At last, the death of the Toba bottleneck

It is no secret that I really don't like the hypothesis that the massive ancient eruption of Mt. Toba, Sumatra, wiped out much of the worldwide human population 74,000 years ago, possibly allowing modern humans to spread in its wake.

Sure, this eruption was the largest known within the last half-million years. If any ancient volcanic event was going to have an effect on human populations and world climate, it would be this one. And it remains quite possible that there were severe climatic effects lasting a millennium or more.

But there has never been any sign of anatomical or archaeological discontinuity outside Africa at this time. Moreover, no genetic evidence suggests a sudden harsh bottleneck at 74,000 years ago -- most genes are consistent with such a bottleneck only because a recent, sudden, and short bottleneck would have almost no effect on gene diversity. Considering that the Neanderthals in glacial Europe continued right on after the Toba eruption without any hiccup, it always seemed like a very shaky idea.

But still, there seemed to be nothing impossible about a more local effect of the eruption. I mean, if a giant megavolcano spouts off right next door, it has to be bad, right?

Well, maybe on Sumatra itself, but apparently not in some other fairly nearby places. This week's paper by Petraglia and colleagues (2007) appears to have sunk the Toba bottleneck entirely. Very simply, they found a Toba ash horizon in India, and found very similar archaeology both below and above the eruption.

Based on some features of the tools, Petraglia and colleagues speculate that the makers may have been a relatively early sample of modern humans:

Analyses of the archaeological industries recovered from the site indicate a strong element of technological continuity between the pre- and post-Toba assemblages. Together with the presence of faceted unidirectional and bidirectional bladelike core technology, these pre- and post-Toba industries suggest closer affinities to African Middle Stone Age traditions (such as Howieson's Poort) than to contemporaneous Eurasian Middle Paleolithic ones that are typically based on discoidal and Levallois techniques (Fig. 3). The coincidence of (i) evidence of hominins flexible enough to exhibit continuity through a major eruptive event, (ii) technology more similar to the Middle Stone Age than the Middle Paleolithic, and (iii) overlap of the Jwalapuram artifact ages with the earlier end of the most commonly cited genetic coalescence dates (21-23) may suggest the presence of modern humans in India at the time of the YTT event. This interpretation would be consistent with a southern route of dispersal of modern humans from the Horn of Africa (24); the latter, however, will remain speculative until other Middle Paleolithic sites in the Indian subcontinent and Arabian Peninsula (25) are excavated and dated.

I tend to discount point (i) about flexibility, since European Neandertals were apparently flexible enough to survive ice ages with large decade-scale swings from warm to cold. But it is hard to get people to Australia by 50,000 years ago unless they were in India before that.

A dispersal of MSA people from Africa would be an interesting twist on the "modern human origins" problem. If the first "modern" humans outside Africa were MSA users, there is no particular reason to assert that they were different from the population represented at Skhul and Qafzeh. The lack of a full Upper Paleolithic technical kit anywhere in Africa before 50,000 years ago makes an MSA-associated disperal seem more credible. The assembly of the Upper Paleolithic in Eurasia would therefore be a local cultural development, possibly associated with further biological change.

I find that association to be the most important reason to continue investigating these sites. In the meantime, we can forget about the cataclysmic effect of Toba on the poor hominids.


Petraglia M, Korisettar R, Boivin N, Clarkson C, Ditchfield P, Jones S, Koshy J, Lahr MM, Oppenheimer C, Pyle D, Roberts R, Schwenninger J-L, Arnold L, White K. 2007. Middle Paleolithic assemblages from the Indian subcontinent before and after the Toba super-eruption. Science 317:114-116. doi:10.1126/science.1141564