The inhabitants of the Pacific came in waves. Aboriginal Australians were the first to cross the area, and they were followed by separate populations that inhabited New Guinea and nearby island chains. Later still, the Polynesians, descendants of early East Asians, spread through the distant islands of the Pacific.
While modern genetics has made these rough outlines clear, it has also made it clear that these different populations sometimes interacted, sharing DNA along with technology and trade goods. Paleontology finds have made it clear that at least three distinct hominin species had occupied some of these islands before modern humans arrived, including the enigmatic Hobbits of Indonesia and a similarly diminutive species in the Philippines.
A recent study of the genomes of Pacific island populations provides a map of some of the major interactions that took place in the Pacific. And it suggests at least one of these involved the introduction of additional Denisovan DNA.
The work started with the sequencing of over 300 genomes volunteered by individuals from 20 different populations throughout the Pacific. The research team grouped these populations according to whether they came from Near Oceania (Indonesia, New Guinea, and the Philippines) or more distant islands of the Pacific (collectively Far Oceania). The latter is largely populated by the Polynesians, who arrived relatively late and had a distinct genetic history. But there were clearly interactions between the two groups, and the border between the areas each occupies is fuzzy in locations.
By comparing the genome sequences with each other and ancestral populations, it’s possible to make estimates of which groups are related to which others, as well as the time at which the different populations branched off. In addition, it’s possible to detect interbreeding among the populations, based on the appearance of stretches of DNA that are found in one population but are more similar to those from another.
The people who live in the highlands of Papua New Guinea have the earliest split, separating from the populations of other islands about 40,000 years ago. The branches of that lineage who inhabit the Bismarck and Solomon Islands separated from each other about 20,000 years ago.
But things get much less neat in Vanuatu, a group of islands out past the eastern end of the Solomons. About a third of their genome comes from Bismarck islanders, and that was a recent arrival, the result of interactions that took place only about 3,000 years ago. The rest comes from a group that started out in Papua but interbred with the Solomon Islands population en route. All of that means that Vanuatu is like a melting pot of near oceanic populations.
Then there are the Polynesians. They seem to have interbred with both the Bismarck and Solomon islanders. The best fit to the data involves one interaction right as the Polynesians arrived in the area about 3,500 years ago and a second interaction that occurred a thousand years later.
All of the populations sampled seem to have roughly similar amounts of Neanderthal DNA, present in similar locations in the genome, suggesting there was nothing unusual about their genetic history compared to other groups in the region. But that was not the case with the Denisovans. The amount of Denisovan DNA varied considerably among the populations, with the highest percentage found in those from the New Guinea highlanders.
Analysis of the Denisovan DNA segments was used to determine two things. The length of the DNA provided a measure of how long ago the interbreeding took place, as the Denisovan DNA segments would get shorter over time thanks to recombination. The sequence itself could be compared to the genome of a Denisovan bone in Siberia, which tells us a bit about how diverse the Denisovan population was.
East Asian populations and the Polynesians appear to have had two different periods of interbreeding with Denisovans, both of which were reasonably closely related to the Siberian population.
The people of Papua New Guinea also showed signs of two periods of interbreeding. But, rather critically, they weren’t the same ones seen in East Asians. The first involved interbreeding around 45,000 years ago with a population that had separated from the Siberian Denisovans by roughly 200,000 years—a genetic contribution shared with the East Asians and Polynesians. But the second interbreeding event took place about 25,000 years ago—after the point where the population was out in the Pacific.
And that’s a bit strange. In terms of fossil evidence, we know that Homo erectus was in the area before modern humans arrived, but its DNA would be substantially different from that of Denisovans. There are two other species—the Hobbits of Flores and an equally odd hominin from the island of Luzon. While these look very different from modern humans (having some traits shared with the earlier Australopiths), we can’t rule out that they are closely related to the Denisovans, which would explain the origin of this DNA.
The researchers checked, and the only signs of distantly related DNA can be accounted for by Neanderthals and Denisovans. So if these island species aren’t Denisovans, then it appears we didn’t interbreed with them in a way that left its mark on modern genomes.
What this tells us
Modern humans reached places that required travel across the open ocean very early during their expansion out of Africa. That would seem to suggest that ocean-going voyages were well within our abilities. But these data indicate that most populations remained relatively isolated from each other once they were established. That suggests that, even though the technology was available to manage this travel, it wasn’t widely used—certainly, there’s no indication of longstanding trade until the Polynesians arrive.
Once the Polynesians did arrive, however, there are indications that they interacted at least twice with the inhabitants of the islands near New Guinea. And Vanuatu, at the border between Near Oceania and Polynesia, seems to have an exceedingly complicated history.
To an extent, it seems that, outside of Vanuatu, these people groups interacted with each other about as often as their ancestors interacted with the Denisovans. The genomic data provides evidence of several distinct periods of interbreeding, including one that for now appears specific to a group that is native to the Philippines. This indicates that some of the interbreeding likely went on after modern humans had migrated out into the Pacific islands.
Since we don’t know of any Denisovan remains in the region, it suggests two possibilities. One is that the Denisovans were in the area undetected—not a huge surprise, given how long their presence in Asia went undetected. But the more intriguing prospect is that one of the species we’re aware of from skeletal remains—Homo luzonensis or Homo floresiensis—represents a branch of the Denisovan lineage. So far, all attempts at extracting DNA from these skeletons have failed, so it’s not clear if or how we’d be able to figure this out.