search this blog

Saturday, February 11, 2017

Yamnaya-related admixture in Bronze Age northern Iberia


The question of when ancient steppe or Yamnaya-related ancestry first entered Iberia is crucial to the Proto-Indo-European (PIE) homeland debate.

If the steppe or Kurgan PIE hypothesis is correct, then we'd expect this to have happened during the Bronze Age rather than, say, the Medieval Period with the migrations into Iberia of Northern Europeans likely rich in Yamnaya-related admixture like the Visigoths. That's because Indo-European languages are attested in Iberia as early as the Iron Age.

And indeed, the earliest Iberian sample in my dataset to show Yamnaya-related ancestry is Iberia_BA ATP9 from Gunther et al. 2015, dated to 3,700–3,568 C14 cal yBP or the Middle Bronze Age. This has not been reported before, but I'm certain that my finding will be confirmed sooner or later in scientific literature.

Let's start with a basic Principal Component Analysis (PCA) featuring ATP9 alongside a wide range of modern-day and ancient samples from West Eurasia and South Central Asia.


Clearly, ATP9 is shifted east, closer to Yamnaya, relative to the earlier Iberia Chalcolithic (Iberia_ChL) group, and almost clusters with Basques, who are known to harbor significant Yamnaya-related ancestry (see here). I can use formal statistics as well as models based on formal statistics to investigate this in more detail.

Mbuti Yamnaya_Samara Iberia_ChL Iberia_BA D 0.0031 Z 0.859
Mbuti Yamnaya_Samara Iberia_ChL Basque_French D 0.0086 Z 5.035
Mbuti Yamnaya_Samara Basque_French Iberia_BA D -0.0044 Z -1.316

Surprisingly, based on those D-stats ATP9 doesn't appear to share more drift with Yamnaya Samara relative to Iberia_ChL (Z<3). But I suspect this might be due to inflated hunter-gatherer ancestry in Iberia_ChL, so let's try something a little different.

Western_HG Yamnaya_Samara Iberia_ChL Iberia_BA D 0.0188 Z 4.987
Western_HG Yamnaya_Samara Iberia_ChL Basque_French D 0.024 Z 13.163
Western_HG Yamnaya_Samara Basque_French Iberia_BA D -0.0063 Z -1.768

OK, that's basically in line with the PCA above, and I can cement this finding with the qpAdm algorithm. Note the nice chunk of Early Bronze Age steppe (Steppe_EBA) ancestry in ATP9.

Outgroups
AG3-MA1
Chukchi
Dusun
Igorot
Iran_Neolithic
Karitiana
Kosipe
Kostenki14
Lebbo
Levant_Neolithic
Mbuti
Satsurblia
Ust_Ishim
Villabruna

Iberia_BA ATP9
Caucasus_HG 0.038±0.063
Lengyel_LN 0.683±0.066
Steppe_EBA 0.177±0.087
Western_HG 0.102±0.044
chisq 5.216 tail_prob 0.876272

Spanish
Caucasus_HG 0.014±0.028
Lengyel_LN 0.607±0.028
Nganasan 0.011±0.016
Onge 0.013±0.022
Steppe_EBA 0.273±0.043
Western_HG 0.059±0.020
Yoruba 0.021±0.006
chisq 1.605 tail_prob 0.978452

Basque_French
Lengyel_LN 0.590±0.027
Nganasan 0.009±0.016
Onge 0.015±0.022
Steppe_EBA 0.285±0.031
Western_HG 0.096±0.019
Yoruba 0.006±0.006
chisq 3.485 tail_prob 0.900346

Of course, Basques are not Indo-Europeans, so the fact that ATP9 has some Yamnaya-related ancestry doesn't necessarily mean she was an Indo-European. However, it's not unreasonable to assume that the ancestors of Basques incurred gene flow from early Indo-Europeans moving into the Iberian Peninsula, and this probably explains their relatively high level of Yamnaya-related ancestry. So ATP9 may well have spoken an Indo-European language, and if not, then like Basques she probably has Indo-European ancestry

See also...

qpAdm tour of Iberia and France

115 comments:

mickeydodds1 said...

Why Lengyel?

Davidski said...

It works across the board for Europe. As far as I can see, other Neolithic samples often don't work for all Europeans.

I suspect much of the Neolithic admixture in Western Europe might actually be from Late Neolithic East Central Europe.

Rob said...

I think Lengyel is good for steppe admixed groups, but not for where early farmer ancestry is more prominent

Gioiello said...

R-V88 from Italy: definitly demonstrated

R-V88 Z30230/Y7770 * V88/PF6279 * PF6332+59 SNPs15400 ybp, TMRCA CI 95% 13400 10200 ybp" class="age"formed 17100 ybp, TMRCA 11700 ybp
R-V88*
⦁ id:YF07201 [ITALY and UK]
R-M18 PF6372 * PF6319 * YP5453+34 SNPs
⦁ id:ERS256975ITA [IT-CA]
⦁ id:ERS256965ITA [IT-CA]
R-Y7777 SK2065/FGC21014/Y7777 * Y7768 * FGC21018/Y8460+10 SNPs10200 ybp, TMRCA CI 95% 11000 8200 ybp" class="age"formed 11700 ybp, TMRCA 9600 ybp
R-Y7777*
⦁ id:YF07902 GBR [and ITALY and France]
R-Y8451 FGC20993/Y7786 * FGC21063/Y7784 * FGC21033/Y8445+15 SNPs8200 ybp, TMRCA CI 95% 8900 6400 ybp" class="age"formed 9600 ybp, TMRCA 7600 ybp
R-Y8451*
R-V35 V35
⦁ id:ERS256961ITA [IT-CA]

The R-V88 in Iberia (7100 YBP) very likely with the migration from Italy Zilhao spoke about.

Davidski said...

Iberia_BA ATP9 also shows good fits with Iberia_EN and LBK_EN. But many modern-day Euros don't seem to like Iberia_EN, and sometimes not even LBK_EN.

André de Vasconcelos said...

What about modern Iberians, do they show a preferrence for Lengyel rather than Iberia_EN aswell, David?

Rob said...

Dave
My point is that Lengyel should only be the EEF associated with whatever steppe ancestry admixed into them. Basques and other Europeans, will also have EEF /MNE ancestry from Baalberg, or MN Iberia, or Neolithic Greece perhaps .
So the above results seem rather broad brushed , as Basques or Spaniards shouldn't be getting all their EEF ancestry from Lengyel

John Smith said...

R1b-M269 most likely and for a fact a very related line was found in Northern Iberia at 3500bc almost certainly as a minority. This has been known for more than a year https://genetiker.wordpress.com/y-snp-calls-for-atp3/. In addition the only ancient groups found dominated by H 7,000 years ago are in Romania if any one is aware of others group at this time that are dominated by H please post.

Matt said...

What do we get with:

Kotias Yamnaya_Samara Iberia_ChL Iberia_BA
Kotias Yamnaya_Samara Iberia_ChL Basque_French
Kotias Yamnaya_Samara Basque_French Iberia_BA
EHG Yamnaya_Samara Iberia_ChL Iberia_BA
EHG Yamnaya_Samara Iberia_ChL Basque_French
EHG Yamnaya_Samara Basque_French Iberia_BA
Samara_Eneolithic Yamnaya_Samara Iberia_ChL Iberia_BA
Samara_Eneolithic Yamnaya_Samara Iberia_ChL Basque_French
Samara_Eneolithic Yamnaya_Samara Basque_French Iberia_BA

Al Bundy said...

Interesting Davidski.But IE in Europe seems to be game over anyway for the Southern folks.You'll remember some people arguing on here that Centum came through Anatolia and Satem through the steppe but Insular Celts have high steppe admixture as you've pointed out.With this and the BB stuff coming out it's good news for those interested in Iberia.

Al Bundy said...

@John Smith Maju seems to know all about Iberia and Basques,and he is very interested in R1B in Western Europe.He and Davidski have talked about R1B origin on here.The details are way over my head though.I wonder what Maju has to say about the latest info.

André de Vasconcelos said...

Al Bundy, Iberia was never a contender for the IE homeland. At most it's related with r1b and/or Bell Beaker origin.

Al Bundy said...

Oh yea I just meant as in PIE coming south of the steppe.As I understand that's the only serious contender to the Kurgan theory and it's on thin ice especially in Europe.The whole r1b and BB thing is another matter which you know a lot more about th
an I do.

André de Vasconcelos said...

Oh sorry, my bad!
Hey, I'm just a blog lurker, I probaby know much less about the issue -BBs and r1b- than you do. Anyway, very interested in reading the coming discussion, always up for that when it comes to Iberia.

Al Bundy said...

in the Caucasus and then Anatolia for Greek and Italic or something.That seems to be what the Max Planck folks are saying and of course IndoIranian from there as well and Yamnaya a secondary homeland for a subset of IE Balto Slavic and perhaps Germanic.As Davidski said Heggarty has trashed the Kurgan theory in the past so maybe it's just a salvage job with ancient d na not going there way.

Al Bundy said...

I've learned a lot from this blog but as far all the spreadsheets and data go I know very little.Despite some people having different theories here the data always comes first or what's in the literature which is as it should be.

Al Bundy said...

The PIE homeland debate definitely inspires a lot of passions.Some people don't want the language to have anything to do with Asia and the steppe is obviously in Europe.Some Southern Europeans hope the language comes from the South.When the next big Reich paper comes out there might be a lot of whining.

Davidski said...

@André and Rob

Early Neolithic farmers from Europe are all very similar, and generally interchangeable for most Europeans in these sorts of tests. It's just that some don't always work for all Europeans in the model that I've set up, while Lengyel LN does. So I just use Lengyel LN nowadays.

I haven't tested yet which Neolithic farmers are specifically preferred by different European ethnic groups.

@Matt

Rather neutral.

Caucasus_HG Yamnaya_Samara Iberia_ChL Iberia_BA -0.007 -1.418 361009
Caucasus_HG Yamnaya_Samara Iberia_ChL Basque_French 0.0012 0.521 813245
Caucasus_HG Yamnaya_Samara Basque_French Iberia_BA -0.0072 -1.641 358661
Eastern_HG Yamnaya_Samara Iberia_ChL Iberia_BA -0.0015 -0.346 337686
Eastern_HG Yamnaya_Samara Iberia_ChL Basque_French -0.0023 -1.228 759975
Eastern_HG Yamnaya_Samara Basque_French Iberia_BA -0.0014 -0.356 332780
Samara_Eneolithic Yamnaya_Samara Iberia_ChL Iberia_BA -0.0003 -0.063 263521
Samara_Eneolithic Yamnaya_Samara Iberia_ChL Basque_French -0.0016 -0.895 592282
Samara_Eneolithic Yamnaya_Samara Basque_French Iberia_BA 0.0018 0.42 248084

Simon said...

Do the Lengyel matches in Yamnaya-rich populations in Europe suggest a Middle PIE homeland in the Late Tripolye culture?

Davidski said...

Do the Lengyel matches in Yamnaya-rich populations in Europe suggest a Middle PIE homeland in the Late Tripolye culture?

You would need Late Tripolye to be rich in R1a, R1b and steppe admixture, which I think is very unlikely.

I'd say Lengyel LN just represents the type of really late Neolithic farmer that had a wide range just before the homogenization of Europe via all sorts of population movements during the Bronze Age.

Davidski said...

Some more D-stats like those above, with Western_HG as the outgroup. Iberia_BA's eastern shift is unlikely to be the result of admixture from the Near East.

Western_HG Yamnaya_Samara Iberia_ChL Iberia_BA 0.0188 4.988 361193
Western_HG Yamnaya_Kalmykia Iberia_ChL Iberia_BA 0.0194 4.561 355695
Western_HG Caucasus_HG Iberia_ChL Iberia_BA 0.0255 4.549 362813
Western_HG Eastern_HG Iberia_ChL Iberia_BA 0.0213 4.347 339025
Western_HG Samara_Eneolithic Iberia_ChL Iberia_BA 0.0204 3.819 263988
Western_HG Iran_Neolithic Iberia_ChL Iberia_BA 0.0132 2.427 305839
Western_HG Iran_ChL Iberia_ChL Iberia_BA 0.0099 2.225 343912
Western_HG Armenia_EBA Iberia_ChL Iberia_BA 0.0029 0.643 345349
Western_HG Jordan_EBA Iberia_ChL Iberia_BA -0.0031 -0.662 319072

Simon said...

Are there any proofs of widespread Lengyel LN-like admixture in Europe prior to Corded Ware, or does it appear to be a mostly local C/E European Neolithic population? Are you suggesting Lengyel superimposed themselves over the other Neolithics like LBK etc?

Is Lengyel the best match for EEF ancestry in CW and Beaker samples as well? If so, I'm very much inclined to believe that Indo-European languages younger than Anatolian and Tocharian have a common origin in C/E Europe and a formative admixture event that occured as the Kurgans conquered Tripolye (/related cultures).

Regardless, it's likely that Tripolye had something to do with the spread of Lengyel-like ancestry as the oldest evidence of wheeled vehicles is from the troubled Late Tripolye culture (dated ca 3800 BC)

epoch2013 said...

@David

Don't delete the old post. Add an addendum to it. Show previous work as it shows how things evolved. It really is necessary.

Davidski said...

Are there any proofs of widespread Lengyel LN-like admixture in Europe prior to Corded Ware, or does it appear to be a mostly local C/E European Neolithic population? Are you suggesting Lengyel superimposed themselves over the other Neolithics like LBK etc?

I don't know, it's hard to check that sort of stuff with genome-wide DNA at this resolution. Maybe looking at ancient mtDNA from Lengyel and related groups might help to reveal something?

Grey said...

the very early atlantic megalith culture centered in southern portugal had maritime connections all the way to sweden and denmark in one direction and sardinia in the other

so any originally steppe descended population that reached either end of that network could potentially find their way to iberia

#

iirc (?) one of the things about basques is they are mostly lacking the chg component that usually follows the IE signal. if i haven't remembered wrong one possible explanation for that (if you believe the chg component is likely the result of IE eventually over-running steppe farmers when they were strong enough) is some yamnaya related folks moved west before that event and somehow ended up in iberia (maybe as a non-conquering minority thereby losing IE language along the way)

#

it's odd how atlantic megalith is ignored - you'd think an east med level civilization in south portugal with connections in all directions would be relevant but maps like the one from Max Planck from the "dead cat bounce" post don't even mention the atlantic coast civ which links the end of their map's green and red arrows

http://eurogenes.blogspot.co.uk/2016/10/dead-cat-bounce.html

#

BB

if "BB" was the result of a common population x who merged with different substrates in different regions that could explain three different regions, Iberia, Germany, Holland claiming to be the origin of BB - similarities from the x population, dissimilarities from the substrate

in which case the substrate for iberia would probably be atlantic megalith

so those people who think BB has a north african connection might have a point if atlantic megalith had a north african connection i.e. maybe the particular substrate of Iberian BB has the north African connection

Matt said...

Hmm... So no shift of Iberia_BA vs Iberia_Chl towards Yamnaya_Samara relative to any of CHG, EHG, Samara_Eneolithic, only WHG. Just for completeness, what happens with Iberia_EN, Iberia_MN, Barcin_N and LBK_EN? Same as WHG?

andrew said...

"Of course, Basques are not Indo-Europeans, so the fact that ATP9 has some Yamnaya-related ancestry doesn't necessarily mean she was an Indo-European. However, it's not unreasonable to assume that the ancestors of Basques incurred gene flow from early Indo-Europeans moving into the Iberian Peninsula, and this probably explains their relatively high level of Yamnaya-related ancestry. So ATP9 may well have spoken an Indo-European language, and if not, then like Basques she probably has Indo-European ancestry."

This is a very unreasonable assumption. There is no reasonable model in which Basque receive more distinctively non-Neolithic genetics from Indo-Europeans.

A far more plausible hypothesis is that there is more than one linguistic population with origins somewhere in the vicinity of the steppe that have Yamnaya-like ancestry. In the North, those were Indo-Europeans who came to be exemplified in the Corded Ware culture.

In the South, there were people with a non-Indo-European language family who were probably ancestral to both the Minoans and in a male dominated migration the Bell Beaker people whose language under the formative substrate influence experienced by the earliest Bell Beaker people gave rise to the Vasconic languages of which Basque is the only surviving representative. The original Bell Beakers probably migrate to their archaeological point of origin in Iberia's major cities to exploit the tin deposits found there which their superior to local populations metalworking skills and mining knowledge allows them to exploit. These men (or at least those of them who participate in Bell Beaker expansion) marry Iberian women with elevated levels of mtDNA H which is what causes its levels in the gene pool of Western Europe to grow in tandem with Y-DNA R1b.

For roughly a thousand years Europe is divided between a Vasconic West known for their archery and cattle and distinctive pottery and metalworking and religious beliefs, and an Indo-European East known for their chariots and cavalry and horses and their distinctive pottery and metalworking and religious beliefs distinct from those of the Vasconic people despite the fact that the two have origins not geographically far from each other on the Steppe.

Then, following the climate events that give rise to Bronze Age collapse. Most of the Vasconic, Y-DNA R1b-M269 people of Western Europe who acquired the language and Y-DNA via the Bell Beaker fell to Urnfield/Celtic Indo-European people in the late Bronze Age/early Iron Age and experienced linguistic shift, but the Basque and some neighboring Vasconic people survived into the Roman era without falling to linguistic shift until eventually only the Basque were left. The later shift from Vasconic to Celtic leaves only a modest genetic impact roughly reflected by the percentage of R1a in Western Europe.

andrew said...

"OpenID epoch2013 said...
@David

Don't delete the old post. Add an addendum to it. Show previous work as it shows how things evolved. It really is necessary"

I agree with epoch2013. Don't delete the old post, just note that it has been updated. Showing the chain of thought that leads to a final conclusion really is valuable.

Samuel Andrews said...

@David,
"Maybe looking at ancient mtDNA from Lengyel and related groups might help to reveal something?"

There is a bunch of Lengyel mtDNA sampled from Hungary. It's the same as earlier Neolithic Hungarians and East Germans. Unlike Iberians and East Germans from the same time they didn't have more U5 than earlier farmers. So like NE7 not a lot of WHG.

So far Neolithic and Bronze age mtDNA from deep in Poland is mostly H. A decent amount of Neolithic mtDNA from Romania is mostly H. Also it looks like the Near Eastern ancestors of Yamnaya and Catacomb were mostly H. So I think a mixture of natural selection, Steppe admixture, and Eastern EEF could have possibly made modern European mHG frequencies different from Neolithic Europe.

Al Bundy said...

From where do you think IE came to Greece?

Romulus said...

Didn't we JUST see a study on Iberian mtdna that showed Steppe like mtDNA showing up in the BA? I'm sure this isn't 10 posts down.

Rob said...

@ Dave, etc

IMO nMOnte can demonstrate where a group derives most of its EEF :

BASQUE
Basque_Spanish
Baalberge_MN:I0560 34.55 %
Hungary_N:I1495 33.35 %
Yamnaya_Samara:I0231 18.05 %
Latvia_HG:ZVEJ32 13.95 %


Yes Lengyel registers, but so does Baalberg, representing an earlier MNE stratum. Curiously, Iberia Chalcolithic or MN isn;t there


SPANISH

Spanish_Extremadura
Iceman_MN:Iceman 36.65 %
Latvia_MN1:ZVEJ26 19.9 %
Barcin_Neolithic:I1099 16.2 %
Anatolia_Chalcolithic:I1584 8.05 %
Iran_Neolithic:I1290 6.9 %

Spanish_Valencia
Baalberge_MN:I0560 51.2 %
Armenia_EBA:I1635 16.5 %
Yamnaya 12.45 %
Latvia_MN1:ZVEJ26 6.45 %
Remedello_BA:RISE489 6.35 %



I think once we get full coverage of Iberia, it'll be more complex than even imagined. Again, Iberia Chalcolithic doesn't really feature (? surprising perhaps, shocking no - as there is very little I2a2 in Spain these days, which appears to have been the dominant lineage in MNE and Chalc of Iberia). Instead, Extremadura shows more Italian MNE acnestry (in line with the fact that ATP 9 ancestry peaks there; whilst Valencia shows more Yamnaya & Baalberg (Central European MNE).


@ Matt

"Rob, why model with Barcin rather than Iberia_EN, Iberia_Chal or Iberia_MN? Unless you did and Barcin was preferred.'

Exactly. I did. The only place in Europe were "Barcin" type ancestry was around in BA was probably the Balkans, and then, no unadmixed. I.o.w. it's showing up becuase the ideal groups has not yet been sampled and included in sources. So I was thinking the Barcin was already mixed with somethign Ukraine Neolithic like (? some Copper Age Balkan culture like Cernavoda)

(I now these aren;t direct tests, but still potentially interesting)

Kurd Dgk said...

David and everyone,

My website, EurasianDNA.com is live now, for anyone interested in South & West Asian population histories. Although I am not a prolific poster, my intention is to have something posted maybe bi-weekly.

Rob said...

@ Matt

Further - even with BB and CWC included, neither they nor Yamnaya appear in ATP9.
Oh well, must be an oddity, and Dave has squeezed from it whatever he can. Hopefully we'll clarify everything soon. (btw I am in no particular 'Beaker camp'. I just find it all very interesting).

Davidski said...

@Matt

Iberia_EN Yamnaya_Samara Iberia_ChL Iberia_BA 0.023 6.636 354819
Iberia_EN Yamnaya_Samara Iberia_ChL Basque_French 0.0397 25.434 799702
Iberia_EN Yamnaya_Samara Basque_French Iberia_BA -0.0185 -5.549 346565
Iberia_MN Yamnaya_Samara Iberia_ChL Iberia_BA 0.0264 6.705 337176
Iberia_MN Yamnaya_Samara Iberia_ChL Basque_French 0.0436 26.694 760791
Iberia_MN Yamnaya_Samara Basque_French Iberia_BA -0.0186 -5.018 318443
Barcin_Neolithic Yamnaya_Samara Iberia_ChL Iberia_BA 0.0211 7.151 361110
Barcin_Neolithic Yamnaya_Samara Iberia_ChL Basque_French 0.0345 25.819 813361
Barcin_Neolithic Yamnaya_Samara Basque_French Iberia_BA -0.0122 -4.4 357592
LBK_EN Yamnaya_Samara Iberia_ChL Iberia_BA 0.021 6.592 360369
LBK_EN Yamnaya_Samara Iberia_ChL Basque_French 0.032 22.672 811669
LBK_EN Yamnaya_Samara Basque_French Iberia_BA -0.0115 -3.813 355579

Samuel Andrews said...

@David,

Do you by chance have DNA from Madagascar.

Davidski said...

No, why?

Davidski said...

Using transversions only. So with an even higher quality sequence and more power we'd see significant Z scores IMO.

Yoruba Yamnaya_Samara Iberia_ChL Iberia_BA 0.0107 1.691 56637
Yoruba Yamnaya_Samara Iberia_MN Iberia_BA 0.0078 1.204 58134

Rob said...

Dave

Could it be that ATP9 shares more drift with Yamnaya cf preceding Iberians due to a 'third source' which contributed to, both, ATP9 and Yamnaya

Davidski said...

I don't know, but it seems like it was the usual type of mixture of mostly CHG and EHG and very similar to Yamnaya Samara.

Rob said...

But wouldn't it be Beta to also perform the above tests with Kotias, Ukraine NL/HG and Samara Eneolithic to confirm ?

alobrix said...

In the Iron Age Galicia was deeply indoeuropean, but not other iberian territories like Valencia. Could you find any differences between the Bronze Age ancestors of these present populations?

Davidski said...

@Rob

Can't check the Ukrainian angle. Not enough markers at this time. But Kotias is in the test above as Caucasus_HG. And when I swap Samara_Eneolithic for Steppe EBA, I get this.

Iberia_BA
Caucasus_HG 0.090
Lengyel_LN 0.681
Samara_Eneolithic 0.111
Western_HG 0.119
chisq 5.454 tail_prob 0.858873

In other words, Caucasus_HG and Western_HG go up a bit, but the fit is about the same. So the steppe population could have been like Samara Yamnaya or Samara Eneolithic, or anything in between, or more CHG-heavy. Impossible to say at this stage. Need more ancient steppe samples.

But if more like Samara_Eneolithic, then the extra Caucasus_HG admix arrived independently.

@alobrix

Nah, not at this stage. Modern-day Iberians are too mixed, with more recent ancestry than the Bronze Age from different parts of Europe which would skew the results.

Matt said...

Comparing compatible D for stats for Iberia_BA vs French Basque:

http://i.imgur.com/tlZVyxq.png

(Z stat obviously confounded for comparison by SNP coverage).

French Basque shifted more than Iberia_BA on:
CHG->Yamnaya, WHG->Yamnaya, LBK->Yamnaya, Barcin_N->Yamnaya and particularly on Iberia_MN->Yamnaya.

French Basque no more shifted than Iberia_BA on:
EHG->Yamnaya, Samara_Eneolithic->Yamnaya

Suggests maybe French_Basque is different from Iberia_BA in having increase of other European_MN ancestry, with similar amounts of EHG? Iberia_BA more of a direct Iberia_Chal+Steppe, while French_Basque includes also some levels of Europe_MN from other regions of Europe.

Possible to run?:

Hungary_CA Yamnaya_Samara Iberia_Chl Basque_French
Hungary_EN Yamnaya_Samara Iberia_Chl Basque_French
Esperstedt_MN Yamnaya_Samara Iberia_Chl Basque_French
Baalberge_MN Yamnaya_Samara Iberia_Chl Basque_French
Sweden_MN Yamnaya_Samara Iberia_Chl Basque_French
Irish_MN / Ballynahatty Yamnaya_Samara Iberia_Chl Basque_French
Hungary_CA Yamnaya_Samara Iberia_Chl Iberia_BA
Hungary_EN Yamnaya_Samara Iberia_Chl Iberia_BA
Esperstedt_MN Yamnaya_Samara Iberia_Chl Iberia_BA
Baalberge_MN Yamnaya_Samara Iberia_Chl Iberia_BA
Sweden_MN Yamnaya_Samara Iberia_Chl Iberia_BA
Irish_MN / Ballynahatty Yamnaya_Samara Iberia_Chl Iberia_BA

Simon_W said...

@ Al Bundy
"From where do you think IE came to Greece?"

We, that is the public who doesn't have secret insider info, have no idea what the genetics of Bronze Age Greece were like. That's still one of the big, hot open questions. Maybe Greek came from the north or from somewhere around Armenia - either theory is 100% compatible with the Steppe theory.

Simon_W said...

@John Smith
"R1b-M269 most likely and for a fact a very related line was found in Northern Iberia at 3500bc almost certainly as a minority. This has been known for more than a year https://genetiker.wordpress.com/y-snp-calls-for-atp3/."

Not really. ATP3 is positive for one out of 105 SNPs that are currently regarded as phylogenetically equivalent to R1b-M269. ATP3 hasn't been proven positive for M269 or any of its other 104 equivalents. Hence the lineage leading to him may have split from the lineage leading to all living R1b-M269 males 10000 years ago or more.

postneo said...

@david
"That's because Indo-European languages are attested in Iberia as early as the Late Bronze Age"

Can you kindly share evidence for this?

Simon_W said...

@postneo

Afaik there are no written documents from Iberia dating to the LBA. What David perhaps refers to is the theory that the Tartessian language, which is attested from the 7th century BC onwards, is Celtic-related IE. And it has been argued on archaeological grounds that this could only be explained with Bronze Age movements.

Simon_W said...

@andrew

I'm not completely against this intellectual game of deriving Basque from somewhere near the steppe. However, assuming that all R1b steppe people were not IE would make it hard to explain the Anatolian branch of IE in Anatolia. As we know, the Poltavka culture was still dominated by R1b. Only with the beginning of the Potapovka culture about 2500 BC we see a change to an R1a dominated population on the Russian steppe. Hittite is attested from the 16th century BC, but its ancestral versions seem to have split from the IE mainstream at a relatively early date, certainly earlier than the split between Balto-Slavic and Indo-Iranian, I'm not sure how you want to reconcile this with the late expansion of R1a people on the Pontic-Caspian steppe.

And although Basques have more steppe ancestry than Italians, Albanians and Greeks, they have relatively little steppe ancestry in a west European context. According to David's recent qpAdm tour of Europe, Scots have 45.5% - in contrast to French Basques with 28.5%, and Highland Scots don't have less R1b. But granted, we don't know when the ancestors of the Scots started to talk in Celtic speech...

Chad Rohlfsen said...

result: Iberia_CA Iberia_BA EHG CHG 0.0015 0.312 19264 19206 388713
result: Iberia_CA Iberia_BA EHG Yamnaya_Samara 0.0010 0.240 18230 18195 385757
result: Iberia_CA Iberia_BA EHG Samara_CA 0.0027 0.486 12437 12371 276012
result: Iberia_CA Iberia_BA EHG Armenia_CA -0.0109 -2.436 17922 18316 366593
result: Iberia_CA Iberia_BA Samara_CA Yamnaya_Samara -0.0007 -0.160 13617 13636 287777
result: Iberia_CA Iberia_BA Samara_CA CHG -0.0016 -0.288 14186 14231 289283
result: Iberia_CA Iberia_BA Samara_CA Armenia_CA -0.0121 -2.573 13504 13836 277313
result: Iberia_CA Iberia_BA Yamnaya_Samara CHG 0.0002 0.051 20538 20530 420080
result: Iberia_CA Iberia_BA Yamnaya_Samara Armenia_CA -0.0123 -3.732 18932 19405 388255

It's pretty clear the population is from North of the Caucasus, in Eastern Europe. Armenians are too Near Eastern to be a source.

Chad Rohlfsen said...

result: Iberia_CA Iberia_BA Ukraine_HG/N1 EHG 0.0325 1.319 432 405 10996
result: Iberia_CA Iberia_BA Ukraine_HG/N1 Samara_CA 0.0024 0.079 318 317 8150
result: Iberia_CA Iberia_BA Ukraine_HG/N1 Yamnaya_Samara 0.0046 0.221 469 465 11615
result: Iberia_CA Iberia_BA Ukraine_HG/N1 CHG -0.0144 -0.610 462 476 11715
result: Iberia_CA Iberia_BA Ukraine_HG/N1 Armenia_CA -0.0214 -0.967 436 455 10792

The top one and bottom two would definitely be significant with higher SNP counts.

Simon_W said...

Oh, and @andrew

Your postulating of a connection between Minoan and Basque seems far fetched. Is there any linguistic evidence? I mean the fact that modern Cretans have R1b cannot be regarded as evidence... Perhaps it would make more sense to connect Basque with Paleo-Sardinian, which would point to a Neolithic, rather than steppic connection (no European pop has less Steppe_EMBA than Sardinians):
https://en.wikipedia.org/wiki/Paleo-Sardinian_language

Romulus said...

@Simon_W

Linear B & Basque:

http://www.faculty.ucr.edu/~legneref/bronze/linerb.htm

Chris Davies said...

If Basques were known for archery, it's no surprise to me that their word for 'arrow' is related to a word for the same in some Chadic languages.

"Arrow"

Basque: 'gezi'

Kwang [East Chadic]: 'kēsé'
Sokoro [East Chadic]: 'kɛsɛ́'
Mbara [Central Chadic]: 'kèsé'

Matt said...

@ Simon_W, off topic, in terms of R1b, I don't have a detailed view and you seem to have good knowledge and English language literacy, so was wondering if I could bounce this general picture off you:

1. Karmin 2015 (http://genome.cshlp.org/content/25/4/459) found that there was a main expansion of the entire R1b clade around 15,000 YBP. This was around the same time frame as the post LGM expanion of I2 (and various other lineages). Then expansion of subclades within more recently.

2. Ponzik 2016 (http://www.nature.com/ng/journal/vaop/ncurrent/full/ng.3559.html), which used the 1000 Genomes whole genomes just found an expansion particularly of R1b-L11 around 5,000 - 7,000 YBP. This is the time frame of mid-Holocene expansions by R1aZ93, R1aZ282 and I1.

(Poznik confirmed Batini 2015 - http://www.nature.com/articles/ncomms8152, also confined to European samples)

Crucial difference between these two: Karmin is sampling more populations, though with many fewer males per pop - "456 geographically diverse high-coverage Y chromosome sequences", inc. Central Asian populations (Bashkirs, Tajiks who have the diverse R1b that predicts the early divergence ~15,000 YBP). Poznik essentially gets all it's R1b from West Europeans - GBR, CEU, IBS, TSI - and admixed American populations who derive it from Iberians and British - ASW, CLM, PUR. Therefore only really able to detect the secondary expansion of R1b with any detail.

Seems possible to me that first expansion of R1b relates to expansion and the mixing of Villabruna clade in Southeast Europe through Eastern Europe and Middle East and steppe (maybe even gets to Africa by this expansion), but doesn't get to Western Europe (WHG = I2 ), while the secondary expansion is the expansion of steppe subclades (along with I1 and main branches of R1a).

Alternatively the first expansion of R1b could relate to ANE populations (outliers in Karmin in Tajik and Bashkirs speaks for this) or EHG populations, but Villabruna seems evidence for the Villabruna clade atm. More intensive whole genome sequencing of R1b males outside of Europe would help with this.
Anything seem obviously wrong with this?

Al Bundy said...

@Simon Linguists like Anthony and Ringe who support steppe PIE group BaltoSlavic and IndoIranian together but others group II with Greek.The former seems to give PIE a more northern origin while the latter a southern one.

Rob said...

Matt

nice overview

We still don't know for sure where I2 or R1 expanded from after the LGM, so labels like "WHG" might be presumptive at this stage
I actually think it's possible that R1b came via Anatolia & Southern Europe rather than via Siberia but this would require a leap frogging through Basal territory, or a timing before BE had reached the caspian region

Davidski said...

That's because Indo-European languages are attested in Iberia as early as the Late Bronze Age.

I'll change this to Iron Age.

Rob said...

@ Al / SimonW

There are many different scenarios possible - as the potential for steppe/ central European influences reaching Greece aren't singular. But there is also evidence for movements from Anatolia.
So it'll be a matter of correlating the evidence in the best way.
As for the branchings and language relations, the fact that there are different positions means that no rigid approach of 'family tree branches' is optimal as of yet.

Al Bundy said...

Davidski has speculated that steppe admixture in Greece could be as high as 30 percent.Then there is the whole debate about how Slavic modern Greeks are which would would have presumably come to Greece after the Yamnaya migrations.So to solve the puzzle we need Ancient Greek DNA which I hear is on the way.

Olympus Mons said...

@Simon,

"Not really. ATP3 is positive for one out of 105 SNPs that are"

So, is that SNP found in any other haplogroup or subclade?
If not, since it was very low coverage, the other 104 are irrelevant if that one found is exclusive to M269. right?

Al Bundy said...

Modern Greeks have a lot of J and E y haplogroups which don't seem to have much to do with Yamnaya.Spanish took hold in Latin America more through elite dominance and English I guess both, elite dominance and a significant AngloSaxon genetic contribution 30 or 40 percent depending on the region in Britain.Even a smallish Yamnaya impact in Greece could have affected language change.

Rob said...

@ Al

Modern colonialist analogies fail for Bronze Age Greece/ Balkans, for what should be obvious reasons: 100 Spanish conquistadors with guns and horses conquered the native American Empires.
The technological differences between Yamnaya pastoralists & early Helladic Greece would not be so skewed, in fact, they'd be the reverse. Anyhow, Yamnaya never reached Greece, so that cannot be an explanation.

Al Bundy said...

Thanks @Rob

Rob said...

I should clarify , .... not directly but could be indirectly

Al Bundy said...

Right

postneo said...

at simon_w:

https://en.wikipedia.org/wiki/Tartessian_language

The language is unknown and need not be celtic or even IE. Of course there may have been IE presense in the iron age, hypothetically but tartessian cannot be taken as proof.

Davidski said...

Kum4 from Bronze Age Troy appears to show significant steppe admixture, so if that's correct then there will be significant steppe admixture in Bronze Age Greece.

John Smith said...

@ Rob

"Anyhow, Yamnaya never reached Greece, so that cannot be an explanation."

Say what? If not Yamnaya, then certainly a closely related IE people reached Greece. Is there really any debate about this? This was known before ancient DNA was ever a thing.

There are good arguments out there that Homer's epics were actually rooted heavily in old IE Steppe folklore as opposed to being any sort of real historical account of the Trojan war. Whatever the truth about Homer, it is clear that the Hellenes completely changed the culture of Greece: from a Neolithic styled "Goddess" culture to a militaristic hero culture led by tribal chieftains that worshiped a "Sky God" pantheon. Odysseus was the personification of a mortal hero (i.e. a man, but probably a fictional man) and Achilles was their idea of a demigod. And, of course, everyone is familiar with their pantheon of Gods which is the same from Greece all the way to Iceland. It changes very little across Europe.

Moreover, the Hellenes (which included 4 tribes: Ionians, Aeolians, Acheans and Dorians) were typically described as fair and blond in contrast to the native Minoans and Pelasgians who were swarthy and dark haired. This would make sense if the Hellenes were a northern Steppe people with lots of EHG heritage. (I believe the EHG's, or at least some sort of HG, was the source of this phenotype that is still common in the Baltic today).

Of course, these invaders didn't completely wipe this old culture away, as it was still prevalent in Athens and the islands long after their arrival. Indeed, the Athenians were most likely primarily of EEF stock. In any case, without the natives, Greece would have never been as great and influential as it was (the IE's weren't literate for instance and they certainly learned a lot about art and architecture from Old Europe).

A few centuries later, we see the Romans (specifically Tacitus) write about the contrast between classical culture and pure isolated IE culture of Germania. Tacitus describes them as lazy and useless for anything except war where he admits they are superior. This high emphasis on war (and not much else) sounds oddly familiar to Sparta which most certainly was almost a pure IE city-state. The Spartans themselves often described themselves as "pure" Dorians and restricted citizenship based on one's lineage to Dorian ancestors.

TL;DR -- I suspect Greece during the classical era was mostly EEF by heritage but with a strong component of Yamnaya ancestry, especially on the mainland in old Spartan territory. I would be absolutely shocked if R1b or R1a were not found in bronze and iron age Greece.

Samuel Andrews said...

@David,
"Kum4 from Bronze Age Troy appears to show significant steppe admixture, so if that's correct then there will be significant steppe admixture in Bronze Age Greece."

Who's Kum4? I thought the youngest Greek DNA sample is Chalcolithic. Is he of high coverage?

Amanda S said...

"@andrew...
This is a very unreasonable assumption. There is no reasonable model in which Basque receive more distinctively non-Neolithic genetics from Indo-Europeans.

A far more plausible hypothesis is that there is more than one linguistic population with origins somewhere in the vicinity of the steppe that have Yamnaya-like ancestry. In the North, those were Indo-Europeans who came to be exemplified in the Corded Ware culture."

Andrew, you've obviously put a lot of thought into this theory. Still I can't quite grasp why you don't think that it's possible that, given the contest between two languages, a local one and one spoken by invaders, you don't think that it was possible that in this part of Europe, the local language (i.e. that spoken by the Neolithic farmer population) survived as the dominant language even whilst the invader's genetics became part of the general population. There are, after all, other examples from history where this has happened.

As I understand it one of the important innovations which came in the Early Bronze Age (and might well have come via the Steppe people) was woolly sheep. Neolithic people, such as Otzi, wore clothes made of animal skins. With wool would have come a revolution in the clothing that people wore and whole new areas of human activity connected to preparing, spinning and weaving wool. Woollen goods could also be traded in a period when we can see an expansion of trading networks i.e. Beaker people.

Woolly sheep became particularly important in upland areas, such as the Basque country, as they are able to thrive in areas with harsh climates. Thus the introduction of these animals allowed for an expansion in the productivity of upland areas and an increase in the number of people that they could support.

Another feature which is associated with upland livestock farming in Europe is transhumescence; the seasonal movement of herders with their flocks (and herds) to utilise summer pastures. Although it's almost died out as a way of life, transhumescence was practiced all over Europe.

https://en.wikipedia.org/wiki/Transhumance

The evidence so far is that Western Europe was not settled directly by people from the Steppe. Instead the intrusion came from intermediate societies formed by Steppe people in association with European farmers. This is what the paper on the Bronze Age people from Rathlin Island showed.
So one can perhaps even suggest another possiblility for the origins of the Basque language which is that the Basque country was settled by transhumescent livestock herders from a region such as the Carpathians, whose ancestry was a mixture of Neolithic European farmers and Steppe herders. This population could have carried the male biased Y chromosomes of their Steppe ancestors dating from the period when they had been militarily dominant but could have spoken a Neolithic farmer language which they brought from the Danube area. This is purely speculative and is just to show that quite a lot of scenarios can be considered.

Aram said...

Simon W
The problem is not only Anatolians but also Greeks and even more Latins. Also most of I and R1a in Britain are recent invasions related to Vikings and Anglo Saxons.
In a Cyprus paper Greek language expansion into Cyprus showed correlation only with E V13, R1b and J2 M67. The R1a was not even included while 99 % of I2 in Cyprus showed connection with Danube, thus Slavs.

Rob said...

@ John Smith

You're completely right :the Indo-European speaking Mycenean Greeks were Indo-European.
But neither the pseudo-scientific beliefs that the archaic Greeks were blonde & blue eyed, nor tales from Tacitus (who lived 3000 years after the arrival of proto-Greeks) change the very specific fact which I stated - the Yamnaya culture did not reach Greece propper.
This is different to stating 'no steppe or steppe admixed people reached Greece", is it not ?
Also, the Homeric sagas were created in the Archaic period of Greece ("Dark Ages", late Iron Age), which , again, is some 2500 years removed from the PIE arrival, also separated by profound social , cultural & perhaps even demographic changes after the BA 'collapse'. By then, a warrior ethos was not restricted to IE peoples, if it ever was.
Also, weren't Yamnaya Swarthy ?
Lastly, which specific R1b & R1a subclades would you link with Myceneans ?

Samuel Andrews said...

Iron age Spanish mtDNA. Published May 2016.

http://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0155342

Table 1 shows the results...

http://dx.doi.org/10.1371/journal.pone.0155342.t001

btw, One is U5a1b1, same type of U5a1 found in Spanish Bell Beaker. One is W, there isn't a single W from Neolithic Spain.

24 samples. 9 of 24 are H1/H3 and only 1 of 24 are K. Of the Neolithic Spanish samples H is only 20% and K is 20-30%. Essentially 100% of Neolithic Spanish H is H1/H3 but the frequency is lower than in these Iron age samples.

As far as I know H is 40% in Iron age Polish and Scandinavian mtDNA aswell while in Neolithic Germany and Hungary it was about 20%. So Iron age mtDNA so far supports the argument natural selection changed mHG frequencies in Europe.

Rob said...

@ David

Kum4 follows the levels from which Kum6 derives after a hiatus. So it could be a new population. I could not find too much information about it, but its burial (flexed position to right, head to south, associated Neonate burials) doesn't recall anything from Yamnaya. BUt it's clearly from somewhere around the Black Sea going from the PCA.

Rob said...

@ SimonW

ATP3 was R1b if not M269 specifically.

Davidski said...

Kum4 looks like it's from the same place as Corded Ware at the moment, which is very telling if true, because Corded Ware is in all likelihood from the steppe.

Ric Hern said...

Vucedol Culture.

Rob said...

Could be, but her grave dates to c. 3800-3400 BC IIRC, so it must be from some Eneolithic / Copper Age group.

Aram said...

Chris Davies

That word could be a IE influence.
In Armenian 'Ges' means long hair, or something long. The same word is present in Iranian. Also the name Caesar is sometimes etymologised deriving from hair.


Aram said...

I forget to add that the word 'ges' in Armenian was also used in ancient times for long things used in home/house furniture. Type of eaves, cornice or ledges. This word is supposed to be a Greek loanword. Geisos in Greek.

Cossue said...

Whilst Tartessian language itself is probably a non-IE language -pace the efforts of professor John T. Koch and of other scholars- Tartessian inscription contain Indo-European names and, maybe, formulas: http://ifc.dpz.es/recursos/publicaciones/29/54/26koch.pdf.

The situation is similar to Latin inscriptions in Hispania, Gallia or Britain containing Celtic names and expressions.

Also I would like to note that by the Roman conquest, the name of the person, gods, rivers, mountains, towns, of all the northwestern quarter and of all of central Iberia, were fully Indo-European, but at the same time very peculiar and very different of what you could find in Gallia, Italia or Britain... And divergence usually implies time.

Cossue said...

As an example, take a name like Coropotus (a composite meaning 'War/Army' *coro-, and *-portus 'Lord'): It's not Celtic, since it preserves PIE *p, and is very much in line with other local names as Coronerus (*nerus 'man/hero')... And then we also have unquestionable Celtic place names like, say, Nemetobriga (*nemeto- 'Sanctuary', *-briga 'fortress, city' = Germanic burg).

Rob said...

@ Cossue
What is your opinion about the arrival of the earliest IE Lects in Iberia ?

Gioiello said...

@ Cossue
"As an example, take a name like Coropotus (a composite meaning 'War/Army' *coro-, and *-portus 'Lord'): It's not Celtic, since it preserves PIE *p, and is very much in line with other local names as Coronerus (*nerus 'man/hero')... And then we also have unquestionable Celtic place names like, say, Nemetobriga (*nemeto- 'Sanctuary', *-briga 'fortress, city' = Germanic burg)".

But this is just the demonstration that these languages were the languages of the migrants from Italy and are linked to the IE languages of Italy, from which we have infinite witnesses. Just *nerus may have come only from Italy.

Cossue said...

@Goiello
Actually this *nerus have cognates in many Indo-European languages
https://en.wiktionary.org/wiki/ner#Old_Irish
https://en.wiktionary.org/wiki/Reconstruction:Proto-Indo-European/h%E2%82%82n%E1%B8%97r

@Rob
If the Old European hidronomy is a thing, and that thing is IE, then we must have a rather young arrival of IE speakers into NW Iberia (https://en.wikipedia.org/wiki/Old_European_hydronymy).

But if we look just at the linguistic situation at the beginning of our era, IMO we still have at least two groups of Western IE dialects locally: ones that preserve IE *p, and others that do not preserve it (Hispano-Celtic). And all of them are heavily divergent of what we can find at the same time in other Western European countries... Just for the numbers, we know hundreds of different IE personal names in Hispania, hundreds of place names, and tens of God names, aside from Celtiberian and Lusitanian inscriptions proper... It's a large linguistic corpus.

I guess that the Late Atlantic Bronze Age (so, +3000 yBP) with its elite culture was very important in the arrival of some/any of these languages; but I would be surprised if it was the fist arrival of IE languages at all.

Cossue said...

Gioiello, not **Goiello, sorry

Gioiello said...

@ Cossue.
"@Gioiello
Actually this *nerus have cognates in many Indo-European languages
https://en.wiktionary.org/wiki/ner#Old_Irish
https://en.wiktionary.org/wiki/Reconstruction:Proto-Indo-European"

Of course, but *nerus < *neros may have come only from Italy, from where in fact came also the DNA (R-V88 and R-M269 found in aDNA in Iberia) and much other.

Cossue said...

@Gioiello
Ah, I understand. But no; most names in Latin inscriptions are latinized in its declination, so the -us in Coronerus, Coropotus, etc... its a render of local -*os. In Celtiberian inscriptions in Iberian script we have names as 'Mezukenos' or 'Letontu' which are rendered in Latin inscriptions as Medugenus (-os represented as -us) or Letondo/Letondonis (-long u in nasal stem represented as -o).

Gioiello said...

@ Cossue
"@Gioiello
Ah, I understand. But no; most names in Latin inscriptions are latinized in its declination, so the -us in Coronerus, Coropotus, etc... its a render of local -*os. In Celtiberian inscriptions in Iberian script we have names as 'Mezukenos' or 'Letontu' which are rendered in Latin inscriptions as Medugenus (-os represented as -us) or Letondo/Letondonis (-long u in nasal stem represented as -o)".

Of course, but also Old Latin had -o. Anyway I am seeing that you are an expert, thus I ask you: doesn't that demonstrate that Tartessian doesn't derive from Celt but the way may have been the other way around: Italy>Iberia>Northward. Of course there had been also a back migrations of other Celt languages to Iberia.

Chris Davies said...

Off topic, and probably controversial. However is it just me or is there a notable correlation between regions of Iberia with highest Yamnaya-related ancestry, and present-day highest regional GDP per capita in Iberia?

André de Vasconcelos said...

Not exactly, but correlation does not imply causation anyway. It's extremelly ignorant to propose that, considering Spanish industrialization and its historical development.

Chris Davies said...

As far as I know, 19th century industrialisation took off first in Basque Country, Catalonia, and Madrid. These being the areas still with highest GDP per capita in the region to the present day. Whether the fact that northern Iberian populations are genetically differentiated from other populations in the region was actually a factor in the take-up of industrialisation, I have no idea. But the correlation exists.

André de Vasconcelos said...

Except the small genetic differentiation is mostly East-West, rather than North-South. No idea why anyone would even think that had anything to do with genetics, I'm flabbergasted at that logic - or lack thereof. Besides, Eastern Europe would like to have a word with you.

What's next, natively speaking a Germanic language makes you more industrious?

Matt said...

You'd find the reverse anyway if you were looking at different regions of Europe anyway - Greece vs Balkans (Balkans more Yamnaya->lower GDP), within Germany (Northeast more Yamnaya->lower GDP), within Britain (North and Celtic fringe more Yamnaya->lower GDP), within the broader Central European region as a whole (Austria vs Czech Republic vs Poland vs Belarus, etc.).

Steppe vs non-Steppe would only really correlate within Italy and Spain and EEF vs non-EEF might end up being the strongest relationship, if you had a good way of accounting for later Neolithic population changes in Spain and Italy (Levant + North African ancestry).

Not that any of this would be causal.

Alberto said...

A bit late to this thread, but on the main topic. This is what I get for ATP9 with Global 10 (vanilla):

Iberia_BA:ATP9
Iberia_MN:I0408 62.9 %
Bell_Beaker_Germany:I0112 28.05 %
Kotias:KK1 6.4 %
Loschbour:Loschbour 1.7 %
Yamnaya_Kalmykia:RISE552 0.7 %
Karelia_HG:I0061 0.25 %
Iberia_EN:I0412 0 %
Iberia_Chalcolithic:I1300 0 %
Hungary_N:I1495 0 %
Esperstedt_MN:I0172 0 %
Barcin_Neolithic:I0707 0 %
Yamnaya_Samara:I0231 0 %
Armenia_Chalcolithic:I1631 0 %
Armenia_EBA:I1658 0 %
Iran_Chalcolithic:I1661 0 %
Latvia_HG:ZVEJ32 0 %
Baalberge_MN:I0560 0 %

Distance 0.003465

And with the same populations, but using weighted values:

Iberia_BA:ATP9
Iberia_MN:I0408 46.7 %
Iberia_Chalcolithic:I1300 23.15 %
Yamnaya_Kalmykia:RISE552 14.6 %
Kotias:KK1 6 %
Hungary_N:I1495 4.8 %
Loschbour:Loschbour 4.75 %

Which looks different but it's actually the same thing. Not far from qpAdm and compatible with the D-stats, so it all seems to make sense.

What will be interesting to find out with more samples is if this extra CHG arrived independently or not. My guess is that there were movements to Iberia both from Central Europe and from the Eastern Mediterranean, so to shift to the east is probably due to both influences.

From Kumtepe4 I only know an admixture run in one paper where it looked like a southern version of Bell Beakers from Germany. Difficult to say how accurate that could be, since it's a very low coverage sample. But that sample is around 3500 BC, so very early to be like that. Though if confirmed, it could represent a good source for southern Europe.

Chris Davies said...

Personality traits are conferred largely genetically, this is well-attested from twin and adoption studies over several decades. Not my area of expertise though.

Chris Davies said...

Massive iron ore deposits in northern Iberia must also have played a major role in facilitating industrialisation earliest in that part of the region.

andrew said...

"assuming that all R1b steppe people were not IE would make it hard to explain the Anatolian branch of IE in Anatolia. As we know, the Poltavka culture was still dominated by R1b. Only with the beginning of the Potapovka culture about 2500 BC we see a change to an R1a dominated population on the Russian steppe. Hittite is attested from the 16th century BC, but its ancestral versions seem to have split from the IE mainstream at a relatively early date, certainly earlier than the split between Balto-Slavic and Indo-Iranian, I'm not sure how you want to reconcile this with the late expansion of R1a people on the Pontic-Caspian steppe."

There is reasonably good archaeological and Akkadian language historical evidence that the Hittite culture expanded from a very small nucleus ca. 2000 BCE to its subsequent dominance, more or less contemporaneously with the appearance of Indo-Europeans in mainland Greece. Based upon this and upon how languages evolve, my hypothesis is that the Anatolian languages look older than they appear primarily because they encountered very different substrate languages than other families of Indo-European languages which led them to be more divergent. The extent to which language contact rather than Brownian motion like random mutation governs language change is greatly understated.

For example, the phonetic similarities between Hittite and previous Hattic and Hurrian non-Indo-European languages is quite visibly obvious (and Minoan also is quite similar based upon Egyptian phonetic reconstructions to Hattic/Hurrian non-Indo-European Anatolian languages).

A Minoan connection is more speculative although there are cultural cues (Chthonic gods, a prominent almost religious role for the bull, a non-Indo-European language, similarity in Y-DNA and mtDNA, approximately correct timing), and the similarities may be due to a common source culture making them sisters, rather than a parent-child relationship.

There is also decent evidence to suggest that the Anatolian languages immediately preceding Hittite were not the Neolithic languages of Anatolia and instead were themselves metal age arrivals . . . to discuss another day.

Davidski said...

I'm not sure how you want to reconcile this with the late expansion of R1a people on the Pontic-Caspian steppe.

Strange comment.

Dnieper Donets R1a
Khvalynsk R1a

Both on the Pontic Caspian Steppe.

Ryan said...

Late to the thread here but...

Iberia_BA ATP9
Caucasus_HG 0.038±0.063
Lengyel_LN 0.683±0.066
Steppe_EBA 0.177±0.087
Western_HG 0.102±0.044
chisq 5.216 tail_prob 0.876272


That extra CHG is interesting. Are you able to check how things shake out if you replace Steppe_EBA with someone from the Latvian Neolithic?

I'm thinking that Steppe_EBA has too much EHG which is depressing its results and inflating CHG. Though obviously the error bands are huge... just curious if having a more "western" source of steppe ancestry improves the fit at all.

On the archery terminology people mentioned before... I'd put money on that being a wanderwort. It's not all that surprising to have archery terms to spread with archery itself.

Davidski said...

Are you able to check how things shake out if you replace Steppe_EBA with someone from the Latvian Neolithic?

Hardly any markers for that.

Adrian TS said...

It's unfortunate that the Iberian non-IE culture of Eastern Spain is almost always left out of these discussions. This culture occupied a much greater extent than historically attested Basque speaking regions, and today those areas are also dominated by R1B D27. The Iberian language was widely spoken at the time of the Roman conquest and it seems to have endured even into the early Middle Ages.

To my knowledge there's been several attempts to link Basque and Iberian, but no definitive success so far.

I think a lot of commenters get a bad rep for suggesting that R1b is not necessarily linked to the IE expansion in Western Europe. But in light of the data this is still a legitimate possibility.

Myself I've got no problem with the Kurgan expansion theory, but in order to account for the R1B in Iberia you still need to explain why large stretches of territory suffered a yDNA turnover (and not just the Basque country as it is often claimed) with no IE language shift and one of them also happens to harbor the highest R1b percentage of the whole of Europe.

Plains Wanderer said...

@Adrian TS

I agree. There are also the Tartessian and Turdetani languages in Iberia, and even Basque-Aquitanian historically had a larger ranged across northern Spain and southwest France. Several non-IE languages survived in Iberia despite being loaded with R1b-DF27 and L21. Although IE languages had great success nearly everywhere they expanded, they failed to replace the languages of Iberia, allowing not just one, but multiple relic languages to survive. And it wasn't due to a lack of steppe admixture; all Iberians have some and R1b reaches very high frequencies precisely in these non-IE groups. The question is why?

I fully agree with Davidski and others about the genetics of Neolithic and Bronze Age Europe. R1a, R1b, and Yamna-like admixture spread from the steppes into Europe. R1b-L51 and L11 was likely spread by the Yamnaya or a Yamna-like population into Europe. However, I have my doubts that this expansion was entirely mediated by IE speakers.

While Corded Ware and R1a seem to correlate well with the spread of IE languages, the same cannot be said about Bell Beaker and R1b. Archaeologists and linguists have long said that Celtic languages and culture spread during the Urnfield-Hallstatt-La Tene periods and that those groups had a relatively small demic impact on western Europe and the British Isles. Bell Beaker is too early to have spread Celtic languages. To make them IE, you have to assume that they spoke an extinct branch of IE such as Nordwestblock, or that Celtic languages spread far earlier than previously believed. They may have just as well spoken a pre-Vasconic language, which would nicely explain the (controversial) Vasconic substratum across western Europe.

In addition, nearly all R1a is associated with IE or later Turkic groups. However, R1b has been found in many non-IE contexts: V88 in Africa, P297 in Villabruna, PH155 in South Asia, M335 in Southwest Asia, and M269 south of the Caucasus.

It is possible that the R1b Yamna-like expansion into Europe was non-IE. Perhaps the Bell Beaker groups adopted some non-IE languages as they migrated across Europe and mixed with local Neolithic populations, or perhaps the Yamnaya were non-IE themselves. The Yamnaya may have spoken a Caucasian-related language due to Maykop influence, while Sredny Stog and related groups in the Dnieper and Don basins spoke PIE. Or maybe Yamnaya and Bell Beaker were IE and there is another explanation for the non-IE languages of Iberia. I'm not married to any one theory, I just want each possibility to be properly explored.

Ryan said...

"To my knowledge there's been several attempts to link Basque and Iberian, but no definitive success so far."

That's because Iberian is poorly attested and mostly untranslated. There's no real information to suggest Iberian and Basque aren't from the same language family - there's just not a lot of solid information at all. Proto-Basque is thought have similar phonology, and virtually all of the vocabulary that we think we know is because it has a Basque cognate.

I agree with you re: R1b though.

Al Bundy said...

@Plains Wanderer Yamnaya spread PIE or a subset into Europe and had a big genetic impact especially on Northern and Central Europe,or Yamnaya like people. There's not much doubt about that.But the PIE question is more about origin than spread.The steppe was a vector for spreading but was it the origin.I wonder if you had some thoughts about that.PIE homeland might very well have been on the steppe so I'm glad we're getting huge stuff from Reich and Mycenean Greece, hopefully Anatolia,and more from the steppe.Crucial areas from where we don't have a lot of Ancient DNA and that are fundamental to PIE origin.

Davidski said...

@Plains Wanderer

Considering the close relationship on several levels between Yamnaya Samara and Kalmykia and Kura-Araxes Armenia EBA, as well as present day Northeast Caucasian speakers, like Lezgins, there's actually a chance that eastern Yamnaya rich in R1b-Z1203 spoke Northeast Caucasian languages.

That's why I think it's very important to locate the earliest instances of R1a-M417 and R1b-L51, and track their spread, because these really look like the markers of the main PIE clans. My guess is, they'll be sitting in the ancient DNA record somewhere between the Don and Dniester until exploding both west and east during the EBA.

Al Bundy said...

As Davidski has said we need those areas to,most likely,seal the deal for a steppe homeland.

Ryan said...

The Caucasus-assimilated Yamnaya idea is an interesting one. How would they have made it to Iberia though - an before the actual IE speakers?

What culture is ATP9 from by the way. Early Cogotas? Late Beaker?

Have there been any samples from Cucuteni-Trypillian sites by the way? I think that'd be a good candidate for a culture that received steppe admixture without adopting IE too.

Cossue said...

@Plain Wanderer
"Archaeologists and linguists have long said that Celtic languages and culture spread during the Urnfield-Hallstatt-La Tene periods and that those groups had a relatively small demic impact on western Europe and the British Isles. Bell Beaker is too early to have spread Celtic languages. To make them IE, you have to assume that they spoke an extinct branch of IE such as Nordwestblock, or that Celtic languages spread far earlier than previously believed."

Just adding to the debate, if you look at the -also usually neglected- linguistic material of Western and Central Indo-European Iberia (which comprehends 2/3 of the total area) you have not just Celtiberian and Hispano-Celtic, but also Lusitanian, which shows a lot of resemblances to both Celtic and Italic languages, being different from both (https://digilib.phil.muni.cz/bitstream/handle/11222.digilib/114048/N_GraecoLatina_11-2006-1_2.pdf?sequence=1).

At the end of the day, either Indo-European personal names exclusive of IE Iberia such as Taurocutius, Amparamus, Coropotus, Coropolla, Pictelancea, Praenia < *Pra(i)geniâ (or divinity names as Paramaeco, Proinetie < *Progenetiâ) are not Celtic, or either they belong to an extremely divergent and conservative Celtic branch which lacks much of the defining characteristics of the Celtic family, even the loss of PIE *p.

Again, even Hispano-Celtic/Lusitanian gods (Reve, Bandua, Navia, Cossue, Crougia, Ataecina, Aerno, Vestio...), personal names (Vesuclotus, Cadroiolo, Andamus, Medamus, Cloutius, Corotures, Coronerus, Celtius, Arceltius, Celtiatis, Boutius, Sunua, Apanus, Eburia, Doviterus, Reburrus, Ambatus, Caturus, Nantius...), and place names (Check for example the Ancient Celtic place-names in Europe and Asia Minor by Patrick Sims-Williams) are highly divergent and frequently exclusive to the peninsula.

So, if Proto-Celtic was a thing 4000 yBP, Lusitanian must have diverged from it earlier, and then attributing it to the (re)flux of Indo-Europeanized beaker people is , I think, a reasonable assertion.

Gioiello said...

@ Cossue

In fact I think that this Language was the Language of the migrants from Italy Zilhao spoke about. 7500 years ago is perhaps too early, but the migrations were many, and at the aDNA we have the witnesses of R-V88-Y7777 clearly from Italy and also the first R-M269 beyond many documented mt-s.

Amanda S said...

Gioiello is an Italian equivalent of the comedy sketch show Goodness Gracious Me's Mr Everything is Indian.

https://youtu.be/8tw7LIykvBw

Gioiello said...

@ Amanda S

Have you ever heard something about Villabruna 14000 years ago and what many are beginning to say about my theory that R-V88 was born in Italy? Your name in Latin doesn mean just "To be f----d", but Amanda Lear was a little desirable from me...

truth said...

@ Andrew

There is plenty of Celtic place names (or Indoeuropean) in the Basque Country, so definately they went there also.

Rob said...

@ Cossue, Alberto, et al

Have a look at Kochs video about PiE and Celtic in the Atlantic fringe

https://m.youtube.com/watch?v=Ub5izFOdtDs

A very good presentation, IMO;; well versed in all arms of evidence. My take home is that the spread of western / Atlantic occurred from single grave area to the North Atlantic, then south to western Iberia; whilst eastern Iberia and the mountains were not part of the "Atlantic Bronze Age Koine ", which is why Iberian and Basque languages survived there...