The Gedrosia component and the origin of R1b M269

[spongetaro]

I was recently looking at the K12a spreadsheet for the participants of the Dodecad proect and I've noticed that R1b frequencies and the Caucasus admixture are totally unrelated in Northwestern Europe. If the proto IE carrying R1b M269 or R1b L11 really originated around the Black sea area we would expect a higher rate of Caucasus component than the Gedrosia one for the British or Basque people for instance. But the Irish in the Dodecad project have 1,5% of Caucasus and 9,5% of Gedrosia and the French Basque have 0,1% of Caucasus vs 7,6% of Gedrosia.
As we know, the Basque do have really low frequencies of "neolithic haplogroup" such as G2a, EV13, J1, T etc. we can conclude that the Caucasus component (only 0,1% among French Basque but 22,3% for Otzi) only represents the neolithic legacy of men carying G2a, EV13, J1 etc.
So we have the caucasus admixture related to Neolithic farmers and R1b somewhat related to the Gedrosia admixture in Northwestern Europe (and in the Basque country).
It really looks like R1b men came to Europe directly from an area East of the Caspian sea without settling around the Black sea area (Maykopp etc) .

 

[ElHorsto]

Very interesting, it makes some sense to me. But still there are problems:
- NW-Europe is still a stronghold of neolithic mtDNA K. Neolithic Ötzi had K too.
- Basques and Sardinians (Ötzi) were supposed to be related, but this seems less likely now.
- The basque origin is probably neither IE or neolithic then, further increasing the mystery of basque origin.

 

[Knovas]

Both K12a and K12b experiments show more noise than ancestry IMO. There's even little or no correlation with the K7b experiment, which was developed at the same time and mathes better what was done before (also present experiments by the Eurogenes Project). There we see for example that Sardinians are 0% West Asian, while the Caucasus component it's over 20% in the others. Honestly, I wouldn't take any conclusions based on this.

 

[spongetaro]

Very interesting, it makes some sense to me. But still there are problems:
- NW-Europe is still a stronghold of neolithic mtDNA K. Neolithic Ötzi had K too.

Otzi was at least partly "mediterranean" since he was 57.7% "Atlantic_Med" (more than its 22,3% caucasus). Northwest european like British people in the K12a experiment shows around 40% of mediteranean admixture. mtDNA K could be an evidence to this mediterranean legacy (or "Atlanto med") as opposed to Ydna G2a which I think represents the Caucasus admixture.

 

[spongetaro]

Both K12a and K12b experiments show more noise than ancestry IMO. There's even little or no correlation with the K7b experiment, which was developed at the same time and mathes better what was done before (also present experiments by the Eurogenes Project). There we see for example that Sardinians are 0% West Asian, while the Caucasus component it's over 20% in the others. Honestly, I wouldn't take any conclusions based on this.

We still have some regularities like the Basque with almost no Caucasus and almost no West Asian admixture which once again reflects the absence of Ydna G2a, EV13, J1 etc.
You probably know more than me in the field of autosomal studies but I just find problematic to place the origin of Proto IE people and the european branch of R1b in the Black sea region when population with high levels of R1b (Basque and British people especially) show almost no Caucasus admixture.

 

[Taranis]

We still have some regularities like the Basque with almost no Caucasus and almost no West Asian admixture which once again reflects the absence of Ydna G2a, EV13, J1 etc.
You probably know more than me in the field of autosomal studies but I just find problematic to place the origin of Proto IE people and the european branch of R1b in the Black sea region when population with high levels of R1b (Basque and British people especially) show almost no Caucasus admixture.

The question is, what does this make of the Basques if they have virtually no Caucasus/West Asian admixture (both usually taken as a signature of the Neolithic farmers)? I mean, essentially things are all the more mysterious. We might interprete it this way that the Basques must be either older (ie, Mesolithic), or they arrived even later (Copper Age / Bronze Age)?

 

[spongetaro]

The question is, what does this make of the Basques if they have virtually no Caucasus/West Asian admixture (both usually taken as a signature of the Neolithic farmers)? I mean, essentially things are all the more mysterious. We might interprete it this way that the Basques must be either older (ie, Mesolithic), or they arrived even later (Copper Age / Bronze Age)?

As you say, it reduces the solution to two options. The mesolithic one seems quite risky, it would mean that I2a people (which I think are the pre IE ancestors of Basque, as showed by the high level of med admixture) were "hidding" somwhere during the neolithic.

 

[Taranis]

As you say, it reduces the solution to two options. The mesolithic one seems quite risky, it would mean that I2a people (which I think are the pre IE ancestors of Basque, as showed by the high level of med admixture) were "hidding" somwhere during the neolithic.

Yeah, I see your point. I certainly agree on the assessment that the Mesolithic option is risky. There's also the linguistic perspective here: if we remove all Indo-European borrowings, we get essentially a Copper Age or Bronze Age language. If Basque really was Paleolithic, this would mean not only that a hunter gatherer society survied the Neolithic and the Bronze Age, but also that Basque borrowed all the words for domesticated animals, for crop plants/agriculture, as well as for metals and metal-working from non-IE sources. That's quite a stretch in my opinion. The flip side, which in my opinion speaks against a later introduction is this: Basque is obviously a language isolate, with no clearly demonstrable link to other language families (Iberian, perhaps? but this is disputed too), and this in turn supports the ancientness of the language.

 

[Knovas]

Ok ok, I partly agree. My point is not to use only this autosomal run to support the argument, it's better to check as much experiments as possible. The Basques are regular in the absence of West Asian/Caucasus components compared to other runs, but no way the British and Irish. Those populations usually show 5-10% West Asian/Caucasus, so they clearly don't match what you say spongetaro exept for a single run.

I recommend to check the last experiments by the Eurogenes Project, they could clarify some things. There are some West Eurasian genetic plots as well, which are very ilustrative.

 

[ElHorsto]

Otzi was at least partly "mediterranean" since he was 57.7% "Atlantic_Med" (more than its 22,3% caucasus). Northwest european like British people in the K12a experiment shows around 40% of mediteranean admixture. mtDNA K could be an evidence to this mediterranean legacy (or "Atlanto med") as opposed to Ydna G2a which I think represents the Caucasus admixture.

Indeed, the elevated mtDNA K does not seriously question your hypothesis. It rather decreases the likelyhood of mtDNA K being neolithic. That in turn fits to my theory that a good part of "Mediterranean" is actually older than neolithic. Initially I thought that a good part of Y-HG I represents "Mediterranean" too, but Sparkey's Paleolithic Remnants map proved this to be unlikely, unfortunately.
BTW, would be great to have a Bayesian Networkat hand for all these theories.

 

[ElHorsto]

... I just find problematic to place the origin of Proto IE people and the european branch of R1b in the Black sea region when population with high levels of R1b (Basque and British people especially) show almost no Caucasus admixture.

If you are right, then the nearly absence of "East-European" (Maciamo's map) in Iberia, Britain and Basque should represent a similar problem then.

I'm not familiar with the autosomal analysis methods, just looking at Maciamos automsomal maps (f.i. "NW-European"). It looks like the Caucasus is a region with skyrocketing genetic diversity and strong genetic boundaries between the peoples, probably as a result of the high mountains. For instance there is plenty of "NW-European" north of the Caucasus (30%-40%) in Russia, but south of the mountains in Turkey it suddenly drops to 5%-10%. Even more so with "East-European". The "Caucasian" Y-HG G is also very unevenly distributed even among Caucasian peoples. And also founder effects should be considered for autosomals, I think. The "NW-European" for instance might represent only a part of the original (centum?) IE, eventually later becoming a unique "NW-European" cluster only by admixture from "Atlanto-Med", for instance. Maybe original IE tribes were much more diverse and only a subset went to certain places of Europe.

 

[spongetaro]

If you are right, then the nearly absence of "East-European" (Maciamo's map) in Iberia, Britain and Basque should represent a similar problem then.

The absence of "East European" admixture is indeed a big challenge for the "post neolithic arrival of R1b" theory. It clearly speaks in favour of the paleolithic continuity unless "East European" is a relatively new admixture that formed after the Kurgan expansion into europe. The other explanation would be that R1b came indeed during the bronze age in Europe and spread without affecting the genome of the population (founder effect) hence the lack of East european admixture in western Europe.

The "NW-European" for instance might represent only a part of the original (centum?) IE, eventually later becoming a unique "NW-European" cluster only by admixture from "Atlanto-Med", for instance. Maybe original IE tribes were much more diverse and only a subset went to certain places of Europe.

The North West European and North European admixtures are obviously a mix of the Paleolithic and IE genetic legacy since it peaks in non R1b area like Scandinavia and Finland.

 

[ElHorsto]

The absence of "East European" admixture is indeed a big challenge for the "post neolithic arrival of R1b" theory. It clearly speaks in favour of the paleolithic continuity unless "East European" is a relatively new admixture that formed after the Kurgan expansion into europe. The other explanation would be that R1b came indeed during the bronze age in Europe and spread without affecting the genome of the population (founder effect) hence the lack of East european admixture in western Europe.

Maybe things are different. Let's once assume that South-West-Europeans population is still primarily of palaeolithic origin (i.e. K12b "Atlantic-Med") and R1b abundance is just a founder effect from Bronze age: Then there might be a similar pattern in eastern europe with regard to R1a:
According to the K12b table, the "North-European" admixture peaks in Finland and the baltic countries (70%), but is also second highest in "Mixed_Slav" (64%), whereas "Mixed_Germanic" has only (44%, Scandinavia about 54%). This "North-European" cluster overlaps much with the previous "East-European" cluster, and the peaks of both have actually not changed(!). The respective percentages for "Atlantic-Med" are exactly complementary. The former "North-West" cluster on the other hand, does not match any cluster in K12b.

To summarize - from K12b I conclude that:

1. former "East-European" = "North-European",
2. former "NW-European" = "North-European" + "Atlantic_med".

Thus let's ignore "NW-European":

Where do K12b "North-European" and K12b "Atlantic_med" peak:

1. Finland / Lithuania (75% North_European) = non-IE + satem IE
2. French Basque (73% Atlantic_Med) = non-IE

These are almost exactly the non-IE languages of Europe. Interestingly, Lithuanian is considered as a very pure IE language, yet it is close neighbor of non-IE Finland. But that's in accordance with the overlap of former "East Euro" and K12b "North-Euro".

The complementary cluster of "North_European" Finnic would be "Atlantic_Med" Basque, not British or Germanic. The British celts, Germans and Scandinavians are mixtures.

Note, this is not a proof that IE is paleolithic. It is rather an evidence that "East-European" (or "North_European") is not a useful IE marker, contrary to what I suggested before. The differences between two models provide much insight, and the East-European scenario (paleolithic finns got satemized) can be a useful complement of what happened in west europe.

 

[ElHorsto]

The differences between two models provide much insight, and the East-European scenario (paleolithic finns got satemized) can be a useful complement of what happened in west europe.

I have found another analogy for a Basque-Lithuanian-Finnic comparison:

Lithuanians have 8% Caucasian and 0% Gedrosian.
Basques have 0% Caucasian and 9.8% Gedrosian - opposite percentages.
Finns have 1.3 Caucasian and 0.9 Gedrosian - almost none.

I'd be tempted to conclude that "Gedrosian" represents Centum-IE and "Caucasian" Satem-IE, but it does not explain why Basques don't speak IE despite having the same Gedrosian percentage as neighboring countries. OTH, one could also ask how could 10% Gedrosian manage to change the native languages. Or maybe this musing is just too simplistic.

 

[spongetaro]

I have found another analogy for a Basque-Lithuanian-Finnic comparison:

Lithuanians have 8% Caucasian and 0% Gedrosian.
Basques have 0% Caucasian and 9.8% Gedrosian - opposite percentages.
Finns have 1.3 Caucasian and 0.9 Gedrosian - almost none.

I'd be tempted to conclude that "Gedrosian" represents Centum-IE and "Caucasian" Satem-IE, but it does not explain why Basques don't speak IE despite having the same Gedrosian percentage as neighboring countries. OTH, one could also ask how could 10% Gedrosian manage to change the native languages. Or maybe this musing is just too simplistic.

I see your point but it only applies to Europe. Iranian and Indo Aryan language are satem language but you would find the highest percentages of Gedrosia admixture in Iran, Northern India Afghanistan etc.

 

[ElHorsto]

I see your point but it only applies to Europe. Iranian and Indo Aryan language are satem language but you would find the highest percentages of Gedrosia admixture in Iran, Northern India Afghanistan etc.

Actually, Iranians and Kurds have much more caucasian than gedrosian. But you are right for Balochi, Brahui, Burusho, Sindi, etc. Also, the majority of Caucasian percentages in Italy, Greece and Balkans is most likely neolithic, not IE.

The absence of centum today in the east remains a challenge in general. Still, there was Tocharian and Hettite, and R1b is visible in Turkey, Armenia and Uyghurs today. But that's not quite Gedrosia unfortunately. In Europe, R1b seem to corellate with both, Gedrosian and Atlantic_Med. It is odd that exactly the east europeans have almost no Gedrosian component, despite that they are both R1a and satem speaking. Not to mention the similarities of Lithuanian and Sanskrit.

 

[spongetaro]

Actually, Iranians and Kurds have much more caucasian than gedrosian. But you are right for Balochi, Brahui, Burusho, Sindi, etc. Also, the majority of Caucasian percentages in Italy, Greece and Balkans is most likely neolithic, not IE.

The absence of centum today in the east remains a challenge in general. Still, there was Tocharian and Hettite, and R1b is visible in Turkey, Armenia and Uyghurs today. But that's not quite Gedrosia unfortunately. In Europe, R1b seem to corellate with both, Gedrosian and Atlantic_Med. It is odd that exactly the east europeans have almost no Gedrosian component, despite that they are both R1a and satem speaking. Not to mention the similarities of Lithuanian and Sanskrit.

There is also the possibility that the Gedrosia admixture is pre IE and came to the Doggerland during the mesolithic carrying maybe mtdna J

 

[ElHorsto]

There is also the possibility that the Gedrosia admixture is pre IE and came to the Doggerland during the mesolithic carrying maybe mtdna J

And you probably think so because Caucasian component might have spread later, in neolithic times. That's a valid idea.

On the other hand, if we still assume an IE origin of Gedrosian, the Pontic Steppe would be indeed a very unlikely route, not only because of lacking Caucasian admixture in Gedrosian, but even more so because of lack of Gedrosian today in Ukrainians and Russians. But a little further south in Caucasus and Turkey there is again a significant Gedrosian component (ca. 13%) and in Chechens there is a peak of even 25%! That could make "Gedrosian" itself an alternative marker for Caucasian origin, since we don't know who settled when in Caucasus itself.

 

[spongetaro]

And you probably think so because Caucasian component might have spread later, in neolithic times. That's a valid idea.

On the other hand, if we still assume an IE origin of Gedrosian, the Pontic Steppe would be indeed a very unlikely route, not only because of lacking Caucasian admixture in Gedrosian, but even more so because of lack of Gedrosian today in Ukrainians and Russians. But a little further south in Caucasus and Turkey there is again a significant Gedrosian component (ca. 13%) and in Chechens there is a peak of even 25%!

And also 25,5% for the Kurds!
The Gedrosian admixture in Turkey could also be from Tukish tribes who came from Central Asia where Gedrosia admixture is high. Anyway, Gedrosia admixture didn't come with the caucasus one to Northern Europe. It was either sole Gedrosia or North European+Gedrosia, or Med+ Gedrosia.
But as the med admixture is centred on western mediterranean and Atlantic it is very unlikely that Gedrosia ad came with the Med one. I personally think that the first men carrying R1b to Europe were 75% North European and 25% Gedrosia.

If you look at this proposed Urheimat below, it is located enough North to lack the Caucasus admixture and enough East to have a little bit of Gedrosia admixture. The question is why are Russian and Ukrainian lacking Gedrosia ad? Maybe, only the branch of IE represented by Tocharian and Italo Celtic (which are closely related)
included this component because of its more eastern initial location.

 

[ElHorsto]

And also 25,5% for the Kurds!
The Gedrosian admixture in Turkey could also be from Tukish tribes who came from Central Asia where Gedrosia admixture is high. Anyway, Gedrosia admixture didn't come with the caucasus one to Northern Europe. It was either sole Gedrosia or North European+Gedrosia, or Med+ Gedrosia.
But as the med admixture is centred on western mediterranean and Atlantic it is very unlikely that Gedrosia ad came with the Med one. I personally think that the first men carrying R1b to Europe were 75% North European and 25% Gedrosia.
If you look at this proposed Urheimat below, it is located enough North to lack the Caucasus admixture and enough East to have a little bit of Gedrosia admixture. The question is why are Russian and Ukrainian lacking Gedrosia ad? Maybe, only the branch of IE represented by Tocharian and Italo Celtic (which are closely related)
included this component because of its more eastern initial location.

Yes, that makes much sense, since "North_European" stretches remarkably far to the east. Even the Mordvins still have 63% of it. I wouldn't be surprised to see similar figures for Bashkirs and Tatars. And the absence of Gedrosian in Russians/Ukrainians can be explained by local diversity at this place of origin. Slavs actually came rather late from the north-west I suppose. Also the Sardinians perfectly support this view by having 0% "North_Euro" and 0% "Gedrosian" but plenty of "Atlantic_Med".
It only bugs me why have the Basque proportionally more Gedrosian than North European then? Maybe this is either noise, or the clustering algorithm had some problems to group them properly, assuming the Basques developed genetic peculiarities over time due to isolation.