Stable Isotopes and Faunal Bones. Comments on Milner et Al. (2004)

Article excerpt

One of the first applications of stable isotope analysis in archaeology was aimed at the investigation of dietary changes between Mesolithic and Neolithic times in northern Europe (Tauber 1981). Since then, important isotopic research related to this topic has been carried out using samples from different regions (i.e. Lubell et al. 1994; Richards & Hedges 1999; Richards et al. 2003). The degree to which the transition from the Mesolithic to the Neolithic implied an abrupt subsistence change, accompanied by an almost complete abandonment of marine resources, has been questioned in recent debate (Bailey & Milner 2002; Hedges 2004; Liden et al. 2004; Milner et al. 2004). In this note, we amplify some of the caveats already expressed there regarding the comparison of faunal and isotopic data (see also Parkington 2001), and we suggest a number of methodological issues that could be taken into account in order to validate the comparisons. Some of these are also important for the integration with other kinds of archaeological data (i.e. distributions of artefacts).

Use of the mineral fraction of bone

One of the methodological problems mentioned by Milner et al. (2004) regarding isotopic analysis is that [delta][sup.13]C and [delta][sup.15]N values on the organic fraction of bone--or also dentine in teeth--are influenced by metabolic processes related to the amount of protein, lipids and carbohydrates consumed, and therefore, do not reflect the total diet of an individual (see Ambrose & Norr 1993). They correctly point out that 'Preliminary models (suggested by Hedges in this volume) show that a diet dominated by plant-food might include up to 20 per cent of marine protein without raising the [delta][sup.13]C values of bone collagen above--21 per mil; this would standardly be interpreted as a diet dominated by terrestrial food' (Milner et al. 2004: 18). A tool ideally suited to overcome this situation is [delta][sup.13]C analysison the mineral fraction of bone-enamel in teeth. The method has been thoroughly studied from the point of view of preparation of the samples (Koch et al. 1997), evaluation of diagenetic processes (Krueger 1991; Lee-Thorp 2000) and dietary interpretation (i.e. Tykot et al. 1996). If such analyses are included, some of the existing controversies may be independently assessed and perhaps solved.

Factors of comparison

When evaluating differences in isotopic and faunal data, Milner and co-authors (2004: 19) soundly suggest that '(...) these two techniques measure palaeodiet at two such different scales of observation and resolution that to bring them into a coherent relationship according to a common scale of values is very difficult. Using one technique as a crosscheck on the other may be misleading'. We could not agree more with this observation, since an uncritical comparison of these types of evidence would certainly lead to a misinterpretation of data on methodological and theoretical grounds. However, there are three theoretical aspects that could be developed, beginning with units of analysis. The obvious level of significance of isotopic data is the individual, whereas faunal assemblages are more difficult to relate to a specific entity that produced them. Depending on the context, they may range from being the product of one discrete event of human occupation--something never easy to establish on taphonomic grounds--to an assemblage which reflects a long term average of human and non-human activities. Related to this is the temporal resolution that each type of evidence represents. Isotopic values on bone reflect the averaged product of food consumed during approximately the last 10 years of life (variability exists in relation to differences in bone turnover rates, and age has also been pointed as another source of variation for the rate of formation of the isotopic signal). Faunal assemblages represent a wider variability than isotopes, ranging from the duration of single occupation event to an assemblage representing thousands of years' duration. …