According to a recent paper a human population in South Africa may have been isolated for a long period of time - until about 1400 years ago. If so, the population may well have included no GAE descendants until that time.
Ars Technica report: Humans in southern Africa were an isolated population until recently - Ars Technica
Paper: Homo sapiens-specific evolution unveiled by ancient southern African genomes | Nature
There does remain the possibility of genealogical ancestry that left no genetic signature. Even one migrant long ago could diffuse ancestry throughout the population.
Yes, we can’t prove it - but it does show that there is reasonable doubt about GAE. Unless the deadline is brought forward a few centuries.
Much more important data in this paper about human evolution than GAE.
Yes, but the probability for that happening would most probably be independent for each isolated population (I cannot imagine why the probability of this happening in South Africa would depend on it happening in Tasmania, for example). This would mean that the probability for it happening in all isolated populations would be multiplicative, and so get low quite quickly as the number of isolated populations increases.
I don’t think this produces reasonable doubt, especially compared to Tasmania. Migration must be completely nonexistent, and a water barrier plus a people without boats make a reasonable bet. All you have here is negligible detected introgression but no impassable barrier, and a migration rate below 1 per generation would be compatible. A sufficient amount of migration for the scenario would be be likely undetectable in the genome.
Note that I don’t believe that GAE existed, but the genetic data aren’t part of the case against them.
That assumes that the probability for any given isolated population isn’t close to 1. And that there are many such populations, though we know only of one or two.
I disagree. Migration is very limited - and the entire population must be expected to be GAE descendants by the deadline. You suggest that the probability of that is “close to 1” but it’s far from clear to me how that can be the case.
It only takes one who has descendants. Diffusion within the population takes care of the rest.
Not quite. It takes one who arrives early enough for diffusion to cover the entire population by the deadline.
For Tasmania, GAE relies on putting the GAE well before the isolation of the Tasmanian population. The isolation of the South African population goes back much further.
You seem to be confusing “before the population becomes isolated” with “early enough for diffusioh to cover the population”. There is no evidence or expectation that the South African population was ever completely isolated, just that there were very few migrants. Tasmania is different in that respect: there seems an effective barrier to migration.
For Tasmania, we don’t know it was completely isolated - it may well have been, but it’s hardly proven. The size/nature of the barrier is often over-stated, as I read the evidence. e.g. The strait was crossed by paddle boarders in 2014, and it has been crossed by kayakers, and even swum across. We know the indigenous Tasmanians and people on the mainland had small boats/rafts.
No, just because I mention both doesn’t mean that I’m confusing them.
The rarity of immigrants makes it a concern that a GAE-descended immigrant would arrive and their descendants spread to include the entire populationbefore the deadline.
The length of isolation is a concern because the main way of dealing with Tasmanian isolation is to place GAE well before the isolation. Which could still be done - but pushes GAE much further back.
We also know that the “small boats/rafts” of Tasmanians and adjacent mainlanders were not sea-going. I covered this in this post in an earlier topic.
My point is that immigrants could easily be common enough to make the scenario credible without showing up in the population’s genomes.
Yes, isn’t the point of GAE that such genealogical ancestors wouldn’t leave genetic evidence? Am I missing something?
We’ve already got past “could be”. The question is likelihood. Something was going on that prevented detectable mingling of populations. How far did it go? How can we know? Maybe the isolation was near-total.
Well, somebody here is missing something.
I’d say that we simply lack information. We don’t know how this isolation occurred or lasted so long. To me the duration suggests a strong barrier of some sort - although it is only suggestion. The fact is we don’t know.
True. The point is that you can’t say, based on the genetic evidence, that the likelihood is low. We don’t have the necessary evidence to determine the likelihood, which would be the migration rate. There’s a lot of distance between the level needed for a genetic signature and the level needed for genealogical ancestry, and between that and zero. All we know is that it’s somewhere between the first and the last.
Sure, there was a strong barrier. But the point is that a strong barrier doesn’t have to result in complete isolation. As you say, we lack information. And that means that you can’t say it’s a problem for GAE.