What happened to CHG Ydna in Yamna

[Polska]

Also didn't both the J2b2 and the J1 Kotias Klde samples share the same maternal haplogroup?
Found in the same cave, with the same dating. Seems like the smoking gun that J2b2 was part of the CHG (which no one doubted anyways). The bigger question now is where was the European J2b2 when L283 first developed. It could have already been closer to the Balkans.

Both NEO281 and the original Kotias, KK1, share the same maternal lineage.

Yes, formation could have been closer to the Balkans, but I still think the L283 Nal’chik sample points to possible origination somewhere in the NW Caucasus, especially with a Mesolithic formation date for J2b L283. Lots of contacts between NW Caucasus and the dry steppe north of the Black Sea stretching to present day western Ukraine and Moldova.

 

[Viramund]

Are there any studies on chalcolithic steppe populations soon to be published? Maybe some more J would show more in pre R1b bottlenecked yamnaya herders.

 

[Ghurier]

We might deal with not just one but multiple mixture events. Like e.g. the initial spread of the CHG component along the coast with those fisher-pastoralist population which reached the Lower Don might have been dominated by CHG on both the male and the female side, but male totally. They could have incorporated local EHG females (no or very few males) and this mixed population with say 75 : 25 (CHG : EHG) could then have been taken over by male foragers from the North.

Having multiple mixture event didn't saves the day for a male-EHG dominated build-up of the WSH signal.

For instance, in a single admixture model, here is the likelihood according to Lazaridis numbers (indeed, not strong constraints, but still definitely in favor of CHG-males) :

I computed the constraints under two extreme hypotheses :
(i) The errors are dominated by the limited amount of data and thus nearly uncorrelated.
(ii) The errors are "model dominated" and therefore ~fully correlated.
It didn't make a big difference anyway.
At least in such setting it shows how disfavored is a male-EHG bias.

Then we can look at a smooth admixture, with "new comers" each generations :

As we see, in such scenario, the X chromosome (adm_x) still receives higher contribution from the female source populations.
Therefore, if you merge EHG-males and CHG/EHG admixed women, chrX will have less EHG than CHG.
Such scenario fails to produce a boost for the sex bias.

However, there is an "easy" way to "enhance" the sex bias derived from X-chromosome admixture, which consits in a sequenced patern of reproduction between two source populations progressively merging into a third new population, here is a small schema (each arrow represent only one generation, and a black line on some individual indicates that they got replaced or excluded from reproduction) :

This mixture scenario "takes advantage" of the fact that first-gen admixed males (in a heavily sex-biased merging scenario) only have admixture from the source population of the females on the X chromosome.
Then, by avoiding first-gen admixed females to participate to the reproduction, you prevent the X chromosome from the male source population to be passed down.
By having the first-gen admixed male mating with female's source population only, you have "cleaned" all contribution on the X chromosome from the male source population. However, you drop at only ~25% of total autosomal contribution for the male source population after two generations.
Then, you can increase again this component with a new wave of males from the male source population.
If you repeat this patern, technically you can have after few centuries a population that will be fully sourced by the male side for all chromosomes but X and fully sourced by the female-side for the X chromosomes.

But as you see, this "bias enhancement" scenario didn't help here, as it "increase" the sex bias ... but it is completely unable to "reverse" the apparent sex-bias.

To my knowledge, you have no way to produce a "reversed" sex-bias.
I don't see any merging pattern that would be able to produce more male-side admixture on the X chromosome (you can do the opposite because you can "flood" males X with females Xs).
(PS: unless you completely replace EHGs X-chromosome with the above scenario by CHGs X-chromosomes, and at the same time you also replace the CHG-X chromosome by an EHG-X chromosome and then you'll be able to reproduced the observed numbers ... but the above scenario is already convoluted, and that would start to become "cumbersome".)

Therefore, whenever you replace CHG males by EHG males (carrying EHG autosomal DNA), you will end up with less EHG on the X chromosome, which is not what is observed.

If we accept the numbers from Lazaridis, the predominance of some Y-lineages have to be related with a later founder effects that occured post-merging.
For instance, R-M269 only started to expand around 4400 BCE, therefore, if the merging creating WSH occured in a small population around 4500 BCE, then it is fairly easy for a random lineage to take over when the population will expand.
Let imagine that the Y-lineages of the WSH post-merging were ~80% CHG and ~20% EHG (completely inside the error bars of Lazaridis numbers), then there is ~20% chances that a founder effect will concern an EHG lineage ... not the most likely, but ~20% probability is not really disturbing.
R-M417 is definitely no big-deal in such scenario, as it could be a lonely lineage that got absorbed and "demographicly flooded" by WSH before expanding around ~3400 BCE. It didn't need to have been part of the WSH population since "founding".

If the above scenario is correct, with a WSH "creation" around 4500 BCE and R-M269 becoming dominant inside a small population arond 4400 BCE, then it might take a long time before any heavily-CHG group related to the root of WSH will be found among ancient DNA.

Also, it is interesting to note that some Y-lineages with a likely origin among CHG did survived, for instance the well known J-CTS1026. Another exemple could be J-Y167175.

 

[Riverman]

I simply don't accept that model reflecting the truth. I can't come up with alternative models, you can ask David, but if we get multiple huge founder events for EHG haplogroups, plus a dominance in ideology and culture, whereas we see a significant contribution from the mtDNA of the CHG population, the case is clear.
Again it is not just one EHG lineage in or one CHG out, its all CHG out and half a dozen of EHG in. And the founder event and diversification starts for those EHG branches right at or pretty close to the final formation moment.

Can't be chance, regardless of hypothetical models which try to tell us otherwise.

The only real evidence acceptable would be an ancient DNA population which fits the bill at the Lower Don, still dominated by J and the EHG lineages being already present. But even that would be not parsimonious, because why should all J branches die out, and not just one but half a dozen EHG ones survive and completely dominate the scene? Makes no sense at all.

Internal restructuring from within rather not.

 

[Ghurier]

I simply don't accept that model reflecting the truth.

Rejecting a quantified results is rarely a good thing.

I can't come up with alternative models, you can ask David

If this David want to speak, he is welcome to join the discussion ...
However, sex-biased chr-X admixture is something quite well studied, there is a paper from A. Golberg in 2015 that is quite interesting, even if a bit "formal".
I don't really see what this David could bring to such discussion.

Unless you contest that WSH is a CHG-EHG admixted population, there is little room for what you propose.

but if we get multiple huge founder events for EHG haplogroups, plus a dominance in ideology and culture, whereas we see a significant contribution from the mtDNA of the CHG population, the case is clear.

Yet, I don't see the "multiple" events you speak of ... the only one needed is R-M269 around ~4400 BCE.
All other haplogroups are likely subsequent absortions during later expansion phases (the first one to face that were EHGs, but later some ENF haplogroups also did).
It is a matter of population size ... the CHG-lineages dominated population was likely quickly overtaken by R-M269, after that ... the few CHG-remaining lineages won't rise again if post-founder effect they are a in severe minority.

Again it is not just one EHG lineage in or one CHG out, its all CHG out and half a dozen of EHG in.

I have to disagree, around 4400 BCE it is one EHG lineage that expand, not half a dozen.
As it is the core of our disagreement, I'll like to know who are these EHG-lineages that were among early WSH and how you can be so sure that they haven't been absorbed later (== after R-M269 first expansion) ?

We have many examples of Y-haplogroup overtaking other haplogroups.
R1a becoming dominant over R1b, I2 over R1a, ...

Y-DNA is clearly overlooked on forums ... it is strongly affected by random process, particularly for small initial populations.

And the founder event and diversification starts for those EHG branches right at or pretty close to the final formation moment.

Formation of WSH is estimated by scientific studies around ~4500 +/- 200 BCE, R1a have nine branches (surviving in modern population) at that time (only 1 really became successfull), none of them is showing anything spectacular around ~4500 BCE.
R-M269 was a single lineage by that time.

Can't be chance, regardless of hypothetical models which try to tell us otherwise.

Yet, I don't see any need of "chance".
You claim that ~half a dozen of EHG lineages where inside a forming WSH population (thus ~4500 BCE according to published estimations).
I see no proof of that.

The only real evidence acceptable would be an ancient DNA population which fits the bill at the Lower Don, still dominated by J and the EHG lineages being already present.

Honestly, if the sex-bias was in favor of EHG-males ... would you reject this result ?
Or whould you use it as some ultimate argument ?

But even that would be not parsimonious, because why should all J branches die out, and not just one but half a dozen EHG ones survive and completely dominate the scene? Makes no sense at all.

Again, who are these half a dozen EHG lineages supposingly present among early WSH ???
Where are the data about them ?
Why do you consider them as part of early WSH rather than later absorbtions during WSH expansions (as WSH absorbed many European lineages later when expanding westward, one of the most obvious being I1) ?

 

[Riverman]

Honestly, if the sex-bias was in favor of EHG-males ... would you reject this result ?
Or whould you use it as some ultimate argument ?

The principle difference would be that yDNA, autosomal and sex-bias data, plus archaeological and cultural analyses, would be in agreement. That would of course be a confirmation. The way it is, methods like the one applied by Lazaridis, can only cause doubts about the EHG dominance, but they can't prove a CHG male dominance in the formation, in the ethnogenesis phase.

And no, R-M269 is not alone, and it doesn't matter which branches were later successful, we have data from the ancient DNA as well. We are still missing the ancestors of a lot of R1a from the steppe, which surely lived in the Western regions and forest steppe, in core Sredny Stog groups, later the corded decorated ware horizon, which resulted in Cernavoda-Usatovo vs. Corded Ware in the West.

But there is another known group of steppe people, which is Khvalynsk, and they branched off from the more diverse Sredny Stog core before Yamnaya did.

Recent genetic studies have shown that males of the Khvalynsk culture carried primarily the paternal haplogroup R1b, although a few samples of R1a, I2a2, Q1a and J have been detected.

Khvalynsk culture - Wikipedia

en.wikipedia.org

Therefore even in the mixed Eastern offshot, we see a dominance of (E)HG lineages. And not just R-M269, but R1a, I2a2 and Q1a as well. J too, but that just proves they were not the dominant element, not in the West (R1a/R1b/I2), not in the centre (R1a/R1b) and not in the East (R1b/R1a/I2/Q).

Khvalynsk was a very early offshot, keep that in mind. We see therefore that the presence or shift of EHG dominated male lineages starts very early in Khvalynsk and the piedmont steppe.

 

[Ghurier]

The principle difference would be that yDNA, autosomal and sex-bias data, plus archaeological and cultural analyses, would be in agreement. That would of course be a confirmation.

Then, when you have conflicting clues, insteed of trying to build a better model to fit all the clues ... you hide the clues that didn't fit your narrative under the carpet ?
I could name some great theories of science that wouldn't have existed if our predecessors acted that way.

Cultural arguments are not that strong from my opinion, peoples are quick to forgot all the so-called "Indo-Europeanized" cultures of Europe that unfolded as ~totally ENF in admixture.

Normally in science you consider all arguments, not only the one supporting your favorite model, doing so is called "cherry picking".

The way it is, methods like the one applied by Lazaridis, can only cause doubts about the EHG dominance, but they can't prove a CHG male dominance in the formation, in the ethnogenesis phase.

For instance, in a single admixture model, here is the likelihood according to Lazaridis numbers (indeed, not strong constraints, but still definitely in favor of CHG-males) :

I computed the constraints under two extreme hypotheses :
(i) The errors are dominated by the limited amount of data and thus nearly uncorrelated.
(ii) The errors are "model dominated" and therefore ~fully correlated.
It didn't make a big difference anyway.
At least in such setting it shows how disfavored is a male-EHG bias

PS: multiple admixture events in "reasonable" scenarii didn't significantly change the above constraints.

And no, R-M269 is not alone, and it doesn't matter which branches were later successful, we have data from the ancient DNA as well.

In fact, it is the main question about the early population that will source Yamnaya.
Ancient DNA, if post CHG/EHG merging and post expansion won't be that helpfull and won't prove what you try to prove ...
In a "small" population Y-DNA can shift totally in few generations (and obviously, the Y-DNA shift that is implied by the sex bias, need to occur very-very-very early in the story of the WSH population that will source Yamnaya).
Emulating a spurious X-chr admixture is more tricky than replacing a Y-lineage without significant admixture change.

For some reasons, you have two probes, you give more credits to the one with the highest statistical variance.
I'm fairly skeptical to such approach, particularly when totally dismissing some data that you don't like.

But there is another known group of steppe people, which is Khvalynsk, and they branched off from the more diverse Sredny Stog core before Yamnaya did.

It would be interesting to check if the same sex-bias is found for Khvalynsk ... It is an open question per-se considering the dating.
The sex bias we are discussing is observed for Yamnaya.
I think you are seing the problematic in a too linear and too simplistic way.

I will give one small scenario :
Imagine CHGs (male and women) entering the steppe.
Once in the steppe, imagine that they formed some mating network with EHGs.
Then, you could have the built up of one sub-population with a CHG-male bias and one sub-population with a EHG-male bias.
Both being able to expand on their own ... autosomally speaking you'll likely find them very similar, but the sex bias would be drastically different.

And this just one exemple among legions.

Therefore even in the mixed Eastern offshot, we see a dominance of (E)HG lineages. And not just R-M269, but R1a, I2a2 and Q1a as well.

The point is that you haven't proved that the proxy you use is a valid proxy for our current discussion.

Khvalynsk was a very early offshot, keep that in mind. We see therefore that the presence or shift of EHG dominated male lineages starts very early in Khvalynsk and the piedmont steppe.

Per se you are trying to prove something about culture "A" based on the history of a sister/cousin culture "B" ...
Didn't sounds reasonable to me. It is full of un-proven assumptions about the exact deep relation of these cultures build-up.

 

[Riverman]

For some reasons, you have two probes, you give more credits to the one with the highest statistical variance.
I'm fairly skeptical to such approach, particularly when totally dismissing some data that you don't like.

I don't consider a hypothetical model data, but just a hypothetical model. Archaeological evidence for local hunter-fisher cultures clans being the dominant force is more like data to me.
I'm not saying that its not possible in the way you try to paint it, I'm just saying that this model from Lazaridis on its own proves nothing, because there are many ways how the observed pattern could have evolved.

If the CHG side was dominant, why should they have accepted the dominant forager clan culture and being totally dominated by a whole set of EHG lineages at the same time? Within a fairly short time span.

We even see how they built their settlements along the sea and Don to some degree, but then there was a cultural-ideological change right at the time we have to assume the PPIE ethnogenesis and takeover of dominant clans. How the X-chromosome analyses, if truly correct, fits into this, I don't know, but the rest of the facts doesn't fit into a CHG-dominant model.

Anyway, like I wrote before, we need ancient DNA to prove it either way.

Yet one thing is for sure: We deal with at least 3-4 different steppe populations, all stem from early roots it seems, and they all being dominated by EHG lineages, and this EHG dominance looks like being present from before the split. And its NOT just R-M269, its the whole range we can observe from the West (Western Sredny Stog and groups), centre West (not sampled yet, but that's the root), centre East (Yamnaya) and East (Khvalynsk).

Once we get DNA from the Lower Don early source group, we know for sure, because whatever happened, it happened there and we should get samples from "in the process" individuals hopefully.

 

[Ghurier]

Erratum :

Yesterday I said (see bold part) :

there is an "easy" way to "enhance" the sex bias derived from X-chromosome admixture, which consits in a sequenced patern of reproduction between two source populations progressively merging into a third new population, here is a small schema (each arrow represent only one generation, and a black line on some individual indicates that they got replaced or excluded from reproduction) :

This mixture scenario "takes advantage" of the fact that first-gen admixed males (in a heavily sex-biased merging scenario) only have admixture from the source population of the females on the X chromosome.
Then, by avoiding first-gen admixed females to participate to the reproduction, you prevent the X chromosome from the male source population to be passed down.
By having the first-gen admixed male mating with female's source population only, you have "cleaned" all contribution on the X chromosome from the male source population. However, you drop at only ~25% of total autosomal contribution for the male source population after two generations.
Then, you can increase again this component with a new wave of males from the male source population.
If you repeat this patern, technically you can have after few centuries a population that will be fully sourced by the male side for all chromosomes but X and fully sourced by the female-side for the X chromosomes.

This is incorrect, after verification of the convergence of such repeated process, you'll be stuck at a 67/33 repartition of the ancestry for the autosomal signal (while X-signal can be 100/0).
Thus, contrary to what I claimed, even with such weird merging scenario, you can't produce a reversed sex-bias.

 

[Vespasianus]

This level of caucasian admixture doesn't seem very plausible to me. Read this:

History and genetics of the Yamna culture

The Yamna (Pit-Grave) culture is considered as one of the homelands of the Proto-Indo-European people. Discover how Yamnayans lived, what they looked like physically, and who they resembled genetically.

www.eupedia.com

Here the modulation for caucasian genome is significantly smaller, many U samples among mtDNA.

People forget the very low concentration of R1a among Yamna samples, R1b and not R1a was the original indo-european languages bearer, only after the spread of R1b we have the indo-european cultures, R1a before that was only hunter gatherers with primitive technology and absent of geometric ceramics. Corded Ware culture is largely influenced by Yamna, but also considerably different, the dominant haplogroup there is R1a, not R1b(only 11%), the ceramic pattern is circular(corded) not geometric(roman and greek ceramics are much more geometric). Linguistically germanic languages are considered less indo-european than Latin for example(taken as the prime late indo-european language).

Most likelt R1b male lineage was from pastoralists coming from the south and not foragers. Since they brought about many military conquest, taming of horses and advanced military structure, they were much closer to the neolithic civilizations, which had domesticated the dog, pigs, goats etc, thus already herders and farmers, R1b people just focused more on the herder side, they shared some agricultural tools with neolithic farmers and vocabulary also(like the word for plow). Pastoralists tame horses not hunter-gatherers, the horse was essential for the brewing of milk and for military power(cavalry). Also, if the R1b came from the Caucasus after the last Ice Age it would explain the high degree of Caucasian admixture, not only does the eupedia website thinks that, but it is the most obvious. As the R1b brought Bronze technology, they had previously mastered Copper, a metal also previously used in the northern Middle East(which would explain the link between the two). The high degree of caucasian admixture(even if likely lower than 50%) would only be explained by R1b coming from the south, otherwise the caucasian hunter gatherers wouldn't have expanded without their male lineage to the north, quite simple.