The selection of non-marine molluscan shells for radiocarbon dating
The use of shells from terrestrial and freshwater molluscs for radiocarbon dating has in the past been viewed with scepticism or even dismissed entirely because the results are prone to distortion from post-depositional diagenesis, or the incorporation of material with a low 14C/12C ratio whilst the mollusc was alive, or both. The thesis attempts to discuss how the relative importance of the two factors can be assessed and the corresponding ages corrected accordingly. Diagenesis was studied by comparing the structure and chemical composition of 39 species of British molluscs from modern specimens with those of fossil samples. Three main effects of diagenesis were studied using light and scanning electron microscopy, mass spectrometry and atomic absorption spectroscopy: the transformation of aragonite to calcite, a reduction in the standard deviation associated with the distribution of major ion concentrations within a shell population, and an enrichment in the lighter isotopes of carbon and oxygen. Only the transformation of mineralogy proved to be sufficiently sensitive to detect diagenetic material present at levels of 1-27%. The effects of diet and ambient carbon dioxide were studied by chemical analysis and mass spectrometry. Differences in stable isotope ratio values between micro-environments were found to be significantly greater than those between and within species. The established relationship between 13C, 14C and the environment can be used to identify shells in which the apparent age effect will be at a minimum. Comparison of 14C dates for charcoal and untreated shell samples showed that the above sources of error led to discrepancies of as much as 1500 years in the shell dates. Seven samples selected and pretreated in accordance with the procedures proposed in this thesis gave ages within two standard deviations of the values accepted for the deposit in question, and four samples gave ages which were being statistically identical with the accepted values. In addition, the screening tests permitted samples destined to give erroneous ages to be identified from the outset.