Hobbit Cave
© Liang Bua TeamLiang Bua Cave on the island of Flores, where specimens of the 'Hobbit' species were discovered.
Anthropologists know of at least two ancient species of tiny humans that lived on the islands of southeast Asia over 50,000 years ago. The origin of these extinct humans is unknown, but new research suggests they're more closely related to Denisovans and Neanderthals — and, by consequence, modern humans — than previously thought.

New research published in Nature Ecology & Evolution has found no evidence of interbreeding between modern humans (Homo sapiens) and two extinct species of short-statured humans, Homo floresiensis (commonly known as the Flores Island "hobbits") and Homo luzonensis (found in the Philippines). Fossil evidence of these two species, described in 2004 and 2019 respectively, suggests these island-dwelling humans stood no taller than around 3 feet and 7 inches (109 centimeters), a possible consequence of insular dwarfism — an evolutionary process in which the body size of a species shrinks over time as a consequence of limited access to resources.

At the same time, the new paper, led by João Teixeira from the University of Adelaide, provides further confirmation of interbreeding between the Denisovans and modern humans, specifically modern humans living in Island Southeast Asia, an area that encompasses tropical islands between east Asia, Australia, and New Guinea. Denisovans — a sister group of Neanderthals — reached the area some 50,000 to 60,000 years ago, but archaeologists have yet to uncover a shred of fossil evidence related to these so-called "southern Denisovans." That's obviously weird, given the overwhelming genetic evidence that they lived in this part of the world, but it means there are important archaeological discoveries still waiting to be found. At least in theory.

So, the new paper, co-authored by anthropologist Chris Stringer from the Natural History Museum in London, suggests modern humans interbred with Denisovans but not H. floresiensis or H. luzonensis. That's an important result, because it could help to explain the presence of the diminutive humans, who died out around 50,000 years ago, in this part of the world. Excitingly, it could mean that these "super-archaics," in the parlance of the researchers, "are not super-archaic after all, and are more closely related to [modern] humans than previously thought," explained Teixeira, a population geneticist, in an email.

In other words, H. floresiensis or H. luzonensis might actually be the elusive southern Denisovans.

Given that present-day human populations in Island Southeast Asia have retained a significant amount of Denisovan DNA, the authors wondered if H. floresiensis and H. luzonensis also interbred with modern humans. It was also possible, though unlikely, that another ancient human called H. erectus, which lived in Java until around 117,000 to 108,000 years ago, might've also contributed to modern human ancestry. Indeed, one possible scenario is that the super-archaics were descended from H. erectus.

To that end, the scientists studied the DNA of 400 modern humans, of which more than half were of Island Southeast Asia ancestry. The team searched for key genetic signatures indicative of interbreeding events related to "deeply divergent hominin species," said Teixeira. Island Southeast Asia is the "most likely geographic region where such events could have occurred due to the aforementioned presence of H. floresiensis and H. luzonensis, and perhaps H. erectus as well," he added.

It's important to note that scientists do not have genomes for the two short-statured species, nor H. erectus for that matter.

"There are no 'first-hand' genomes of the kind we have from Neanderthals and Denisovans, but there are 'second-hand' bits of DNA in the Denisovan genomes that seem to come from them having interbred with a super-archaic population," explained Stringer in an email. "These can be recognised by their greater-than-average divergence within the genome and also, if there has been recent interbreeding, the strands of DNA will have been shuffled up less, and hence found in larger and more 'pristine' chunks."

To be clear, the scientists are not looking for specific species-related genomes, but evidence of interbreeding, which leaves a pronounced genetic signature across the entire genome.

Results showed that modern humans did not interbreed with the two small human species, but the team did confirm Denisovan ancestry among individuals from Island Southeast Asia. As Stringer put it, "the DNA of local populations shows signs of ancestry from the Denisovans, who are currently only known from fossils in Asia, but no genetic evidence deriving from the ancient humans whose bones have actually been found in the area."

Indeed, fossil evidence of Denisovans is non-existent in Island Southeast Asia, and the evidence that does exist elsewhere is sparse. Aside from genetics, the presence of this human species is known from a finger bone, several teeth, and skull fragments found in Siberia, as well as a 160,000-year-old jawbone found in a cave on the Tibetan Plateau.

The new research confirms that the two super-archaic species "did not contribute ancestry to modern human populations," or if they did, they're "not so divergent as currently assumed based on morphological comparisons," said Teixeira. These short-statured humans may seem very different from modern humans, and thus very divergent, but that could be an illusion, as their DNA may actually be very similar to ours and especially to that of Denisovans, according to this line of thinking.

For Teixeira, the absence of this interbreeding combined with the widespread Denisovan ancestry means the two super-archaic species might represent the missing Denisovans in Island Southeast Asia, or some kind of offshoot.

"The ISEA fossil hominins are thought to represent a much older split (approximately 2 million years ago). But those estimates rely on morphological comparisons to, and the assumption they descend from, H. erectus," he explained. "Our results show that such super-archaic species did not interbreed with modern humans in ISEA — but what if we're wrong? What if hominin occupation in ISEA was not continuous? What if Denisovan ancestry in ISEA comes from these groups?"

To which he added: "No one knows for sure what a Denisovan is supposed to look like nor how much morphological variation existed within different Denisovan populations," he explained. "If that is the case," the revelation that the super-archaics are actually the southern Denisovans "could have serious implications for paleoanthropology."

Stringer, on the other hand, isn't so sure, as his interpretation of the evidence suggests a different lineage for the tiny human species.

"The known fossils of H. erectus, H. floresiensis, and H. luzonensis might seem to be in the right place and time to represent the mysterious 'southern Denisovans,' but their ancestors were likely to have been in place in Island Southeast Asia long before the Denisovan lineage had evolved," and possibly as long as 700,000 years ago, Stringer explained.

"George, co-authors do not always agree on everything," Teixeira told me when I queried him about this apparent inconsistency.

Regardless, theco-authors believe that interbreeding between southern Denisovans and modern humans happened in Island Southeast Asia.

"The presence of the largest amounts of Denisovan-like DNA in regions like Papua New Guinea and Australia suggests that the interbreeding occurred in ISEA or, much less likely in my opinion, a place like Papua New Guinea," explained Stringer in his email. "My guess is that Sumatra, Borneo, and Sulawesi were the homelands of the missing 'southern Denisovans' and will most likely yield up their fossils."

Stringer said these results depend on the samples analyzed and that more samples are likely to provide a fuller picture.

The new paper, while illuminating, raises some very important questions. First and foremost, where are the Denisovan fossils in Island Southeast Asia? And, as Teixeira asks, "have we already found them but assumed these fossils represented much more distant relatives?" In other words, maybe the "hobbits" were the southern Denisovans all along.