With the latest paper, the story on European origins is becoming clearer. Three populations account for European ancestry: the Mesolithic hunter-gatherers of western Europe (dubbed WHG in the paper), early European farmers derived from somewhere in the Mideast (EEF), and a third group more closely related to ancient Siberians (ANE) than any existing population. Those Sibermen also contributed a third of Amerindian ancestry, the rest being similar to modern East Asian populations.
The key facts are the dates, derived from radiocarbon dating, and statistical tests on ancient and existing genomes, tests that can show admixture, tree structure, etc. WHG and ANE are sister populations, but distinct. Here’s a graphical representation of the model:
These methods are powerful, but not all-powerful. They can detect population changes if the intrusive group is sufficiently genetically different – basically, if it has drifted enough. But I don’t think they can detect movements of and/or replacements by more closely-related populations., and it looks to me that there must be more to the story than we see in this graph. The uniparental lineages hint at more dramatic events than these autosomal analyses. G2A was apparently very common in the Impressed-Ware populations along the Mediterranean coast, but now it’s rare, found mainly on islands and up in the hills. Yet this analysis suggests that southern European populations such as Tuscans are 75% EEF farmers – where did the Y chromosomes go? Northern European populations are, according to these calculations, something like 35% descended from WHG populations. Those WHG populations seem to have had exclusively U mtDNA lineages (U,U4,U5, and U8) but those only add up to a frequency of 10 or 11% in those northern European populations today. Where did those mtDNA lineages go?
Could be selection. When you look at the explosive expansion of R1b and R1A (probably originating in the Sibermen), you have to wonder. Or maybe it was more like the story of the Golden Family, the Jenghiz-Khanites, or Neil of the Nine Hostages – maybe everybody who was anybody in early Bronze-Age Europe had to be a descendent of a certain legendary (but real) king, and this distorted Y-chromosome transmission
Could be near-replacement by some population that was autosomally similar but fairly different in its uniparental lineages. Could be that we’re missing something in the analysis.
The Sibermen look to be the Indo-Europeans, or a major component thereof. And so you must have an Urheimat that is fairly far east – not in Europe. Before their advent, most of Europe must having been speaking EEF languages of some kind, of which Basque is the only survivor. Interesting that this paper say absolutely nothing about the Indo-European expansion. The word “language” isn’t even used. I wonder if there is a paper on that in prep, one that makes use of all this new ancient DNA information. Or, maybe the authors feel it isn’t their pidgin and are leaving that up to someone else.