Unravelling the ketone story. Secondary vessel use at the Ness of Brodgar, Orkney, Scotland

The formation of ketones in archaeological contexts is often associated with the heating of animal fats within a ceramic vessel, resulting in ketonic decarboxylation. It is specifically assumed that the long-chain ketones K31 and K35 reflect the original distribution of C16:0 and C18:0 fatty acids within an archaeological sample, allowing reconstructing past vessel content. In our study, we demonstrate that the interpretation of ketones using a sole lipid biomarker approach needs to be considered with caution. However, by applying compound specific isotope analysis combined with Bayesian modelling to ketones, a secure lipid source can be identified. We also show that our methodology encompasses the possibility to securely identify secondary vessel use via the detection of different phases of use within a ceramic pot. Furthermore, our heating experiments demonstrate that the palmitic (C16:0)/stearic acid (C18:0) (P/S) ratio of archaeological samples is not a reliable tool to reconstruct past vessel content, neither from fatty acids nor from acyl chain composition of the ketones if samples were exposed to heat.