA is European-only (or mostly), Basques and Sardinians display clearly different polarities (typically Sardinians vs Russians in PC1 and Basques vs Caucasians in PC2). It would be very interesting to observe how these ancient samples behave in a Europe-only PCA.
ReplyDeleteMaju with this, bear in mind though, that even on these European only PCA, the spatial patterning of populations is still substantially similar, and this is more clear if the PCA is rotated.
http://i.imgur.com/7ThU2yB.jpg
If you look at the attached PC where a European only PC like you describe is rotated, the rotations still map fairly well onto their West Eurasia PCA. So it's not a totally different thing.
The big difference in the rotation is, yes, the Basques are more separated out and less like other Europeans. Whether this is drift, Basques showing more isolation by distance than other Europeans, or something else is hard to say.
On PCA, with fewer populations, you better capture the distances that make those few populations different (because you are not compromising the PCA to explain variation between other populations), but you can be less certain that the phenomenon isn't driven by drift specific to one or two populations. So they offer different advantages.
But what happens with NE2 populations? Where do CW and Yamna and LNE/EBA samples fall on?
DeleteThe characteristic of this Europe-only PCA is that Basques and Sardinians decouple quite radically, as Basqueness (vs Eastern Mediterraneanness) takes over the PC2 axis. We know from Skoglund that Atlantic Megalithic farmers (from Sweden) cluster with Basques, Iberians and French, not Sardinians nor Italians. This may be hidden in the All West Eurasia PCA, where Basques and Sardinians almost overlap but should be more apparent in the Europe-only PCA. Sure EN1/EEF do cluster with Sardinians but what about EN2/MNE and others?
For me it is an important question that hope to see answered soon. In fact I commented on the study itself requesting that analysis.
... "the rotations still map fairly well onto their West Eurasia PCA."
It does not: the near-overlap Basques-Sardinians vanishes and a better defined gap appears. In fact the PC2 of the WEA PCA becomes PC1 (Russia vs Sardinia) and a new PC2 shows up (Basques vs Eastern Mediterraneans). In the WEA PCA Europeans form almost a single line along the CW-NE2 (~Russia-Sardinia) axis but in the Europe-only PCA they form a cross with two well defined axes. I'm quite interested in the second axis, which seems to correspond to WHG or Atlantic Neolithic influence.
"Whether this is drift, Basques showing more isolation by distance than other Europeans, or something else is hard to say".
Precisely is what I would like to find out because it's not about Basques only: all SW Europeans tend in that direction and that seems to correspond to Megalithic Atlantic farmers (per Skoglund).
Maju, "fairly well" is a subjective term; I think the similarities between a) the rotated Europe only PCA with b) the European portion of the West Eurasian PCA, and the distinctions between them are both notable (from your snippy and somewhat insecure sounding tone I assume you do not).
DeleteHere is another example of the same from rotation from- http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1002453 - "Inference of Population Structure using Dense Haplotype Data", including the Adygei
http://i.imgur.com/xhizzx4.png - the lesser degree of West Asian highlands distinction from the Europeans compared to the pan-West Eurasia PCA is more evident here.
One of the things i've wondered about for a while is the idea there were two early waves of farmers - a relatively fast maritime one to Iberia that led to Atlantic Megalith and a second wave that became Cardial with this second wave displacing the first wave along the coast from Greece to Italy but not as far as Iberia - or at least not as far as the Atlantic facing parts.
Delete@Grey: what we see in the aDNA is that there were no two waves of farmers but almost surely just one. This wave however became more and more Western (WHG) as time passed and this shift seems linked to Megalithic expansion and the peculiarities of Atlantic Neolithic. It's likely that the Oceanic climate was not precisely the best for early farmer expansion, and also that the area served as "refuge" for retreating aboriginals or their mixed offspring. All I'm saying is: please, pay attention to Atlantic Neolithic, because it clearly adds a complexity twist to the otherwise too simplistic interpretations.
Delete@Matt: They are different PCAs. PCAs (like other algorithmic analysis) are not absolute but highly dependent on samples. If you added North Africans, you'd get a different PCA, if you add East Asians another, etc. All I say is: West Asians are irrelevant for European analysis, fear not and remove them, if only to see how ancient samples behave in a Europe-only context.
Southwest Asian populations such as the Saudis are the ones responsible for the change in West Eurasian PCA that makes Sardinians one end of Dimension 2 and places Basques to the "northeast" of them, while without SW-Asians Sardinians and Basques define dimensions 1 and 2. Volga-Ural populations or West Asian populations don't cause this change, as seen below Sardinians and Basques are at the ends of separate dimensions when those are included:
Deletehttp://s24.postimg.org/zd5ja4qet/Engl_Europe_2_1.png
BedouinB are extreme enough to take over both dimensions 1 and 2, but this doesn't change the West Eurasian PCA's shape when it comes to other populations compared to one where Saudis and Sardinians define dimensions 1 and 2. When Sardinians take dimension 2 from Basques difference is somewhat more visible, but as Matt showed not that much when rotated. Perhaps this means the "Atlantic" and "Near Eastern" dimensions are closely related?
I think Lazaridis et al's European PCA gives us enough information to figure where Corded Ware etc would land on an European plot. More interesting would be to see a SpaceMix plot, looking at the results from the preprint it seems to differentiate variation, especially global variation, much better than a PCA.
http://oi61.tinypic.com/2vlofw7.jpg
Different samples create different PCA plots. If we don't understand that, we understand nothing about statistical analysis of autosomal genetics (or whatever other massive dataset considered). Typically PCAs only evaluate the two main axes of variation (so-called principal components, sometimes dimensions or eigenvectors, very similar if not identical) and in this they'd be similar to a K=3 or K=4 in ADMIXTURE or similar analysis, which is usually not too informative.
DeleteI wouldn't blame Lowland West Asians alone for the changes between the West Eurasian and European-only PCAs but it's clear that West Asian excessive weight takes from Europe-specific insight, so they are better out of the sample if what we mean is to analyze Europe (or check both PCAs, there's no reason not to do it: that's how we gain a perspective of things: by looking at the puzzle from several angles and not always from the same one).
Your first graph clearly illustrates the issue but lacking the ancient samples, we don't know how they could behave. Or rather we do know, more or less how they behave in the South-North axis (because that is dimension 2 in the WEA PCA, becoming dimension 1 in the European-only one) but we don't understand how they behave in the West-East axis, where Basques are outstanding in "westernness" (or "non-easternness") and what does that axis mean ultimately in terms of ancient ancestry.
We know how some ancient samples behave in the Europe-only PCA because Lazaridis did publish such analysis (see: http://forwhattheywereweare.blogspot.com/2014/04/revised-lazaridis-study-on-ancient.html - 4th graph). WHG/SHG tend to be "extreme Northwest", while EEF/ENE tend to cluster with Sardinians or Italians ("South tendency"). But otherwise we have only seen pan-WEA graphs, even in Skoglund and Dasakali (see: http://forwhattheywereweare.blogspot.com/2014/04/diversity-and-legacy-of-ancient.html).
Something that has me spooked anyhow is how Gökhem clusterst systematically with Basques and Iberians (even in one case between Basques and Orcadians) in these other two studies but in Haak's it clusters rather with Sardinians. This is very odd and to my eyes, it suggests, that Haak's analysis is not well done (it's 1 vs 2 studies).
The question with PCA dimensions isn't whether they are identical over two PCA's made using different samples (of course they aren't) but rather how much information they represent overlaps, or how much is hidden by their differences however you want to put it. If the relative positions of Basques, Poles and Sicilians don't show much change between two PCA's, one of which has dimension 2 defined by Sardinians and other by BedouinB, what should we figure from that?
DeleteIf we compare the PCA I linked and the Lazaridis et al. Europe-only PCA, the West Asian-Basque dimension and Sicilian/Maltese-Basque dimension don't differ that much when it comes to Europeans' placement relative to each other along it. West Asians do compress Europeans so such a PCA may be easier to read without them, but what information available on an Europe-only PCA can we assume actually is lost if West Asians are present?
Swedish farmer clusters northwest of Sardinians in Haak et al PCA, as it does in this Eurogenes PCA done using the dataset Lazaridis used.
https://drive.google.com/file/d/0B9o3EYTdM8lQUGRmcURIV3BiOFU/view?pli=1
I'd say this is more or less accurate. It clusters with Basques or Spanish on a West Eurasian PCA only when projected or on a merged PCA with less ancient samples. So, to me Haak's PCA seems to be the best of the bunch.
I think that there is a radical difference because Sardinians and Basques stop near-overlapping and instead Basques become polar, not vs. Eastern Europeans but vs. Eastern Mediterraneans and I do suspect that a lot of the answers to the riddle of European genesis are related to that second PC axis, only visible in the European-only PCA, and that it is directly linked to Megalthic Atlantic farmers of the likes of Gökhem.
DeleteIn any case, why not? It's just a few hours of PC time if you have the data and it does potentially show more information. Or complementarily introduce North Africans (who are also West Eurasian in essence) in sufficient numbers to see how the ancient samples behave with them in: maybe we get some info from that analysis or maybe not but worth trying.
DeleteWhat I really want to know anyhow is why Gökhem samples behave so differently in Haak than in Skoglund and Dasakali. That is very strange, really.
Any thoughts on the mtDNA? I noticed the Karelian HG's mtDNA was C1g - C1 is a common haplogroup in the Americas, so potentially another ANE marker? Section 8 also leaves it as somewhat of a open question whether phylogenically EHG is closer to Karitiana or MA-1.
ReplyDeleteNative Americans have four different East Asian lineages (and a less common West Eurasian one: X2). C1, like other C, and also A, B and D haplogroups were absorbed as they moved eastwards through (probably) Mongolia, Manchuria and Siberia-east-of-Baikal. The main patrilineage Q1 remained dominant (patrilocality) but the mtDNA and much of the autosomal baggage became East Asian.
DeletePrevious studies of Karelian Epipaleolithic mtDNA show that C1f was only one third (3/9) of the lineages found there, the rest being West Eurasian (U2e, U4, U5a and H). I.e. the process of "westernization" was already very active.
Personally I'd rather attribute ANE to West Eurasian lineages of the kind of U. But it's almost impossible to make a simple association because these lineages are also found in absence of ANE and vice versa. It's important to understand that autosomal affinity works differently than haploid founder effects. You can carry Y-DNA R1b and be 99.99% African and you can carry R1b and be 99.99% European, same for mtDNA. So there are no haploid markers of autosomal affinities, at least not in any simple way, much less in long chronologies such as the one attributed to ANE (Ma1 is 24 Ka old!!!)
Yes, I realize all that, and that autosomal DNA is much more informative than mtDNA or YDNA. However, they can still tell us something about the migration history of different peoples. Also I'd point out though that based on the aDNA they couldn't tell if Karelia HG's ANE heritage is closer to MA-1 or Karitiana.
DeleteY haplogroup P also traces back to East Eurasia (or at least South Asia), so at the time depths we're talking about I'm not sure one can call R or Q really "West Eurasian."
"Personally I'd rather attribute ANE to West Eurasian lineages of the kind of U. But it's almost impossible to make a simple association because these lineages are also found in absence of ANE and vice versa."
I think you misunderstand me - I'm not suggestion ANE was mostly C1. I'm suggesting that older European C1 is mostly ANE. Like you said, U has roots in Europe independent of ANE. Do you think C1 exists in Europe independent of ANE?
Also, I'm suggesting that like the peopling of the Americas, there may have been structure in the movement of ANE into Europe. In the Americas, X2 and C4 are mostly specific to the northern part of North America (north of the Rio Grande). Also, C1 is mostly specific to Central and South America.
"(Ma1 is 24 Ka old!!!)"
Which makes him not ancestral to Karitiana.
My question is: did C4 and C1 follow the same paths to Europe? Yamna is a fusion of EHG and Armenian-like farmers, both of whom carry some ANE. Is the Armenian ANE the same as the EHG ANE? Did the arrive in each population along the same route?
Sorry, Ryan, I seem to have missed this comment for some reason.
DeleteRe. what you sat about associating ANE in America and Europe and even attribute mtDNA structure in America to mtDNA structure in Europe, on first look I find it far-fetched: you want or imagine too tight relations between haploid lineages and autosomal DNA and that's not realistic, more so when Amerinds have a very clear single founder effect in Beringia that forfeits their previous relations to Siberia and West-South Asia.
The closest living thing to "ANE" are the ailing Yeniseian peoples (Kets, etc.), who are linguistically related (it seems now) to Na-Dene speakers but carry different main Y-DNA pools, with Kets being dominated by Q1 and Na-Dene by C2. So IMO Kets are more related patrilineally to Amerinds and Eskimo-Aleuts, even if linguistically they are closer to Na-Dene. Kets live right there in West Siberia, at the very gates of Europe, so IMO people obsessed with "ANE" should probably pay more attention to them.
However you may be onto something because browsing for documentation, I found this on mtDNA C and D, and the authors of the study think that C1 may have coalesced in Beringia, because most descendant branches are Native American. Nevertheless at least C1a has also a Siberian distribution.
This is not the case of C4, which seems much more Asian-centric (NE Asia/Siberia). One unnamed basal lineage seems European (although there's also some other European C4 that is derived from Asian lineages, as happens with C5). This could suggest two different times for flows to Europe of C4: one very old (unnamed lineage) and others more recent (Asian-derived ones).
It is surely of interest to you this entry on ancient West Siberian mtDNA, where it is apparent that C and A are present since very early on in times of Ust-Tartas = Combed Pottery culture, i.e. early Uralics. There's also some D (n=1) but this one actually increases its presence later on, in the Bronze Age, suggesting a different origin related to the Seima-Turbino phenomenon and maybe a transitional Tungusic-related immigration, later reversed.
On the other hand in the Siberian steppe belt or in Central Asia, we see no East Asian mtDNA before the Iron Age, what strongly suggests that the route of migration of these oriental lineages to Europe was via the Taiga and not further south. It's possible that it was not just proto-Uralics who carried them (Yeniseians first and maybe Tungusics later could have been involved too) but mostly it fits with the scenario perceived for N1 and the proto-Uralic migrations.
Another reference of interest is probably this sample of Siberian mtDNA. It has some limitations but still better than most.
No worries. I can't promise my responses will be quick either - work and whatnot. Just have time for a quick response now, will follow up more later.
Delete"Re. what you sat about associating ANE in America and Europe and even attribute mtDNA structure in America to mtDNA structure in Europe, on first look I find it far-fetched: you want or imagine too tight relations between haploid lineages and autosomal DNA and that's not realistic, more so when Amerinds have a very clear single founder effect in Beringia that forfeits their previous relations to Siberia and West-South Asia. "
I realize that the evidence doesn't support this - just highlighting it as a possibility to watch for. People model/consider groups like WHG, ANE and EEF as monolithic, homogeneous blocks because that's makes the analysis simple. Unfortunately it doesn't really reflect reality. So I'm asking what sort of substructure we can infer from the existing evidence for each group.
Re: the Kets - I think there's a very good case to be made that their language (and a good chunk of their genetics) comes from a back migration from Alaska to Siberia coinciding with the expansion of the Arctic small tool tradition. This would also explain the link with Eskimo-Aleut, and the presence of "Mongolian" Y-DNA in some parts of North America.
I'll try to dig up some more sources after work - there's some you'd find interesting too. That Derenko et al paper is one I have bookmarked and refer to often. :3
I agree with you on the Taiga as a route into Europe for those eastern lineages completely. I suspect that also contributed significantly to the Karelian HG's ANE. Yamna got ANE also from its Armenian-like side. My suspicion is that this side would also have been free of much in the way of eastern lineages, and that that ANE came by way of central Asia, as the Underhill paper suggests R1a did.
This is the paper I was looking for: http://www.pnas.org/content/110/35/14308.long
DeleteIt has a pretty good overview on the mtDNA structure in North America and its connection to Siberia.
Any thoughts on this paper: http://scholarworks.gvsu.edu/cgi/viewcontent.cgi?article=1004&context=theses&sei
Frankly I find it hard to follow, but it mentions quite a few C4a samples from Neolithic Ukraine and later.
This paper also has some information on Mesolithic Karelia. I think it the same sample as Haak used. http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1003296#pgen-1003296-g006
3/9 are C1. There are no C4, but not a large sample size. If only there was more ancient DNA already available! :3
I agree completely with you re: the value of looking at the Kets.
"Any thoughts on this paper (...) Frankly I find it hard to follow, but it mentions quite a few C4a samples from Neolithic Ukraine and later".
Delete→ http://forwhattheywereweare.blogspot.com/2013/09/ukraines-neolithic-and-bronze-age.html
There's two C* and one C4a2'3'4 in Neolithic Ukraine (low Dniepr, two sites). All the rest is Western mtDNA (H, T, U3 and U5a). It's notable the high frequency of H (6/14), I must say (although a couple of samples are unclear).
In the "Bronze Age" Kurgan sites of SW Ukraine, all three samples are C4a3. It is unclear if this lineage could derive from the one found in Neolithic Dniepr (and I would think it unlikely on first sight, but who knows!) but it may be somehow related in the wider picture of C migrations westwards.
IMO these "oriental" lineages would have been incorporated from contacts with proto-Uralic peoples further North. One could even imagine that these contacts brought pottery to Ukraine, from where it spread to other places (Thessaly, West Asia) but this issue requires further research for all I know.
Re. the Kets, I rather tend to consider them a survival of Paleo-Siberian cultures, related to the Beringian founder effect but not necessarily nor probably back-migrants from Beringia itself. They may have originate further East than their current geography but I don't think that so far East. In fact the Na-Dene (who are related at least by language to Kets) are a "recent" secondary migration to America, and they very possibly originated in Siberia rather than Beringia.
DeleteHopefully I didn't post a half completed response already. Apologies if I did. :/
DeleteSo keep in mind in some of those links they don't distinguish between A1 and A2. A2a is the one associated with both Dene and Eskimo-Aleut peoples. I believe the Kets have A1, not A2, which would be a para-Beringian haplogroup along the lines you suggest. Selkups, Chukchi, Koryak and Evenk peoples have A2a though, and A2a is nested within A2 which is throughout the Americas.
Here's a paper on the linguistic evidence for a back migration (admitably somewhat weak) - http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0091722#pone.0091722-Tamm1
In terms of the Y-DNA of the Dene and the Kets, keep in mind that the Dene are not majority C-P39 - it's still outnumbered by Q, though only barely in some groups. It's not specific to the Dene either, but also found among Wakashian, Algonquian, Sioux and Haida groups, and not just those groups that live close to Dene. The Navajo and Apache migrations are relatively recent to be an explanation for this in my opinion. There's also a couple of strange C(xP39) in SE Alaska and Columbia of all places. There are some very strange results in a couple Columbian groups actually. C-P39 is too old to be a Dene marker too actually.
Paper on Inuit and Dene Y-DNA: http://www.pnas.org/content/109/22/8471.full
Do you know of any good ones on how Ket Y-DNA Q fits into the phylogeny of the haplogroup?
Whether the ultimate origin is west of the Bering Sea or east of it, may not be 100% possible to determine, and I think the most likely there were contributions from both sides. Inuit from Alaska to Greenland and Chukchi on the Bering Sea coast are more closely related to one another than either is to Siberians or the Dene.
My thinking is that there were two original streams of ancestry into the Americas - one along the coast all the way to Patagonia, and a second on the interior that did not reach South America. The two mixed in North America. Meanwhile gene flow continued across the Bering Strait in both directions, creating a mixed neo-Beringian population intermediate between Siberia, East Asia and the Americas. From the coast, that group later expanded outwards, giving rise to groups like the Saqqaq, and spreading their DNA along the coasts of both Asia and North America. A2a marks this expansion.
These para-Saqqaq then mix with some Amerind groups somewhere in Alaska. This mixed population gives rise to the proto-Dene, who then expand again into the interior of North America and across the Bering Strait into Siberia. They give rise to the Kets and Dene, but also bring for the first time Amerind DNA into proto-Inuit and proto-Chukchi groups, and across the Bering Strait.
Then, much later at around 800 AD, the Thule culture arises on the Alaskan coast. It spreads quickly across the Beiring strait to form the Yupik, but also east, quickly replacing the older Dorset/Saqqaq related groups, giving rise to the modern Inuit. It is still mostly "neo-Beringian", but unlike Saqqaq it also contains some Amerind ancestry.
I'm not sure how else one can reconcile the Kets and Inuit having Amerind affinities but Saqqaq not. http://www.nature.com/nature/journal/v463/n7282/full/nature08835.html#f3
Don't know if I shared this before, but well worth the read too: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0000829
Well, it's indeed possible that the C1 lineage of North America is not truly from a secondary migration but a diversity remnant that just didn't make it further south (and was swamped by Q1 in Beringia and other Siberian areas (like among Kets), although the Na-Dene migration seems pretty much demonstrated on its own terms anyhow.
DeleteThere's some C1* in some populations of Ecuador that could be supportive of such notion, although it's not the former C3b (C1b now?), which is the one present in the North, so it's even more mysterious. → http://forwhattheywereweare.blogspot.com/2013/04/southern-native-american-y-dna-no.html
"Do you know of any good ones on how Ket Y-DNA Q fits into the phylogeny of the haplogroup?"
Nope, sorry. Only generic stuff. They are a very interesting population that deserves more attention, no doubt.
"My thinking is that there were two original streams of ancestry into the Americas - one along the coast all the way to Patagonia, and a second on the interior that did not reach South America".
That would be Clovis, right? However latest I read from Clovis strongly suggested a rather southernly origin in North America, not supporting the continental route. It's just possible, I guess, that C1 was a less dominant lineage that just didn't make it to the Meso-South-American founder effects, remaining restricted to the North. I think there was a recent paper that found important differences between Canada's and Mexico's Natives, with the first ones being much more diverse in terms genetic, however the big USA blank in between did not allow for any sort of nuanced reconstruction.
It's also a fact that the only Clovis aDNA to date (Anzik) is not C1 but Q1a2a1. Another aDNA from the Alaskan "panhandle", of roughly similar age, is also Q1a2a1. See:
→ http://forwhattheywereweare.blogspot.com.es/2014/02/ancient-dna-from-clovis-culture-is.html
→ http://forwhattheywereweare.blogspot.com/2014/05/12000-years-old-mexican-girl-confirms.html
These results still allow for C1 to have been boosted by some sort of migration.
...
...
Delete"These para-Saqqaq then mix with some Amerind groups somewhere in Alaska. This mixed population gives rise to the proto-Dene"...
But then how do you get Dene-Yeniseian to be found both in North America and West Siberia? Wouldn't be more logical to imagine a Na-Dene migration from Siberia that had a larger genetic impact than the areas of Na-Dene lenguages? Either some other languages are cryptically related to Dene-Yeniseian or, more simply, these Na-Dene groups lost their languages by contact, admixture and alliances with other populations.
I don't easily accept that Kets migrated from North America to West Siberia, really. My thinking about the Kets is that they are true "Paleosiberians" of the kind of Ma1 but obviously with some admixture with West Eurasians particularly but also with other Siberians. A different branch of these Paleosiberians, accidentally dominated by C1, would migrate to North America through the Esquimo-Aleut area of former Beringia but without making a major impact there.
"I'm not sure how else one can reconcile the Kets and Inuit having Amerind affinities but Saqqaq not".
That's an interesting link, thank you. However I do not see Kets in that graph being at all Amerind-related but rather like Saqqaq and other "Paleosiberians", including the Koryak-Chukchi, just that with about 1/3 of West Eurasian admixture.
Inuit are half-way between this Paleosiberian zone and Amerinds, while Na-Dene are obviously more admixed with Amerinds. It's somewhat difficult to explain admittedly but all points to Saqqaq, Inuit and Na-Dene being secondary migrations into America. On the other hand, the Kets (and Saqqaq) have no Amerind affinity so explaining the Kets as back-migrants from America does not make any sense to me. It's rather something Eastern Siberian IMO.
As for your last link: it's old and well-known (at least to me) but it is no doubt a very informative study.
Apologies, only have time for a partial response so far. Work. :/
Delete"However I do not see Kets in that graph being at all Amerind-related but rather like Saqqaq and other "Paleosiberians", including the Koryak-Chukchi, just that with about 1/3 of West Eurasian admixture. "
Take a closer look. On the PCA, the Kets and the Selkups (who were recently Yeniseian speakers) are pulled down noticeably from other Siberian populations, towards Amerind groups. If it was purely Paleoeskimo/Neoeskimo (ie Chukchi, Koryak, Saqqaq, etc) it should be pulling them to the left. Though it's hard to make out given the high levels of recent European admixture throughout the modern populations.
The ADMIXTURE run is clearer though. The dark maroon component is elevated in the Kets and Selkups compared to other Siberian and Beringian groups, excepting the Chukchi, and the Chukchi have been in direct contact with Alaskan groups for over 6,000 years. Obviously that's a small component, perhaps 10%, but it's the only thing that distinguishes the Kets/Selkups from their neighbours. I don't see how that could possibly be explained as some Paleosiberian component either - it rises to 100% in Karitiana. That's the Amerind component I'm referring to. It's a small amount, but quite striking.
Clovis is actually the coastal route. Reich calls the interior route "Algonquian." He has the Dene as 90% para-Algonquian and 10% some sort of cryptic East Asian group unrelated to the Inuit/Saqqaq. His choice of Dene groups is poor though - he chose a group from the youngest Dene branch (Athabaskan) that probably absorbed a lot of Algonquians recently, so that 90% number may be wrong.
The old remains from the Northern BC/ SE Alaska coast seem to be from that coastal migration, but some sort of discontinuity happens around or before 8ky, when the Hecate Strait is formed, isolating the Haida on Haida Kwai. That discontinuity seems unrelated to the Algonquian though - the Haida or almost entirely A2, and the Dene groups on the coast are mostly A2 with a bit of recent Siberian D2 likely courtesy of groups like the Saqqaq. mtDNA X, B, D1, C and D4h3a are strangely absent. Neither the modern nor ancient populations seem to be well sampled though - neither Rasmussen nor Reich include samples from there.
The Haida used to be considered part of Dene, but that's now disfavoured. I could buy some sort deep connection there though.
I can buy the Dene and the weird yDNA C3* in Ecuador and Columbia having a common origin too. There's a bunch of C3* along the Alaskan panhandle too. I agree that the Dene likely have more recent ancestors in Asia than Clovis or "Algonquians." All I'm suggesting is the Kets' linguistic and genetic distinctiveness actually come from a secondary expansion of these Proto-Dene out of the Alaskan Panhandle. That may seem convoluted, but it fits the linguistic and genetic evidence.
I'm not arguing that there was strictly one way flow out of Alaska either. Just that it was a two-way process. Just as D2 in North America is pretty clearly of recent (6ky?) Siberian origin, A2a in Asia is pretty clearly of recent North American origin.
What you call "noticeably" for me is "not significant", nearly zero. But we see that better in the ADMIXTURE graph, where Kets and Selkups show up as mostly mainline Siberian (orange: 60-70%), somewhat Western (blue: 20-30%) and only very slightly Beringian (cream: 5% at most) and even less so Amerind (maroon). Does that justify the back-migration hypothesis? Certainly not for the bulk of the ancestry (>90%) but there is a very minor element which might be back-migrant. However an alternative possibility is that they actually have an element that is rather proto-Amerindian (but still Siberian) and manifests in the analysis as that mix of cream+maroon. I think it is as plausible and surely more parsimonious.
DeleteMore interesting maybe is that the Na-Dene actually have almost exactly the same structure (cream+marooon) as the minor para-Beringian element in Yeniseians, so almost certainly that "double" minor component in Yeniseians speaks of some sort of relation very specifically with the Na-Dene (what is coincident with recent linguistic research).
One very likely possibility is that ADMIXTURE, at least at that K value, cannot properly solve the "double component" (dominant in Na-Dene and very minor in Yeniseians) and that further K-rounds would show it as a single Dene-Yeniseian component distinct from both the Amerind and the Beringian but related to both. Do you follow my reasoning?
"All I'm suggesting is the Kets' linguistic and genetic distinctiveness actually come from a secondary expansion of these Proto-Dene out of the Alaskan Panhandle. That may seem convoluted, but it fits the linguistic and genetic evidence."
DeleteI still think that the origin of the Na-Dene is East Siberian although within the ancient "proto-Amerind" macro-population (the true Paleosiberians, as in pre-LGM Siberians). Within this model the Yeniseian would have a small Na-Dene element (minus some extra Amerind admixture in this population), a major mainline Siberian element (which is not truly Paleosiberian but rather Neosiberian, i.e. post-LGM and even Holocene Siberian, possibly incorporated in their migration westward) and then the remaining Western component (incorporated at the end of their migration mostly).
My opinion anyhow. My impression is that the longer that true Paleosiberians spent in Eastern Siberia the more they became admixed with something like modern East Asians, so the less admixed are Ma1 and Native Americans (whose East Asianness is clear in mtDNA but seems very "primitive" in autosomal genetics), so Eskimo-Aleut and Na-Dene are more that way than Amerinds but they still remained distant enough, it seems, to remain relatively unadmixed. But other Siberians are much more admixed with mainline East Asians (see fig. 1b of Mal'ta paper for a complementary perspective) and Yeniseians mostly belong to that group, even if they seem to retain something from the true Paleosiberian population as represented by Na-Dene and Eskimo-Aleuts.
"A2a in Asia is pretty clearly of recent North American origin".
That can well be just a Beringian forth-back flow. A2a specifically shows three branches: first an Eskimo-Aleut one and then another that splits in two: one heading to Asia and the other to North America. So to me it looks like A2a was all the time in Beringia and not further inland in America. Anyhow, do Kets have this lineage? Most Asian A is not A2 but belongs to other sublineages.
Kets do not have A2 (that we know of) - they are A1. Selkups do. I'm not sure which one though. I would share your view that A2a didn't penetrate deeply into the interior of North America though, but it is still nested in a pretty star-like A2 tree. Interestingly, the Kets lack D3 too (I'm not sure if Selkups have it), which is the recent Siberian lineage ubiquitous among the Eskimo-Aleuts, but much rarer among Dene (and absent among Haida).
DeleteRe: the magnitude of the admixture - well, 5-10% ANE is the difference between between French people and Basque people, so I think these small differences can actually have a lot of meaning. Reich had the Dene as 90% Algonquian-like and 10% something specific to Dene (though as I mentioned, that`s probably a low estimate), so somewhat symmetric.
Re: on-going admixture, for sure. Probably both from East Asia and Siberia really - meaning as time progresses people on the West side of the Bering Strait. That's really central to the scenario I outline. I'm just describing the usual isolation-by-distance forces creating a cline that spans Siberia, East Asia and the Bering Strait. Under the scenario you outline, what mechanism keeps the Kets isolated and "pure" while other populations around them continue to exchange genes with each other? And what ends that constraint? And why is their impact so much more pronounced in Alaska than in Siberia?
I also think both genetic and archaeology suggest that the Paleosiberians/Paleoberingians were never actually a homogeneous group.
I agree that the sampling and other issues though make the ADMIXTURE run less than ideal. You'd think all these people studying the peopling of the Americas would you know, actually sample the people along the path that those first migrations followed, and scrutinize them to death.
Something that jumped out at me from one of the old posts you linked btw - the Oroqen are a group with a high rate of yDNA C3*. I wonder if there's something to that.
Meant to mention this before - the you were talking about cryptic linguistic relations. Wakashian here on the South Coast as being a distant relative of Eskimo-Aleut. I haven't looked at the proposal in touch much detail though.
When I get a chance I'll see if I can find anything at about the archaeology of canoes in the region. Whichever direction the Dene went it would have been pretty intrusive and noticeable (assuming it was by canoe). The coastal Dene in Alaska (and other groups all the way down the BC coast) built dugout canoes that could fit up to 60 people at a time. One of the benefits of living somewhere where trees grow to 300 feet. The Sitka spruce in particular grows extremely tall and has extremely strong and light wood. It was used to make the Wright brothers' plane even. http://qmackie.com/2010/04/27/a-tlingit-spruce-canoe/ The Eskimo-Aleuts have few trees and rely on driftwood skeletons to make seal skin boats. So a giant 60 person dugout would be noticeable you'd think.
According the data of Tamm 2007, A2a is found among Selkups and Evenks. Importantly neither is Yeniseian speaker. Although it seems apparent that Selkups probably have a Yeniseian background, they are Uralic speakers of the Samoyedic sub-group. Evenks are Tungusic (Altaic) speakers. Both families probably have East Siberian/Manchurian origins, even if Uralic represents a much older westwards migration: Tungusic probably expanded mainly in the Siberian Bronze Age, with the Seima-Turbino phenomenon, while Uralic should be rooted in the Late UP and Epipaleolithic migrations of Y-DNA N1. It's possible I guess that Yeniseians (Kets) also followed this general route westwards through the taiga but less clear when it happened.
DeleteIn any case what we see is a Beringian (or back-migrant North American if you wish but always via a Beringian node) element hitchhiking these Northern specialist migrations to the West. We do not see anything that can be described as specifically Yeniseian in that, although it might be related to the Dene-Yeniseian connection, it's only "might", nothing solid.
"... 5-10% ANE is the difference between between French people and Basque people"...
Every day I doubt more that the ANE estimates per Lazaridis are meaningful as such raw figures, they seem too high considering the ancient background, even Yamna ANE affinity itself, which was not higher than 25-30%. Basically to get a 15% ANE figure (+10 pts rel. to Sardinians), you need a 40% Yamna admixture and that is just too high. Even Basques would need to have almost a 20% Yamna admixture (what means like double "Celtic" admixture) and nothing else in the genetics seems to support that. I would like to see direct Yamna or CW admixture estimates instead and anyhow I still want to see more Atlantic and North Sea ancient genetic data before I make up my mind.
"Algonquian-like"
Is it possible that Algonquians are Na-Dene who have lost their original language?
"I also think both genetic and archaeology suggest that the Paleosiberians/Paleoberingians were never actually a homogeneous group".
What's "homogeneous"? What we see very clearly is that mode 4 technologies spread eastwards from Altai before the LGM and that flow can be quite robustly associated to the First Paleosiberian population, which was (roughly) Ma1-like originally (and Kets are the most Ma1-like of all modern populations, even more than Native Americans surely).
As these "First Paleosiberians" reached to East Siberia (and Mongolia/Manchuria/North China) they mixed with the early East Asians quite heavily, so there you have the main breach of homogenization, surely segregating the populations living West of Baikal (Ma1-like) and those East of it (proto-Native-Americans). Of course then there would have been many subdivisions but for what interests us right now, that would be "splitting hairs", a bit too subtle.
"the Oroqen are a group with a high rate of yDNA C3*. I wonder if there's something to that."
The Oroqen are peculiar (along with the Tujia). Not sure what is up with them but check the section on "Tianyuan affinity mystery" in this entry. They are not significantly more Native-American-like than Europeans or Australasians, they also have extremely low Tianyuan affinity (similar to non-Papuan Melanesians, much lower than any other Eurasian or Papuan populations). Not sure how to explain this, maybe you have some ideas? A possibility could be that they have been relatively isolated since very long ago but that's weird, more so as both populations seem relatively unrelated to each other, be it by geography or language (Oroqen are Tungusic from Inner Mongolia/Manchuria, while the Tujia are Sino-Tibetan from Central-South China).
On the rest I don't have anything to add.
Re: the ANE percentages, agreed, but I wouldn't view the Ket signal as any different. I'm just saying that the Ket signal is on the order of the signal we see with deep linguistic divisions of Europe. Whether that ANE actually means ANE... yah, it's probably a proxy for something. Or more than one thing.
DeleteThe paper only says the Evenk mtDNA is A2, but yes I think its safe to assume it is A2a. It's interesting that the Selkup A2a and Apache-Dene A2a cluster together in the A2a tree.
"Is it possible that Algonquians are Na-Dene who have lost their original language?"
More likely that there could just be a deep linguistic relationship there. We're talking migration events that are at least 14,000 years old - so likely quite a few linguistic relationships that we just can't reconstruct. Maybe they could have absorbed some similar groups though? My understanding is Algonquian expanded relatively recently (~2,500-3000 BP). The Athabaskan subgroup of Dene languages found in the interior of the continent only expanded ~2,000 BP. So we're definitely blind to whatever was going on in the northern half of North America before that. Dene groups seem to lack the C4c and X2a though, while these are relatively prevalent among Algonquians. It seems complex to say the least. One thing to keep in mind though is that while the Algonquian language expanded recently, the distinctive Algonquian DNA did not - its expansion occurred close to the same time as the other main migration.
...
...
Delete"Tungusic probably expanded mainly in the Siberian Bronze Age, with the Seima-Turbino phenomenon, while Uralic should be rooted in the Late UP and Epipaleolithic migrations of Y-DNA N1. It's possible I guess that Yeniseians (Kets) also followed this general route westwards through the taiga but less clear when it happened."
I seem to recall there being archaeological evidence for this in Yenesei basin too, and that prior to the Bronze Age remains in the basin all seem to be overwhelmingly West Eurasian. Hard to tie that specifically to the Kets though I guess.
"What's "homogeneous"? What we see very clearly is that mode 4 technologies spread eastwards from Altai before the LGM and that flow can be quite robustly associated to the First Paleosiberian population, which was (roughly) Ma1-like originally (and Kets are the most Ma1-like of all modern populations, even more than Native Americans surely)."
A decent chunk of their ANE could come from their West Eurasian side though. Indo-Europeans made it to the Yenesei before being pushed back by later migrations I believe?
Also, looking at the original Ma-1 Paper's Supplementary Info (http://www.nature.com/nature/journal/v505/n7481/extref/nature12736-s1.pdf - Figures 21 and 22), Kets are actually less close to Ma-1 than any population in North or South America. They are second only to the Naukans in closeness to Ma-1 among Eurasians though, and the Naukans speak an Eskimo-Aleut language and lived at the closest point to North America in Eurasia.
I'm not strong on the archaeology side at all (and most sources I find on this seem to be behind a paywall) but from what I can understand, there the early Aurignacian migration that you mention, but there's also a later input of microblade technologies just a bit before sea level rise cuts of east and west Beringia from one another. Apparently the microblade and Clovis tool kits are quite distinct from one another, though they overlap in some sites, and the significance of this seems under debate.
That spread of microblades seems to originate from Lake Baikal too (though much later), and also spreads to SW Asian and to the Pontic Steppe (among other places). Perhaps this coincided with the spread of C4 and X2? I suspect that much to Davidski's chagrin we will find ANE predating the Indoeuropean expansion in SW Asia in the future, and this would be a possible vector that would also explain why X2 has its present distribution.
I agree with you about the subtlety point, but I think these subtle differences if could explain a lot about the genetic variation of both Europe and the Americas. I wish NA was better sampled.
Re: Oroqen - yah, that's the post. No idea what it means, but I guess it's something to keep in mull. It would have to be some sort of admixture from something unrelated to Tianyuan, right? But that would require being isolated in East Asia for a very long time. How do the Oroqen place on archaic admixture?
"The paper only says the Evenk mtDNA is A2, but yes I think its safe to assume it is A2a."
DeleteI think it's safe to assume that it is NOT A2a, why would it be?
"We're talking migration events that are at least 14,000 years old"
Only for the Amerind bloc (rather 17 Ka old according to the latest data, older dates like a Brazilian 22 Ka one are still not regarded "credible" but the ones from SW USA and the Andean area are indeed quite credible and strongly support a barely post-LGM expansion in America from Beringia.
However the Na-Dene migration should be much much much more recent. Late Holocene even!
" One thing to keep in mind though is that while the Algonquian language expanded recently, the distinctive Algonquian DNA did not - its expansion occurred close to the same time as the other main migration."
If you trust molecular-clock-o-logy only. I just do not, unless very well done, what is the exception and not the rule. What "distinctive" marker are you thinking of anyhow?
"Kets are actually less close to Ma-1 than any population in North or South America."
You may be right. They are still the most relevant Eurasian population, having more Ma1 affinity than Inuits, never mind Russians and Uralics. Naukans do not beat them at all, check fig. 23, please. They do beat Selkups and Khanty, so they are close enough to the Ket level.
Re. the Aurignacoid expansion, this is a qualitative different level than the differences between microblades, Clovis, etc. It's a whole brand new technological mode that spread from somewhere in West or Central Asia (or maybe even India but no data to support that yet) and that corresponds to the early Upper Paleolithic (i.e. the emergence of "mode 4" with stone blades of non-levalloise, i.e. not flake-based, design). It's a true "stone age revolution" in technological concepts, and also comes with needle and very likely domestic dogs.
DeleteI use the term "Aurignacoid" (taken from the literature) to underline its similitude to mainline Aurignacian, even if most of these earliest UP techs are not "true Aurignacian" (but usually similar enough to see a shared pattern). This earliest UP spread to West Asia, Central Asia (incl. parts of Siberia), South Asia, Europe and NE Africa (LSA) very early (50-40 Ka BP, 40-30 Ka at the latest). Then a branch from Altai migrate through Mongolia and North China (and possibly nearby parts of Siberia) to the East, bringing "mode 4" to that region (since 30 Ka BP only) and founding the proto-American population.
Some available references:
→ Altai Aurignacian: http://www.geology.cz/sbornik/antropozoikum/no23/23-17-the%20middle%20and%20upper...pdf
→ Siberian early UP chronologies and proto-Americans: https://journals.uair.arizona.edu/index.php/radiocarbon/article/view/4122/3547
→ West Eurasian UP colonization: http://prehistorialdia.blogspot.com/2014/03/la-expansion-de-homo-sapiens-hacia.html (I thought I had made an English language synthesis of this but seems not, so you will have to use a translator or just look at the very nice maps).
"I think it's safe to assume that it is NOT A2a, why would it be?"
DeleteA2a is present elsewhere in NE Asia. Really only A2a and A2b are. Also, I half-remembered seeing another paper reporting A2a there, but couldn't find it. :3 Now I have!
http://www.ddl.ish-lyon.cnrs.fr/fulltext/Pakendorf/Duggan_2013_Tungusic_PLoSONE.pdf
A2a is found among Taimyr Evenks but no other Tungusic groups. The Taimyr peninsula is directly adjacent to the Yenisei river. So pretty plausible.
"Only for the Amerind bloc (rather 17 Ka old according to the latest data, older dates like a Brazilian 22 Ka one are still not regarded "credible" but the ones from SW USA and the Andean area are indeed quite credible and strongly support a barely post-LGM expansion in America from Beringia."
I agree. Only setting 14Ka as a lower bound.
"However the Na-Dene migration should be much much much more recent. Late Holocene even!"
If we accept a genetic and/or linguistic relationship with the Haida though, it's at least 8Ka old. After that the sea separates Haida Gwai from the mainland, with even greater isolation from the continent than what exists at the Bering Strait.
"If you trust molecular-clock-o-logy only. I just do not, unless very well done, what is the exception and not the rule. What "distinctive" marker are you thinking of anyhow?"
I don't trust their calibration, but I think they can at least give a good idea of the relative age of different groups. I'm referring to C4c and X2a and their diversity. This paper has X2a at 14-17 Ky compared to D4h3a's 14-18 Ky estimate. D4h3a was Anzick's mtDNA haplogroup. X2a may be younger than D4h3a, but not by a lot.
http://www.sciencedirect.com/science/article/pii/S0960982208016187
I'll take a look at those sources. All I was trying to get at though is that there was later cultural (and perhaps genetic) flow into Beringia that may created some structure in the initial peopling of the America.
Interesting that you mention India as a possible source. That would maybe explain the possible "basal East Asian" admixture in Ma-1, and how the heck haplogroup P got involved.
Spanish is fine for me. I don't speak it, but I'm pretty fluent in French, and so long as I have an online dictionary for the occasional word. Put me in front of someone actually speaking Spanish and I'd be useless though.
In the supplemental table 1 (link to HTML format article), A2a is only listed among Taymirs. Another A2 variant, A2b1 is found among the Koryaks.
DeleteOn the other hand, A4 and A8 are found in various populations, so it is not just A2 what we see of A in Siberia. In fact mtDNA A is common in NE Asia (Mongol, Japanese, several Siberian groups not studied in that paper, Manchurian Chinese, as well as among Tibetan populations (Yunnan, Tibet). While I do not know of any detailed study on the underlying structure of this haplogroup, it's easy to imagine that it spread from either the Tibetan region or NE Asia and had some sort of burst in the Paleolithic because it shows star-like structure at the pre-A2 level particularly and then also at A2 but this is probable related to Native-American expansion already (see PhyloTree). The short "molecular" difference between this pre-A2 and A2 stars suggests a post-LGM chronology, what fits well its relative Northerner specialization and may still be related, albeit indirectly (via East Siberia) to the Native American formation process.
Incidentally I must note that A4 seems to have been dropped from the latest PhyloTree nomenclature. It'd be interesting to see how the newest phylogeny and older data converge (too much work though).
"The Taimyr peninsula is directly adjacent to the Yenisei river. So pretty plausible."
Believe what you want but "plausible" is not fact and the rarity of A2a plays against the notion.
Anyhow, please chew on what I just mentioned on the pre-A2 (A1"22) star-like structure and the subsequent A2 one. They are separated just by a mere HVS-1 mutation so we are probably looking here at the very same right-after-the-LGM expansion, just that at different sides the American-Asian isthmus.
"If we accept a genetic and/or linguistic relationship with the Haida"...
The same linguists who are applauding the Dene-Yeniseian theory are rejecting that Haida is part of Na-Dene.
" After that the sea separates Haida Gwai from the mainland, with even greater isolation from the continent than what exists at the Bering Strait".
→ http://en.wikipedia.org/wiki/Haida_people#History
Apparently the Haida have occupied Haida Gwaii since the very first NA colonization some 17 Ka ago. Also they were not really isolated and are credited for the introduction of the totem pole and other NW coast traditions. They were skilled seamen (within the limitations of their pre-colonial technological stage).
...
...
Delete"This paper has X2a at 14-17 Ky compared to D4h3a's 14-18 Ky estimate. D4h3a was Anzick's mtDNA haplogroup. X2a may be younger than D4h3a, but not by a lot."
I would not consider those lineages "Algonquian" for the very reason that they are very old in time. If the Algonquian language family would be that old it would be impossible to recognize. In fact, the most likely language family (albeit hotly debated) for that chronology can only be Amerind.
If X2a is restricted to Native Americans (AFAIK the Siberian sister clade is X2g or maybe X2h, while there is more X2 in West Asia and, since Neolithic, in Europe as well) then it's only logical that it has a 17 Ka BP age (roughly). All founder NA haplogroups should.
"Interesting that you mention India as a possible source. That would maybe explain the possible "basal East Asian" admixture in Ma-1, and how the heck haplogroup P got involved."
If it's truly "East Asian" it should rather mean some admixture with ancient East Asians, more apparent in Native Americans - but there's no reason why Ma1, so close to Baikal, should be protected from East Asian genetic backflows.
However the route of P1 is pretty much unmistakable today: K probably migrated to SE Asia either soon after or just before the Toba catastrophe and K2 then expanded from that region: first the NO branch northwards and then K2a to Papua and P1 to India, where it left a legacy in form of P1* in the Lower Ganges and in form of upstream branches of Q and R in North India, Pakistan and Iran. A similar trail can be followed for mtDNA R, which is clearly coupled with Y-DNA K2.
Scholastic molecular-clock-o-logists may disagree but for me it is unmistakable, even if the association of some mtDNA N subclades and that or some R with Y-DNA IJ in West Eurasia may slightly blur the issue - but nothing that can't be explained: (y) IJ was incorporated into the P1 westward trail, becoming important for some reason, while (mt) N seems to have expanded not too long before the expansion of R and from around the same SE Asia-Bengal arch.
Oh, re: Figure 23 - that one is Han vs Ma-1, no? So more a measure of ANE vs East Asian than actual closeness to Ma-1.
DeleteYes, you're right, my bad. The figure I was searching for was fig. S21 but got confused when looking at both S22 maps. In any case you're right: Naukans are more Ma1-like than Kets.
DeleteOn the other hand they are from Chukotka (Asian half of Beringia) and belong to the Inuit-Yupik (or Eskimo) wider group so it's not particularly fair to consider them Asian above American.
It's interesting to understand the Eskimo genesis, from Wikipedia "Eskimo" entry: "the earliest positively identified North American Eskimo cultures (pre-Dorset) date to 5,000 years ago. They appear to have evolved in Alaska from people related to the Arctic small tool tradition in Asia who probably had migrated to Alaska at least 3,000 to 5,000 years earlier. There are similar artifacts found in Siberia going back perhaps 18,000 years."
The latest sentence is rather meaningless as such because Siberia is huge but seems to underline a deep Asian origin of the macro-population. However it's clear that the proto-Eskimo pop. must have been living in extreme NE Asia (a branch of the proto-Amerinds of Beringia surely) since "always", migrated to Alaska c. 10-8 Ka ago (within early Holocene context therefore) and back to Chukotka c. 5 Ka ago.
"I would not consider those lineages "Algonquian" for the very reason that they are very old in time. If the Algonquian language family would be that old it would be impossible to recognize. In fact, the most likely language family (albeit hotly debated) for that chronology can only be Amerind."
DeleteI agree, and even today it's not just confined to Algonquian peoples - you can find the same haplogroups in Sioux, Iroquois, Salish... So "Algonquian" is not the correct word. Para-Algonquian? It's the word Reich used though, which isn't helpful.
I disagree with just lumping it under Amerind though, as that would include the groups that do not contain this North American specific ancestry.
Re: Haida - I'm not suggesting they are part of Na-Dene, but simply that they might have a more distant relationship with them. Based on this paper, it seems like the Haida and Dene share a common origin, but went through a period of prolonged isolation. http://labs.icb.ufmg.br/lbem/pdf/schurr12AJPA-southeastAlaskapops.pdf
I'd note too that the oldest DNA from the Alaskan Coast is from On Your Knees Cave around 10 Ky with D4h3a - http://forwhattheywereweare.blogspot.ca/2014/05/12000-years-old-mexican-girl-confirms.html
Haida completely lack D4h3a, but their A2 is very diverse (in more diverse than the coastal Dene tested in that study). The arrival of microblades to the SE Alaska / NW BC region at 10-8 Ky would fit as been the source of a lot of the ancestry of both Dene and Haida IMHO. That deep a linguistic relationship would be hard to reconstruct though, so we'd be left with only a vague similarity.
Agree with all your thoughts re: P1 and Toba.
Re: the Inuit and other Eskimo groups, I'd just keep in mind that they come from a subsequent and much more recent expansion (1-2Ky). The Paleoeskimos were mostly wiped out by this expansion. :(
There was a recent study on Native American genetic diversity in Canada and Mexico that showed very clearly that the former are much more diverse than the latter. Can't find the paper now, sorry. The USA was not studied, so there was a huge blank left in the middle of North America but, in any case, it seems obvious that either, in Mesoamerica or in a previous stage further North, there was some sort of bottleneck effect affecting at least Mexico and everything south of it.
DeleteThat does not deny the purported unity of Amerinds (as first Native Americans out of Beringia), it just speaks of a particular "mega-branch" from Mexico towards the south with a reduced genetic diversity. So I would not make an issue of lineages or other markers that were just left behind in the march to the South.
Rather than thinking in "Algonquian" markers, I'd think of left-behind diversity of Northern North America but always within a wider Amerind concept.
"... it seems like the Haida and Dene share a common origin, but went through a period of prolonged isolation".
Alright. Interesting.
[Haida's] "A2 is very diverse".
That's also very interesting and may explain the origins of the A2 lineage, as it seems in NW North America, not far from Beringia.
Re. the Eskimos: they seem part of the Beringian left-behind population. They should not be considered "Asian" in any meaningful way but very specifically Beringian and closely related, at least in genetics, to the Amerind founder effect. Less clear to me is the origin of the Dene-Yeniseian: although they are somehow related to the wider Native American roots, they might well be more "Siberian" in origin, something partly hidden by admixture with other NAs (or with other Eurasians in the case of Kets).
How many of these 69 tested where new, it seems the bulk where old samples retested.
ReplyDeleteLike the G2 and T1 in central germany
I wouldn't be able to say how many but it's true that many of these samples were previously used in ancient DNA papers (mostly mtDNA, in some cases also Y-DNA). The novelty here is sequencing and comparing their autosomal DNA.
DeleteI think that the truly novel samples come from Eastern Europe (Epipaleolithic, Yamna) and in some cases from Iberia.
I agree
DeleteIt seems to me that after the fertile crescent the next favourable place for farming was yamnya who where a mix of farmers and hunters. From Yamnya it seems they moved west.
The iberian I marker is what I am more interested in , rather the the R marker ( in iberia)
You mean the patrilineages, right? In fact, with the current data on hand we can't state that Y-DNA I was present in Iberia before the Early Neolithic (I2a seems part of the Cardial genetic package). The same re. France (and all Western Europe except Luxemburg). However it's indeed probable that some was already there previously. How much? No idea.
DeleteA detail that may be relevant is that, while Y-DNA I is associated in the Epipaleolithic to mtDNA U (U5a, U4), in Iberia we find high frequencies of mtDNA H and quite less of U, so it's possible that I was relatively infrequent in Iberia or parts of it back then and instead there was something else: something that tends to go along with mtDNA H (particularly H1 and H3), i.e. R1b-L51. But so far the only Y-DNA sequence we have is an oddball: C1a2, which is not very helpful.
thanks
DeleteHave you changed your mind on H, with this Haak article as the H mtdna is the exact same H in the link below, but more refined.
http://forwhattheywereweare.blogspot.com.au/2013/04/brotherton-2013-cherry-picking-evidence.html
My problem with Brotherton's study was mainly on the "molecular clock" thing, which in mtDNA I'm extremely familiar with. IF they used the evidence from Magdalenian Cantabria (as they must), they could never have reached to those conclusions, so I accused them of "cherry-picking" the evidence in order to push for their conclusions in a pseudoscientific fashion.
DeleteIn this study I have not played almost any attention to mtDNA because it's mostly the same as before: no news except for the Yamna samples and a few others.
Erratum "in this study" should read "on this study" or "when discussing this study"...
DeleteThe thing is, when Loschbour is added to an Europeans only PCA it starts to appear that the Basque and Sardinian dimensions are both just WHG-related dimensions, although less informative than ones in a West Eurasian PCA regarding WHG ancestry since Sardinians who do have WHG are doing Bedouin's job as the anti-WHG Near Eastern proxy. I can't find a place for Gokhem2 here that would make it instrumental instead of various other mixes of WHG, Near Eastern, ANE etc
ReplyDeletehttps://drive.google.com/file/d/0B8XSV9HEoqpFMVNnd2RkSXNiTUE/view?usp=sharing
West Eurasian PCA with North Africans makes them "polar" but it still has the same shape as they do without them with BedouinB or Sardinians defining dimension 2 so much of the represented information is the same.
https://drive.google.com/file/d/0B9o3EYTdM8lQaHhna1ZsSnNJem8/view?usp=sharing
The WHG element is critical but not directly. It is critical via the Atlantic Neolithic/Megalithic element (so far only studied in Gökhem) and also Kurgans (which is actually EHG). Personally I'd like to discern what is WHG/Atlantic and what is EHG/Kurgan in the various European peoples and I believe that the Basque prominence in European-only PCAs actually points to that difference.
DeleteThe Basque polarity is probably not WHG as such but rather Gokhem-like and that's the interesting part, particularly regarding the so-called MNE samples, which in Haak's paper behave rather unexpectedly (different than Gok behaves in previous analysis, what raises all kinds of doubts).
The same that we can now basically scrap ANE and replace it with a more direct Corded Ware reference (which is ANE carrier but also EHG and probably also West Asian to some extent), we can in essence scrap WHG and replace it with Gokhem or other Atlantic Neolithic samples to be analyzed in the future.
...
Re. North Africans, it depends on sampling strategy. In your graph there's just a handful of North Africans but what happens when North Africans are sampled in large numbers? This: http://leherensuge.blogspot.com/2010/05/new-paper-on-mediterranean-genetics.html
... which is completely different. And a good example on how sampling strategy does matter and a lot. Different sampling strategies will give us different perspectives on the same issue. If you sample a lot of Finns and Swedes, as for example Anders Paisen does rutinely, you get PCAs that basically work around Baltic differences, if you sample a lot of Mediterraneans and less North Europeans you get something quite different, etc.
And that's because PCAs are sample-sensible and just bidimensional: http://leherensuge.blogspot.com/2008/04/delusion-of-principal-components.html
Most of the r1b had non-H mtdna so this says to me these are att he periphery of the r1b spread. Since there's now r1b found in spain at about the same (along with the fact we've found many concurrent types) I'd say this means europe was already full of r1b. Probably just not as greatly concentrated yet. If nothing else maybe we can end this talk about how young r1b is, obviously it's been around in some form all over europe for a long time.
ReplyDeleteI insist: I'm not talking of R1b in general but of R1b-L51. Just because we seem unable to remember nomenclature beyond three characters does not mean that all is the same, as happens with E1b or other massively widespread lineages.
DeleteWe only have one sample with R1b-L51 in the Chalcolithic: it's not very informative if it is associated with H or not. In general terms however modernly R1b-L51 and H (particularly H1 and H3 also in the case of R1b-S116) are very tightly associated, the main exception being the Baltic and North Africa, where we see abundant H with no or low R1b, and where other processes surely were at play.
I mean r1b in general. It is hard to believe that it would come from an area that did not have rich H and V.
DeleteJust like LBK is obviously a mixing pot and not the source of anything today, this is the case with samara, and moreso it doesn't so far show all the required elements. Some today try to show V coming from near east now, probably to prove this steppe hypothesis for r1b, but it would take some serious data wrangling to find any way to support this. At least there's V found in LBK but it's obviously less than what's found to the west.
Almost everyone who claims this hypothesis ignores mtdna completely and seems to think that they killed off everyone else or something. Even if that complete replacement were possible they would have to have brought a large amount of their original mtdna as well.
So if the recent finds prove anything, it proves that the genetics in europe today have been there a long time. Perhaps in different proportions, but perhaps not. They could well simply be more thoroughly mixed together now. Genetics show that all the seemingly huge migrations of germanics in roman era had surprisingly little effect on genetics, yet for these (possibly totally imaginary) migrations complete replacement is imagined.
R1b in general is another animal. I tend to treat R1b as four main clades: African, Central Asian, Anatolian and European. Most of them display no particular relation with H, let alone V. Only R1b-L51 does and even there there are exceptions. H is also a wide catergory: surely H1, H5 and H8 have little in common beyond a remote common ancestor, probably in the early Upper Paleolithic.
DeleteBut I do agree with your last paragraph very much. It was probably not just Germanics but in general the aristocratic conquests of the Metal Ages, with exceptions though - and discerning when an invasion was a shallow aristocratic layer like Goths, Normans or Arabs or was a more serious colonization attempt as apparently some founder effects in Anatolia/Armenia (attributable to Phrygians, which in turn seem to originate in SE Europe) is a difficult task that requires all the attention of well-done genetic research.
Here and in Lazaridis 2014, we do see that in the Early Neolithic and Late Chalcolithic parts of Europe (namely: East Germany but probably also many other nearby regions) there were important (but not total) population replacements. What we don't see too clearly is what happened in between, particularly in the previous Chalcolithic phases ("Middle Neolithic" in the papers sloppy and anglosaxonish terminology). And we don't see very well (or even at all) what happened in other parts of Europe like France, Iberia, Britain, Italy or the Balcans (what is a huge area).
Worth noting that the R1b found in Neolithic Iberia is ancestral to the African R1b. That doesn't really distinguish between a parallel dispersion or transmission from Iberia to North Africa or visa versa, but interesting none the less.
DeleteNope, Ryan. As far as I can discern from the paper the R1b found in Neolithic Aragon, exactly like the R1b found in Samara is R1b* (M343). Unless you know something I do not, it is not ancestral to anything, but a parallel branch to the rest that just happens to be rare enough that does not have a name (private lineages).
DeleteSome DNA aficionados confuse the "asterisk" paragroups with "ancestral" or "underived". That is simply a misunderstanding caused by their own ignorance. Please do not heed to such claims unless they are well documented.
I realize that it doesn't mean it's the direct ancestor of African R1b.
DeleteHere's what Haak's says about Samara R1b:
"The hunter-gatherer from Samara belonged to haplogroup R1b1 (L278:18914441C→T), with upstream haplogroup R1b (M343:2887824C→A) also supported. However, he was ancestral for both the downstream haplogroup R1b1a1 (M478:23444054T→C) and R1b1a2 (M269:22739367T→C) and could be designated as R1b1*(xR1b1a1, R1b1a2)."
Here's what they say about Aragon R1b:
"We determined that this individual belonged to haplogroup R1b1 (M415:9170545C→A), with upstream haplogroup R1b (M343:2887824C→A) also supported. However, the individual was ancestral for R1b1a1 (M478:23444054T→C), R1b1a2 (PF6399:2668456C→T, L265:8149348A→G, L150.1:10008791C→T and M269:22739367T→C), R1b1c2 (V35:6812012T→A), and R1b1c3 (V69:18099054C→T), and could thus be designated R1b1*(xR1b1a1, R1b1a2, R1b1c2, R1b1c3). "
"The occurrence of a basal form of haplogroup R1b1 in both western Europe and R1b1a in eastern Europe (I0124 hunter-gatherer from Samara) complicates the interpretation of the origin of this lineage."
That's from page 44 of the paper, and the last bit seems like the significant part.
Seems like they are saying that Samara is downstream towards R1b1a from the R1b1a/R1b1c bifurcation, while Aragon R1b is not. Am I correct?
The term "ancestral" is not properly used there and that's a minus for Haak's study that I hope they correct prior to formal publication. Unless they have tested the whole chromosome and found that it is not derived by a third (or nth) branch, they can't claim it is "ancestral". It's sloppy terminology and I lament the confusion they cause with such poorly chosen wording.
DeleteWhat they mean as "ancestral" is that it is not included in the definition of the subhaplogroups tested for, as the technical terminology clearly indicates: "R1b1*(xR1b1a1, R1b1a2)" and "R1b1*(xR1b1a1, R1b1a2, R1b1c2, R1b1c3)". It's not "ancestral" but "excluding" or "excepted" what it actually reads (that's what the "x" means).
When they use the term "basal", while less confusing than "ancestral", they again mean that the lineage is just "hanging from the root node", nothing else. They do not provide further details on those two lineages, unlike what happened with the Ma1 boy's Y-DNA, carefully mapped versus other haplogroups, so... nothing to see here, just random R1b1* lineages that have no relation whatsoever almost certainly with the big R1b1 sublineages.
If anything it can be underlined that the lineages are oddly overrepresented in these ancient samples in comparison with modern ones, particularly in Iberia (in Russia there's so little R1b, excepted some Volga minorities, that it's hard to compare).
Are they incorrect when they refer to Samara as a basal form of R1b1a then?
DeleteIf anything it can be underlined that the lineages are oddly overrepresented in these ancient samples in comparison with modern ones, particularly in Iberia (in Russia there's so little R1b, excepted some Volga minorities, that it's hard to compare).
DeleteMaybe that's because recently successful lineages like R1b-P312 didn't even exist in the early Neolithic, much less in Neolithic Iberia? It's not so "odd" that you tend to find basal lineages like the R1b1 Iberian from this paper, or e.g. the K-M526 carried by Ust'-Ishim, once you consider that they were found in time periods close to the conventional age estimates of these lineages, before the lineages commonly seen in modern populations had the chance to disperse, or even come into being.
One of the authors of the study has said that the reference to the Samara forager as R1b1a was an error. He was just R1b1, with no reliable result for P297.
Delete@Ryan: Capra explained it already (better than I could).
Delete@Lank: "Maybe that's because recently successful lineages like R1b-P312 didn't even exist in the early Neolithic"...
One has to really force a lot the "molecular clock" to imagine R1b-P312 not existing in the early Neolithic. It also forces a lot the logic behind the reconstructed geographic expansion, which is impossible (Occam Razor does not allow by any means) to originate in Eastern Europe. My doubt is whether R1b-L51 coalesced within some sort of (as of yet undetected) Neolithic founder effect, expanding then within Megalithism, or it was already there in the late Paleolithic/Epipaleolithic and just re-expanded to present day levels of dominance with Megalithism, as probably did mtDNA H.
K2* from Ust'-Ishim can't be from "time periods close to the conventional age estimates". UI is more recent than Ma1 and this one had already a P1* (pre-R) lineage which is very much derived, and must be more recent than R1b-M343. K2 is much much older (I'd say that c. 70 Ka BP, 60 Ka at the latest). Just that early diversity is reduced by expansive founder effects and other such demographic phenomena. Many lineages that in the past were somewhat common (apparently, judging on aDNA data), now are too rare to even deserve a name or have gone extinct altogether.
Ah, thanks for the clarity Capra, Maju. So much for that then lol. I guess all we can say for sure is that R1b1 had a pretty wide range at the time of the early Neolithic.
Delete
DeleteK2* from Ust'-Ishim can't be from "time periods close to the conventional age estimates". UI is more recent than Ma1 and this one had already a P1* (pre-R) lineage which is very much derived, and must be more recent than R1b-M343.
UI is from 45 kya and MA1 24 kya, so I'm not sure what you're talking about. UI's Y-DNA K-M526 and mtDNA R* are not only "basal" or from branches parallel to modern lineages, but also sit near the root of these major lineages (Y-DNA K-M526 and mtDNA R). This is not unexpected considering conventional age estimates (largely confirmed from analyses of branch shortening from ancient samples), as well as the fact that UI's Neanderthal segments form relatively large chunks, dating to roughly 10,000 years before he lived. So, his proto-Eurasian ancestors, who likely lived just before Y-DNA K-M526 and perhaps even mtDNA R existed, mixed with Neanderthals ~55 kya.
Sure, my bad. :(
DeleteBut still K2 could not have coalesced in or near Siberia, could it? It looks rather an erratic that traveled west with P1 and was later displaced by it (mere drift would explain that if P1, i.e. R1 and Q1 in that area, were originally dominant, as dominant lineages are statistically helped to become more dominant by drift mechanics, which are stronger in low density areas like Siberia).
"... as well as the fact that UI's Neanderthal segments form relatively large chunks, dating to roughly 10,000 years before he lived. So, his proto-Eurasian ancestors, who likely lived just before Y-DNA K-M526 and perhaps even mtDNA R existed, mixed with Neanderthals ~55 kya".
That part is intriguing indeed. Neanderthals lived in Altai for sure until maybe 33 Ka BP, so it's perfectly possible that UI remixed with Neanderthals at that location. Were there also other Neanderthal chunks that can be attributed to older dates, i.e. the original "pan-Asian" admixture event proposed by Reich 2010?
But still K2 could not have coalesced in or near Siberia, could it?
DeletePerhaps not, but Ust'-Ishim lived during a period of great mobility in Eurasia, so his ancestors may very well be recent migrants into the region.
That part is intriguing indeed. Neanderthals lived in Altai for sure until maybe 33 Ka BP, so it's perfectly possible that UI remixed with Neanderthals at that location. Were there also other Neanderthal chunks that can be attributed to older dates, i.e. the original "pan-Asian" admixture event proposed by Reich 2010?
I doubt their method can exclude the existence of earlier admixture. But they did find that the average Neanderthal segment in UI is rather large, and consistent with an admixture date of 55 kya. Their thinking was that the Neanderthal admixture in early proto-Eurasian dates to <60 kya, which fits with the recent Eurasian dispersal model that has been popular among geneticists for many years now.
Look, I know you've been a staunch critic of mutation rates for a long time. But even if we disregard the mutation rates, we now have a large amount of ancient samples (from Europe at least). And relatively recent lineages like R1b-M269 are clearly lacking in the early/mid-Neolithic heart of Europe. It's first seen in the Bronze Age steppes, with the more recent R1b-P312 appearing in later Europeans, who also happen to show an autosomal "steppe" shift. The natural conclusion would be that R1b-P312 ultimately derives from later migrations, reaching farther into Europe. Yet, from what I understand, you strongly disagree with this. So where do you propose R1b-P312 was supposedly hiding for all this time?
You may continue to think the aDNA data is "odd", but the sample sizes are growing ever larger, and eventually you'll need to reconcile the aDNA lineages with other lines of genetic evidence (like the mutation rates that have been confirmed by branch shortening data, Neanderthal admixture date, etc.), supporting more recent patterns of migration in general than what you've been advocating.
More is not better. Large samples in East Germany mean nothing regarding Portugal or France.
DeleteRe. R1b1a2a1a2 (P312/S116), the geostructure very strongly points to either Southern France or the Iberian Peninsula as origin. That's all I can say about this major haplogroup. So maybe it was hiding in Portugal or France.
Something we see in Chalcolithic Upper Ebro data is that there were quite apparently two distinct populations (one lactose intolerant and another one lactose tolerant) beginning to mix at that very time. We know that mainline Neolithic peoples of LBK/Cardium were lactose intolerant but we also know that British Neolithic peoples were heavy dairy consumers, so I'm all the time thinking in an Atlantic (Megalithic) zone as major contributor to Western and North-Central European genetics in the Chalcolithic and maybe also later in some areas.
So I say: sample Portugal, West France, Britain... make some of these wonderful ancient autosomal and Y-DNA studies also in Atlantic Europe, where the answer of the puzzle was hiding all the time almost certainly.
Whether R1b-L51 arrived to Atlantic Europe in the Neolithic (founder effect from West Asia with a trampoline in Portugal surely) or was already there in the Upper Paleolithic, I remain relatively agnostic (and have remained so since many years ago, because I can't trust "molecular clocks" at all, much less above the geostructure). Only wider pre-Neolithic sampling with Y-DNA results in the Magdalenian area can clarify (so far only two samples).
I want to emphasize this last: for some time (certain) people argued heavily about mtDNA H not being native to Europe and the results from Central Europe seemed to support this, while HVS-1 results from Portugal and other areas were dismissed as "unreliable". However since 2011-12 mtDNA research has demonstrated that there was indeed mtDNA H in North Iberia and Karelia before the Neolithic.
DeleteI believe that the same will likely happen with R1b-L51 when more samples from the right geographies are available. Partly because there is a relationship between these two haploid lineages in Western Europe and partly because the geostructure seems more supportive of pre-Neolithic expansions centered in the Franco-Cantabrian Region and Doggerland than of Neolithic/Chalcolithic expansions, which should be centered in places like Portugal or Brittany. But, as said above, I remain cautious.
Maju
DeleteU made some interesting comments elsehwere on the lengyel data . Care to elaborate on the profile vis-a-vis that from steppe sites; and what u meant by "hyper-modern" ?
Not sure about all the details of your question (what's "the profile vis-a-vis that from steppe sites") but on the notion of "hyper-modern": it's an mtDNA concept. Neolithic populations from Germany and other places had way too low levels of mtDNA H (~25%), what is clearly not "modern". On the other hand Neolithic Portugal (Chandler 2005) had LOTS of H, making it "hyper-modern" (i.e. a possible source for the correction towards more H up to the 40-60% modern levels), while Paternabidea (Neolithic Navarre) had a quite modern-like mtDNA pool (and roughly the same can be said of Neolithic and Chalcolithic "Basques" when pooled, but not so much when considered site by site, with the exception of Paternabidea).
DeleteThe new Cuyavian late LBK (not really Lengyel, which only reached to Southern Poland but something related anyhow) shows similar "hyper-modern" (very high) levels of mtDNA H (7/11, assuming we trust the HVS-1 assignment, which is not good enough for European mtDNA unless further tests are made) and suggests another possible source of mtDNA pool "modernization" (increase of H, decrease of "Neolithic" lineages like T, W, K1, N1a, etc.) That's about it.
My only other remark was that it's risky to rely too much on samples from only certain areas (particularly NE Germany has been hyped quite a bit), while much of ancient Europe remains unsampled.
Some relevant ancient mtDNA studies:
Delete→ http://forwhattheywereweare.blogspot.com/2013/09/basque-and-other-european-origins.html (my own)
→ http://forwhattheywereweare.blogspot.com/2014/08/chalcolithic-mtdna-from-atapuerca-still.html (a very similar pattern detected by Sánchez & Olalde)
Thanks
DeleteI meant "compared to steppe sites "
Krefter was saying that modern european mtDNA likeky dervived from yamnaya and catacomb culture groups ; in large part
Well, I said already in the main entry that: "An interesting detail is that there is no or nearly no mtDNA H within the Kurgan (IE) samples, strongly suggesting that their migration was largely male-biased, at least initially".
DeleteSo basically the mtDNA at least is not Kurgan. What else do you want me to say?
"Krefter was saying that modern european mtDNA likeky dervived from yamnaya and catacomb culture groups ; in large part".
His opinion. What is clear is that Corded Ware has a tendency towards Yamna and that such is a likely vehicle of part of the European Ma-1 tendency anomaly (rel. to neolithic and some epipaleolithic people) and detected already in Lazaridis 2014. However it's probable that Motala-like affinity is also responsible of the other half or so. Finally there's a factor of noise (uncertainty) in all these measures that, coupled with the lack of sufficient ancient samples for most regions, makes measuring ancestry fractions almost impossible.
So, his proto-Eurasian ancestors, who likely lived just before Y-DNA K-M526 and perhaps even mtDNA R existed, mixed with Neanderthals ~55 kya.
DeleteThere is now also a mtDNA R from a ~40 kya Protoaurignacian specimen from northern Italy. It is another basal R lineage that sits close to the R root, much like Ust'-Ishim's R* lineage. This new mtDNA sample has a slightly longer branch, however, and was estimated to be a couple of thousands of years younger than Ust'-Ishim's mtDNA lineage, which is consistent with the archaeological context.
We now have early R lineages from the period 40-45 kya spanning a geographic range from Europe (Fumane 2), to Siberia (Ust'-Ishim) and East Asia (Tianyuan, who belonged to a basal lineage within the B sublineage). Clearly indicating a relatively recent common ancestry for these groups before 45,000 years ago, prior to the wide dispersal of anatomically modern humans in Eurasia that is seen in the fossil record.
Good to see you back online my friend
ReplyDeleteThank you but as you can see, my activity is most limited. Getting too old and grumpy to put up with humankind-as-it-is, if you get my point.
Deletehttp://www.sciencedaily.com/releases/2015/03/150304075334.htm
ReplyDeleteOf what interest is that link? I do read SD often but this is the ranscript of the press release (as SD articles usually are) of the paper under discussion here, so what 's the point?
DeleteThe "Future research" bit on Iberia ?
ReplyDeleteAh, ok. It's better to explain your point when you post a link, rather than a naked link.
DeleteI hope so but now I'm even more interested in ancient British and Irish samples. That SHG tendency and the practical impossibility of 100% replacement, as claimed by "ANE alchemists", make the matter very intriguing.
I agree !
DeleteMaju
DeleteAs I stated elswhere, I too am suspect of the figures and narrative generally advanced for northern and western Eruope.
i also agree that using ANE is taking a step back, rather than forward. In fact, I think David's K15 analysis which breaks up the ancient genomes into 'Atlantic', "EE", "North Sea", "West Asian", etc if more informative.
If you look at the graph I made of Davids K15:
https://drive.google.com/file/d/0B1vtTHobiXwVY0dqSGxzZ3g4d0E/view?usp=sharing
Then it is clear, IMO, that Labschour, La Brana and esp Motala already posessed some "EE" and "baltic" components which otherwise characterise Samar, Karelia, and later Yamnaya. The only truly novel components are the "West Asian' and 'south Asian" ones.
What this means in the big scheme of things requires clarification.
It's difficult to reach to conclusion based on "zombie" components taken from modern people who make ancestors seem our descendants rather than the other way around, and worse: mediated by artificious components and not free running ones.
DeleteHowever I do see your point: actually the BB and CW samples are almost identical in that graph to Braña/Lochsbour but with "something else" as admixture: mostly West Med in BB (attributable to EEF/ENE/MNE) and West Asian in CW (attributable to the Yamna roots: Caucasus-Zagros component). So CW look Yamna but BB look HG with some ENF (with one exception).
Anyhow, I'm thinking that the global K=16 that is "optimal" in the global sample is just equivalent to K=3 in the European sample (blue-orange-dirty green), which is clearly much suboptimal. Even K=20 is just equivalent in Europe to K=4, what can only be a draft.
Ryan, as you are interested in interaction between America and Eurasia, you are surely interested in this very recent paper on Greenland Inuit yDNA: Peopling of the North Circumpolar Region – Insights from Y Chromosome STR and SNP Typing of Greenlanders http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0116573
ReplyDeleteThey say that ”Two Y-chromosomal lineages, Q-M3 (xM19, M194, L663, SA01 and L766) and Q-NWT01 (xM265) were found to be prominent in Greenlanders. The Q-M3 lineage has previously been found among Native American populations throughout North and South America, whereas Q-NWT01 (xM265) was previously reported at high frequency in an Inuit population of the Canadian Northwest Territories.
The estimated TMRCA of the Q-M3 (xM19, M194, L663, SA01 and L766) lineage in Greenlanders is lower than that estimated for the same lineage among the Inuit populations of the Canadian Northwest Territories. The same was observed for the haplotype diversity and the MPD. These observations are consistent with the theory that the derived M3 variant appeared in Beringia or Alaska and that individuals carrying the derived stated at M3 migrated towards the East across North American and eventually reached Greenland. It is likely that individuals carrying the derived state at M3 are descendants of the Thule culture, which has its origin in North Alaska.
However, the older TMRCA of the Q-NWT01 (xM265) lineage in Greenland than in the Canadian Northwest Territories could indicate that the Q-NWT01 (xM265) lineage spread from an eastern region of North America towards the north and west. It is possible that the Q-NWT01 (xM265) lineage originated in individuals who later developed into the Dorset culture. The detection of Y-HGs of possible Dorset origin in the current Inuit populations would indicate that there was interbreeding between the individuals of the Dorset and Thule cultures. In light of this hypothesis, a back migration of the Q-NWT01 (xM265) lineage and/or its derived lineages into Asia needs to be considered as the Q-M120 lineage (see Figure A in S1 File) is wide spread at low frequencies in Asian populations.
Overall, our results are in agreement with previous studies, indicating that the populations of the North Circumpolar Region, including Greenland, share a common origin with Native American populations. However, the male Inuit population in Greenland appears to be a mixture of individuals belonging to two Y-chromosomal lineages within Y-HG Q-M242. Males carrying the derived state at the M3 locus appear to stem from an expansion of humans from the northwest region of North America. Most likely, these individuals are descendants of the Thule culture that reached Greenland approximately 800 y. a. Individuals carrying the derived state at the NWT01 locus may have been part of the same expansion [22]. However, our results indicate that the Inuit individuals of the Q-NWT01 (xM265) lineage may be descendants of the Dorset Paleo-Eskimo culture. Therefore, our results points to the possibility that there has been gene flow between the Dorset and Thule Inuit and that the current male Inuit population, at least in Greenland, bears traces of both Thule and Dorset descent.”
New information on Q-L54 lineage is available also in this Russian paper http://rjgg.molgen.org/index.php/RJGGRE/article/view/150/174 . Figure 1 (Рисунок 1.) and Table 2 (Таблица 2.) present the branching and age estimates of Anzick lineage and its related lineages, including Q-L330 which is typical of Kets.
ReplyDelete