We know of several examples of a huge expansion of a paternal lineage, and several other cases seem likely to be the same thing. It’s worth taking a close look at the first one found, the paternal lineage of Genghis Khan.
8% of the men in Central Asia carry this y-chromosome, half a percent of all men. One of the authors of that study made an interesting mistake: Chris Tyler-Smith, in an interview, said “We don’t think that Genghis Khan was the common ancestor, because our best estimate of the time when the common ancestor lived was a few generations before he was born.”
Now that’s silly, because are big error bars in that kind of TMRCA, not least when you’re doing Y-chromosomes with STRs, the state of the art at that time. If you found a piece of trinitite at Alamagordo and came up with a date of 1943 from some kind of radioactive dating, forget it: it was July 16th, 1945, 05:29:21 MWT (plus or minus 2 seconds). In much the same way, we know when Genghis Khan made his historical moves, and we know who the rewards flowed to – his sons, and a smaller fraction to his full brothers, not his cousins or whatever. Genghis’s sons got countries, had harems, founded dynasties: he (or maybe his father, depending upon how successful Hasar and Hachiun were) is the TMRCA .
It seems to me that this error stems from geneticists thinking that genetic data is the only real data: sloppy genetic time estimates trump precise historical dates. In much the same way, people (using the old too-high mutation rate) estimated that the split with Neanderthals was ~300,000 years ago, even though the fossil record clearly showed hominids in Europe shambling towards Neanderhood half a million years ago. The new, lower estimates of the mutation rate have reconciled genetic and paleontological evidence on the split time – but the geneticists should have realized that there was an inconsistency.
Unfortunately other disciplines have the exact same problem.
Anyhow, the existence of the Genghis effect should have us thinking more generally, about Clark’s survival of the richest, or about populations that grow from hundreds to millions in a few centuries, after chancing on a favorable ecological niche (moneylending, or being pacifist farmers in America, or perhaps a ruling class after a conquest). A ruling class could end up replacing the original conquered population without even intending to, if it stayed fairly endogamous while having higher fertility than the serfs.