Complex Signals for Population Expansions in Europe and Beyond
Kristiiana Tambets et al. 2003
From the article:
"Case study: U4U4 is even more 'European' than U5: while U5 is relatively frequent all over western Eurasia, U4 is, with a few interesting exceptions, more frequent in eastern Europe and is either absent or very rare in the Near East and elsewhere. In the European north, an interesting exception is the Saami mtDNA pool, where U4 is virtually absent. We have constructed a HVRl-based phylogenetic tree for U4, using information from ~80 populations comprising a total of ~400 U4 genomes (Fig. 35.2). The topology of the U4 cluster is relatively simple, revealing the presence of a limited number of sub-founders. Of these, U4a and U4b are likely monophyletic, while U4c, determined by a transition at np 16,362, might be polyphyletic, at least in a pan-western Eurasian context. The highest frequencies of U4 (both in absolute terms and as a percentage of Hg U) can be observed actually not in Europe, but among Obi-Ugric Khantys and Mansis, living in northwestern Siberia. It is also frequent among the Finnic-speaking populations and in Volga Basin Turkic speakers, where, in some instances, its frequency exceeds that of U5.In spite of this, we have not found any U4b mtDNA genomes among Finno-Ugric and Volga region people (N > 1000). This sub-cluster is largely, though not solely, typical for Germanic-speaking populations, being yet another highly characteristic example of a steep cline in the distribution of maternal lineages in Europe. The coalescence age of U4 is around 16,000-24,000 BP (Richards et al. 2000). With geographically more representative data at hand, it is interesting to estimate coalescence ages not only for each sub-division of U4 (i.e. U4*, U4a and U4b), but also for different linguistic/geographic entities within a sub-cluster. We found the answers intriguing. For the Baltic Finno-Ugric and Volga people (note that Hungarians differ here from the Finnic-speaking people), the coalescence ages both for U4* and U4a are around the maximum of the LGM, at 20,000-22,000 BP. furthermore, taking U4c tentatively as monophyletic for this particular region, the corresponding sub-clade lineages in the FU-Volga area coalesce at about 19,000 BP — coinciding within the limits of error with U4* and U4a. Postulating the beginning of expansion during the LGM seems strongly counter-intuitive at first glance. Here, however, comes an equally unexpected archaeological finding (Dolukhanov 2000), that actual population density (calculated from the number of precisely dated settlements) rose considerably after about 25,000 BP in a periglacial area of northern Ukraine-southern Russia, reaching its maximum around the peak of the LGM. It is of course highly speculative, but nevertheless tempting to bring these two completely independent findings together and to suggest that the eastern 'Periglacial refugium' postulated by archaeological data, and beginning of theexpansion of U4 among eastern Europeans, can be attributed to the same prehistoric people. Next, we calculated coalescence ages of U4*, U4a and U4b for the Germanic speaking people (Germans, Norwegians, Swedes, Icelanders, Scots, German-speaking Swiss). Here the other interesting observation came. As indicated above, they share U4* and U4a with FU-Volga people, whereas the latter lack U4b. Nevertheless, the coalescence ages for all three indicated clades/sub-clades for the Germanic-speaking people are close and lie around 10,000-14,000 BP, suggesting that the beginning of their expansion was:a ) in the late Pleistocene, corresponding to the period of fast regression of continental ice cover in northern Europe and the general 'improvement' of climate;b ) much later (for U4* and U4a) than for people living in the adjacent geographic area of northeastern Europe.One may ask about U4 in Mediterranean Europe. While U4a is so rare there that no meaningful calculation can be performed, the coalescence age for U4* for the Mediterranean isagain about 13,000 BP. Notice that this time scale (late Pleistocene) overlaps with that which one observes for the majority of the nicely star-like sub-clades of U5, discussedabove. There is, of course, a profound difference between the spreads of U4 and U5 in western Eurasia. U5 is one of the major pan-western Eurasian maternal lineage clades, present in northwestern Africa, in the Near and Middle East and in Central Asia, while U4 is largely a northeastern-central European variety of mtDNA, found also in western Siberia/Altai and, in low frequencies, in Mediterranean Europe and the Near East. Interestingly, we have found a few U4 lineages even in India (see Kivisild et al. this volume).