Steppe Ancestry in western Eurasia and the spread of the Germanic Languages

[Parastais]

The late Neolithic migration from the East Baltic proposed by the authors cannot be associated with any archaeological movement previously posited as being Uralic (post- “continuation theory” paradigm). The earliest visible wave of materials from the Russian forest zone were Luganuse and Asva coarse grain ware: ~1200 BC in-land open settlements; ~9th century BC coastal Stone-cist graves and hillforts. N1a is notably missing from Stone-cist graves, so an associated movement of people is perhaps dubious. However, by ~600 BC, N1a had arrived with early-Tarand graves.

The situation is of course different in Finland, where Asbestos ceramics and Textile ware both may have arrived from the east in the early 2nd millennium BC. Although, the chronology and provenance of both phenomenon are nebulous.

For those interested, there are a few new N1a samples from this study:

CGG_2_106751 N-L550 from early Roman Iron Age Slusegård, Denmark. Most of the samples at this site show evidence of violence. The sample in question was found with an iron “point” embedded in his skull, but still possessed grave goods. Some individuals at the Slusegård site have excessive IBD sharing with East Baltic populations (this sample included).

CGG_2_24147 N-L550 from early Roman Iron Age Oland, Sweden. Isotopic analysis indicates that he was a local. There was evidence of perimortem SFT and BFT to his skull. IBD results are similar to other samples in the Central Swedish cluster.

Seven new samples from Lithuania; four of which are male.

CGG_0_17685 G-Z31461 Roman Iron Age Marvele (Flat burials of central Lithuania).
CGG_0_17690 N-L550 400-700 AD Berciunai (North Lithuanian Barrows Culture).
CGG_0_17691 N-L550 400-700 AD Kaireneliai (Central or North Lithuanian Barrows Culture?).
CGG_0_17685 N-L550 8th-9th century Maudžiorai (Samogatian with some Curonian influence).

Quite interesting that there were no R1a samples; especially considering the previously published samples from Lithuania.

Seven samples from the Mazunino culture, five are male.

CGG_0_211459 N-L1026 Roman Iron Age Boyaski
CGG_0_211460 N-L1026 Roman Iron Age Boyaski
CGG_0_211461 N-Z1936 Roman Iron Age Boyaski
CGG_0_211462 N-Z1936 Roman Iron Age Dubrovsky
CGG_0_211464 R1a-Z2124 Roman Iron Age Dubrovsky

The Mazunino culture is derived from the Pyanobor culture and was probably at least partly Uralic speaking (usually associated with Udmurt).

Ok, so now Lithuanian Iron Age is looking like N-L550 population. First Millennium Lithuania samples:
6 N-L550 (75%)
1 G
1 R1a

That G from Marvele was from Caucasus? Modern Georgian and nearby Russians are under it.

Do we have G25 coordinates from this article?

[Codaman]

That G from Marvele was from Caucasus? Modern Georgian and nearby Russians are under it.

Do we have G25 coordinates from this article?

I can't quite make out all of the the sample IDs in Extended Data Figure 9 and 10 but one of the samples from Lithuania is completely outlying so perhaps this is the one with G? North Caucasian origin looks possible.

The two new samples from Finland look to be a stark contrast to the old ones from the same area, the Saami-like samples and the Scandinavian outlier. Käldamäki and the new Levänluhta sample both seem to be broadly albeit not precisely similar to samples from western Estonia but the former actually has slightly more of the Baltic IBD component which suggests there was little admixture within Finland at this point and place.

[Fabrice E]

Amazing number of sample from France during IA.
Few samples from Austrian-IA too ... those data are great, they fill many hole in ancient DNA coverage !
Right after the new samples from South-Western Germany IA, its a lucky week !

Yes
We'll have samples from VIX !
     
CGG023310
Vix,_tumulus_1,_Dame_de_Vix
WesternEurope
France
Bourgogne_Franche_Compt
750BCE-450BCE
mt T2k

[miquirumba]

Amazing number of sample from France during IA.
Few samples from Austrian-IA too ... those data are great, they fill many hole in ancient DNA coverage !
Right after the new samples from South-Western Germany IA, its a lucky week !

Yes
We'll have samples from VIX !
     
CGG023310
Vix,_tumulus_1,_Dame_de_Vix
WesternEurope
France
Bourgogne_Franche_Compt
750BCE-450BCE
mt T2kultra rare T2k maternal haplogroup in this sample from Vix l https://www.yfull.com/mtree/T2k/ I checked FTDNA mt Haplotree where there are 2 more samples from Germany and Ireland. Basal clade found in Azerbaijan and Arabia

[Jaska]

That G from Marvele was from Caucasus? Modern Georgian and nearby Russians are under it.

Do we have G25 coordinates from this article?

I can't quite make out all of the the sample IDs in Extended Data Figure 9 and 10 but one of the samples from Lithuania is completely outlying so perhaps this is the one with G? North Caucasian origin looks possible.

The two new samples from Finland look to be a stark contrast to the old ones from the same area, the Saami-like samples and the Scandinavian outlier. Käldamäki and the new Levänluhta sample both seem to be broadly albeit not precisely similar to samples from western Estonia but the former actually has slightly more of the Baltic IBD component which suggests there was little admixture within Finland at this point and place.

There seems to be an error in the supplementary table S1: Käldamäki gets a dating 450-0 BCE, but in the archaeological supplement it is dated to the migration period. Perhaps these two new samples with majority Baltic ancestry were early Finnic speakers? 

The IBD clustering of the three Saami-like samples are well in line with e.g. the qpAdm results of Peltola et al. 2023, where Levänluhta people were a mixture of BOO people and farmer ancestry. There the BOO ancestry was Siberian + EHG with little farmer ancestry, in this study all the three share Amur + EHG with little Iran ancestry (another half+ consisting of Baltic and Scandinavian ancestry).

[Anglesqueville]

That G from Marvele was from Caucasus? Modern Georgian and nearby Russians are under it.

Do we have G25 coordinates from this article?

I can't quite make out all of the the sample IDs in Extended Data Figure 9 and 10 but one of the samples from Lithuania is completely outlying so perhaps this is the one with G? North Caucasian origin looks possible.

The two new samples from Finland look to be a stark contrast to the old ones from the same area, the Saami-like samples and the Scandinavian outlier. Käldamäki and the new Levänluhta sample both seem to be broadly albeit not precisely similar to samples from western Estonia but the former actually has slightly more of the Baltic IBD component which suggests there was little admixture within Finland at this point and place.

There seems to be an error in the supplementary table S1: Käldamäki gets a dating 450-0 BCE, but in the archaeological supplement it is dated to the migration period. Perhaps these two new samples with majority Baltic ancestry were early Finnic speakers? 

The IBD clustering of the three Saami-like samples are well in line with e.g. the qpAdm results of Peltola et al. 2023, where Levänluhta people were a mixture of BOO people and farmer ancestry. There the BOO ancestry was Siberian + EHG with little farmer ancestry, in this study all the three share Amur + EHG with little Iran ancestry (another half+ consisting of Baltic and Scandinavian ancestry).

Peltola's qpAdm models for the two levänluhta individuals are hardly convincing, perhaps because Peltola used the original genomes, which suffer from very low coverage. With their versions imputed in a very convincing manner by Allentoft we can easily obtain numerically excellent and historically entirely plausible models, at least it seems to me.

left pops:
Finland_IA.imputed_allentoft_234
Sweden_IA.imputed_allentoft
Russia_Minino_IronAge.ial


best coefficients:    0.516    0.484
totmean:      0.516    0.484
boot mean:    0.515    0.485
      std. errors:    0.032    0.032
     
fixed pat  wt  dof    chisq      tail prob
          00  0    12    6.169        0.907327    0.516    0.484
          01  1    13  245.659              0    1.000    0.000
          10  1    13  161.650    9.13301e-28    0.000    1.000
best pat:          00        0.907327              -  -
best pat:          10      9.13301e-28  chi(nested):  155.481 p-value for nested model:    1.09914e-35


left pops:
Finland_IA.imputed_allentoft_238
Sweden_IA.imputed_allentoft
Russia_Minino_IronAge.ial


best coefficients:    0.166    0.834
totmean:      0.166    0.834
boot mean:    0.165    0.835
      std. errors:    0.041    0.041


  fixed pat  wt  dof    chisq      tail prob
          00  0    12    10.713        0.553645    0.166    0.834
          01  1    13  549.946              0    1.000    0.000
          10  1    13    24.722      0.0251016    0.000    1.000
best pat:          00        0.553645              -  -
best pat:          10        0.0251016  chi(nested):    14.009 p-value for nested model:    0.000181944

[Jaska]

That looks good. But does the BOO as the source give much weaker results?

[Anglesqueville]

That looks good. But does the BOO as the source give much weaker results?

For those who read my post before I edited it, I did not give the correct models, but the old ones (probably rotten by the presence of kra001 on the right). Here are the good ones:

left pops:
Finland_IA.imputed_allentoft_234
Sweden_IA.imputed_allentoft
BOO.ial

best coefficients: 0.573 0.427
totmean: 0.573 0.427
boot mean: 0.573 0.427
std. errors: 0.042 0.042

fixed pat wt dof chisq tail prob
00 0 11 10.361 0.498224 0.573 0.427
01 1 12 108.576 1.14404e-17 1.000 0.000
10 1 12 162.844 1.38221e-28 0.000 1.000
best pat: 00 0.498224 - -
best pat: 01 1.14404e-17 chi(nested): 98.215 p-value for nested model: 3.75342e-23


left pops:
Finland_IA.imputed_allentoft_238
Sweden_IA.imputed_allentoft
BOO.ial


best coefficients: 0.272 0.728
totmean: 0.272 0.728
boot mean: 0.272 0.728
std. errors: 0.046 0.046

fixed pat wt dof chisq tail prob
00 0 11 7.176 0.784656 0.272 0.728
01 1 12 254.357 0 1.000 0.000
10 1 12 40.843 5.20493e-05 0.000 1.000
best pat: 00 0.784656 - -
best pat: 10 5.20493e-05 chi(nested): 33.667 p-value for nested model: 6.53973e-09

Personally I wouldn't risk choosing between the two models from a strict mathematical point of view. But between Bolshoy and Minino_IA, if it is true that they found Tapiola pottery on the site (which I have not been able to find confirmation of), I don't hesitate too much.

edit edit: for all those models the right list was:

Russia_Ust_Ishim.DG
Cameroon_SMA.DG
Italy_North_Villabruna_HG
Czech_Vestonice16
Belgium_UP_GoyetQ116_1
Russia_MA1_HG.SG
Iran_GanjDareh_N
Russia_Kostenki14.SG
Indian_GreatAndaman_100BP.SG
Israel_PPNB
Georgia_Kotias.SG
Turkey_N_I0707
Sidelkino.ial

[Uintah106]

This one intrigues me ' During the Migration Period, we detect a previously unknown northward migration back into Southern Scandinavia, partly replacing

earlier inhabitants and forming the North Germanic-speaking Viking-Age populations of
Denmark and southern Sweden, corresponding with historically attested Danes'.

[Rodoorn]

From the supplements (S6.7.2)

The Bell Beaker sub-cluster (..) located primarily from the Eastern North Sea (ENS) region (present day the Netherlands) is unique in its high NWHG ancestry, low European Farmer, and inability to be modelled primarily as Bell Beaker ancestry, like most others from Bell Beaker sub-clusters (Figure S6.3.6). When using early representatives of this cluster as a source, we see a large degree of genetic  continuity from 3700 - 1700 BP. 
(...) 
A transition by at least 1612 BP is apparent, Frisian individuals are modelled primarily as Southern Scandinavian ancestries, but possessing small amounts of the local ENS ancestry.

An interesting cluster; probably Hilversum-culture related (or Elp?). Its continuity until West-Germanic migrations is remarkable.

Indeed very interesting Pylsteen, also from the paper:

"We further find that the IA Southern Scandinavians that arose from admixture between Bronze Age Southern and Eastern Scandinavians are central to understanding the Germanic dispersal. After the Pre-Roman Iron Age, around 2000 BP, Proto-Germanic diverged into North, East and West Germanic. The spread of West Germanic to Germany, the Netherlands  and Britain, appears to be closely related to populations migrating from the Jutland Peninsula.

In these regions, we see the transition from Bell Beaker-related to the Corded Ware-related  Southern Scandinavian ancestry. For Germany and Britain, where Celtic was known to be  spoken, this period also saw a linguistic transition to Germanic. In the Netherlands, IA Southern Scandinavians’ ancestry became dominant in the place of a distinct Eastern North Sea population. The linguistic affiliation of this population is unknown. According to the linguistic ‘Nordwestblock’ hypothesis, the Netherlands may have harboured a language distinct from both Celtic and Germanic80. Given that ENS is a Bell Beaker subcluster, which is associated with Celtic languages in Britain and France, our results can alternatively be  brought in line with theories of Celtic speakers, perhaps including the Frisii of the Roman Period, inhabiting the Dutch North Sea coast during the Early Iron Age 81. 

Although no unadmixed ENS populations are found during the migration period, the incoming Southern Scandinavians carry small proportions of ENS ancestry, indicating the migrations were not a complete replacement. Dutch coastal areas see a habitation hiatus around 1600 BP and subsequent appearance of a new material culture that is often referred to as Anglo-Saxon in nature 82, mirroring the genetics and timing of the Late Iron Age, linguistically West- Germanic Frisians in this dataset. In addition, we find that the Southern Scandinavian  ancestry of these migrating populations is better modelled by individuals near Southern rather than the Northern Jutland, and that the migrating populations often carry varying but minor proportions of ENS ancestry, inherited from the earlier people who previously lived in the region. In contrast to previous studies, which relied on Scandinavian samples postdating the Migration Period 47, we can now reject the Danish Isles and Sweden as a source area for the  Anglo-Saxons in Britain, as these were dominated by Eastern Scandinavian ancestry prior to831 the Viking Age (Figure 6)."

I presume that the ENS factor could be bigger in the inland area's of Netherlands and the least in Westergo/ Friesland which was most probably depopulated about 400 AD. Would need specific further research. The ENS factor in Dutch population?  Still some Bell Beaker heritage?  

The "Germanization" of the Netherlands was most probably a fact after the migration ages!?

[Uintah106]

This one intrigues me ' During the Migration Period, we detect a previously unknown northward migration back into Southern Scandinavia, partly replacing

earlier inhabitants and forming the North Germanic-speaking Viking-Age populations of
Denmark and southern Sweden, corresponding with historically attested Danes'.

So maybe the notion that Danes are related  the Svear is mostly Physical  resemblance .Danes Back Back Migration ahead of the Viking period is huge.

[Jaska]

This one intrigues me 'During the Migration Period, we detect a previously unknown northward migration back into Southern Scandinavia, partly replacing earlier inhabitants and forming the North Germanic-speaking Viking-Age populations of Denmark and southern Sweden, corresponding with historically attested Danes'.

Apparently this migration did not bring a whole new language, because we can follow the development from Northwest Germanic to Old Scandinavian in the Runic inscriptions, starting already ca. 100 CE. But the great Scandinavian changes began ca. 500 CE, before which Proto-Scandinavian was still quite close to Proto-Germanic. So it is possible that migration of closely related speakers from the south acted as a catalyst, causing these qualitatively great changes. At least one traditional rule of thumb is that languages in intense contacts change faster than isolated languages (compare to Icelandic).

EDIT: More about the topic over here:
https://genarchivist.com/showthread.php?...3#pid12913

[AimSmall]

U152>L2 samples

CGG022418 Aisne_Haut_de_France 480-450 BC IronAge LaTene
CGG022430 Aisne_Haut_de_France 514-397 BC IronAge LaTene
CGG107756 Province South-Holland 42 BC to 106 AD IronAge Roman

Hopefully someone is able to drill down further.

The supplement lists more U152, but I'm only seeing two L2.     CGG022430 isn't listed in the supplement that I can find.

   

EDIT: Found him on a subsequent sheet.

[Rodoorn]

"In the Netherlands, IA Southern Scandinavians’ ancestry became dominant in the place of a distinct Eastern North Sea population. The linguistic affiliation of this population is unknown. According to the linguistic ‘Nordwestblock’ hypothesis, the Netherlands may have harboured a language distinct from both Celtic and Germanic80. Given that ENS is a Bell Beaker subcluster, which is associated with Celtic languages in Britain and France, our results can alternatively be  brought in line with theories of Celtic speakers, perhaps including the Frisii of the Roman Period, inhabiting the Dutch North Sea coast during the Early Iron Age 81. "
...

Dutch coastal areas see a habitation hiatus around 1600 BP an dsubsequent appearance of a new material culture that is often referred to as Anglo-Saxon in nature.

.....

In addition, we find that the Southern Scandinavian  ancestry of these migrating populations is better modelled by individuals near Southern rather than the Northern Jutland, and that the migrating populations often carry varying but minor proportions of ENS ancestry, inherited from the earlier people who previously lived in the region. "


In add to this there are indications that this is supported by archeology. For example: the Hondsrug area in Drenthe (inland North Dutch) was one of Europe's most dense Bell Beaker hotspots.

In Eelde- on the Hondsrug, beneath the city of Groningen- we find a settlement that can be traced back to Bell Beaker times, a very rich BB grave is found there. The settlement was left in....400 AD. In other words: with the incoming Germanics.

It's likely that there was some BB related ENS that left its traces in (North) Dutch population. On the Hondsrug in Drenthe more than in Westergo Friesland, which was about 400 AD depopulated (so no ENS left).

I guess they have a strong case.

[alanarchae]

kind of supports the idea of a Celts or NW block group in the Netherlands until pretty late on.