Message #10850

[Europa]

Results and Discussion

To represent the indigenous population of the British Isles, we have selected a site in central Ireland that has had no known history of contact with Anglo-Saxon or Viking invaders (Castlerea, see Figure 1). Given the demonstrated similarity of Celtic and Basque Y chromosomes [4 and 5] (p = 0.6, using haplogroups), these sample sets were combined [8 and 9] to provide a representation of the Y chromosomes of the indigenous population of the British Isles. Norwegian invaders were represented by two sites in western Norway (Bergen and Trondheim), Danes were represented by a general Danish collection, and Anglo-Saxons were represented by samples from their historical homeland in Schleswig-Holstein (North Germany). Linguistic and historical investigations seem to suggest that internal migrations were minor and have not unduly blurred the genetic landscape of North Germany and Denmark in the last 1500 years [10]. We also note that some historians view the Anglo-Saxons themselves as Germanic invaders from what is now North Germany/Denmark. Population differentiation between the continental and indigenous British Isles groups was assessed by using an analog of Fisher's exact test calculated by using haplogroup (hg) frequencies, as implemented by the Arlequin software package [11]. There was no significant difference between the Trondheim and Bergen samples or between the Danish and North German samples (p = 0., while the Norwegians were different from the other northern European samples (p < 0.05). We therefore clustered these source populations into two continental groups, referred to from now on as the Norway and German/Danish sample sets. Note that the similarity of the Danish and North German Y chromosomes means that, at the hg resolution, we cannot distinguish the genetic contributions to the British Isles of the two component groups. All continental populations, however, show significant differences from the indigenous group (p < 0.01), and Norway can be distinguished, though to a lesser degree, from the German/Danish sample (p < 0.05). Sampling in the British Isles was mainly undertaken to conform to a systematic 3 × 5 grid (Figure 1).



Figure 1. British Isles Sampling Locations MapThe location of the sampled small, urban areas and the 3 × 5 grid of collection points are shown. For each grid point, we selected the closest town within a 20-mile radius. Only towns with 5–20,000 inhabitants were chosen. Individuals were, with the exception of one location, then selected if their paternal grandfather's birthplace was within a 20-mile radius of the selected center. Midhurst samples were collected up to 40 miles from the respective grid point. When the grid point was at sea, the nearest point on the coast was used (Morpeth and Stonehaven). We also added additional points to cover important geographic regions not covered by the grid (Shetland, York, Norfolk, Haverfordwest, Llangefni, Chippenham, Cornwall, Channel Islands) and included two Irish samples, Castlerea and Rush (North of Dublin). The total number of points sampled in the British Isles was 25.