Message #10854

[Europa]

To aid interpretation of this plot, we simulated admixed populations by drawing varying proportions of individuals at random from each of the source populations and plotting them on PC plots that also included the source populations themselves (Figure 4). The simulated populations show that the position of a population on the first two axes provides a sensitive indicator of the degree of continental input into an indigenous background, with Norwegian input moving populations strictly along axis 1 and German/Danish input moving populations at an angle through both axes. Inspection of Figure 3, in light of these simulations, shows that Orkney and Shetland have significant Norwegian input and little to no German/Danish input, that the English and Scottish sites all have German/Danish influence, and that the Western Isles and Isle of Man have German/Danish influence, presumably due to English immigration. In addition to these relatively clear patterns, the PC plots also provide a suggestion of more subtle differences. For example, there is a group of populations that appears shifted to the left from the main angle of German/Danish influence, and this is consistent with some degree of Norwegian input. It is not surprising that the Western Isles and Isle of Man are in this group, but the inclusion of Penrith is of particular interest given the Scandinavian influence on dialect in this region [15]. Similarly, Rush appears to be shifted slightly toward the Norwegian pole on PC1, but there is no shift toward the German/Danish position. In addition, the mainland Scots are somewhat closer to the indigenous type than any English sets, except Cornwall. The sites with the highest degree of German/Danish input are York and Norfolk, followed by Southwell and Llanidloes. All of these except Llanidloes are historically in regions where the Danes are known to have had a significant presence. The peculiar position of Llanidloes might reflect recent migration in the past two centuries [1]. The remaining samples are closer to the indigenous group; for these populations, this finding suggests a lower demographic impact by North European populations. This can also be seen in the frequency of AMH+1, which is always above 33% in British populations but remains below 26% in the continental source populations; these data are consistent with the presence of some indigenous component in all British regions.



Figure 4. Principle Components Plot of Simulated PopulationsA Principle Components plot of simulated populations of N = 50 comprising, at admixtures of, respectively, 20%, 40%, and 60%, the indigenous British and the German/Danish and Norwegian sets. The circles group simulated populations with the same continental proportions. The arrows indicate the directions along which the simulated population tends to move according to the relative proportion of Continental input.