The analysis of protein residues recovered from archaeological artifacts provides a unique opportunity to reveal new information about past societies. However, many scientists are currently unwilling to accept protein-based results due to problems in method development and a basic lack of agreement regarding the ability of proteins to bind to, and preserve within, artifacts such as pottery. In this paper, I address these challenges by conducting a two-phase experiment. First, I quantitatively evaluate the tendency of proteins to sorb to ceramic matrices by using total organic carbon analysis and spectrophotometric assays to analyze samples of experimentally cooked ceramic. I then test a series of solvent and physical parameters in order to develop an optimized method for extracting and preparing protein residues for identification via mass spectrometry. Results demonstrate that protein strongly sorbs to ceramic and is not easily removed, despite repeated washing, unless an appropriate extraction strategy is used. This has implications for the future of paleodietary, conservation ecology and forensic research in that it suggests the potential for recovery of aged or even ancient proteins from ceramic matrices.

Yes / Distinguishing animal fats from plant oils in archaeological residues is not straightforward. Characteristic plant sterols, such as ¿-sitosterol, are often missing in archaeological samples and specific biomarkers do not exist for most plant fats. Identification is usually based on a range of characteristics such as fatty acid ratios, all of which indicate that a plant oil may be present, none of which uniquely distinguish plant oils from other fats. Degradation and dissolution during burial alter fatty acid ratios and remove short chain fatty acids, resulting in degraded plant oils with similar fatty acid profiles to other degraded fats. Compound specific stable isotope analysis of ¿13C18:0 and ¿13C16:0, carried out by gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS), has provided a means of distinguishing fish oils, dairy fats, ruminant and non-ruminant adipose fats but plant oils are rarely included in these analyses. For modern plant oils where C18:1 is abundant, ¿13C18:1 and ¿13C16:0 are usually measured. These results cannot be compared with archaeological data or other modern reference fats where ¿13C18:0 and ¿13C16:0 are measured, as C18:0 and C18:1 are formed by different processes resulting in different isotopic values. Eight samples of six modern plant oils were saponified releasing sufficient C18:0 to measure the isotopic values, which were plotted against ¿13C16:0. The isotopic values for these oils, with one exception, formed a tight cluster between ruminant and non-ruminant animal fats. This result complicates the interpretation of mixed fatty residues in geographical areas where both animal fats and plant oils were in use. / AHRC