The Seima-Turbino Phenomenon

[Zelto]

Aside from I6717, none of the Krasnoyarsk_mlba samples seem to want KRA001 ancestry, but those all date inbetween 1900-1500 iirc. Same thing with the Karasuk samples which are available G25. I get 5% on RISE495 for example, compared to the ~40% of Siberian ancestry from other populations in that sample. With the rest its 0.

I checked the supplementary, I6717 was buried in a kurgan with other Krasnoyasrk_MLBA samples. Nothing extraordinary about him in particular. I wonder if his elevated kra001 is just the result of admixture from a population farther north? Is the "western" WSHG ancestry present in other Krasnoyasrk_MLBA samples?

Karasuk isn't generally held to have been derived from Fedorovo. it has gone from being attributed to "north mongoloids" to being a Siberian Andronovo-influenced culture, to being considered to be an Andronovo derivative. A lot of this is due to divergent ceramic traditions, Kuzmina distinctively called them genetically/ethnically unrelated to Andronovo because of this and even tried to hint at a Turkic connection due to the pottery practises having similarties with those of Siberian Turkic ethnic groups. Another issue is that with the old chronology you only had Fedorovo in Siberia from about 1500 BC to 1200 BC, and this was held to be the only Andronovo migration into the region. Nowadays more archaeolgists seem to connect the early Karasuk sites with Alakul, which would make sense with the Alakul expansion in the central Asian steppes in the 2nd millenium b.c, and would provide a nice linguistic connection between the Andronovo groups in Southwestern Central Asia.

I'll make note of that, I've seen Andronovo/Fëdorovo in connection to Karasuk, even in fairly recent publications. Here for example, the chronology goes from Andronovo(Fëdorovo) to Karasuk:
https://journals.uair.arizona.edu/in...File/3487/3003

[CopperAxe]

I checked the supplementary, I6717 was buried in a kurgan with other Krasnoyasrk_MLBA samples. Nothing extraordinary about him in particular. I wonder if his elevated kra001 is just the result of admixture from a population farther north? Is the "western" WSHG ancestry present in other Krasnoyasrk_MLBA samples?

THe hard part is that because we have such little from the far north its kind of hard to establish what kind of population clines there were going on. In South Siberia you can sort of see from Altai to the western shores of lake Baikal you have a noticeable WSHG -> ESHG cline, but who knows what it ws like further north in the Taiga? By western 'WSHG" I mean a pure WSHG profile a la Sosonivoy or Tyumen.

Most of these Krasnoyarsk samples seem purly Andronovo or with a small amount of WSHG related stuff. For I1617 even Rise677 (Okunev sample without any Afanasievo or central asian admix) wasn't a preferred option over that WSHG+Kra001 combo and that sample is close to being 50% WSHG 50% Baikal_N. And also came from a bit further north on the Yenisei than the other Okunev samples. So one pathway would be that if you go north enough you had WSHG populations without Baikal_N type admix. Another would be that a Kra001 population travelled along the Ob and came across people with a proper WSHG profile in the Taiga region, and admixed with them. THen that mixed offspring mixed with steppe_mlba people further south. Or a different pathway such as Kra001+steppe_mlba+wshg or wshg+steppe_mlba+kra001.

Btw that site (Orak) contains loads of burials, also from different periods and apparently not everything was properly labeled either. Did it share a grave with other samples? reburials in kurgans were quote common for example.

To contextualize what I was babbling about earlier here I have the average of Krasnoyarsk, Karasuk and their respective outliers, Rise677, as well as the Mansi, Nenets and three of the Ket samples from Altvred (which are far better than those on G25) in a messy G25 sheet:

 

In particular pay attention to the difference between Sosonivoy, Munkhairkhan, Baikal_N and Baikal_BA ancestry. It would be interesting to see if that same pattern holds up for that Karasuk sample.

I'll make note of that, I've seen Andronovo/Fëdorovo in connection to Karasuk, even in fairly recent publications. Here for example, the chronology goes from Andronovo(Fëdorovo) to Karasuk:
https://journals.uair.arizona.edu/in...File/3487/3003

The chronology doesn't imply that Karasuk is derived from Fedorovo, but that it succeeds it. Which is pretty much what happened. Depending on which author you read, some envision the Karasuk as having driven out the earlier tribes and/or assimilated what was left of them, others invisioned the Karasuk as an invasion of Asiatic nomads who ultimately were dissolved into the either Iranic or "Dinlin" populations.

[Zelto]

THe hard part is that because we have such little from the far north its kind of hard to establish what kind of population clines there were going on. In South Siberia you can sort of see from Altai to the western shores of lake Baikal you have a noticeable WSHG -> ESHG cline, but who knows what it ws like further north in the Taiga? By western 'WSHG" I mean a pure WSHG profile a la Sosonivoy or Tyumen.

There are a lack of samples for sure. But we know kra001-like populations must have crossed the Taiga in an unadmixed form, eventually ending up in West Siberia and maybe even Europe. I also don't think kra001 is directly ancestral to the group he proxies in Uralics, so some of this ancestry must have stayed behind in Krasnoyarsk (and presumably farther north). A band of pure kra001 across the Taiga seems too simplistic, but I guess we'll have to wait and see. The Arctic/sub-Arctic zones where the westward spread of Ymyyatkh is well documented, almost certainly carried this ancestry though.

Most of these Krasnoyarsk samples seem purly Andronovo or with a small amount of WSHG related stuff. For I1617 even Rise677 (Okunev sample without any Afanasievo or central asian admix) wasn't a preferred option over that WSHG+Kra001 combo and that sample is close to being 50% WSHG 50% Baikal_N. And also came from a bit further north on the Yenisei than the other Okunev samples. So one pathway would be that if you go north enough you had WSHG populations without Baikal_N type admix. Another would be that a Kra001 population travelled along the Ob and came across people with a proper WSHG profile in the Taiga region, and admixed with them. THen that mixed offspring mixed with steppe_mlba people further south. Or a different pathway such as Kra001+steppe_mlba+wshg or wshg+steppe_mlba+kra001.

What I was getting at, is if all the Krasnoyarsk_MLBA samples have "western" WSHG, there is nothing to suggest the kra001-like in I16717 derives from the west. However, the elevated Sosonivoy _HG in I16717 could indeed hint at a migration from from the west, like you said.

Btw that site (Orak) contains loads of burials, also from different periods and apparently not everything was properly labeled either. Did it share a grave with other samples? reburials in kurgans were quote common for example.

The supplementary is very vague and poorly organized, IMO. The site consists of several kurgans with square stone enclosures on top. Most burials are classified as Fedorovo or Karasuk. It also says, the number of burials in the museum catalog and publications are inconsistent.

I3389, I3390, I3391, I3392, I6718 and I6716 all have individual gender/age descriptions but no other archeological context is given. I3393, I3394 and I3395 all belong to kurgan 2 (Andronovo). I3396 was found in a different kurgan "kurgan on the swamp" and finally, I6716 is listed with his date and gender. No further information is given and it is presented in that order.

The chronology doesn't imply that Karasuk is derived from Fedorovo, but that it succeeds it. Which is pretty much what happened. Depending on which author you read, some envision the Karasuk as having driven out the earlier tribes and/or assimilated what was left of them, others invisioned the Karasuk as an invasion of Asiatic nomads who ultimately were dissolved into the either Iranic or "Dinlin" populations.

I'm not disagreeing, but in some areas (like Orak), there is direct succession between the two, in the same burial sites. With that being said, Karasuk has been seen as related to Fedorovo in the past, perhaps anachronistically. Here is an example:

They seem to have come southern Central Asia with the Yaz I culture (fig. 34) and to southern Siberia in the 13th century BC in the closing Fedorovo phase of the Aodronovo cultural complex. Here the Andronovo culture was succeeded by the at least partly genetically related Karasuk culture (1200-1000 caIBC), which flourished around the upper Yenisei, Mongolia and the Ordos region of China.

https://www.sgr.fi/fi/files/original...2353038397.pdf

In the future I'll just say Andronovo to avoid confusion.

[Huck Finn]

But we know kra001-like populations must have crossed the Taiga in an unadmixed form, eventually ending up in West Siberia and maybe even Europe. I also don't think kra001 is directly ancestral to the group he proxies in Uralics, so some of this ancestry must have stayed behind in Krasnoyarsk (and presumably farther north). A band of pure kra001 across the Taiga seems too simplistic, but I guess we'll have to wait and see. The Arctic/sub-Arctic zones where the westward spread of Ymyyatkh is well documented, almost certainly carried this ancestry though.

East Asian looking Avars might be an example of a group which indeed quickly arrived into Europe in an unadmixed form. But, a more slow expansion through Arctic or possibly even Sub Arctic should according to my understanding probably result in a mixture with ANE type of features only, leading into something like BOO. Ancient extreme North, just like West Siberia before the migrations from the West, was an ANE biased country.

Slide 25/27 here https://bedlan.net/wp-content/upload...e-siivottu.pdf shows at K=9 the so called Khanty-Mansi component (enriched especially in Khantys). It is a complex mixture of both eastern and western (i.e. Non Siberian, eastern or western) features and probably a good candidate for the Proto Uralic marker. Then there's the question related to kra001/Nganasan. The same slide 25/27 at K=6 shows that Nganasan are basically East Asians with some western features, including the Khanty-Mansi component. I'd guess that Pre Proto Uralic speakers were East Asians but unlike Samoyedic speaking Nganasans, most East Asians did not mix with the sort of back migrating Proto Uralic speakers. In other words: Nganasan are a good proxy for the eastern features of the Uralic speakers because a. they are East Asians like the Pre Proto Uralic speakers were but also because b. they have some more western Proto Uralic genes and hence also the language based on Proto Uralic. K=6 shows that the East Asian of Nganasan is the same North East Asian that fex Yakuts (having most N1c in the world, in relative terms) have. Kra001 shouldn't basically have any Khanty-Mansi component yet, but maybe there's something in the mixture (mostly East Asian, some West Siberian) that makes it suitable, even without any Non Siberian in it?

Slide 16/27 shows the probable location of Proto Uralic homeland, still around Ural mountains, though the leeway is admittedly quite extensive. The Samoyedic homeland is BTW worth noticing, also the apparent rather southern expansion channel of Uralic. Pre Proto Uralic, different question, but maybe somewhere around the upper reaches of Ob and Irtysh?

[Zelto]

East Asian looking Avars might be an example of a group which indeed quickly arrived into Europe in an unadmixed form. But, a more slow expansion through Arctic or possibly even Sub Arctic should according to my understanding probably result in a mixture with ANE type of features only, leading into something like BOO. Ancient extreme North, just like West Siberia before the migrations from the West, was an ANE biased country.

I don't think PU or the N-L1026/kra001-like in Uralic was spread via the Arctic. However, I'm almost certain this ancestry formed out of a very specific admixture event north of Baikal. The Steppe, Taiga and Arctic zones were all home to non-kra001-like populations prior to their expansion and the WSHG in BOO was probably only picked up in West Siberia, for the same reason CopperAxe believes I16717 came from the west.

A Mongolian orgin isn't very convincing to me either. N1c has been found in BA Mongolia, however it is the wrong subclade. This could change once the Avar samples are analyzed by YFull, but right now N-Y16323 is split between N-B202 (TMRCA 2500ybp) found in Chukchi and N-F4205 (TMRCA 2300) found throughout the Steppe, peaking in the east (especially in Buryats). These TMRCAs are too young to suggest anything meaningful for N-L1026*. If anything, the geographic center between Buryatia and Chukotka would be along the Lena.

Slide 25/27 here https://bedlan.net/wp-content/upload...e-siivottu.pdf shows at K=9 the so called Khanty-Mansi component (enriched especially in Khantys). It is a complex mixture of both eastern and western (i.e. Non Siberian, eastern or western) features and probably a good candidate for the Proto Uralic marker. Then there's the question related to kra001/Nganasan. The same slide 25/27 at K=6 shows that Nganasan are basically East Asians with some western features, including the Khanty-Mansi component. I'd guess that Pre Proto Uralic speakers were East Asians but unlike Samoyedic speaking Nganasans, most East Asians did not mix with the sort of back migrating Proto Uralic speakers. In other words: Nganasan are a good proxy for the eastern features of the Uralic speakers because a. they are East Asians like the Pre Proto Uralic speakers were but also because b. they have some more western Proto Uralic genes and hence also the language based on Proto Uralic. K=6 shows that the East Asian of Nganasan is the same North East Asian that fex Yakuts (having most N1c in the world, in relative terms) have. Kra001 shouldn't basically have any Khanty-Mansi component yet, but maybe there's something in the mixture (mostly East Asian, some West Siberian) that makes it suitable, even without any Non Siberian in it?

Kra001-like alone, seems to be the common denominator (see altvred's recent posts). Either way, the K9 component clearly carries a lot of kra001-like in it. Look at how it mirrors the yellow components in K3 and K6 among Mansi in that slide. This is also the ancestry N-L1026* probably carried judging by kra001 and the Ymy samples.

Slide 16/27 shows the probable location of Proto Uralic homeland, still around Ural mountains, though the leeway is admittedly quite extensive. The Samoyedic homeland is BTW worth noticing, also the apparent rather southern expansion channel of Uralic. Pre Proto Uralic, different question, but maybe somewhere around the upper reaches of Ob and Irtysh?

Honestly, the correlation between modern Uralic speakers as a whole and N-L1026 isn't great. Kra001-like admixture also probably reached modern Uralics in multiple waves. Rather, Uralic speakers universally have elevated frequencies of N-L708 and N-P43 lineages, but N-L1026 specifically, is fairly low in Samoyedics, Komi, Udmurts, Mordvin, Khanty, etc. who all have higher frequencies of the aforementioned (N-L1026x) subclades.

With that being said, N-L1026 has been connected to the spread of Seimna-Turbino in multiple studies. Proto-Uralic has also been connected to Seima-Turbino in multiple studies. It follows logically then, that there may have been a connection between N-L1026 (and thus kra001-like ancestry) and Proto-Uralic. When/where PPU ends, when/where PU begins and when/where/how they entered the Seima-Turbino network are things I don't think anyone can answer now.

Here are my arguments against an Arctic expansion:

1. PU has a lexicon that most closely resembles southern Taiga HGs.

2. There were probably not any PIIs in the Arctic.

3. Even BOO probably has some southern admixture, partially evident by his mtDNA.

4. Ymyyakhtakh sample N4b2 (4091-4348ybp) has a related, but non-ancestral subclade https://www.yfull.com/tree/N-Z1979/

5. I'm not aware of a major demographic decline in the Russian forest-zone cultures between 1900-1000BC. Maybe in the Steppe, but that seems irrelevant to N-L1026. Even in the event of a demographic decline in the Taiga zone, I don't see how their population could have ever been any lower than that of their Arctic contemporaries.

6. No social/economic/technological way of explaining the success of N-L1026 in the Taiga zone.

7. N-L1026 rumored in Rostovka and kra001 found well below the Arctic zone.

[Kale]

Unfortunately this one isn't available on G25. Is it similar to Krasnoyarsk_MLBA_o I6717 by any chance? That sample was a bit younger than the others, from around 1400 BC which would be contemporary to Karasuk and has quite some kra001 related ancestry as well.

Whats also weird about I6717 is that it WHSG-ESHG ancestry seems very western, rather than something more proximate (Okunev, Glazkovo, Khovgol, Munkhairkhan etc.) which you also see amongst Siberian Uralics. But this is just on G25 and I wonder if qpadm replicates this as well.

Target: RUS_Krasnoyarsk_MLBA_o:I6717
Distance: 2.0158% / 0.02015779
44.8 RUS_Sintashta_MLBA
29.2 RUS_Sosonivoy_HG
26.0 RUS_Krasnoyarsk_BA
0.0 MNG_Munkhkhairkhan_MBA_1
0.0 RUS_Baikal_BA
0.0 TKM_Gonur1_BA_o

With F3 I'm getting about 38% Steppe_MLBA, 29% Something very intermediate between WSHG and EHG, 21% Serovo, 12% Uralic

The "Nganasan-like" component is probably the result of fairly recent admixture, between Trans-Baikal_N and the Neolithic Lena population. This event itself, is the basis of the Ymyiakhtakh culture. I would wager, the Neolithic Lena population (Belkachi culture), is where this slight increase in ANE comes from.
It would be interesting to see if ~3000BC suggested by the C14 dates of early Ymyiakhtakh materials, is substantiated by ALDER/DATES.
Back to Seima-Turbino, I found a more recent article I had lost, which has an allusion to the "East Siberians".
https://www.sgr.fi/sust/sust265/sust265_chernykh.pdf

Interestingly enough, the 'Kra001 cluster' has representatives from geographically close to Kolyma_meso and to 'Russia_LenaRiver_MiddleN.SG', and has lower ANE than both. A recent migration from Transbaikal wouldn't account for that either because the Kitoi-Serovo transition ~4500-4000BC was accompanied by an uptick of ANE, putting them also higher than the Kra001 cluster.

I know some of the proposed origins of the Karasuk culture are quite complex, however there is usually some resurgent local element described, alongside the dominant Fëdorovo component. I remember a few Karasuk samples having at least some "Nganasan-like" admixture. That was quite a while ago though and probably in G25.
Rise493 https://www.yfull.com/tree/Q-YP832/ and I6717 https://www.yfull.com/tree/Q-F2019/ belong to old/divergent branches. They don't seem to be related (in their uniparentals at least).

Karasuk samples have fairly diverse eastern ancestry. I can see 5 major Siberian clusters potentially coming into play here...
'Kitoi' = Baikal_EN, Mongolia_N, ZabayKalsk_meso/N
'Serovo' = Main cluster ~4000-2000BC Baikal
'Uralic' = Kra001 & co.
'CentralSteppe' = Kazakhstan_Central_Steppe_EMBA.SG (they actually have unique internal drift)
'ANE-rich' = Botai, WSHG

Here's the breakdown for each Karasuk sample. Note in small quantities Serovo can be deconstructed as Kitoi + ANE-rich.
RISE497: 64% Serovo, 26% Kitoi, 10% ANE-rich - exact match to RUS_Irkutsk_UstIda_LN.SG
RISE495: 47% Steppe_MLBA, 42% Serovo, 9% CentralSteppe, 2% Central Asian
RISE496: 71% Steppe_MLBA, 17% CentralSteppe, 12% Serovo
RISE493: 42% Steppe_MLBA, 23% ANE-rich, 19% Uralic, 8% Serovo, 6% Kitoi, 2% Central Asian
RISE502: 53% Steppe_MLBA, 19% ANE-rich, 15% Kitoi, 10% Serovo, 3% Central Asian

RISE494: 83% Steppe_MLBA, 4% CentralSteppe, 3% ANE-rich
RISE499: 78% Steppe_MLBA, 10% Kitoi, 8% ANE-rich, 4% CentralSteppe
These two actually have noticeable Steppe_EBA also, the CentralSteppe & ANE-rich also suggest Okunevo input.

[Huck Finn]

Interestingly enough, the 'Kra001 cluster' has representatives from geographically close to Kolyma_meso and to 'Russia_LenaRiver_MiddleN.SG', and has lower ANE than both. A recent migration from Transbaikal wouldn't account for that either because the Kitoi-Serovo transition ~4500-4000BC was accompanied by an uptick of ANE, putting them also higher than the Kra001 cluster.

This is indeed an interesting observation. According to Kilinc et al North East Asians were expanding into ANE based areas in the West and North already since Neolithic and in the BA. Also, the location of kra001, in this respect ie. having for some reason a limited amount of ANE, is rather western. I wonder if both Lena and Kolyma are reflections of Ymyakhtak expansion, kra001 then being a related but parallel expansion?

[Zelto]

Interestingly enough, the 'Kra001 cluster' has representatives from geographically close to Kolyma_meso and to 'Russia_LenaRiver_MiddleN.SG', and has lower ANE than both. A recent migration from Transbaikal wouldn't account for that either because the Kitoi-Serovo transition ~4500-4000BC was accompanied by an uptick of ANE, putting them also higher than the Kra001 cluster.

I'm confused. The Kitoi-Serovo/Glazkovo transition occurred in Cis-Baikal. The Trans-Baikal samples with N-L708 (like brn008 and brn003), were only 0.8-3.0% ANE according to your post on page 10.

[Kale]

I'm confused. The Kitoi-Serovo/Glazkovo transition occurred in Cis-Baikal. The Trans-Baikal samples with N-L708 (like brn008 and brn003), were only 0.8-3.0% ANE according to your post on page 10.

O my bad, sorry I didn't realize the 'Trans' in Trans-Baikal meant 'beyond', I thought it meant 'circum'. I suppose that can't be ruled out then, though I'm not aware of any samples 4000-2000BC from the area.

[tipirneni]

With F3 I'm getting about 38% Steppe_MLBA, 29% Something very intermediate between WSHG and EHG, 21% Serovo, 12% Uralic



Interestingly enough, the 'Kra001 cluster' has representatives from geographically close to Kolyma_meso and to 'Russia_LenaRiver_MiddleN.SG', and has lower ANE than both. A recent migration from Transbaikal wouldn't account for that either because the Kitoi-Serovo transition ~4500-4000BC was accompanied by an uptick of ANE, putting them also higher than the Kra001 cluster.



Karasuk samples have fairly diverse eastern ancestry. I can see 5 major Siberian clusters potentially coming into play here...
'Kitoi' = Baikal_EN, Mongolia_N, ZabayKalsk_meso/N
'Serovo' = Main cluster ~4000-2000BC Baikal
'Uralic' = Kra001 & co.
'CentralSteppe' = Kazakhstan_Central_Steppe_EMBA.SG (they actually have unique internal drift)
'ANE-rich' = Botai, WSHG

Here's the breakdown for each Karasuk sample. Note in small quantities Serovo can be deconstructed as Kitoi + ANE-rich.
RISE497: 64% Serovo, 26% Kitoi, 10% ANE-rich - exact match to RUS_Irkutsk_UstIda_LN.SG
RISE495: 47% Steppe_MLBA, 42% Serovo, 9% CentralSteppe, 2% Central Asian
RISE496: 71% Steppe_MLBA, 17% CentralSteppe, 12% Serovo
RISE493: 42% Steppe_MLBA, 23% ANE-rich, 19% Uralic, 8% Serovo, 6% Kitoi, 2% Central Asian
RISE502: 53% Steppe_MLBA, 19% ANE-rich, 15% Kitoi, 10% Serovo, 3% Central Asian

RISE494: 83% Steppe_MLBA, 4% CentralSteppe, 3% ANE-rich
RISE499: 78% Steppe_MLBA, 10% Kitoi, 8% ANE-rich, 4% CentralSteppe
These two actually have noticeable Steppe_EBA also, the CentralSteppe & ANE-rich also suggest Okunevo input.

Do you have f3 with Indian population both Swat as well as Roopkund or moderns. the RISE493 seems to be enriched in vast majority of people in India including myself. on 1cM I was getting 14.2% match which goes to 68,8% at .25cM