With the detection of Neanderthal admixture in Eurasians (Green et al), evidence for two admixture events in an upcoming paper from Jeffrey Long’s group (probably Neanderthals and erectus), and analysis from Jeffrey Wall and Vincent Plagnol suggesting that some African populations (Pygmies and Bushmen) admixed with other archaic populations, it seems that we are on the verge of a new paradigm.* In the picture, anatomically modern humans arose in East Africa and then spread globally – but mixed (to a degree) with the archaic humans that already already occupied most of the Old World.
Of course this is the surprise-free prediction, much more plausible than total replacement, which would have required an unlikely biological incompatibility.
Admixture at the few-percent level would have given most favorable archaic alleles a good chance of reaching high frequency in modern human populations, and some must have done so – alleles that conferred regional adaptations, and perhaps some that had general advantages. Svante Paabo thinks, or any rate says, that this Neanderthal admixture probably didn’t have any biological significance, but he’s almost certainly wrong.
This pattern has been seen in some other invasive species: the cosmopolitan species assimilates favorable alleles from local sister species. Resistance is futile.
There may be more in the pipeline: The Max Planck people have that strange finger bone from the Altai, and if it’s a good-enough sample, they could run the same general analysis and check for admixture from that archaic population, whoever they were. And there’s a possibly-significant line in their Science paper, concerning Wall and Plagnol’s high estimate (14%) of archaic admixture in Europeans:
“almost an order of magnitude greater than our estimates, suggesting that their observations may
not be entirely explained by gene flow from Neandertals. ” Maybe from somebody else?
*Mike Hammer and others have also been saying this.