The Arrival of Steppe & Iranian Related Ancestry in Islands of West Mediterranean

Ygorcs said:
If the authors claim only ~56-62% of the modern Sardinian genetic makeup is explained by ancient Sardinian EEF up to the Iron Age, then I wonder why they are still overwhelmingly EEF (ANF+WHG) in ancestry (the samples they modelled as shown in the picture above are less so, but still the vast majority is made up of ANF+WHG). Does that mean that all the foreign inputs were also very rich in ANF and/or WHG, so that they added "exotic" admixtures, but mostly extra ANF and WHG from a different and already somewhat drifted source? If that be what really happened, then I would guess most of the input was from people similar to Mycenaeans and Minoans and/or modern North Italians and Central Italians.
Click to expand...

That's my thinking for now.

Wish they had made it all a bit clearer, but perhaps they're saving the more proximate sources for a later paper.

Another thing that struck me in the PCA is how close one of the Natufian samples is to both Greek and Anatolian Bronze Age. This might be part of what is muddying up some of the amateur analyses which try to model Sicily and Southern Italy using Natufian as one of the sources.
 
The PCA from Marcus, Posth, Ringbauer et al 2020, also published this week (70 ancient individuals from 21 Sardinian sites spanning 6000 years, with additional historic samples (Punic, Roman, Medieval) added since their pre-print.



41467_2020_14523_Fig2_HTML.png

Same as in Fernandez et al, strong continuity from Neolithic farmers until the Nuragic Bronze Age, but beginning from the Iron Age there is dynamic (and transient) contact. Interestingly, some of the 500 BCE Punics (from Villamar, an inland rural Punic site) go clearly toward North Africa, the Imperial Romans toward more Eastern Mediterranean (similar to Antonio et al Roman paper) and then in the Medieval some Sardinian individuals shift towards Italian and possible Iberian sources (these are individuals from a church in the North West of the island, so Iberia makes sense). The present-day individuals form a cline towards Neolithic-Nuragic, with, surprise surprise, Ogliastra being closest to Nuragic (some individuals overlapping in PCA!), and Gallura (in the North where a Corsican dialect is prevalent) the most admixed.:

41467_2020_14523_Fig5_HTML.png


(left Fig is the relevant one and the zoom-in of the big PCA, right panel is within Sardinia PCA which is driven by recent Ogliastra drift, as these individuals are often from the same village which drives PC components)

Considering the wide range of sources and the general European post-BA similarities it is hard to pinpoint the exact admixture fractions, but ~30-70% turnover depending on the region in Sardinia seem about right. There are also some Neolithic Y chromosomes (most importantly R1b-V88, that story is written up in the supplement) left on the island at low but substantial frequency, which are found only extremely rarely in the rest of Europe at present-day, so the old view that Sardinia is reservoir of Neolithic ancestry holds up (but aDNA also showing more historical turnover than previously suspected).

The paper can be found at https://www.nature.com/articles/s41467-020-14523-6; it is open access.
 
Thanks for the post. We have a dedicated thread on the published paper, and I had been meaning to review it and post, but life got in the way.:)

In the discussion on this site of the pre-print of this specifically Sardinian paper, I pointed out that the Reich Lab pre-print was a bit misleading in that, first of all, they didn't do a separate analysis of the Ogliastra region. This group did a better job, and even in the pre-print it was pretty clear that Cavalli-Sforza was very good at picking the samples from Sardinia, and that indeed the people of Ogliastra are pretty close to ancient Sardinians, although Nuragic Sardinians instead of perhaps Neolithic .
Sardinians.

The other way in which the Reich Lab paper being discussed here went wrong imo is that like the Antonio et al paper they assume that all the samples they are testing are "locals" who stayed in the area and whose genes must have had a substantial impact on the population genetics of the area. Their analyses showing a big change in modern peoples of the area then is attributed to yet other large migrations. I don't think it's justified.

That's why I think your admixture estimate of 30-70% for Sardinia is much too large. Even the Reich Lab paper sees a 62% continuance from the Nuragic Era as an average for the whole island if I remember correctly. In Ogliastra it might even be more than 70%.

Even in the case of the extreme Northwest, settled by the Spanish, and the Corsican speaking people of the North, or the areas where there's been some migration within even the last century from mainland Italians I highly doubt there's been 70% replacement.

Maybe a fruitful approach would be to do a separate analysis based on language. In Italy, not only is there genetic variation across very small distances, but it often tracks with language. Italy is not like, for example, Denmark, where the population is "extremely" homogeneous. You wouldn't find that in Italy even within a province, so averages are pretty useless most of the time.
 
Hi Angela, I just thought it makes sense to discuss the two papers together, or at least the Sardinia aspect, and was curious to hear your thoughts!

From Fernandes et al.: `Even this range of estimates of ~56–62% is an upper bound, as we are forcing in pre-Iron-Age Sardinians as the only source of European farmer ancestry on the island in all of these models`. In the pre-print they discussed working models of almost complete turnover (which was a bit too much as working models don't mean is that's what actually happened), and then seeing the other paper they tuned it down somewhat, but between the lines it is clear that they think this is a "upper bound" (most Med. populations have 50%+ EEF ancestry, so go figure what they actually mean).

The exact amount of turnover is hard to pinpoint, but taking the most extreme "plausible" sources (for closer ones one would need even more) one would need at least ~30% new ancestries to explain the shift on the PCA, qpAdm etc.. And keep in mind that 30% is cumulative, there likely wasn't a one time turnover, but many small ongoing mixtures over a long time.

And also consider that evidence: R1b-M269 seems completely absent or very low frequency until including Nuragic Sardinia, in all regions and in total for several dozen individuals.
But nowadays R1b-M269 is at ~20% in all Sardinian regions without too much substructure, including Ogliastra. Not all new arrivals had R1b-M269, so the amount of "male ancestry" turnover seems to have been significantly more than 20 % (and even other haplogroups seem to be new arrivals, including E1s, with the majority of Nuragic haplogroups being at very low frequency at present-day, pushing it even higher). Of course that is just one part of the tree (the male one) and there are also "random" fluctuations, but that seems to clearly point toward new ancestries, in all parts of the island.

That doesn't mean that Sardina is not unique or became just like everyone else, it is still the most "EEF" like population in Europe, and in particular Ogliastra stands out.



 
IceMummy said:
Hi Angela, I just thought it makes sense to discuss the two papers together, or at least the Sardinia aspect, and was curious to hear your thoughts!

From Fernandes et al.: `Even this range of estimates of ~56–62% is an upper bound, as we are forcing in pre-Iron-Age Sardinians as the only source of European farmer ancestry on the island in all of these models`. In the pre-print they discussed working models of almost complete turnover (which was a bit too much as working models don't mean is that's what actually happened), and then seeing the other paper they tuned it down somewhat, but between the lines it is clear that they think this is a "upper bound" (most Med. populations have 50%+ EEF ancestry, so go figure what they actually mean).

The exact amount of turnover is hard to pinpoint, but taking the most extreme "plausible" sources (for closer ones one would need even more) one would need at least ~30% new ancestries to explain the shift on the PCA, qpAdm etc.. And keep in mind that 30% is cumulative, there likely wasn't a one time turnover, but many small ongoing mixtures over a long time.

And also consider that evidence: R1b-M269 seems completely absent or very low frequency until including Nuragic Sardinia, in all regions and in total for several dozen individuals.
But nowadays R1b-M269 is at ~20% in all Sardinian regions without too much substructure, including Ogliastra. Not all new arrivals had R1b-M269, so the amount of "male ancestry" turnover seems to have been significantly more than 20 % (and even other haplogroups seem to be new arrivals, including E1s, with the majority of Nuragic haplogroups being at very low frequency at present-day, pushing it even higher). Of course that is just one part of the tree (the male one) and there are also "random" fluctuations, but that seems to clearly point toward new ancestries, in all parts of the island.

That doesn't mean that Sardina is not unique or became just like everyone else, it is still the most "EEF" like population in Europe, and in particular Ogliastra stands out.
Click to expand...

I agree it makes more sense to discuss them together.

Perhaps I misunderstood you. I thought you were saying there was up to a 70% "replacement" in Sardinia. That's virtually impossible imo.

A 30% average for the whole island may be in the right ballpark. First thing that should be done is, as I say, to remove the areas settled by Spaniards and the area speaking a Corsican dialect and then see what it looks like.

YDna is not going to be in a one to one ratio with autosomal inheritance, especially not on Sardinia.
 
The Catalans didn't had a great impact. R1b DF27 is very low or 0% They settled mostly in Cagliari and Alghero but they were replaced by native over the course of the centuries

utilizzando Tapatalk
 
I reworked my post to better correspond to my thinking on the subject. Here it is:

I agree it makes more sense to discuss them together.

Perhaps I misunderstood you. I thought you were saying there was up to a 70% "replacement" in Sardinia. That's virtually impossible imo. I also think you're doing a lot of "reading between the lines" to reach conclusions for which there is absolutely no evidence. I read the supplement carefully. 52-60% continuity is a decent estimate, although I think if they looked at the majority of the island it would be closer to 70%.

Many of those ancient samples from the Punic Era, the Roman Era, and Late Antiquity represented genetic trails which disappeared in Sardinia not because there was massive influx from the mainland in the modern era necessarily, but at least partly because they were a small percentage of the population and either left or their genetic input was washed out.

So many researchers just haven't absorbed the fact that when elites come in, as is the case we're seeing in Sardinia, it's their burials which you are going to find. You can't assume everyone in the area you're studying had that dna signature, which means you don't have to find some actually non-existent huge migration in later eras to explain current autosomal signatures. They did the same thing in their analyses of Central Europe. There are these very large percentages for "steppe", and then there's this big "resurgence" of farmer ancestry, representing about 45-55% in some areas. There wasn't a "HUGE" resurgence caused by Romans or anyone else. Those people were no doubt still there, but they didn't have graves whose skeletal remains we were able to test.

At any rate, the first thing that should be done is, as I say, to remove the areas settled by Spaniards and the area speaking a Corsican dialect and then see what it looks like.

YDna is not going to be in a one to one ratio with autosomal inheritance, especially not on Sardinia.

@Cato,
Do you have a study which does a survey of the y in the Catalan speaking areas? Not that it's all that consequential seeing that it's such a small area.

Speaking of yDna, are the "E1b1b" clades in Sardinia, other than E-V13 and immediate upstream clades, still predominantly in the SW? Are the R1b M269 predominantly in the Corsican speaking areas and a few other areas where we know there was modern migration from the mainland?

Anybody have the link to the big Ydna study of Sardinia which would provide that info, and hopefully by area?
 
No but there is a study about the surnames of Alghero, in the XVI century those of catalan origin were already only 15%

the Corsican influx was much greater

utilizzando Tapatalk
 
Angela said:
I reworked my post to better correspond to my thinking on the subject. Here it is:

I agree it makes more sense to discuss them together.

Perhaps I misunderstood you. I thought you were saying there was up to a 70% "replacement" in Sardinia. That's virtually impossible imo. I also think you're doing a lot of "reading between the lines" to reach conclusions for which there is absolutely no evidence. I read the supplement carefully. 52-60% continuity is a decent estimate, although I think if they looked at the majority of the island it would be closer to 70%.

Many of those ancient samples from the Punic Era, the Roman Era, and Late Antiquity represented genetic trails which disappeared in Sardinia not because there was massive influx from the mainland in the modern era necessarily, but at least partly because they were a small percentage of the population and either left or their genetic input was washed out.

So many researchers just haven't absorbed the fact that when elites come in, as is the case we're seeing in Sardinia, it's their burials which you are going to find. You can't assume everyone in the area you're studying had that dna signature, which means you don't have to find some actually non-existent huge migration in later eras to explain current autosomal signatures. They did the same thing in their analyses of Central Europe. There are these very large percentages for "steppe", and then there's this big "resurgence" of farmer ancestry, representing about 45-55% in some areas. There wasn't a "HUGE" resurgence caused by Romans or anyone else. Those people were no doubt still there, but they didn't have graves whose skeletal remains we were able to test.

At any rate, the first thing that should be done is, as I say, to remove the areas settled by Spaniards and the area speaking a Corsican dialect and then see what it looks like.

YDna is not going to be in a one to one ratio with autosomal inheritance, especially not on Sardinia.

@Cato,
Do you have a study which does a survey of the y in the Catalan speaking areas? Not that it's all that consequential seeing that it's such a small area.

Speaking of yDna, are the "E1b1b" clades in Sardinia, other than E-V13 and immediate upstream clades, still predominantly in the SW? Are the R1b M269 predominantly in the Corsican speaking areas and a few other areas where we know there was modern migration from the mainland?

Anybody have the link to the big Ydna study of Sardinia which would provide that info, and hopefully by area?
Click to expand...

@Angela
You won! I did not put too much looks on this thread before and after I read some of the last posts, I was going to speak about these wrong effect of ancient elite burials on statistics. We all (almost all of us at least) know that very often, after colonisations and changes of powers with foreign elites the local ancient pop tends to take again the strong side concerning demography. It occurred a lot of times everywhere. It is not purely scientific but I think Sardinian current pop, one of the most static or stable one in Europe, did not underrun a turnover more than up to 10/15%, so 30 to 70%, gulp! I don't swallow it. Sardinia never was New York, Moscow or Pekin, nor even Roma.
 
In the paper by Marcus there are the % of replacement..eastern mediterranean and northern mediterraneanScreenshot_20200301-145359_Drive.jpg

utilizzando Tapatalk
 
MOESAN said:
@Angela
You won! I did not put too much looks on this thread before and after I read some of the last posts, I was going to speak about these wrong effect of ancient elite burials on statistics. We all (almost all of us at least) know that very often, after colonisations and changes of powers with foreign elites the local ancient pop tends to take again the strong side concerning demography. It occurred a lot of times everywhere. It is not purely scientific but I think Sardinian current pop, one of the most static or stable one in Europe, did not underrun a turnover more than up to 10/15%, so 30 to 70%, gulp! I don't swallow it. Sardinia never was New York, Moscow or Pekin, nor even Roma.
Click to expand...

Thank you Moesan. :)

As I mentioned above, I think the 58-60% continuity figure settled upon by the authors is already pretty low, anything below that is highly unlikely. Certainly, anything over 50% replacement is absurd. The authors specifically point out that modern Sardinian samples can all be modeled with over 50% Nuragic ancestry.

It continues to amaze me that PHD candidates or research associates, supervised by big names in the field, can be so blind to the fundamental factor we are discussing. As I said, anyone can take a look at the steppe in some German Corded Ware individuals and then look at modern Germans, who are close to 50% EEF if I remember correctly. There absolutely wasn't some huge migration from Imperial Italy into Germany to explain it.

In the case of Italy, which has a very complicated genetic history, and just history in general, the problem is compounded by the fact that these researchers know nothing of Italian history and linguistics. You can apply all the advanced statistical methods available, but if you don't know the fundamental historical facts, and use some common sense, for goodness' sakes, your conclusions will be wrong, or, at least, we won't know if you're right or wrong.

If we take a look at where the samples were taken, and at their PCA, some interesting conclusions can be drawn.
vA8sImJ.png



PzuYzAD.png

It's very clear that the "Punic" individuals are far outside modern Sardinian variation. In terms of admixture, as I have always maintained, to vociferous opposition from the usual quarters, I might add, the Punic settlements in Sardinia represented trading marts, NOT FOLK MIGRATION, and therefore had very little impact on the "host" genetics.

To quote the paper: "the inferences of N.African ancestry are negligible". The amount is less than in Sicily and Spain, fwiw.

In terms of the "North Mediterranean" admixture, they should be using Corsicans or at least Tuscans as a source as that is how much of the more "North Mediterranean" like ancestry would likely have entered the island. The authors don't seem to realize that Sardinia was the back of beyond. It wasn't that easy, given the navigation systems of prior eras, to get from mainland Italy to Sardinia because of ocean and wind currents. Much, if not most, of the migration would have taken the route used since the Mesolithic, which was through Corsica.

As for the "Eastern Mediterranean" ancestry, their name for the "orange" cluster in their admixture runs, it is also present in Spain, France, and Tuscany. Did the Druze and Palestinians migrate to France as well? I think the 5% share for J1-L862 tells that tale. It's much more likely that the "orange" component, representing more Iranian Neolithic like ancestry, arrived with ancestry making its way from the mainland through Corsica, and also from migrations from Southern Italy or Greece.

The steppe ancestry is extremely small as well.

As I've also been saying forever, there is substructure in Sardinia, with Campidano in the SW shifting more toward "Eastern Mediterranean", and the northeast toward "Northern Mediterranean". Of course that's where the Northeast shifts: that's where we find the Corsican dialect speakers. It's also where we find the most M-269.

Really, can't they do a little more research before analyzing the data?

As for the admixture chart in the paper, why on earth are they showing us only Cagliari in the south, and to top it off, the capital. Jeez!











 
IceMummy said:
Hi Angela, I just thought it makes sense to discuss the two papers together, or at least the Sardinia aspect, and was curious to hear your thoughts!
[... And also consider that evidence: R1b-M269 seems completely absent or very low frequency until including Nuragic Sardinia, in all regions and in total for several dozen individuals.
But nowadays R1b-M269 is at ~20% in all Sardinian regions without too much substructure, including Ogliastra. Not all new arrivals had R1b-M269, so the amount of "male ancestry" turnover seems to have been significantly more than 20 % (and even other haplogroups seem to be new arrivals, including E1s, with the majority of Nuragic haplogroups being at very low frequency at present-day, pushing it even higher). Of course that is just one part of the tree (the male one) and there are also "random" fluctuations, but that seems to clearly point toward new ancestries, in all parts of the island...]
Click to expand...
I'm late, I liked your post, but you know that male lineages inherited in the Chalco to Iron age specific era were elite 's haploG with a social/political "drift" - at least concerning Y-R1b6M269 downstreams - with a strong discrepancy in their % compared to the autosomes % - Once in place, they could expand progressively into the whole pop, even into more remote places, sometimes helped by pure hazard drift.
 
Dear Angela, I only now saw your reply, sorry for not checking back earlier. Here a reply to some of the points brought up by you.

In the case of Italy, which has a very complicated genetic history, and just history in general, the problem is compounded by the fact that these researchers know nothing of Italian history and linguistics.
Sardinian Italian is one of the most conserved Latin languages. Why do you believe linguistics does not support turnover? In fact it does, the language the Nuragic people spoke is not existant at all anymore. By the way, we had several experts for Sardinian history and archeology as (active) co-authors, and claiming that they know nothing of Italian history and linguistics is a bit disrespectful I believe.


As I've also been saying forever, there is substructure in Sardinia, with Campidano in the SW shifting more toward "Eastern Mediterranean", and the northeast toward "Northern Mediterranean". Of course that's where the Northeast shifts: that's where we find the Corsican dialect speakers.
We mention that point in the discussion, and briefly describe Gallurese and its historic context (immigration of Corsica). But thanks for reiterating that point.

As for the admixture chart in the paper, why on earth are they showing us only Cagliari in the south, and to top it off, the capital.
We do not use the capital, as we explain in the introduction we use province labels (and Cagliari is the name of the province). We use it as Cagliari is in the center of present-day Sardinian genetic variation (this is with people where all for grandparents are from Sardinia).

remove the areas settled by Spaniards and the area speaking a Corsican dialect and then see what it looks like.
Removing Cagliari (as you suggest above), Northwest and North East would remove a large fraction of the Sardinian population. The paper is about the whole of Sardinia after all. And we did look at Ogliastra only, and show that also Ogliastra needs 40% new ancestry in a conservative estimate (btw there is 20% R1b-M269 there too, this is not published but I looked at this data). The other, independent, Fernandez paper used exclusively Ogliastrans, and claims even more turnover for them! Do you think we both independently arrive at the same conclusions by mistake?

So many researchers just haven't absorbed the fact that when elites come in, as is the case we're seeing in Sardinia, it's their burials which you are going to find. You can't assume everyone in the area you're studying had that dna signature, which means you don't have to find some actually non-existent huge migration in later eras to explain current autosomal signatures.
We point that out in the discussion, that buried individuals are unlikely to present a random cross-section of the population. However, the 40% turnover estimate is based on a large number of Neolithic to Nuragic individuals, who all tightly cluster (and we are reasonably sure about that component), as one source, and modern individuals with four grandparents from Sardinia as target. The Punic, Roman and Medieval individuals do not enter that analysis, and as you say, can all be transient - and that's why we did not use them in the turnover estimates.

Wonderful. I need something to take my mind off this madness.
I am sorry to upset you, but hard evidence is hard evidence. If you think our analysis ignores some evidence or think our conclusions are unwarranted in light of the data, please write a rebuttal, that would help Scientific progress!
 
^^Since when do you need population replacement levels of 40-50% to change a language? Look at Hungary, or even France.

Fine, you're looking at the genetics of the whole island. What I am most interested in is not the areas where you have more recent population spread from mainland Italy or Corsica (let's leave aside the Catalan speaking areas since the ydna shows little change) but the core areas of Sardinia from which samples were taken for population genetics comparison purposes to see why Sardinians from those areas are so close to the Middle Neolithic inhabitants of Europe and have conserved so much of the original yDna signatures of those people.

It is highly improbable that those people have experienced a 40-50% replacement but still so closely plot with Middle Neolithic people from the Balkans, mainland Italy, etc. The 20% number for R1b-269 there actually supports my point. Usually, autosomal figures are going to be less than the yDna if it is infiltrating a populated area. Look at the percentages for steppe ancestry in areas in northern Italy which have 60-70% of it. In a lot of cases it's 25%.
 
IceMummy said:
Dear Angela, I only now saw your reply, sorry for not checking back earlier. Here a reply to some of the points brought up by you.

In the case of Italy, which has a very complicated genetic history, and just history in general, the problem is compounded by the fact that these researchers know nothing of Italian history and linguistics.
Sardinian Italian is one of the most conserved Latin languages. Why do you believe linguistics does not support turnover? In fact it does, the language the Nuragic people spoke is not existant at all anymore. By the way, we had several experts for Sardinian history and archeology as (active) co-authors, and claiming that they know nothing of Italian history and linguistics is a bit disrespectful I believe.


As I've also been saying forever, there is substructure in Sardinia, with Campidano in the SW shifting more toward "Eastern Mediterranean", and the northeast toward "Northern Mediterranean". Of course that's where the Northeast shifts: that's where we find the Corsican dialect speakers.
We mention that point in the discussion, and briefly describe Gallurese and its historic context (immigration of Corsica). But thanks for reiterating that point.

As for the admixture chart in the paper, why on earth are they showing us only Cagliari in the south, and to top it off, the capital.
We do not use the capital, as we explain in the introduction we use province labels (and Cagliari is the name of the province). We use it as Cagliari is in the center of present-day Sardinian genetic variation (this is with people where all for grandparents are from Sardinia).

remove the areas settled by Spaniards and the area speaking a Corsican dialect and then see what it looks like.
Removing Cagliari (as you suggest above), Northwest and North East would remove a large fraction of the Sardinian population. The paper is about the whole of Sardinia after all. And we did look at Ogliastra only, and show that also Ogliastra needs 40% new ancestry in a conservative estimate (btw there is 20% R1b-M269 there too, this is not published but I looked at this data). The other, independent, Fernandez paper used exclusively Ogliastrans, and claims even more turnover for them! Do you think we both independently arrive at the same conclusions by mistake?

So many researchers just haven't absorbed the fact that when elites come in, as is the case we're seeing in Sardinia, it's their burials which you are going to find. You can't assume everyone in the area you're studying had that dna signature, which means you don't have to find some actually non-existent huge migration in later eras to explain current autosomal signatures.
We point that out in the discussion, that buried individuals are unlikely to present a random cross-section of the population. However, the 40% turnover estimate is based on a large number of Neolithic to Nuragic individuals, who all tightly cluster (and we are reasonably sure about that component), as one source, and modern individuals with four grandparents from Sardinia as target. The Punic, Roman and Medieval individuals do not enter that analysis, and as you say, can all be transient - and that's why we did not use them in the turnover estimates.

Wonderful. I need something to take my mind off this madness.
I am sorry to upset you, but hard evidence is hard evidence. If you think our analysis ignores some evidence or think our conclusions are unwarranted in light of the data, please write a rebuttal, that would help Scientific progress!
Click to expand...

are you talking about

https://www.nature.com/articles/s41598-019-49901-8


or

https://www.nature.com/articles/5201205


or the links in this post

http://www.khazaria.com/genetics/corsicans.html



below is the last I can recall
[url]https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0200641
[/URL]
 
The area with higher steppe (10%) is the Carbonia province according to Marcus 2020, so it's not only the north that have experienced post nuragic immigration
 
Ygorcs said:
If the authors claim only ~56-62% of the modern Sardinian genetic makeup is explained by ancient Sardinian EEF up to the Iron Age, then I wonder why they are still overwhelmingly EEF (ANF+WHG) in ancestry (the samples they modelled as shown in the picture above are less so, but still the vast majority is made up of ANF+WHG). Does that mean that all the foreign inputs were also very rich in ANF and/or WHG, so that they added "exotic" admixtures, but mostly extra ANF and WHG from a different and already somewhat drifted source? If that be what really happened, then I would guess most of the input was from people similar to Mycenaeans and Minoans and/or modern North Italians and Central Italians.
Click to expand...

This is what happened most likely, a mixed steppe-EEF population caused the turnover on Sardinia. Probably similar to Iberian BB, which did, in fact, enter the island as well.

IceMummy said:
Hi Angela, I just thought it makes sense to discuss the two papers together, or at least the Sardinia aspect, and was curious to hear your thoughts!

From Fernandes et al.: `Even this range of estimates of ~56–62% is an upper bound, as we are forcing in pre-Iron-Age Sardinians as the only source of European farmer ancestry on the island in all of these models`. In the pre-print they discussed working models of almost complete turnover (which was a bit too much as working models don't mean is that's what actually happened), and then seeing the other paper they tuned it down somewhat, but between the lines it is clear that they think this is a "upper bound" (most Med. populations have 50%+ EEF ancestry, so go figure what they actually mean).

The exact amount of turnover is hard to pinpoint, but taking the most extreme "plausible" sources (for closer ones one would need even more) one would need at least ~30% new ancestries to explain the shift on the PCA, qpAdm etc.. And keep in mind that 30% is cumulative, there likely wasn't a one time turnover, but many small ongoing mixtures over a long time.

And also consider that evidence: R1b-M269 seems completely absent or very low frequency until including Nuragic Sardinia, in all regions and in total for several dozen individuals.
But nowadays R1b-M269 is at ~20% in all Sardinian regions without too much substructure, including Ogliastra. Not all new arrivals had R1b-M269, so the amount of "male ancestry" turnover seems to have been significantly more than 20 % (and even other haplogroups seem to be new arrivals, including E1s, with the majority of Nuragic haplogroups being at very low frequency at present-day, pushing it even higher). Of course that is just one part of the tree (the male one) and there are also "random" fluctuations, but that seems to clearly point toward new ancestries, in all parts of the island.

That doesn't mean that Sardina is not unique or became just like everyone else, it is still the most "EEF" like population in Europe, and in particular Ogliastra stands out.
Click to expand...

I think it is a given that R1b-M269 and variants of E1b1b, especially E-V13, entered Sardinia together. It wouldn't have been the first turnover on Sardinia, because Sardinia was colonised, one way or another, by more WHG shifted I2a agro-pastoralists from continental Europe before, which pushed the G2a community back big time.

The next continental colonisation of probably BB-like people (?) from Northern Italy brought steppe ancestry. Concerning the regions with more or less new ancestry, we have to think about the importance of a site. The refuge areas in the mountains were isolated places of low importance and hard to access by newcomers. Yet in later times they would have been better protected from any sort of attack or influence coming from the sea. It should be obvious that those which controlled the big coastal centres were the socio-politically dominant people. Therefore its absolutely no coincidence that the central coastal regions have more steppe ancestry and more of the later uniparentals.

Sardinia is still remarkable as it has less non-EEF ancestry than any other part of modern Europe, making them the best proxy on every level, especially phenotype wise imo, for which the new input was even less momentous.

For the study I found the admixture estimates particularly bad, just look at how they compose Corded Ware. Not even that's right imho, yet they could have just copied better models from the literature. Still great to have more Sardinia.
 
Last edited:
You must log in or register to reply here.

This thread has been viewed 104885 times.

Back
Top