Ancient proteins name Denisovan bones in a Chinese cave
Two Nature papers on 9 September identify Denisovan remains from Bianfu Cave in Yunnan, mostly through ancient protein analysis when DNA does not survive. Among more than 60,000 bone fragments, just seven were human, and they are between 167,000 and 134,000 years old.
Denisovans are the most elusive character in the human story. Known first from a finger bone and some teeth in a Siberian cave, and inferred from traces left in the genomes of living people across Asia and Oceania, they have always been harder to actually find than to argue about. Two papers published in Nature on 9 September, working from a cave in Yunnan in southwestern China, add a firm set of physical remains and, just as importantly, show how such remains can be identified at all.
The site is Bianfu Cave, where excavators recovered more than 60,000 bone fragments. Sorting through them for human remains is the kind of task that defeats ordinary archaeology, because most of the pieces are animal, and the few that are hominin are often too broken to recognize by shape. Out of that mountain of fragments, the team led by Qiaomei Fu and colleagues identified only seven as human: part of an arm bone, two skull fragments and four teeth. The bones date to between about 167,000 and 134,000 years ago.
The identification rests on ancient proteins rather than DNA. In the warm, wet caves of southern China, genetic material simply does not survive the way it does in the cold of Siberia, so aDNA sequencing is rarely an option. The researchers instead looked at the amino-acid sequences of collagen extracted from the bones, and found a specific protein variant that distinguishes Denisovans from Neanderthals and from Homo sapiens. That molecular fingerprint is what allows a shard of apparently unremarkable bone to be assigned to a group whose skeletons are barely known. It is a work-around for the absence of DNA, and it is likely to be the method that unlocks the tropical fossil record, which is exactly where so much of human evolution happened.
The second paper, from Fahu Chen and colleagues, widens the picture from bones to behaviour. Working with the same material plus animal and pollen remains and more than 1,300 stone artefacts spanning roughly 190,000 to 70,000 years, it reconstructs the world these people lived in: a forested, conifer-dominated environment, with specialized hunting of medium-to-large prey and simple bone tools that look surprisingly close to the Neanderthal tool kit. The arm bone itself shows a mix of features, some archaic, some closer to modern humans, which is precisely the sort of ambiguous morphology that protein analysis is needed to resolve.
Why this matters is the map it starts to draw. Genetic clues in living populations long suggested that Denisovans ranged through southeast Asia, but fossils confirming it were almost nonexistent. Bianfu Cave supplies that confirmation in the east, at the southeastern edge of the Tibetan Plateau. Combined with the protein method, the result suggests the Chinese fossil record, long resistant to genetic answers, is now open to question.
The cave is one of the few places anywhere that has yielded Denisovan remains in a clear archaeological context, with tools and prey in the same layers as the bones. That pairing, of a molecularly named hominin and the debris of its daily life, is what lets a species be reconstructed not just from its bones but from its behaviour.