An analysis of genetic differentiation (based on pairwise Fst) indicated that the population of Sweden's southernmost counties are genetically closer to the HapMap CEU samples of Northern European ancestry than to the populations of Sweden's northernmost counties. [. . .] We have shown that genetic differences within a single country may be substantial, even when viewed on a European scale.The paper is in PLoS ONE (i.e., it's open access). More:
The first principal component showed the presence of a north-south genetic gradient that was mainly driven by each northern county being different from other counties. [. . .]
The second principal component did not display a constant gradient across Sweden. The county with the lowest mean was Dalarna, with samples in the north-western parts of this county contributing with particularly low values. These patterns may be linked to Finnish or Norwegian ancestry, although the component most probably encompasses other signals arising from factors such as isolation and drift in the more remote parts of Dalarna county. A large wave of migration did take place from parts of today's Finland to Dalarna, Värmland and Gävleborg counties during the 16th and 17th centuries, with some settlers moving further west to Norway. The migrants, known as the “Forest Finns”, have today become fully assimilated into Norwegian and Swedish society with almost none speaking Finnish. As we conservatively removed samples that had Finnish ancestry to reduce the effect of ethnic Finns on the Swedish population stratification, we removed samples that had a substantial Finnish contribution. Indeed, the samples that had the most extreme values with respect to principal component 2 fell close to our cutoff used to define Finnish ancestry (Figure S10). What we see may therefore be the remnants of early Finnish migrations to Sweden. It could be that even after the removal of samples with Finnish ancestry, Finnish influx leaves its mark, consistent with a long history of Finnish migration to Sweden coupled with a high degree of admixture. However, when we applied the component to a set of Norwegian samples, they too had negative component values with a small number of samples being clear outliers in the negative direction (Figure S6). It is therefore possible that a portion of the component predicts Norwegian ancestry, or isolation and drift in the border areas that has left its mark in both populations, or a combination of these factors.
Population pairwise Fst values
The observed genetic differences between the southern national areas and counties are similar to what has been seen in the UK, when comparing southern England to Scotland, and the genetic differences between the southern national areas and North Middle Sweden (the approximate geographical center of the country) are similar to what has been observed between the UK and Ireland . They are much smaller than the ones observed between regions in Finland where the Fst has been shown to be about an order of magnitude larger than what we observed among the southern Swedish national areas . The genetic differences between Upper Norrland, the northernmost national area, and the southern and middle national areas were much larger than the genetic distances we observed between the southern national areas and the HapMap CEU founders, highlighting the extent of the distances. The genetic differences observed between the southern national areas and the northernmost one were larger than what has previously been observed between Swedes and Danes, and Swedes and the Dutch . This puts the observed intra-Sweden genetic differentiation into perspective, showing that the genetic differences between the southern and northern parts of Sweden are of larger magnitude than those between a general sample of Swedes and samples from some other European countries and Americans of Northern European ancestry. [. . .]
The observed patterns of homozygosity further reinforce the picture of a genetic divide between the southern and northern parts of Sweden. While the differences in the number of extended homozygous segments were mostly small and non-significant for the southern national areas, the counties corresponding to the three northernmost national areas had a larger number of extended homozygous regions, suggestive of an increase in autozygosity. The northern national areas are much more sparsely populated than the southern with vast geographical distances between towns and villages, a fact that may have contributed to the observed loss of heterozygosity. [. . .]
The main south-north pattern detected in the principal component analysis may be the amalgamation of many separate factors. Immigration of continental Europeans, and Britons to the southern parts of Sweden may be one such factor, together with isolation and consecutive allelic drift and admixture with the indigenous Sami people in northern Sweden. [. . .]
The picture of the Swedish genetic structure presented here can only be considered as a historical snapshot. Immigration and movements around Sweden can alter the picture we have described considerably within the span of only a few generations. Within country population movement creates genetic diversity and breaks up existing structures. Immigration from other countries leads to new stratification of a vastly complex nature. With increased movements, homozygosity, due to a high degree of kinship with geographically local possible partners, may swiftly decrease.