Putting the flesh on the bones of archaeology

Chemistry can illuminate the age of a specimen, build up a picture of prehistoric diet and lifestyles and can even probe the genetic makeup of long-extinct populations and species.

The tools of the 21st century archaeologist are as much about cutting-edge chemistry as digging holes. Careful excavation is, of course, a crucial part of reconstructing the past, but post-excavation analysis is yielding answers to questions that are only just beginning to be asked.  

You are what you eat 

There are frustrating gaps in the archaeological record but chemistry is helping to fill them in. One example is how natural variation in the abundance of stable isotopes and marker molecules in different tissues can be used to infer what an individual was eating during their lifetime - clues that are otherwise missing from the archaeological record. 

Plants take up N₂ from the atmosphere and at each step that nitrogen moves up the food chain, there is a perceptible increase in the abundance of ¹⁵N relative to ¹⁴N. Suckling infants feeding from their mothers are one trophic level above adults, and therefore have some of the highest ¹⁵N:¹⁴N ratios. This has been used to work out the age at which children are likely to have been weaned.  

So far, this kind of analysis of stable isotopes has been possible on a handful of Neanderthal specimens where collagen has been extracted from bone. Apart from telling us something about the kinds of food they were eating, some have hoped that it might be possible to infer the reason why the Neanderthals became extinct around 28000 years ago. If, for example, it became apparent that Neanderthals had changed the sorts of food they were eating, this might suggest increasing competition with other predators, such as the ancestors of modern humans.  

But all the Neanderthals that have had their diet reconstructed paint the same picture. High ¹⁵N:¹⁴N ratios indicate that these hominids were top-level predators, hunting large herbivores like woolly rhinos and mammoths rather than scavenging on whatever they could get their hands on. And no obvious change in diet has been detected. 

The isotope signature in other tissues can also be telling. For example, teeth are formed during childhood but bone is continually remodelled throughout an individual’s life. Since isotopes of elements like strontium and lead pass into such structures in approximately the same ratio as they exist in the geological environment, any difference in the isotope signature in the teeth and bone of the same individual may suggest that he or she moved from one place to another at some stage in their life. 

This appears to be the case for the Iceman - the approximately 5300 year old human preserved in an alpine glacier in Northern Italy. The balance of ⁸⁷Sr to ⁸⁶Sr in his teeth is different from that in his bone, suggesting that he probably spent his boyhood at low altitude before moving up into the mountains where he eventually met his death.  

The dating game 

One of the most important contributions of chemistry to archaeology is the development of dating methods. By far the most common technique is radiocarbon dating of organic remains like wood, charcoal and bone, using accelerator mass spectrometry to measure the abundance of radioactive ¹⁴C relative to its stable cousin ¹²C.  

The carbon in a sample is first converted to graphite and then ionised by bombarding it with caesium ions. The carbon ions are accelerated through a magnetic field and deflected at an angle proportional to their atomic weight, allowing each ¹⁴C and ¹²C atom to be counted individually. This means that only a few milligrams are needed and results can be obtained within hours. It’s a powerful technique. 

But, as with most dating techniques, it does have its drawbacks. Most notably, it can only date specimens back to about 50 000 years ago, after which very little ¹⁴C remains. Then there are situations where it’s not practical to take samples of organic material. Scraping away charcoal to date a cave painting, for example, destroys the very archaeological remains that are of interest. There is also concern about contamination from more recent sources of carbon. Some have argued, for example, that radiocarbon analysis indicating the Turin Shroud dates from the 13th century might simply be the result of microorganisms that colonised the garment in the Middle Ages and that the shroud did cover Jesus’ crucified body after all.  

Radiocarbon dates even run the risk of being confounded by the sort of food an individual ate. This problem is illustrated by a bunch of men mysteriously murdered on the upper Danube around 8000 years ago. Radiocarbon analysis of human bone at the scene suggested that the men were up to 700 years older than the spear tips that killed them. This apparent nonsense is explained if they ate a lot of freshwater fish, which is particularly low in ¹⁴C because of the limestone that’s dissolved in the river. This would have the effect of making them look older than they really were. 

Thankfully, there is a suite of other dating methods that can be brought into play when, for whatever reason, a reliable radiocarbon date cannot be obtained (see box). At the other end of the spectrum from radiocarbon dating is potassium-argon dating. ⁴⁰K, like ¹⁴C, is radioactive but decays to ⁴⁰Ar with a much longer half-life - 1250 million years compared to just 5730 years for ¹⁴C. This means there is still enough ⁴⁰K in a geological sample to be detected many million of years after it formed.  

Up and coming 

In between these two extremes, there are several up and coming methods for dating archaeological remains. One that can be used in all sorts of settings is uranium series dating. This is particularly useful for dating inorganic precipitates like stalagmites or stalactites that might have snapped off and been preserved alongside archaeological remains.  

Uranium is relatively soluble so it makes it into the calcium carbonate-loaded water that seeps into a cave and runs over such structures. By contrast, a couple of key long-lived daughter isotopes of uranium decay - ²³⁰Th and ²³¹Pa - are insoluble, so don’t. In time, however, as the uranium in the stalagmite decays, these daughter isotopes begin to appear. So the ratio of uranium to its daughter isotopes gives an indication of the time passed since the stalagmite formed. 

In April 2003, archaeologists discovered engravings of humans and animals etched into the Creswell Crags in Nottinghamshire, UK. This was thought to be the oldest rock art found in the UK, but because the drawings are etched into the inorganic bedrock, radiocarbon dating could not be used to confirm this hunch. Fortunately, in a couple of places, stalagmites had formed across the rock surface obscuring part of the engraving. 

’Analysis of the uranium series isotopes in the very earliest bit of the stalagmite layer suggests that the deposit formed 12 800 years ago, providing a minimum age for the rock art,’ says Alistair Pike, an archaeologist at the University of Bristol. ’This tallies with a burst of human activity in the cave that radiocarbon dating puts at between 13 000 and 16 000 years ago,’ he says. Furthermore, it shows that the Creswell rock art is indeed the earliest known in the UK.  

In addition, there are three further dating techniques - electron spin resonance, thermoluminescence and optically stimulated luminescence (OSL) dating - based on the principle that radioactive elements in an archaeological deposit release electrons that are taken up and stored by certain minerals in the surrounding sediment. The longer the exposure to this ionising radiation, the more trapped electrons these minerals will contain. 

Of these three techniques, OSL dating is experiencing the greatest surge of interest because of the abundance of suitable sediments at archaeological sites. When crystalline minerals like quartz and feldspar are exposed to light, electrons trapped in pockets within the crystal are evicted, which has the effect of ’zeroing’ the mineral. But once they are buried, electrons released from radioactive uranium, thorium and potassium are taken up and stored in these crystal pockets once more. So the longer such minerals have been buried, the more electrons they will contain. And by measuring the luminescence produced by this stored charge, archaeologists can gauge how long it is since a sample saw the light of day. 

OSL dating can currently date minerals that were laid down hundreds of thousands of years ago. But work in progress may extend the reach of this technique even further into the past. ’We think we can go, maybe, a million years and beyond,’ says Jean-Luc Schwenninger, head of the luminescence dating and research group at the University of Oxford, UK.  

Residue analysis 

Chemistry is also being brought to bear upon some rather intriguing archaeological artifacts. When, in July 2003, archaeologists uncovered a sealed cylindrical tin at the site of a Roman temple in London, they called in chemists to characterise the granular cream it contained. The long list of chemical techniques being used on its contents is impressive. 

Fatty acids of animal origin made up a significant proportion of the cream. Partial gas chromatography of these extracted lipids revealed an abundance of saturated, straight-chain fatty acids but also some iso- and anteiso- branched-chain fatty acids. This lipid signature suggests that whoever manufactured the cream used adipose fat from sheep or cattle, says Richard Evershed, an archaeologist at the University of Bristol who masterminded the analysis. 

Another key ingredient turned out to be starch. In addition, gravimetric analysis, involving heating to 850?C, revealed an inorganic residue. X-ray fluorescence suggested this residue was dominated by a tin compound and x-ray diffraction identified the compound as SnO₂. ’The addition of SnO₂ to the starch/fat base confers a white opacity, which is consistent with the cream being a cosmetic,’ wrote Evershed and his colleagues in Nature last year.  

Similar techniques can identify any animal fats and plant waxes left on ceramics that were used to prepare, cook and store food. It has been possible to characterise resins used by North American Indians to waterproof bottles and baskets; and others have described the resins and tars used to seal vases in ancient Greece, repair jars in ancient Rome and waterproof ships like Henry VIII’s Mary Rose. 

Digging up the family tree 

Within the last few years, there has been an explosion of studies into the genetics of ancient human populations. It soon became obvious that this would be a great way to question the relationship between ancient and modern humans. 

The study of so-called ancient DNA (aDNA) has, for example, successfully laid to rest the idea that the ancestors of modern humans and Neanderthals got it together. If such a liaison had occurred in the 10 000 or so years that they lived alongside each other in Europe, there could be tell-tale signs of Neanderthal-like DNA sequences contained in the modern human genome. There aren’t. It turns out that there isn’t even much genetic similarity between Neanderthals and early modern humans, ruling out the possibility that there had been some genetic exchange between Neanderthals and our ancestors that has since been masked by time.  

But getting DNA from really old specimens like Neanderthals is not straight-forward. DNA doesn’t survive too well. Once an organism dies, nucleases set about digesting it. What’s more, radiation, oxidation and hydrolysis will all take their toll and quite quickly a genetic sequence is reduced to a four-letter alphabet soup. Even if a specimen has been perfectly preserved, there’s unlikely to be much DNA left after about a million years. And if it is present, there will often be just a few short sequences to work with. These are amplified using the polymerase chain reaction, but stringent laboratory controls are needed and results must be replicated to be confident that contamination with modern human DNA has not occurred. 

Unfortunately, many archaeological sites are in harsh environments where DNA is unlikely to survive more than a few thousand years. It would have been nice, for example, to collect aDNA from the remains of 75000 year old Neanderthals excavated from Shanidar Cave in Iraq: a genetic comparison between European and Middle-Eastern Neanderthals would almost certainly tell an intriguing story about the evolution of these hominids. But no DNA from these Shanidar specimens survives. 

Ancient protein 

So scientists have resorted to the next best thing: ancient protein. The sequence of amino acids in a protein can inform us about the DNA sequence that instructed the protein synthesis. There’s just a chance that there could be revealing differences in the amino acid sequence of a protein. 

Some proteins can survive for a long time, particularly osteocalcin - a small negatively charged molecule present in bone. ’We’re finding it surviving in much older specimens and in much more hostile environments than collagen,’ says Christina Nielsen-Marsh, a biomolecular archaeologist at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. 

Protection from degradation 

Osteocalcin contains three gamma-carboxyglutamic acid residues that bind calcium ions and allow it to attach to the surface of the hard-wearing bone mineral hydroxyapatite. It’s this association that helps to protect osteocalcin from degradation, says Nielsen-Marsh. There is also a possibility that osteocalcin is cross-linking to collagen, she says.  

The osteocalcin sequence is highly conserved within the central portion of the molecule, but the N-terminus displays considerable variation. Nielsen-Marsh and her colleagues managed to extract osteocalcin from two of the Shanidar Neanderthals and used Maldi-TOF/TOF mass spectrometry to determine its sequence. 

It turns out the string of amino acids that make up the N-terminus of Neanderthal osteocalcin is identical to that found in osteocalcin from modern humans. But this kind of proteomic analysis is in its infancy and attempts are under way to extract and sequence other well-preserved proteins that might yield more evolutionary information.  

One intriguing finding that did come out of the Shanidar study was that Neanderthal osteocalcin doesn’t appear to undergo a key chemical alteration that often occurs to the protein following its synthesis. In most other mammals where the osteocalcin sequence is known, the ninth amino acid proline is hydroxylated to hydroxyproline. But in Neanderthals, as in modern humans, this reaction doesn’t occur. 

This may be because the enzyme prolyl-4-hydroxylase that drives the hydroxylation requires adequate concentrations of ascorbic acid or vitamin C (amongst other things). Since most mammals can only get vitamin C from their diet, this indicates that there may have been a period in hominid evolution where it was necessary to conserve vitamin C and limit the hydroxylation of proline, says Nielsen-Marsh. This may have occurred as species like Homo sapiens and Neanderthals shifted from a herbivorous to an omnivorous diet, she says. 

There’s little doubt that archaeologists will continue to borrow techniques from chemists to dig deeper into the past and to revisit old questions and visit new ones. And the division between the fieldwork and laboratory analysis is less obvious than it’s ever been. Archaeologists need no longer be confined to the stereotype that kneels in a dusty quarry, carefully exposing some buried skull or shard of pottery. Increasingly, they can afford to spend fruitful hours embracing a second stereotype - that of the labcoat clad, safety goggled, fume cupboard dwelling chemist. 

Henry Nicholls is a freelance science writer 

Further Reading

• T C O’Connell and A W G Pike Mass spectrometry: archaeological applications. In Encyclopaedia of Analytical Science (2nd edn), eds. P J Worsfold, A Townshend and C F Poole, Elsevier Science   
• R P Evershed et al, Nature, 2004, 432, 35 
• D Serre et al, PLoS Biol, 2, e57 
• C M Nielsen-Marsh et al, Proc. Natl. Acad. Sci. USA, 2005, 102, 4409 
• M Krings et al, Nature Genetics, 2000, 26, 144