| Region | I1 | I2*/I2a | I2b | R1a | R1b | G | J2 | J*/J1 | E1b1b | T | Q | N |
| Russia | 5 | 10.5 | 0 | 46 | 6 | 1 | 3 | 0 | 2.5 | 1.5 | 1.5 | 23 |
| Lithuania | 6 | 6 | 1 | 38 | 5 | 0 | 0 | 0 | 1 | 0.5 | 0.5 | 42 |
| Latvia | 6 | 1 | 1 | 40 | 12 | 0 | 0.5 | 0 | 0.5 | 0.5 | 0.5 | 38 |
| Estonia | 15 | 3 | 0.5 | 32 | 8 | 0 | 1 | 0 | 2.5 | 3.5 | 0.5 | 34 |
| Finland | 28 | 0 | 0.5 | 5 | 3.5 | 0 | 0 | 0 | 0.5 | 0 | 0 | 61.5 |
| Sweden | 37 | 1.5 | 3.5 | 16 | 21.5 | 1 | 2.5 | 0 | 3 | 0 | 2.5 | 7 |
| Norway | 31.5 | 0 | 4.5 | 25.5 | 32 | 1 | 0.5 | 0 | 1 | 0 | 1 | 2.5 |
| Denmark | 34 | 2 | 5.5 | 15 | 33 | 2.5 | 3 | 0 | 2.5 | 0 | 1 | 1 |
(Y chromosomal haplogroups)
A few weeks ago I saw the Y chromosomal haplogroup group distribution in Finland and Sweden. I’d know this disjunction for a while, but it really struck me. I got the numbers above from Eupedia, but you can find them elsewhere. Most of you probably know that Finland has a high fraction of N (they keep changing the nomenclature, so I’ll leave the number off). What’s curious to me is how low the fraction of N in the rest of Scandinavia is. Much of the N we see in Sweden may even be historical era migration of Finns into Sweden when the two nations were in political union (Finland was basically a Swedish colony). Another notable fact is that N is very common among Baltic people, whether Finnic in the language (Estonian) or Indo-European (Latvian and Lithuanian).
Another strange thing is that while the Indo-European lineages of R1 are both at very low frequency in Finland, I1, which is common to the west in Scandinavia, is not. The latest ancient DNA makes it clear that Finnic languages seem to have arrived in the Baltic in the period between 1000 and 500 BC. Before then Corded Ware/Battle Axe people seem to have been dominant in the East Baltic. These people usually carried Y chromosome R1a.
The fact that N is so high in the Baltic nations shows that newcomers arrived, and in the northern region language shifted happened, but in the south, it did not. Meanwhile, further north in Finland almost all the R1a lineages disappeared. Not so with I1. There are all sorts of tortured explanations for this pattern, so I won’t offer one.
Genome-wide the Finns aren’t that different. The largest proportion of their ancestry is still Yamnaya/steppe:
I only post this to illustrate how strong “male-mediated” dynamics can be. The proportion of Siberian ancestry in Finns is rather low, but > 50% of their Y chromosomes are N. I think it is plausible that one of the reasons for the massive reduction in R1 in Finland might be due to climate change and massive population collapse among the Battle Axe people of southern Finland, and the later arrival of Siberians.
Maybe the other thing to note is that there is what seems to be quite a large minority of Swedish speakers in southern Finland who don’t seem to be well integrated. So presumably they migrated into Finland when it was a Swedish colony. They don’t seem to show up in the figures, or not much.
As a quick complement to the stacked chart above, some recent qpAdm results I’ve run for European populations, trying to use a large extent of current publicly available adna to form outgroups and test: https://imgur.com/a/MikY8NJ
Also using the Steppe_Eneolithic samples from the Caucasus piedmont who model purely EHG+CHG and have less Anatolian/EuroHG related ancestry from Europe than the later Yamnaya look to have at about 1/8 of their ancestry (so Steppe_En provide just a total baseline of the non-Anatolian+EuroHG ancestry).
Result for Finnish is a bit more pushing qpAdm (wanted to allow the WSHG related and East Asian ancestries they receive to vary freely, which might confound things a bit) and so I’ve given an alternative fit of only slightly higher Chi-SQ score just using the Mongolia_N that contributes most Nganasan ancestry, but I think the other results make sense. Not including many of the Uralic speaking populations in the same table from the main post, wasn’t too focused on that bit.
Not sure the p-values are the best but the proportions all seem to make sense, and the results produce FEASIBLE outcomes, so this may be something to do with my choices of outgroups.
(Only tested Finnish with Mongolia_N and WSHG related ancestries, as find that there is some pseudo-East Asia thing of a 1-2% that we don’t find in ADMIXTURE etc when testing many other European samples with Mongolia_N, not just modern Europeans but also when I model some pretty early EEF farmers and such too in about the same amount, where there’s no plausible correlation or mechanism. Think this is some data issue and quirk of qpAdm and not real but not sure why. Visible in the figure from the above post, and it doesn’t seem like new adna has solved it. That might reduce baseline of Finnish Mongolia_N down a couple of percent in reality.)
The Finns are a heavily drifted population, with secondary waves of severe drift associated with their internal migrations to the Center-East of Finland, and strong extant population structure. I am afraid that this makes it too hard to dissect their population history just from the allele and haplogroup frequencies…
About half my ancestry is Swede-Finn (i.e. Swedish speaking Finns). I was surprised to learn upon getting 23andMe results that:
1. Despite being derived from culturally the Swedish colonist influenced community, I have almost no non-Finnish autosomal ancestry from elsewhere in Scandinavia, despite the fact that Swedish rule was from about 1150 CE to 1809 CE (i.e. Finland was part of Sweden less than a century before my Swedish speaking ancestors left). Many generations of marriages to local women has diluted Swede-Finn Swedish ancestry to nearly nil.
2. Finland, in addition to being very distinctive from the rest of Europe, also has immense substructure. 23andMe accurately pointed to the correct county sized region within Finland from which my ancestors derived with surprisingly low uncertainty (the genealogy is so solid I have documented genealogical ties to about 1/6th of the Finns alive today). Different county sized areas in Finland are more genetically distinct from each other than Northern France is from huge swaths of Germany.
Also historical populations of Finland (per Wikipedia sources):
1150: 20,000–40,000
1550: 300,000
1721: 250,000 (at end of Famine of 1696-1699 that killed 1/3rd followed by twenty year long Great Northern War)
1750: 428,000
1770: 561,000
1790: 706,000
1810: 863,000
1830: 1,372,000
1850: 1,637,000
1870: 1,769,000
1890: 2,380,000
1910: 2,943,000
1930: 3,463,000
1950: 4,030,000
1970: 4,598,000
1990: 4,977,000
2010: 5,375,000
And, confirming the OP speculation about historic era migration: “By 2000 about 6% of the population spoke Swedish as their first language, or 300,000 people. However, since the late 20th century there has been a steady migration of older, better educated Swedish speakers to Sweden.”
@John Massey: Hint: not all English speakers in Ireland (or Kenya) have in recent centuries migrated from England, even though some have. Also, the Swedish is spoken in Finland with quite a different accent than that of Riksvenska spoken in Sweden. Linguistic substratum, perhaps?
@A. Karhukainen – Thanks. I got your point.