Sunday, June 11, 2017

Recalibrating Archaic Admixture in Homo sapiens

I was looking at the data in the just published Hublin et al. paper, and the paper published by Rightmire back in January.  I realized that the reason that archaic admixture in H. sapiens is estimated as being low (less than 4%) is that the recent Neandertal and Denisovan samples used by Svante Pääbo's and David Reich's research groups are ill chosen in terms of the time periods they look at.  This only shows that H. sapiens, who had diverged genetically from Neanderthals and Denisovans 400,000 and 600,000 years before present, approximately, was sufficiently diverged that admixture was a low probability event.  The Neanderthal and Denisovan samples used to test admixture are less than 100,000 years old.   They are not 600,000-200,000 years old, and cannot tell us about the gradual divergence, or not, of hominins in Africa and Eurasia.

Here's an incomplete list of hominin crania:

The relationship is not well established, but it is increasingly apparent that an immediate ancestor of H. sapiens is H. heidelbergensis.  The range of H. heidelbergensis stretched from Germany to Ethiopia between 600,000 and 400,000 years ago.  This is based on the small number of H. heidelbergensis samples that exist for that period:  the Mauer Mandible, Petralona, Arago 21, and Bodo.  The range of H. heidelbergensis seems to shift further southward into Africa after 400,000 years ago, but that is difficult to assert when there are so few samples, and these are not securely dated.

What is clear is that archaic H. sapiens appears between 300,000 and 200,000 years ago in Africa, and probably derives mostly from H. heidelbergensis in Africa. What is very much not clear is the degree to which H. sapiens derives ancestry from Neanderthals, Denisovans, H. heidelbergensis, and H. erectus outside Africa in the period between 300,000 and 100,000 years ago.  We have no autosomal ancient DNA samples for Neanderthals, Denisovans, H. heidelbergensis and H. erectus from this period. 

Scientists who make these bold assertions about limited archaic admixture should qualify that their samples are very late in terms of the time that a high degree of admixture between H. sapiens and other humans would have been occurring.   I think this is especially true when talking about Asia, where it is somewhat apparent from the Dali, Jinnuishan, and Xuchang crania, that early archaic H. sapiens probably had a much broader range than Africa after 300,000 years ago.

It is frequently asserted, based on the distribution and characteristics of contemporary human DNA, that most humans left Africa no earlier than 80,000 years ago, and those that did went extinct.  There is no basis for these statements based on the admixture outside Africa argument.  The Denisovan and Neandertal samples used to make this assertion are too late and too sparse to test archaic admixture outside Africa.  These tests assume low mobility and little intermixing between African and Non-African groups between 400,000 and 80,000 years ago.  I doubt that is the case.

The process of the early formation of H. sapiens is probably much more complex than we've imagined.

No comments: