Q&A: Where did the Neanderthals go?

From the fossil record, we know that the Neanderthal population started to decline around 40,000 years ago. There is some disagreement about how long they coexisted with humans—the species might have completely disappeared as long ago as 39,000 years ago, but some radiocarbon dating studies have suggested that Neanderthals might have survived in Asia as recently as 24,000 years ago. Many different factors probably contributed to their extinction, but their decline conspicuously coincides with the movement of anatomically modern humans into Europe and Asia [5].

Many different hypotheses have been proposed to explain the disappearance of Neanderthals. One possibility is competition with modern humans for resources or direct warfare between Neanderthals and humans. Humans were more technologically advanced than Neanderthals and might have been more efficient hunters. Anatomical evidence also suggests that humans were more efficient runners. It has also been proposed that Neanderthals were adapted to colder climates, potentially making it hard for them to thrive in the warming climate. Parasites and pathogens may likewise have contributed to the decline of Neanderthals. Like the conquistadors in the New World, humans may have carried diseases to which Neanderthals had built up no immunity.

There is strong genetic evidence that Neanderthal communities tended to be small and inbred by human standards, with little migration between populations that were situated in different parts of Europe. The Altai Neanderthal genome, which comes from Siberia, happens to be the genome of a girl whose parents were related at the level of half siblings. Such inbreeding may have left Neanderthals vulnerable to high rates of genetic diseases, which can threaten population health further and lead to death spirals of mutational meltdown. The structure of Neanderthal genetic variation tells us that the Altai girl’s inbred family history was not just a one-off event—rather, she came from a society that had maintained a small population size for tens of thousands of years. This is analogous to the way that human genomes contain evidence of an ‘out-of-Africa bottleneck’—African populations living today have more genetic diversity than non-African populations because the human population that first left Africa was relatively small and inbred. However, this migratory human population appears to have bounced back quickly. Neanderthals never rebounded like this; they just kept dwindling in population size until disappearing altogether from the fossil record.

We can see evidence for selection against Neanderthal DNA by comparing the most important regions of the human genome to the ‘junk DNA’ that takes up space in our genome without helping to keep us alive. When Sriram Sankararaman, David Reich, Benjamin Vernot, and Joshua Akey used computational techniques to identify fragments of Neanderthal DNA in large modern human genome repositories, they found that these bits of Neanderthal DNA were much less likely to overlap important genes than we would expect to see if Neanderthal DNA was distributed at random [8, 9]. This suggests that people with Neanderthal DNA in certain important parts of their genomes have historically had fewer children on average than people whose Neanderthal DNA is harmlessly confined to ‘junky’ regions of the genome [10].

A few days after we published our theoretical analysis of the dynamics of Neanderthal introgression, a team led by Qiaomei Fu and David Reich [11] published an empirical confirmation of one of our predictions! Specifically, they looked at Neanderthal ancestry in 51 anatomically modern humans who lived between 45,000 and 7000 years ago and discovered that the older individuals had more Neanderthal ancestry than the younger individuals did. Their study provides strong evidence that our ancestors had more Neanderthal ancestry than people have today, adding support to the idea that natural selection has been slowly purging Neanderthal DNA from our gene pool.

We need more estimates of the proportion of Neanderthal DNA in old human remains. The more of these data points we have, the better we will be able to estimate how much Neanderthal DNA was present in the human gene pool during the period of human–Neanderthal interbreeding, as well as the period immediately after this interbreeding stopped.

See references [5–8, 11].