I’m not a population geneticist, nor do I play one on TV but of course I’m quite interested in this attempt. I’m competent to evaluate this stuff to a point - but for the nitty-gritty, a real population geneticist would need to weigh in. Paging @glipsnort, and perhaps @T.j_Runyon (or Steve) would like to invite my friend Joe Felsenstein over to BL. There’s also a conversation going on at Peaceful Science, so that’s a thread to watch as well.
So, time for a hot take.
In many ways, it has the same results as the conversation we had here, some time ago, with Richard Buggs. And it looks to have the same basic features, especially that it has to have exactly the right variation in two people (all four alleles different, for everything) and requires exponential population growth, especially early on, to not lose the precious variation from just two people.
In the Gauger paper, we see that they need a population growth rate of doubling every 10 generations in order to get this to work. That is a FAST rate for a prehistoric hunter-gatherer population. That’s a doubling time around every ~250 years. Measured rates are more like doubling every 1,700 years. Now, that paper shows that you can have a rate like what Gauger is proposing for short periods - but for the Gauger model to work you need this rate, uninterrupted, until the population swells to 16,000 individuals. Hmm. That’s a stretch.
I think Ann meant “mortality” there, and that’s a heck of a typo.
No, the rate that they chose is not “parsimonious”. It’s as high as has been observed over very short timeframes assumed to hold over long periods.
There’s also the issue of how you stop interbreeding, just like we discussed with Buggs (and no one was able to propose a mechanism that obliterated every other hominin on the planet (Africa, Asia) except two. I see that Ann has touched on this in a piece featured at the ID site evolution news:
Really? That needs more fleshing out. The pair stays isolated in a gorge until the population hits 16,000 with exponential growth? That’s a pretty well-stocked gorge. An event that wipes out all but two across two continents (data not shown)? Evidence, please. Same issue we had with the Buggs idea.
So, it seems like a case in special pleading to me. And it shows just how far you have to push things to try to shoehorn present-day variation into just two people. But I’m welcome to be corrected by someone with deeper skills in pop gen. And I’d love to hear Joe, or Steve, weigh in with their thoughts.
Edit to add: the thing about bumping this towards two people at 100,000 years ago - I’m not seeing anything to support that, but it’s probably going to be some attempt to accelerate mutation and recombination rates (since that is what it would require). But I see no attempt to justify that in the paper.