Assessing the uncertainties of δ13C- and δ15N-values determined by EA-IRMS for palaeodietary studies

At the beginning of a program on palaeodiet, interest arose in how various uncertainty components propagate throughout the process of δ-value determination and how they affect the final uncertainties of the δ-values. The uncertainty components considered in this investigation arise from the precision of the measurement and from the uncertainties of the used isotope standards’ δ-values. In the uncertainty analysis, correlations were also taken into account, since they most often lead to an increase of the final uncertainties. The applied procedure permits not only to calculate the overall uncertainty of the normalized δ-values, but also to estimate the contributions of the various sources of uncertainties to the final uncertainty value. It was therefore possible to trace the different uncertainty components throughout the entire evaluation process. The uncertainties of the δ13C-values determined in this study were mainly caused by the respective statistical components resulting from the measurement process, since the uncertainties of the certified δ13C-values of the used international reference materials are small compared to the measurement uncertainty. In contrast, the δ15N values of the available nitrogen reference materials are less precise than those of the carbon isotope standards and thus more strongly affect the uncertainties of the determined δ15N-values. Consequently the uncertainties of δ15N-values can be underestimated, when only the statistical component is considered. Nevertheless, the final uncertainties obtained in this analysis are small in magnitude compared to the variation in δ-values relevant for diet interpretation. Although the investigated correlations had an enlarging effect on the uncertainties of the normalized δ-values, it could be shown that this effect is small and can be neglected in most palaeodietary studies.