Scientist found 20 ancient hominins in a cave, and a mystery: Why are they all female?

Around 300,000 years ago, a cave in South Africa became the final resting place of at least 20 human cousins from the ancient species Homo naledi. Today, researchers have determined that all of those individuals appear to be missing a Y chromosome—the telltale genetic marker for males—and are likely female.  

Using a new method for sampling ancient teeth, the team was able to collect chemical clues about the sex of each skeleton without destroying it. The technique has the potential to “transform paleontology for several decades,” says Lee Berger, who is the National Geographic Society's Explorer in Residence, paleoanthropologist at Wits University in South Africa, and a co-author on the new study. Berger and his colleagues first discovered Homo naledi’s bones in the cave system called Rising Star in 2013. Following a tip from two cavers, the team uncovered around 1,500 fossils in a chamber deep underground, only reachable through a narrow chute at the end of a twisting 100-foot passage. 

At the time, the team assumed that some of the larger skeletons were males, and that the species had relatively equal numbers of males and females—the common pattern in modern mammals. Finding a massive trove of ancient bones from only one sex is almost unheard of in paleoanthropology. But the samples from Rising Star cave that had previously been identified as male now appear to be female, researchers reported on June 24 in the journal Cell.

“These are very exciting results, pointing to the cultural, and perhaps even symbolic, significance of the Rising Star site,” says Katerina Douka, an archaeological scientist at the University of Vienna in Austria who wasn’t involved in the study. Barring alternative biological explanations, she sees the find as “a unique, and perhaps perplexing, prehistoric discovery.” 

This isn’t the first time that Homo naledi has challenged assumptions. The ancient cousin’s strange mix of traits has long confounded paleoanthropologists: a brain that’s one-third the size of our own, shoulders optimal for climbing like those of older human relatives, and long legs capable of strides, similar to modern humans. Berger’s team has argued that at Rising Star, Homo naledi intentionally buried their dead, stoked fires, and carved symbolic crosshatchings into the cave walls with stone tools—achievements often associated with big-brained species like Homo sapiens.  

(Read more about the team’s explorations of the Rising Star cave system.)  

The new finds deepen the mystery around what the ancient hominins—the umbrella term for humans and our ancient relatives—were doing in that cave, and Berger thinks it strengthens the evidence for advanced abilities in Homo naledi. “It should give you almost goosebumps,” says Berger. “You're looking at a non-human species with a brain the size of a chimpanzee that is saying females have a separate place in this world and an important place, in that this place belongs to them in death.”  

How to determine sex in hominin fossils

DNA is the most surefire way to determine the sex of a skeleton, but it degrades over time, especially in warmer climates. While researchers continue to push the limits of recovering ancient DNA, the current record-holder comes from an environmental sample of 2.4-million-year-old ice in Greenland. The oldest DNA from a hominin is around 400,000 years old. Even when DNA survives, holes in the sequence can limit the tests researchers might run on ancient samples.  

When DNA isn’t an option, researchers examine the bones themselves—the shape of the pelvis and skull can vary between males and females, as can the size of teeth. “These morphological features are fairly reliable, but one can rarely be 100 percent confident in them,” says Paul Constantino, a biologist at Saint Michael’s College in Vermont who was not affiliated with the new analysis. But ancient hominin fossils are often so rare that these variations are hard to discern, leading to debates over whether a skeleton represents a different sex or even an entirely different species.  

Because of the limitations of bones and DNA, scientists are increasingly turning to ancient proteins stuck to teeth and bones to answer some of these big evolutionary questions. Sequences of proteins can tell researchers where a species might fit in the human family tree. Molecular biologist and study co-author Enrico Cappellini’s lab at the University of Copenhagen in Denmark has used such sequences to reveal that modern orangutans are related to a 1.9-million-year-old giant ape Gigantopithecus and shed light on where ancient rhinos and human relatives fit in their respective family trees.  

These protein sequences can also illuminate the ratio of males and females in samples of bones. A protein in tooth enamel called amelogenin looks slightly different depending on whether it comes from a gene on the female X chromosome or a gene on the male Y chromosome. Last year, an amelogenin analysis by the same team showed that the females of the two-million-year-old hominin species Paranthropus robustus weren’t necessarily smaller than males, bucking a common primate stereotype.

A remarkably homogenous hominin

Based on that track record, Berger, Capellini, and their colleagues wanted to use ancient proteins to tackle a number of questions around Homo naledi. They hoped sequences might hint at where the species fits into our own lineage, as well as how much the individual sequences differed across skeletons, which can tell them how diverse the species was.   

First, the team sampled just four individuals from the cave, removing a piece of each tooth to test the enamel and see if they could sequence its proteins. But they noticed something odd: All four lacked the Y chromosome variant of amelogenin.  

So the researchers tested more teeth and even tried an alternative method that was less destructive, involving scraping the enamel off instead of cutting out a chunk of tooth. The etched sample is “no greater than what a dentist would do to roughen up your teeth to attach braces,” says Berger.  

Out of 20 individuals, the Y chromosome was definitively missing in 19 of the samples and likely missing in one outlier. Even a tooth belonging to an individual dubbed “Neo,” whom they had previously identified as a male, was missing the Y version of amelogenin.

“We were all shocked,” says Palesa Madupe, a molecular biologist at the Max Planck Institute for Evolutionary Anthropology in Germany and lead author of the work. Only one other ancient site, a cave in Portugal, has a female majority among skeletons, which researchers have linked to the mortuary practices of local Neolithic humans.

While the H. naledi skeletons included 10 adults and 10 juveniles (including an infant and some toddlers), none of the children were male. That finding is particularly curious: Children rely on their mothers so much, one would expect to see some male babies. Even most Homo sapiens societies that separate dead individuals based on their sex keep male infants with their mothers, says Berger.  

(Here’s what we know about Homo naledi’s youth.)  

Unlike other species, such as P. robustus, H. naledi’s bones have a similar shape across individuals. For this reason, George Washington University paleoanthropologist Bernard Wood, who did not work on the new analysis, says he wasn’t surprised that they might all belong to one sex. He points to a hand bone in particular: “They look as if they've been made in some sort of stamping machine,” he says. “There is no variation at all in the morphology, which is very strange.”  

The protein sequences of each individual were also remarkably similar. Wood suspects that this population may have been isolated in South Africa, perhaps descending from just a few individuals, and thus retaining less variation in their genes. 

From biology to culture

The probability of finding all females among 20 random humans are staggeringly low—specifically 0.0000954 percent, according to the team’s calculations. The question now is how this could have happened in H. naledi’s case.  

In very rare instances, the amelogenin gene on the Y chromosome has been known to vanish. A small fraction of males in some modern human populations have lost the Y version of the protein due to mutations. The rates are low—0.18 percent in an Austrian population and 8 percent in a Pakistani population.  

There’s also the possibility that the Y chromosome has degraded more than the X chromosome over time. Female false positives can happen in those cases, notes Carles Lalueza-Fox, a molecular biologist at the University of Barcelona in Spain who was not affiliated with the study. For that reason, he remains unconvinced that all of the skeletons are truly female. Only additional testing could rule out false positives. However, the authors argue that the detection of the male version of amelogenin in both P. robustus and another ancient hominin called Australopithecus afarensis—both older than H. naledi and thus, more likely to experience degradation—makes this unlikely.  

If the Y chromosome hasn’t gone missing and the H. naledi individuals are indeed female, then some researchers wonder if some unknown ancient hominin behavior could be at play.  

Constantino speculates that groups of Homo naledi may have had fewer males, similar to the single-male social structure seen in mountain gorillas or hamadryas baboons today. Or the species may have divided labor based on sex, and perhaps the males primarily hunted or patrolled the territory. Scientists don’t yet have enough clues to fully comprehend how Homo naledi might have used the depths of the Rising Star cave systems. The one certainty is that its dark chambers became their home in death.  

Wood, who doesn’t believe that Rising Star contains intentional burials, where individuals might have dug graves or arranged bodies, wonders if H. naledi may have taken female remains to the cave, perhaps dropping them in after death. (Others have suggested that the bones may have been transported by water or predators to the cave, but each theory has limitations.) Aside from some form of cultural behavior, “there is really no other explanation for why you would have just one sex in the cave system,” he says. At least, not a natural one.  

Berger thinks the exclusion of male skeletons is evidence of a practice that would have required a level of complex thinking only seen in humans. “We are sitting with evidence of separation at death by sex, and thus likely gender,” he says. “That is clearly evidence of culture in a non-human species.”  

But Beatrice Demarchi, a biomolecular archaeologist at the University of Turin in Italy who did not work on the analysis, is still cautious about drawing cultural conclusions. She notes that beyond evolutionary patterns, it's incredibly difficult to interpret biological data from the past, especially with limited or unique context, as in the case with Homo naledi. “In this sense, Homo naledi is a classic case in which the boundaries between nature and culture are blurred,” she says.  

Nothing but the tooth

More broadly, the study shows just how valuable proteins can be in unveiling information that remains ambiguous in bones or absent in degraded DNA. Constantino called the less-destructive sampling method “a breakthrough” in paleoanthropology. Much of the hominin fossil record consists of teeth, in sizable numbers at some sites, he says, and knowing how many individuals were male and female could point to how these ancient groups were structured and even mating trends.  

Multiple researchers expressed hope that the technique could help sample ancient hominin specimens from East Africa, like Paranthropus boisei, which has also faced questions around its legitimacy as a species. It could also validate preconceptions of sex for famous fossils like “Lucy” or settle debates over the sex of enigmatic fossils like Mr./Mrs. Ples, the most complete Australopithecus africanus skull ever found. “It is wonderful to see how ancient proteins open up the window of prehistory a little wider, bringing us closer to our hominin cousins,” Douka says.  

As for Homo naledi, the mysterious human genus still has no clear ancestors or descendants. The case of the missing Y chromosomes still plagues Madupe, who wonders if there’s some explanation for the absence of the male protein signal lurking in the unknown realms of molecular biology. If so, “that will be a great discovery,” she says. Only a definitively male skeleton could answer that question. Berger estimates that there could be dozens more skeletons in the cave system ripe for discovery. “Clearly, there had to be males, otherwise the species would not exist,” Constantino says. “So where were they?” 

For now, the new revelations have a strange serendipity to them. The tight squeezes required to enter the present-day Rising Star cave system made it harder for male researchers to enter, so the initial group of paleoanthropologists to excavate the bones were all women. That these female fossils were discovered by women, Berger notes, “I think that's one of those great, beautiful ironies of science and exploration.”