When seafaring modern humans ventured onto the island of Java some 40,000 years ago, they found a rainforest-covered land teeming with life—but they weren’t the first humans to call the island home. Their distant ancestor, Homo erectus,had traveled to Java when it was connected to the mainland via land bridges and lived there for approximately 1.5 million years. These people made their last stand on the island about 100,000 years ago, long after they had gone extinct elsewhere in the world, according a new study assigning reliable dates to previously found H. erectus fossils. The finding suggests a trace of H. erectus DNA could live on in modern Southeast Asian populations, thanks to complex intermingling among the diverse humans who have lived in the region.

The newly dated fossils also bookend the existence of a remarkably long-lived human species, says Patrick Roberts, an archaeologist at the Max Planck Institute for the Science of Human History in Jena, Germany, who wasn’t involved with the study. “With this date, the duration of Homo erectus occupation in Southeast Asia is nearly three times as long as our [own] species has been on the planet,” he says. “There is no doubt it was successful.”

H. erectus arose in Africa about 1.9 million years ago. These toolmakers with relatively large brains migrated out of Africa and across Asia, crossing into Java by land bridges about 1.6 million years ago, when savannalike open woodland covered much of the land. Later, sea levels rose, isolating these ancient Javans on an island. Meanwhile, in Africa and mainland Asia, H. erectus disappeared by about 500,000 years ago. 

In the 1930s, a team of Dutch explorers excavated a site by Java’s Solo River, near the village of Ngandong. They unearthed a rare trove of fossils: tens of thousands of animal bones—and 12 partial skulls and two leg bones identified as H. erectus. But the Dutch team couldn’t date the bones with any certainty. Later scientists also struggled, despite more sophisticated dating methods, because these require material from the same sediment layers as the fossils—and nobody knew exactly where the original excavation took place.

“[The fossils] had been an enigma,” says the new study’s lead author, paleoanthropologist Russell Ciochon of the University of Iowa in Iowa City. “Many people had tried to date them, but there was no way to accurately do so.”

O. Frank Huffman, an archaeologist at the University of Texas in Austin and a study co-author, spent 5 years poring over the Dutch explorers’ photos and notes; he even met with their grandchildren. He and colleagues deduced that the 1930s excavation was located near what is now a sugarcane field abutting a dirt road. In 2008 and 2010, Ciochon’s team re-excavated the site, turning up 867 new fossils belonging to deer, wild cattle, and an extinct, elephantlike animal called a stegodon. Based on photographs and documents from the original excavation, they established that some of the newly found animal fossils came from the same rich bone bed as the H. erectus fossils. The researchers applied five types of radiometric dating, including a new method that provides both minimum and maximum dates, to those animal fossils and the sediments around them. The team concluded that the bones were buried between 117,000 and 108,000 years ago, the researchers report today in Nature.

It’s doubtful H. erectus lived on much longer, Ciochon says. A warmer, wetter climate turned Java’s open woodlands into dense rainforests about 100,000 years ago, and Ciochon suggests H. erectus would have struggled to survive in such a transformed landscape. When modern humans arrived on Java, apparently about 40,000 years ago, H. erectus was probably long extinct, he adds.

Aida Gómez-Robles, an anthropologist at University College London who wasn’t involved with the study, says the authors did great detective work in finding the original excavation sites, and that they have laid out a likely scenario. “We can never be certain that we have found the first or the last representative of any species,” she says, “[but] a last appearance date of approximately 100,000 years ago for H. erectus looks reasonable.”

H. erectus left an impressive legacy. Many researchers think it splintered into at least two additional species as it traveled throughout Southeast Asia—H. floresiensis, found on the Indonesian island of Flores, and H. luzonensis, found on the island of Luzon in the Philippines—and may have interbred at some point with the Denisovans, extinct close cousins to Neanderthals. In turn, Denisovans may have mated with modern humans in Indonesia and New Guinea, perhaps as recently as 30,000 years ago. Those pairings, the authors argue, could have introduced a smidgen of H. erectus DNA into the genomes of some modern Southeast Asians, whose DNA contains a trace—about 1%—of genetic material that doesn’t appear to come from modern humans, Neanderthals, or Denisovans.

“[The new study’s] date certainly adds support to this scenario,” by suggesting H. erectus was still around in Java when Denisovans may also have been moving through the region, Roberts says, but, he adds, there’s far too little evidence to confirm it. “Either way, Southeast Asia is clearly now one of the most exciting places to be working in human origins.”