Laboratory cooking and degradation experiments to support interpretations of absorbed lipid residues arising from the processing of broomcorn millet (Panicum miliaceum) and animal products in archaeological pottery vessels II. Isotope analysis

While animal fats are by far the most widespread class of lipid residue preserved in archaeological potsherds, the possibility of mixing foodstuffs in vessels is evident from the detection of plant lipid residues in some sherds. Such ‘mixed’ extracts are usually treated cautiously when determining the origins of the animal fat components. This is particularly relevant in regions where broomcorn millet ( Panicum miliaceum ) was introduced by pioneer farmers alongside domesticated animals. Broomcorn millet processing has been identified in lipid extracts of archaeological potsherds based on the detection of miliacin, typically accompanied by abundant animal fats. Degradation experiments have been conducted to explain the abundance of miliacin observed in archaeological lipid extracts. We now explore the effects of processing varying proportions of millet grains with animal products on the δ 13 C values of C 16:0 and C 18:0 fatty acids used to assign the origins of animal fats. A combination of theoretical mixing calculations and laboratory experiments were used to test different scenarios of mixing animal products and millet grains in archaeological pottery vessels. In all instances the millet fatty acids were found to be isotopically ‘invisible’ when processed with meat or dairy products due to ‘overprinting’ by the abundant animal acyl lipids and the low extraction efficiency of water for millet lipids. Thus, the δ 13 C values of C 16:0 and C 18:0 fatty acids can be used confidently to determine the origins of animal fats even when millet processing is detectable through the miliacin biomarker.