Feature

Rating the Dating

Nancy R. Beavan and Rodger J Sparks

In his account of the controversy surrounding the radiocarbon dating of the kiore bones, Bruce McFadgen has raised a number of important questions concerning the interpretation to be placed on the dates. If the rat bones really are as old as the radiocarbon ages indicate, then a lot of what we thought we knew about early Polynesian contact will have turned out to be wrong.

At the Rafter Radiocarbon Laboratory we agree with McFadgen when he says that scientific results that fly in the face of accepted knowledge must be examined closely, to see if alternative explanations that preserve the status quo are possible. Scientific evidence is only accepted as true when it fits as part of a larger picture that has already been established, or forces changes to that picture to make an even better fit than existed previously. Being "different" or "exciting" does not qualify as a good criterion for advancing existing knowledge.

As the laboratory that supplied Holdaway's dates we are very aware of the necessity of being able to interpret the data correctly. There are basically two questions to be answered, and they underlie all radiocarbon ages:

• is the carbon that was analysed the same carbon that was deposited in the bones when the organism was alive, or has carbon from other sources replaced or added to the original material?
• if the carbon is actually the original carbon, do we properly understand where it came from in the first place?

Bone is a notoriously difficult material to date with radiocarbon. Although the mineral and protein components of bone are much more robust than soft tissue, and can persist for remarkably long times even when buried in the ground, chemical and biological processes can attack the bone so that external carbon can attach to the structure, while the original carbon can degrade and be lost. Whole scientific reputations, not to say careers, have been built on studying the nature of the changes that can occur in buried bone and devising ways of purifying the "native" constituents in order to be able to obtain a reliable radiocarbon age.

It is not surprising then that the first questions to be asked about the Holdaway kiore ages centred on the procedures used to prepare the bones for dating. McFadgen discusses the matter of rat bones from the Shag River site and the inconsistencies they seem to show.

After first noting that none of Holdaway's bones came from Shag River, let us say at once that Shag River poses problems. On the one hand, we have an archaeological site that seems to be well understood, and has been reliably dated using charcoal and shell to around 600 years. On the other hand, rat bones from the same site yield ages up to 2,000 years. A Shag River bone dated at Oxford University came out at around 900 years old -- younger than those measured at the Rafter Laboratory, but still much older than expected for this site, and older than conventionally accepted ideas about Maori occupation of New Zealand.

A feature of all these bones is that they were highly degraded and contained significant quantities of contaminants that had to be removed.

Other Examples

Accepting that there are questions to be answered at Shag River, what of other examples? Rat bones have been dated at the Rafter Laboratory from another well studied archaeological site, Pauatahanui, near Wellington. In this instance we have two cases of a bone that can be compared with a shell taken from the same midden, and which would be expected to be the same age. The measured ages, including a correction for the marine origin of the shells, for the bone and shell pairs are 452 69 and 379 51 years for one and 439 71 and 406 35 years for the second bone-shell pairing. The agreement between the bone and shell ages is good, and the ages are consistent with what is known about this site.

Another example is a sample set submitted by Holdaway for a cross-check of preparation techniques. The sample pair consisted of a piopio bone found in association with fragments of an eggshell below the Taupo Ash layer at the Hukanui site, where a number of Holdaway's rat bones came from. Eggshell does not have the same problems as bone for radiocarbon dating, and indeed is prepared for analysis by quite different chemical methods. The close association of the materials makes the assumption of contemporaneity a reasonable one.

The radiocarbon age obtained for the eggshell was 2905 88 BP (Before Present) and for the bone 2995 72 BP. Once again, agreement is very good, and both ages are older than the ash layer which was deposited around 1,760 years ago.

What this tells us is that when we have a tight association of a bone with a different material that can be reliably dated, we get good agreement. We do not think that the ages found for Holdaway's rat bones can be simply explained as due to inadequate laboratory preparation of the bones, as has been suggested.

Dietary Details

What about the effects of diet? Could the rats have been eating food that was already depleted in radiocarbon, and hence responsible for the bones showing a false age?

McFadgen suggests some possible ways for radiocarbon ages to become biased to older values. This possibility is the main thrust of research at present going on at the Rafter Laboratory. It has been necessary to study the processes by which ingested carbon is routed to different body tissues, in particular to the bone collagen. The only means for depleted carbon from a food source to be integrated into bone portion would be via the portion portion of the diet.

We have investigated these dietary pathways using modern populations of kiore. The work is still in progress, but already some comments can be made. McFadgen suggests that the 900 year age found by Oxford for the Shag River bone can be explained by the rat having a diet of fish, since shallow-dwelling marine organisms have radiocarbon ages 300-400 years greater than terrestrial organisms of the same chronological age.

However, two issues arise. One, the kiore is an omnivore which must include terrestrial foods such as grass seed in its diet, foods which will likely have non-depleted carbon. Secondly, the carbon in bone protein is made up of amino acids which come from the protein portion of the diet, the carbohydrate portion of the diet, or are metabolised from same.

In fact, because of the distribution of carbon over different types of amino acids from different food sources, the shift in age, even if the rat's protein input was entirely fish, is not likely to be more than about 100 years at most, so at best the rat is still about 200 years "too old".

By consideration of how bone collagen is constituted it is possible to set lower limits on the amount of radiocarbon depletion needed in the diet to produce ages in the range observed for the kiore, and these limits tend to be around 7,000 years or greater. It is not inconceivable to construct situations in nature where such depletions in particular dietary items occur, but whether or not they occur over the range of habitats or food supplies observed has yet to be established, and does not seem likely.

It is particularly unlikely, if not metabolically improbable, that kiore can ingest limestone and take up its depleted carbon.

What would be useful is a means of distinguishing depleted and "normal" food sources within an individual specimen. We have a project now under way based on the idea that it might be possible to make this distinction by studying the fundamental components of bone collagen, and if so it could be the key to obtaining reliable radiocarbon ages from bone without dietary bias. The outcome of this ongoing work will be reported in the near future.

McFadgen's article raises questions that have to be asked about the radiocarbon ages measured in kiore bones. The matter is not yet finally resolved, but enough has been learnt that some possible explanations can be put aside, while required parameters for a dietary explanation have been more closely defined. We see this issue as too important to let go with either easy acceptance or dismissal of the evidence offered by the kiore bones, and believe that research is justified to put the interpretation either for or against the antiquity of the kiore on a sound basis.

Nancy Ragano Beavan is a radiocarbon researcher at the IGNS Rafter Radiocarbon Laboratory in Lower Hutt.
Rodger Sparks is Section Leader of the Rafter Radiocarbon Laboratory, Institute of Geological & Nuclear Sciences.