Tuesday, January 19, 2010

293. Aftermath 8: The Later Migrations

Let's take yet another look at the fascinating "isofrequency maps" produced by Mait Metspalu et al:


Maps B and C especially tell a remarkable story, which, I must confess I was not quite prepared for. Given the Toba scenario I've been exploring (which remains hypothetical, as I hope everyone reading here understands), I'd assumed that the resulting "bottlenecks" would have led to fundamental changes, away from the typically African characteristics of HMP, toward those more typical of what we now see among most (though not all) of the various "races" and large-scale "ethnic" subdivisions of Asia, Europe, etc. I still see this as a likely possibility.

However, since India would have borne the brunt of the disaster, it seemed likely that India would have been the principal staging ground for the migrations that would have spread the newly altered genetic/cultural lineages to the four corners of the world. This was the scenario I presented in my "Echoes" essay. But maps B and C tell a different story. According to map B, India has remained relatively isolated, while the scenario implied in map C suggests a massive migration based to the east of India, and spreading both north and northwest from there, with the Himalayas as a significant barrier, channeling the migrants away from India and in the direction of Central Asia and, ultimately, Europe.

Another map, Figure 5 from the same paper, is explicitly devoted to the migration pathways:



From the caption:
Peopling of Eurasia. Map of Eurasia and northeastern Africa depicting the peopling of Eurasia as inferred from the extant mtDNA phylogeny. . . [T]he initial split between West and East Eurasian mtDNAs is postulated between the Indus Valley and Southwest Asia. Spheres depict expansion zones where, after the initial (coastal) peopling of the continent, local branches of the mtDNA tree (haplogroups given in the spheres) arose (ca. 40,000 – 60,000 ybp), and from where they where further carried into the interior of the continent (thinner black arrows). Admixture between the expansion zones has been surprisingly limited ever since.
While the authors seem principally intent on demonstrating the likelihood of the "southern route" over the northern one (and imo their demonstration is extremely convincing), the picture they provide here of the later migrations is equally compelling, in my view.

Though it's extremely difficult to account for every aspect of the genetic picture in terms consistent with the Toba hypothesis (or any other hypothesis, for that matter), I'd like to propose the following, provisional, scenario:

1. 85,000 to 75,000 ya: Exit of HMP from Africa, and migration eastward through South Asia to Southeast Asia and beyond, following the southern route.

2. 74,000 ya: Toba explosion, decimating or completely destroying all migrant settlements in South Asia.

3. 74,000 ya: Population bottlenecks produced by the disaster, with varying degrees of intensity depending on how close each population is to the ash plume. The effect of each bottleneck will be different, depending on completely unpredictable circumstances associated with each group affected. In each case, either the group or its lineage does not survive at all, or, if it does survive, its character, both physical and cultural, will be determined by the unique qualities of each new founder group.

(to be continued . . .)

3 comments:

Maju said...

Your last description of a Toba scenario (different from Oppenheimer's) is an idea I have flirted with some times in the past.

However mtDNA structure would seem to to play against it because, while M and N have medium sized stems (4-5 mutations, 3-5 if we consider only coding region ones), suggesting medium length coalescence times, many or even most of their sublineages have very short stems of one or two mutations only, suggesting that they expanded shortly after their formation.

Y-DNA could be more favorable: the stems leading to F and C, and even DE and D, seem very long. But the full knowledge of the mutational structure is nearly impossible for Y-DNA, while is instead relatively easy and often readily available for the much shorter mtDNA molecule. So mtDNA is in principle more reliable.

Anyhow, one still has to account for the formidable push of macro-haplogroup F (that I compare with mtDNA M) that again has a clear highest basal diversity in South Asia.

Would I go with the Toba catastrophe scenario, I'd have to think either as you are doing in this post (many local lineages surviving, what means not much of a bottleneck after all) or force M and N to have expanded only after Toba (so the bottleneck would be at their stems, not in Arabia but already in South and SE Asia).

In this last case, again, South Asia appears as the main source of early expansion (post-Toba now), having in any case the highest basal diversity for the most widespread and diverse macro-haplogroups, both in the mtDNA (M, also R) and Y-DNA (F).

DocG said...

Maju: "Anyhow, one still has to account for the formidable push of macro-haplogroup F (that I compare with mtDNA M) that again has a clear highest basal diversity in South Asia."

Again, thank you for a very thoughtful and informative comment, Maju. But assuming India had been devastated by Toba, what we now see there could not have originated there. In other words, if the migration had preceded Toba, all the many M's we now see in South Asia could have originated among Toba survivors east of Bengal (or possibly also among Toba survivors in the north and west).

So could the picture of these particular M haplogroups as having arisen in South Asia be a kind of mirage?

"or force M and N to have expanded only after Toba (so the bottleneck would be at their stems, not in Arabia but already in South and SE Asia)."

That's an interesting idea, but I don't see it as likely, because whatever would have preceded M and N (such as some form of L3) would now be present in the region to the east of the Toba plume, and so far nothing like that has been found (though New Guinea and much else in Melanesia has not yet been sufficiently researched imo).

Maju said...

So could the picture of these particular M haplogroups as having arisen in South Asia be a kind of mirage?.

It might be but is quite strongly against the logic of population genetics (greatest diversity tends to indicate origin). Whatever the case the diversity is high in both South and East Asia, so the expansion must have been nearly simultaneous in both regions.

My understanding is one of somewhat complex foundational flows forth and back within a super-region that I call Southern Asia or Tropical Asia (i.e. South Asia and SE Asia, including southern China). Followed by de facto separation of the two subregions, probably after the population reached some critical density.

The greatest basal diversity of the most important macro-haplogroups is clearly in South Asia but some lineages seem to have a SE Asian core instead.

That's an interesting idea, but I don't see it as likely, because whatever would have preceded M and N (such as some form of L3) would now be present in the region to the east of the Toba plume.

This objection is not too valid if you argue for a bottleneck (M and N are 2 out of 7 L3 sublineages, already "too many" in a sense, even without bottleneck) and the colonization of SE Asia was at the moment of Toba limited.

But it's largely a guess game because we have just no data between the L3 and the M and N nodes. And nobody is going to come out of heaven or hell to tell you the correct answer (nor we have time machines to go back in time and find out).