Ancient Fennoscandian genomes reveal origin and spread of Siberian ancestry in Europe

[Jovialis]

Abstract

European population history has been shaped by migrations of people, and their subsequent admixture. Recently, ancient DNA has brought new insights into European migration events linked to the advent of agriculture, and possibly to the spread of Indo-European languages. However, little is known about the ancient population history of north-eastern Europe, in particular about populations speaking Uralic languages, such as Finns and Saami. Here we analyse ancient genomic data from 11 individuals from Finland and north-western Russia. We show that the genetic makeup of northern Europe was shaped by migrations from Siberia that began at least 3500 years ago. This Siberian ancestry was subsequently admixed into many modern populations in the region, particularly into populations speaking Uralic languages today. Additionally, we show that ancestors of modern Saami inhabited a larger territory during the Iron Age, which adds to the historical and linguistic information about the population history of Finland.

https://www.nature.com/articles/s41467-018-07483-5

 

[Angela]

Abstract

European population history has been shaped by migrations of people, and their subsequent admixture. Recently, ancient DNA has brought new insights into European migration events linked to the advent of agriculture, and possibly to the spread of Indo-European languages. However, little is known about the ancient population history of north-eastern Europe, in particular about populations speaking Uralic languages, such as Finns and Saami. Here we analyse ancient genomic data from 11 individuals from Finland and north-western Russia. We show that the genetic makeup of northern Europe was shaped by migrations from Siberia that began at least 3500 years ago. This Siberian ancestry was subsequently admixed into many modern populations in the region, particularly into populations speaking Uralic languages today. Additionally, we show that ancestors of modern Saami inhabited a larger territory during the Iron Age, which adds to the historical and linguistic information about the population history of Finland.

https://www.nature.com/articles/s41467-018-07483-5

Thanks, Jovialis. I was getting extremely bored with this discipline.

I'll read it carefully later. Just looking at the admixture graph, I was immediately struck by the changes in the Northeast over time. Also, it clearly tells the same story as all analyses of the genomes of modern Europeans do. In western and central and certainly in southern Europe, although they don't show the latter, if you consider that perhaps forty percent of Yamnaya is "southern" the WHG and EHG portion is very much a minority one.

 

[halfalp]

What was the Bolshoy Oleni Ostrov individual used / or are they similar? Might be related with those R1b and C mtdna? Or R1a was the first wave of Siberian ancestry before, but related with N1? How Motala was found to have EDAR but apparently SHG doesn't have any Siberian, so EDAR was already with ANE and ultimately EHG? So many questions!

 

[Jovialis]

Thanks, Jovialis. I was getting extremely bored with this discipline.

I'll read it carefully later. Just looking at the admixture graph, I was immediately struck by the changes in the Northeast over time. Also, it clearly tells the same story as all analyses of the genomes of modern Europeans do. In western and central and certainly in southern Europe, although they don't show the latter, if you consider that perhaps forty percent of Yamnaya is "southern" the WHG and EHG portion is very much a minority one.

Indeed,

I believe it would look something like this; for example the modern English, and Norwegian samples in the graphic:

I made this edit that splits it 60% EHG (Cyan), and 40% CHG (Red). I did it by selecting the green component in paint, and reducing it's size horizontally to 60% to get the EHG-sized percentage.

 

[Angela]

So, it seems the Saami were probably there first, speaking an unknown language, adopted a Finno-Ugric language from another group, and to some extent mixed with people coming north from the direction of Estonia also speaking a Finno Ugric language? Those people and I1 people from Scandinavia pushed the poor Saami far north?

"The Finno-Ugric branch of the Uralic language family, to which both Saami and Finnish languages belong, has diverged from other Uralic languages no earlier than 4000–5000 years ago, when Finland was already inhabited by speakers of a language today unknown. Linguistic evidence shows that Saami languages were spoken in Finland prior to the arrival of the early Finnish language and have dominated the whole of the Finnish region before 1000 CE16,17,18. Particularly, southern Ostrobothnia, where Levänluhta is located, has been suggested through place names to harbour a southern Saami dialect until the late first millennium19, when early Finnish took over as the dominant language20. Historical sources note Lapps living in the parishes of central Finland still in the 1500s21. It is, however, unclear whether all of them spoke Saami, or if some of them were Finns who had changed their subsistence strategy from agriculture to hunting and fishing. There are also documents of intermarriage, although many of the indigenous people retreated to the north (see ref. 22 and references therein). Ancestors of present-day Finnish speakers possibly migrated from northern Estonia, to which Finns still remain linguistically close, and displaced but also admixed with the local population of Finland, the likely ancestors of today’s Saami speakers23.?

" Our results suggest that a new genetic component with strong Siberian affinity first arrived in Europe at least 3500 years ago, as observed in our oldest analysed individuals from northern Russia. These results describe the gene pool of modern north-eastern Europeans in general, and of the speakers of Uralic languages in particular, as the result of multiple admixture events between Eastern and Western sources since the first appearance of this ancestry component. "

So, could we infer that the Siberian/East Asian component came with the Finno-Ugric speakers?

"Ten of the eleven ancient individuals from this study fall on this Uralic cline, with the exception of one individual from Levänluhta (ID JK2065, here named Levänluhta_B), who instead is projected closer to modern Lithuanian, Norwegian and Icelandic populations. Specifically, two Levänluhta individuals and the two historical Saami from Russia are projected very close to the two previously published modern Saami (Saami.DG)32 and the new Saami shotgun genome generated in this study (as well as the previously published genome of the same individual, here labelled Saami (WGA)1), suggesting genetic continuity in the north from the Iron Age to modern-day Saami populations. In contrast, the six ancient individuals from Bolshoy are projected much further towards East Asian populations, and fall to an intermediate position along the Uralic cline and close to modern-day Mansi."

"Interestingly, results from uniparentally-inherited markers (mtDNA and Y chromosome) as well as certain phenotypic SNPs also show Siberian signals in Bolshoy: mtDNA haplogroups Z1, C4 and D4, common in modern Siberia24,34,35 are represented by the individuals BOO002, BOO004 and BOO006, respectively (confirming previous findings24), whereas the Y-chromosomal haplotype N1c1a1a (N-L392) is represented by the individuals BOO002 and BOO004. Haplogroup N1c, to which this haplotype belongs, is the major Y-chromosomal lineage in modern north-east Europe and European Russia. It is especially prevalent in Uralic speakers, comprising for example as much as 54% of eastern Finnish male lineages today36. Notably, this is the earliest known occurrence of Y-haplogroup N1c in Fennoscandia. Additionally, within the Bolshoy population, we observe the derived allele of rs3827760 in the EDAR gene, which is found in near-fixation in East Asian and Native American populations today, but is extremely rare elsewhere37, and has been linked to phenotypes related to tooth shape38 and hair morphology39 (Supplementary Data 2). Scandinavian hunter-gatherers from Motala in Sweden have also been found to carry haplotypes associated with this allele4. Finally, in the Bolshoy individuals we also see high frequencies of haplotypes associated with diets rich in poly-unsaturated fatty acids, in the FADS genes4,40,41."

"Our results show that all of the test populations are indeed admixed, with the most negative values arising when Nganasan are used as the Siberian source (Supplementary Data 3). Among the European sources, Lithuanians gave the most negative results for Estonians, Russians and Mordovians. For modern Hungarians, the European source giving most negative results was French, while both Bolshoy and Nganasan gave equally negative results when used as the Siberian source. With Finns as test, modern Icelanders were the European source giving most negative statistics. Finally, Icelanders and Nganasan used as the European and Siberian sources, respectively, yielded the most negative result for the present-day Saami as a Test. This result is still non-significantly negative, either due to the low number of modern Saami individuals in our dataset (n = 3), or due to post-admixture drift in modern Saami. A high degree of population-specific drift can affect f3-statistics and result in less negative and even positive values42"



Not the first time this correlation of Hungarians and the French has shown up, to the incomprehension of some of our posters.

"In contrast to previous models for European populations using three streams of ancestry2,3, we found that some populations modelled here require two additional components: a component related to modern Nganasans, as discussed above, and additional EHG ancestry, not explained by Yamnaya (who have been shown to contain large amounts of EHG ancestry themselves3). "

Not most, however.

"While the Siberian genetic component presented here has been previously described in modern-day populations from the region1,3,9,10, we gain further insights into its temporal depth. Our data suggest that this fourth genetic component found in modern-day north-eastern Europeans arrived in the area before 3500 yBP. It was introduced in the population ancestral to Bolshoy Oleni Ostrov individuals 4000 years ago at latest, as illustrated by ALDER dating using the ancient genome-wide data from the Bolshoy samples. The upper bound for the introduction of this component is harder to estimate. The component is absent in the Karelian hunter-gatherers (EHG)3 dated to 8300–7200 yBP as well as Mesolithic and Neolithic populations from the Baltics from 8300 yBP and 7100–5000 yBP respectively8. While this suggests an upper bound of 5,000 yBP for the arrival of this Siberian ancestry, we cannot exclude the possibility of its presence even earlier, yet restricted to more northern regions, as suggested by its absence in populations in the Baltics during the Bronze Age. Furthermore, our study presents the earliest occurrence of the Y-chromosomal haplogroup N1c in Fennoscandia. N1c is common among modern Uralic speakers, and has also been detected in Hungarian individuals dating to the 10th century44, yet it is absent in all published Mesolithic genomes from Karelia and the Baltics3,8,45,49.The large Nganasan-related component in the Bolshoy individuals from the Kola Peninsula provides the earliest direct genetic evidence for an eastern migration into this region. Such contact is well documented in archaeology, with the introduction of asbestos-mixed Lovozero ceramics during the second millennium BC50, and the spread of even-based arrowheads in Lapland from 1900 BCE51,52. Additionally, the nearest counterparts of Vardøy ceramics, appearing in the area around 1,600-1,300 BCE, can be found on the Taymyr peninsula, much further to the East51,52. Finally, the Imiyakhtakhskaya culture from Yakutia spread to the Kola Peninsula during the same period24,53. Contacts between Siberia and Europe are also recognised in linguistics. The fact that the Nganasan-related genetic component is consistently shared among Uralic-speaking populations, with the exceptions of absence in Hungarians and presence in the non-Uralic speaking Russians, makes it tempting to equate this genetic component with the spread of Uralic languages in the area. However, such a model may be overly simplistic. First, the presence of the Siberian component on the Kola Peninsula at ca. 3500 yBP predates most linguistic estimates of the spread of extant Uralic languages to the area54. Second, as shown in our analyses, the admixture patterns found in historic and modern Uralic speakers are complex and in fact inconsistent with a single admixture event. Therefore, even if the Siberian genetic component partly spread alongside Uralic languages, some Siberian ancestry may have been already present in the area from earlier admixture events."

EDAR may have come with a bit of prior admixture quite early, but then disappeared with admixture with western and southern genomes.

For those interested in reading tea leaves:

"The steppe ancestry component within modern Europeans (green), which is associated with the Yamnaya population, is maximised in ancient Iranian populations and to a lesser extent Caucasus hunter-gatherers (CHG). This ancestry component is also present in modern Armenians from the Caucasus, Bedouins from the Arabian Peninsula and South Asian populations."

 

[berun]

It would be superb to know if this Siberian migration was done by deer herders colonizing HG almost empty lands.

 

[berun]

By the way, the Siberian/Nganasan mixture in modern British and French is from the Celtic expansion or the Germanic one?

 

[Ailchu]

"The steppe ancestry component within modern Europeans (green), which is associated with the Yamnaya population, is maximised in ancient Iranian populations and to a lesser extent Caucasus hunter-gatherers (CHG). This ancestry component is also present in modern Armenians from the Caucasus, Bedouins from the Arabian Peninsula and South Asian populations."

there seems to be something off with this "steppe" ancestry. how can it be maximised in CHG but not in yamna? maybe steppe is not clearly defined and everyone who has CHG or EHG admixture no matter what proportions gets "steppe" ancestry? edit: i think the "green" component they are talking about here is for another graphic and only includes the CHG. but then it would make no sense to call it steppe ancestry. can someone explain?

 

[Angela]

"The steppe ancestry component within modern Europeans (green), which is associated with the Yamnaya population, is maximised in ancient Iranian populations and to a lesser extent Caucasus hunter-gatherers (CHG). This ancestry component is also present in modern Armenians from the Caucasus, Bedouins from the Arabian Peninsula and South Asian populations."

there seems to be something off with this "steppe" ancestry. how can it be maximised in CHG but not in yamna? maybe steppe is not clearly defined and everyone who has CHG or EHG admixture no matter what proportions gets "steppe" ancestry? edit: i think the "green" component they are talking about here is for another graphic and only includes the CHG. but then it would make no sense to call it steppe ancestry. can someone explain?

I think they're speaking of Figure 2, Ailchu, and not the admixture chart in the OP. Therefore, the "green"component to which they're referring is the "Iran Neolithic" component in these groups, which is highest in Yamnaya of all the "European" samples. The green doesn't represent the total Yamnaya component, as it does in the first chart.

If we look at the English, just by eyeballing the first chart it looks like they are about 40% Yamnaya. At 40% CHG, which already had a bit of LBK and SHG like ancestry, that's 16%.

As for the Basque, they have the additional EHG/SHG like ancestry, but not, apparently, the Iran Neo like ancestry. There's been hints of this before. I'm puzzled by it too. Were some IE groups travelling west much lower in Iran Neo and it sort of diluted?

 

[Ailchu]

I think they're speaking of Figure 2, Ailchu, and not the admixture chart in the OP. Therefore, the "green"component to which they're referring is the "Iran Neolithic" component in these groups, which is highest in Yamnaya of all the "European" samples. The green doesn't represent the total Yamnaya component, as it does in the first chart.

If we look at the English, just by eyeballing the first chart it looks like they are about 40% Yamnaya. At 40% CHG, which already had a bit of LBK and SHG like ancestry, that's 16%.

As for the Basque, they have the additional EHG/SHG like ancestry, but not, apparently, the Iran Neo like ancestry. There's been hints of this before. I'm puzzled by it too. Were some IE groups travelling west much lower in Iran Neo and it sort of diluted?

maybe basques already had higher amounts of WHG admixture and then mixed with people who were similar to modern spanish or french?

 

[Ailchu]

By the way, the Siberian/Nganasan mixture in modern British and French is from the Celtic expansion or the Germanic one?

it seems like it wasn't present in yamna and also not in cwc. also figure 2 in this study has no ngasan related ancestry in icelandic people, english and scotts. maybe there is ANE ancestry that is recognized as EHG in the model for figure 2 and as ngasan in the model for figure 4?

 

[Angela]

it seems like it wasn't present in yamna and also not in cwc. also figure 2 in this study has no ngasan related ancestry in icelandic people, english and scotts. maybe there is ANE ancestry that is recognized as EHG in the model for figure 2 and as ngasan in the model for figure 4?

I think that may be right. When there's no EHG sample to "absorb" it, it shows up.

For ease of comparison:

[/IMG]

 

[Ygorcs]

Oh thank God for this paper! I was getting really annoyed by the paucity on aDNA from Northeastern Europe and Western Siberia while the Carlos Quiles' indo-european.eu insists endlessly and with increasingly "labored" speculations on the supposed "Uralic CWC" (and the even more "innovative" idea that it was a steppe language from Sredny Stog II and derived from a Neolithic Indo-Uralic that would've formed the starkly different Repin/Yamna PIE and Sredny Stog Uralic in, what?, 2000 years?!), and that Uralic had "nothing" to do with the spread of Siberian ancestry and N1c, being instead the language spread by R1a-M417 and much of the steppe ancestry in Northern/Northeastern Europe...

Of course, though, IMO the presence of Siberian admixture does not mean the Proto-Uralic speakers came from Siberia, but it does indicate they were placed somewhere in the cline from EHG to WSHG and/or ANE that existed back then and caused them to have a Siberian admixture (enriched in ANE, I presume) that most other Europeans lacked. My personal impression for now (totally open to changes and objections) is that the Proto-Uralic population (which I think either was or lived near the Pitted Ware and the Volosovo cultures) was a mainly EHG population (possibly with some WHG since Narva culture was not very far from there) and that received Western Siberian input and later got more BA Steppe and EEF when CWC expanded and Late Proto-Uralic and particularly Proto-Finno-Ugric).

 

[Ygorcs]

What was the Bolshoy Oleni Ostrov individual used / or are they similar? Might be related with those R1b and C mtdna? Or R1a was the first wave of Siberian ancestry before, but related with N1? How Motala was found to have EDAR but apparently SHG doesn't have any Siberian, so EDAR was already with ANE and ultimately EHG? So many questions!

Wasn't Siberian at least up to the BA rich in ANE by definition? And EHG also had received some ANE, some of it may have come with EDAR, but for some reason it was not selected favorably as much as in East Asia and the Americas. I think that was what they found in the WSHG samples, Botai (that's on the steppes, but pretty northerly), Okunevo and others, that West/Central Siberia preserved a mostly ANE ancestry until really late (though I presume some East Asian ancestry had already introgressed into the region's gene pool).

 

[berun]

I think that may be right. When there's no EHG sample to "absorb" it, it shows up.

For ease of comparison:

[/IMG]

ok, I see, but then the problem is with the Yamna share (EHG + CHG) of British and Frenchs, which is not covering so all the EHG share so that it is lumped by the Siberian component (?). If so it clearly points that IE was not coming from Yamna but from another place with extra EHG, from Kievan Rus area I guess.

 

[markod]

ok, I see, but then the problem is with the Yamna share (EHG + CHG) of British and Frenchs, which is not covering so all the EHG share so that it is lumped by the Siberian component (?). If so it clearly points that IE was not coming from Yamna but from another place with extra EHG, from Kievan Rus area I guess.

ADMIXTURE always makes it look like Yamnaya admixture outside of some Balkanic countries was actually quite low due to a relative paucity of CHG in most of Europe, most notably in Basques and their neighbours. Happens in every unsupervised run, I've tested it myself. I still don't know why it does this or if it is real, but I wouldn't read too much into it.

 

[berun]

The case can be seen looking at Baltic_CCC showing 20% Yamna. It's a nonsense from an steppist point of view.

 

[markod]

The case can be seen looking at Baltic_CCC showing 20% Yamna. It's a nonsense from an steppist point of view.

I think there are only two explanations:

1. ADMIXTURE models are bad

2. EHG + EEF together create a pseudo-steppe/CHG component which formal models cannot distinguish from the real thing

I don't know which is more likely, and it's never addressed in the papers.

 

[halfalp]

Wasn't Siberian at least up to the BA rich in ANE by definition? And EHG also had received some ANE, some of it may have come with EDAR, but for some reason it was not selected favorably as much as in East Asia and the Americas. I think that was what they found in the WSHG samples, Botai (that's on the steppes, but pretty northerly), Okunevo and others, that West/Central Siberia preserved a mostly ANE ancestry until really late (though I presume some East Asian ancestry had already introgressed into the region's gene pool).

There was at least 2 waves of ANE going into Europe that we can be sure of, the one who gives birth to EHG and the one who is related with Nganasans genetic signature. But there could be more, Dzudzuana had some ANE so if we are not talking about very archaic shared ancestry, there was probably an ANE-like migration in Eastern Europe even before the creation of WHG-EHG. Also Lola shows in early BA north caucasus a western siberian signal, we dont know if it is exactly the same as the " Nganasan " one. All those different waves of ANE could have bring phenotypically and genotypically different peoples.

 

[berun]

@markod, this case is nor bad program neither fake steppe signals but is about a ghost population yet unsampled, option C