For years, archaeologist Huw Groucutt and his team had driven one particular stretch of desert on their way to dig sites in Saudi Arabia. As they drove they caught glimpses of what looked like bones, emerging from the slowly eroding sand. Finally, in 2014, the team decided to explore the array of bones at Al Wusta. Within two years, amidst more than 800 fossilized animal bones and nearly 400 stone artifacts, they discovered something remarkable: the middle digit of a finger bone, from what appeared to be a modern human.

Anatomically modern, that is. The fossilized finger dated to at least 85,000 years ago.

“It’s strange, isn’t it? Almost all bones will not be preserved, and there’s nothing special about the finger bone in terms of how hard it is. It just got lucky,” Groucutt says. After all, fossilization on land is very rare ; the water and wet sediments of the paleolake must have offered just the right protection from oxygen to preserve the bone.

If the bone was “lucky,” however, the researchers were doubly so. Forget the needle-in-a-haystack cliché: Finding a human bone in the Nefud Desert—a windswept oval patch of sand dunes the size of Kentucky—is perhaps the world’s most impressive example of an unlikely find. Their analysis of the finger bone and the prehistoric environment it came from appears today in the journal Nature Ecology & Evolution . If the finger is indeed human, the bone may be one of the oldest examples of any Homo sapiens remains found outside of Africa.

The discovery is “a dream come true, because it supports arguments that our teams have been making for more than 10 years,” said archaeologist Michael Petraglia , another co-author of the study, in a press conference. “This find together with other finds in the last few years suggests that modern humans, Homo sapiens , are moving out of Africa multiple times during many windows of opportunity in the last 100,000 years or so.”

The question of how humans left Africa has been debated ever since it became widely accepted that Homo sapiens did indeed evolve from ancestral species in Africa, rather than Asia. (That latter hypothesis was proposed by scientists like Ernst Haeckel, and preferred by many anthropologists until as recently as 60 years ago; some modern researchers still argue for multiple evolutionary jumping off points, based on fossil finds in China ). In the past decade, some geneticists have argued for a single dispersal event from Africa around 60,000 years ago, based on the decreasing genetic diversity in populations that are farther from Africa.

But others believe that the order of events was a bit more complicated.

“Our previous work found that multiple dispersals, with the first one being older than the 50,000 to 70,000 [years-ago] migration, are most compatible with the pattern of both cranial and genetic variation observed among people today,” said Katerina Harvati , director of paleoanthropology at the University of Tubingen, Germany, by email.

Harvati, who wasn’t involved in the research, said she would be cautious in definitively assigning the finger fossil a Homo sapiens identity due to the fact that its shape overlaps with other hominin species. But the fossil does fit the larger pattern of discoveries made in the region. Skulls belonging to Homo sapiens found in Qafzeh and Skhul in Israel have been dated back to 100,000 years and 120,000 years respectively, and the discovery of a human jawbone from Misliya Cave was dated to around 177,000 years earlier in 2018.

All of these fossils suggest humans left Africa much earlier than 60,000 years ago. But the new finger bone suggests some populations continued moving, beyond the Levant and into the Arabian Peninsula.

Groucutt and the rest of the team used a number of dating methods to confirm the likely age of the Al Wusta finger. For the finger itself and the tooth of an ancient hippopotamus found nearby, they applied U-series dating. Like radiocarbon dating, the method works by looking at radioactive decay in the preserved materials. The age of the sediments around the bones was calculated using optically stimulated luminescence—a technique that reveals the last time rocks and sand were exposed to sunlight.

Geochronologist Norbert Mercier , who wasn’t involved in the study, confirmed over email that the date sounded likely. “The combined results obtained from the different mediums, through various types of dating, strongly support the estimated lifespan of the Homo sapiens fossil,” he said in French.

But the question remains: How did humans manage to survive in a desert environment nearly 100,000 years ago?

One possibility is that, at the time, it wasn’t a desert. While the Nefud is all sand and rock today, at the time of the Al Wusta fossil, the region was a savannah, covered in lakes and rivers thanks to summer monsoons. The multitude of animal bones found in the same location, from wild cattle to antelopes, suggest that game was plentiful. The lake itself lasted year round and offered a source of fresh water, though it may have come with risks as well as rewards: many of the faunal bones bore marks of carnivores’ teeth.

“Excursions of Homo sapiens into the Near East, Arabia, and as far as eastern Asia would have been assisted by favorable rainfall, which created well-watered zones that had previously (and subsequently) been quite arid,” says Smithsonian Institution paleoanthropologist Rick Potts, who has extensively studied paleoclimate in Africa, by email. “The finds in Arabia extend the known geographic area of this early dispersal of Homo sapiens , and it indicates that a distinctly sere area of Arabia today was sufficiently ‘green’ and wet to support human populations.”

But researchers behind the recent Misliya discovery argue that the bone’s location on the surface rather than in the lake deposits below mean it might not correspond to that climactic period. Aarchaeologist Mina Weinstein-Evron and paleoanthropologist Israel Hershkovitz , although they generally admire the study, wondered whether the bone belonged to an earlier humid fluctuation period. “The authors rightly highlight the challenges of precisely constructing regional paleoclimates and relating them to human demographic and behavioral change, but their study does not seem to contribute to meeting this challenge,” said the researchers, who weren’t affiliated with this study, by email.

The finger gestures to another question as well: What happened to the population that made it all the way to Arabia? Were they forced to move forward, or retreat when the environment became inhospitable once more within centuries after they arrived?

“It will be interesting to learn if the Al Wusta population came directly from Africa (by crossing the Bab-el-Mandeb strait ) or if it’s related to the fossils from Qafzeh, which would suggest they turned towards the south,” said Mercier.

Although the team will analyze the finger bone for any remnants of DNA, they’re doubtful anything will emerge, given the harsh environment it came from. ( Most ancient DNA comes from cold climates and fossils younger than 50,000 years.) The next steps will be conducting more digs in the Arabian Peninsula and in Africa to piece together the broader picture.

Groucutt does hope that this discovery might spur more research in regions that are usually given less attention when it comes to human evolution. “There’s still a big focus on a few small areas like Europe and South Africa,” Groucutt says. “These are very important areas, but the world’s a big place. We have to be careful not to assume that everything happened where we happened to find it.”