Assessing Raman Spectroscopy as a Prescreening Tool for the Selection of Archaeological Bone for Stable Isotopic Analysis

Stable isotope analyses for paleodiet investigations require good preservation of bone protein, the collagen, to obtain reliable stable isotope values. Burial environments cause diagenetic alterations to collagen, especially in the leaching of the organic bone content. The survival of bone protein may be assessed by the weight % collagen, % carbon and % nitrogen yields, but these values are achieved only after destructive chemical processing. A non-destructive method of determining whether bone is suitably preserved would be desirable, as it would be less costly than chemical processing, and would also preserve skeletal collections. Raman analysis is one such potential non-destructive screening method. In previous applications, Raman spectroscopy has been used to test both the alteration of the mineral portion of bone, as well as to indicate the relative amount of organic material within the bone structure. However, there has been no research to test the relationship between the Raman spectroscopic results and the survival of bone protein. We use a set of 41 bone samples from the prehistoric archaeological site of Ban Non Wat, Northeast Thailand, to assess if Raman spectroscopy analysis of the organic-phosphate ratio has a significant correlation with the weight % collagen, and carbon and nitrogen yields obtained by isotopic analysis. The correlation coefficients are highly statistically significant in all cases (r = 0.716 for collagen, r = 0.630 for carbon and r = 0.706 for nitrogen, p≤0.001 for all) with approximately or close to half of the variation in each explained by variation in the organic-phosphate ratio (51.2% for collagen, 39.6% for carbon, and 49.8% for nitrogen). Although the Raman screening method cannot directly quantify the extent of collagen survival, it could be of use in the selection of bone most likely to have viable protein required for reliable results from stable isotope analysis.