UNIVERSITY PARK, Pa. — Over centuries, the shores of a small Himalayan lake became the final resting place for hundreds of individuals, so much so that today the lake is locally known as Skeleton Lake. Skeletal remains of these ancient people are scattered around the lake, partly due to rockslides, as well as visitors handling and moving the bones around.

Until recently, almost everything about these ancient people was unknown — where they came from, why they were there, how old they were. The prevailing modern theory was that the remains belong to one group of people, perhaps traveling together from the same geographical area, who died about 200-300 years ago. It was proposed that all of these people were killed by a single catastrophic event, perhaps a rockslide, or maybe a deadly epidemic.

Roopkund Lake — the lake’s officially recognized name — is located near a shrine to the mountain goddess Nanda Devi. Local folklore tells the story of a pilgrimage to the shrine that made Nanda Devi unhappy, and so the pilgrims met their demise due to her wrath.

Recently, with the help of Penn State’s Radiocarbon Laboratory, researchers discovered that the mystery of Roopkund Lake runs deeper than originally thought.

Unexpected results

About two to three times a week Brendan Culleton, a laboratory scientist for Penn State's Institutes of Energy and the Environment, runs samples through the accelerator mass spectrometer (AMS) housed at the University Park campus. The AMS is used for radiocarbon dating, which calculates the age of an object by measuring how much carbon-14 is in it. Only material that is carbon-bearing, such as bones, teeth and wood, can be dated using an AMS.

As he does most days, Culleton ran several samples without knowing where they were geographically from or what was expected to be found.

“On this particular run there were skeletal samples from all over the world — Mongolia, Sudan, Spain, Turkmenistan and elsewhere,” Culleton said. “Most of them come with the site name. The descriptions I had for the Roopkund samples were simply codes like ‘R66. Bone powder. India.’ I really had no sense at all of the locale or the archaeological context.”

The AMS calculated that the seven or so samples tested from Roopkund didn’t all date to the same time period. In fact, the samples were about 1,000 years apart.

Without any context, the results didn’t mean anything significant to Culleton. He sent the results back to David Reich, a geneticist at Harvard University who had ordered the radiocarbon testing. Culleton assumed his role in this research was finished.

It turns out that Reich was expecting these samples to all date back to the same time period. For decades, the running theory was that the remains were the result of a single catastrophic event that had occurred in just the last 200 to 300 years.

Curiouser and curiouser

Meanwhile, the DNA of the same skeletons was being tested through collaborative work at the Centre for Cellular and Molecular Biology in India and at Harvard Medical School. These labs were learning that there were three distinct genetic groups among the 38 samples being tested — 23 from present-day India, 14 from the eastern Mediterranean area, and one from southeastern Asia.

When Reich received the radiocarbon results, he asked Culleton to double check the AMS results.

“David told me that the individuals were all supposed to have died in a single event, and so the results really pointed to some kind of mix-up with the samples or maybe contamination at some point,” Culleton said. “We looked through all the lab notes and the AMS run and saw nothing unusual. After that, David was very excited because the genomic data correlated these two time periods. He sent me a table with the carbon-14 dates and population labels ‘Roopkund_A’ and ‘Roopkund_B.’ Still none the wiser, I looked up Roopkund.”

Culleton was astounded when he understood the context and why researchers thought this might be a single event.

So Reich sent more samples to Penn State for radiocarbon dating to complement the genomic data.

By combining the radiocarbon data from Penn State with the ancient DNA results, as well as with stable isotope measures — which reveals dietary profiles — the researchers found that there were in fact three distinct populations at Roopkund Lake who died at two very different times.

The results suggest a group with Indian-related ancestry died at the site between 1,100 and 1,400 years ago. Later, people from the eastern Mediterranean and southeastern Asia perished there between 100 and 400 years ago.

Collaborative discovery

“Projects like this, especially with ancient DNA, are inherently multidisciplinary,” Culleton said. “The challenge is always for everyone to be able to communicate openly and effectively within and between our areas of expertise. We have to be able to educate each other across domains and be willing to be educated as well. It can be threatening for a scientist to have their hypotheses and theories challenged by an ‘outsider.’ But when a team brings together multiple lines of evidence to bear on a problem, everyone reaps the benefits. In the case of Roopkund Lake, the whole is far greater than the sum of the parts, and the great collaboration is very gratifying.”

Due to modern technology and collaborative work across several disciplines, researchers now know more than they expected to about the individuals who perished at Roopkund Lake.

This research, titled "Ancient DNA from the skeletons of Roopkund Lake reveals Mediterranean migrants in India," recently appeared in Nature Communications.

The Radiocarbon Lab is one of several shared multi-user instrumentation facilities that make up the Energy and Environmental Sustainability Laboratories (EESL), which provide cutting-edge research equipment in the areas of energy and the environment.

EESL is a part of the Institutes of Energy and the Environment (IEE), which works to build teams of researchers from different disciplines to see how new partnerships and new ways of thinking can solve some of the world’s most difficult energy and environmental challenges.