Archaeology can be complicated. Sometimes you get beautifully stratified layers of remains making it easy to determine exactly what happened and when. And then sometimes you get a situation like that at Roopkund lake.
By lake standards Roopkund is small, only around 40m across and it is buried high in the Himalayan mountains, more than 5000m above sea level. Even today it’s best described as inaccessible, which makes the mystery surrounding this particular body of water all the more bizarre. The lake is colloquially known as “Skeleton Lake” and for good reason. The shoreline is littered with human remains.
The existence of the skeletons has obviously been known to the locals in the area for hundreds of years but they came to wider international attention only during the 1940s. At the time British authorities in the region were concerned that these might be Japanese soldiers from a previously unsuspected invasion force but it quickly became obvious that the remains were very, very old. And it wasn’t just skeletal remains that were found. The icy conditions also preserved wooden artefacts, iron spearheads, even some of the unfortunate victim’s leather shoes. However, until recently archaeologists had done relatively little investigation of the region, or the remains, and this meant there was very little information for researchers to go on. The most prevalent theory has been that the remains represented a single, large group of people who died together in a hailstorm, although National Geographic suggested they could also have been victims of an epidemic. In an effort to finally get to the bottom of exactly who these people were, and what they were doing in the mountains, archaeologists decided to use a thoroughly modern technique; DNA testing.
The principle was sound. By analysing the DNA from the bones discovered at Roopkund Lake the researchers hoped to find the living population that best matched the remains. This would at least give archaeologists an idea of where the people who died at the lake came from originally. It would also give a good idea of how the victims related to each other. Theoretically this should have given us a wealth of information and helped to solve the mystery. Except that nothing about Roopkund would ever be that simple.
Instead of solving the mystery the DNA testing deepened it. Nearly 300 individuals have been identified in the remains at Roopkund lake so to get a good cross-section the archaeologists analysed the full genomes of 38 individuals and mitochrondial DNA from 71. What they found was fascinating. The remains belonged to a mixture of men and women, and they didn’t show a particularly high degree of relatedness suggesting these aren’t the remains of a group of families as had been suggested previously. More interestingly still the remains fell into three distinct genetic groups. One group, which the researchers referred to as Roopkund A seemed to have ancestry stemming from south east Asia. So far so expected. The second group though, known as Roopkund B, were totally different. Their ancestry wasn’t south east Asian. In fact it wasn’t Asian at all. Instead the population they most resembled was from western Europe specifically the areas around mainland Greece and the island of Crete. The final group, Roopkund C, comprised of a single individual who appeared to be of Chinese descent.
Having discovered that the individuals around Roopkund lake appeared to belong to several distinct groups the researchers then tried to determine if those groups all died there at the same time. Radiocarbon dating was done on the remains and the archaeologists discovered that Roopkund A was by far the oldest group with the remains dating to between the 7th-10th centuries CE (common era). The remains in Roopkund B were found to be far more modern, dated as having died some time between the 17th–20th centuries CE. The single individual from Roopkund C appears to also date to this later period. Just to add to the confusion there was a lot of variability in the predicted age of the remains from Roopkund A suggesting they probably didn’t all die at the same time. This means we have a situation where, over a period of potentially hundreds of years during the 7th-10th centuries, dozens of different individuals all made their way to their independently to an out-of-the way lake high in the Himalayan mountains and died. Then nearly 1000 years later a totally unrelated group of individuals whose ancestors came from Greece, possibly specifically the island of Crete, all travelled on mass to the same lake and there died by unknown means.
So what happened? Well the researchers did offer a plausible explanation for the individuals in Roopkund A. Today the lake is not on a major trade route, but it is on the route of an important pilgrimage. The Nanda Devi Raj Jat pilgrimage is observed by Hindus every 12 years and involves participants gathering to celebrate along the route of the pilgrimage. The researchers point out that documentary evidence of the pilgrimage doesn’t appear until the late 19th century but there are some suggestive inscriptions from nearby temples that could put the origin of the pilgrimage back to the 8th-10th centuries, well within the radiocarbon dates for Roopkund A.
Roopkund B are far more mysterious. Isotope analysis of the bones suggest these individuals were eating a mainly terrestrial diet, rich in wheat, barely and rice but little, if any, seafood. This suggests they lived somewhere inland, despite their nearest relatives being based on the island of Crete. How and why they came to be together at Roopkund Lake, however, is still a total mystery.
Harney, E, et al. 2019. Ancient DNA from the skeletons of Roopkund Lake reveals Mediterranean migrants in India. Nature Communications. 10(3670).