How we can tell mummy ate fish

August 11, 2000

David Thomas meets a woman whose far-flung hair samples can reveal the diets of ancient peoples.

You are happily minding your own business, making the dinner or doing the washing, when a woman speaking an unintelligible language comes up to you and, through an interpreter, asks to take a lock of your hair.

Unnerving, and certainly unusual, but this has happened to many people, from fishermen in Vietnam and villagers in The Gambia to reindeer herders in Lappland and gauchos in Argentina. The scissors-wielding scientist is Tamsin O'Connell from the Research Laboratory for Archaeology in Oxford, who travels by helicopter and canoe and wades through quagmires in her quest for hair.

Dr O'Connell uses high-powered mass spectrometry to relate the composition of carbon and nitrogen stable isotopes in body tissues with those in the food eaten. By knowing the ratio of these isotopes in hair, nail clippings, bone and urine, we can fairly confidently say what was a person's predominant diet. These techniques are being honed in a multidisciplinary approach to archaeology to produce a powerful tool to reconstruct the diets of past populations.

"Looking at hunting tools, cooking methods and food remains are indirect, and they can provide only qualitative evidence," she said. "If a population subsisted on 50 per cent bananas and 50 per cent shellfish, after 3,000 years, the banana skins would have rotted away, whereas the oyster shells would remain. From that evidence one might think that they only ate shellfish."

O'Connell's investigative chemistry has been helping to solve archaeological riddles such as that of the 2,500-year-old Pazyryk mummies. The superbly preserved bodies were found buried with equally well-preserved horses in a land-locked region of Siberia.

The obvious conclusion was that these people herded horses and probably ate lots of horse meat. However, the results of stable-isotope analysis of hair from two women and two men were more akin to a mixed diet of fish and meat.

The conundrum was substantiated by a closer look at the bodies. They were covered with remarkable tattoos, and one of the men sported a design on his shin that is clearly a leaping salmon. The chemists were responsible for a complete rethink of the lifestyles of this group of horse herders.

The diet of archaeological peoples implies much about their ways of life. If the diet is mostly fish, the society is likely to have been a fishing one. If, however, it is based on grain, the people are likely to have been farmers.

This may seem obvious, especially if the fish-eaters were found on the coast and the grain-eaters in a fertile river valley.

But what if you find a population that is based on the coast but eats no fish? This is what happened at the start of the neolithic period in Britain, when people, including those living by the sea, abandoned fishing and turned to land-based foods, possibly grain farming, according to Mike Richards, another of the Oxford team.

To get a true picture of the past using these techniques, the relationships between isotope chemistry and diet must be firmly established. Hence the work of Dr O'Connell, which is funded by the Wellcome Trust's bioarcheaology programme.

Her schedule for the three-year project sounds like a travel writer's dream: "I am collecting hair samples from populations with archaeologically interesting diets, including diets high in animal protein, high in marine foods, low in meat but high in cereals. Almost by definition, the populations with these diets tend to be the indigenous 'non-westernised' people of a region.

"I have sampled hair from the Sapmie (Lapps) in Sweden (who eat a diet with lots of meat), cattle farmers in Argentina (lots of meat), Melanesian islanders (some fish), Gambians (mainly cereals and rice) and Vietnamese fishermen (lots of fish). Other 'target groups' are the Aleuts in Alaska (seals and fish), the Hadza in Tanzania (lots of zebra) and Jains in India (vegetarians)."

Bones are the most likely remains to be found at archaeological sites. Up to now, much of the work has been done using the stable isotope carbon and nitrogen composition of bone, and comparing them with isotopic values from both modern and archaeological animal and plant food.

A better way to interpret the data from archaeological humans is to compare it to isotopic data from modern humans, but bones from living humans are not easy to sample. Fortunately, the hair and nail proxies can provide equally useful information.

But there can be other complications, O' Connell says: "In some societies, taking samples such as hair and nails from an individual is taboo, and witchcraft is suspected. However, where there is no such cultural restriction, most people are quite happy to give me the samples - they just think I am barking mad."

David Thomas is senior lecturer in biological oceanography at the University of Wales, Bangor.


Hair ointments, lotions and potions have been used throughout history, and these may distort the isotope values retrieved from archaeological remains. The Romans were fond of smearing goat's grease and ash in their hair.

Tamsin O'Connell has used her own locks in experiments to compare hair dyes and shampoos. "It is only when you use heavy-duty bleach that any changes occur. There is no difference between grey and normal hair from the same head."

She has also tested personal relationships by using her partner to supply different types of body hair. Again the isotope signature between hair types was the same.

Dr O'Connell's work has extended to the animal world to survey a range of digestive processes and diets. She is comparing the hair from dogs, cats, sheep, pigs, horses and, perhaps, tigers. With Rus Hoelzel from Durham University, she is studying the whiskers of Argentine and Scottish seals to track seasonal changes in diet.

Several unlikely research links have been set up. In conjunction with Hamish Simpson at the Nuffield department of orthopaedic surgery, hair and bone samples from patients undergoing orthopaedic surgery are being analysed to see if they are isotopically comparable. This will show whether archaeological (bone) isotopic values and modern (hair) isotopic values can be directly related.

With Andrew Prentice at the London School of Hygiene and Tropical Medicine, individuals of a wide age span in a village in The Gambia were sampled to see if children and pregnant women have different nitrogen isotopic values from other adults even though they have the same diet.

Please Login or Register to read this article.

Register to continue

Get a month's unlimited access to THE content online. Just register and complete your career summary.

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments