Haplogroup correlations in Germans

I used the haplogroup data (Table 2) from Kalevi Wiik's new paper to see whether there were any correlations between haplogroup frequencies, similar to my recent analysis of Russian Y chromosomes. I considered only the most frequent haplogroups R1b, R1a1, I, E3b, J2.

R1b: +0.80 with J2, -0.73 with I, -0.50 with R1a1
R1a1: -0.72 with J2, -0.50 with R1b
I: -0.73 with R1b
J2: +0.80 with R1b, -0.72 with R1a1

The main points of interest are:

• R1b and J2, vs. R1a1
• I vs. R1b

The association of J2 with R1b is probably the most interesting and unexpected result.

Next, I tried to find spatial patterns of variation, by correlating haplogroup frequencies with longitude and latitude.

Longitude (west-to-east): J2 (-0.82), R1b (-0.59), R1a1 (+0.72)
Latitude (south-to-north): R1b (-0.56), R1a1 (+0.53)

The southwestern association of R1b is consistent with Iberian or Italian origins; the northeastern association of R1a1, in the direction of Poland and the Baltic states also makes sense.

The strong western (-0.82) but weak southern (-0.37) orientation of J2 is unexpected if J2 came to Germany from the Balkans, but is consistent with a maritime mode of propagation of this haplogroup.

Interestingly, the J2 frequencies in French (5-17.3%), Dutch (6.2%) and Belgian (5.0%) samples (Table 17) all exceed the German average (4.0%), so they are probably consistent with this interpretation.

We really need to differentiate between J2a and J2b clades in this area, since J2a may hold the promise of reflecting maritime colonization (as its high frequency in coastal and island southern Europe suggests) or Roman descendants, while J2b may hold the signal of an expansion out of the Balkan area (See also Some aspects of J2 distribution). In terms of the latter, it is interesting to note that haplogroup E3b, which includes the common Balkan subclade E-V13 has a south-to-north spatial correlation in Germany (-0.41), but no detectable association with longitude