δ18O-inferred salinity from Littorina littorea (L.) gastropods in a Danish shell midden at the Mesolithic–Neolithic transition

© The Author(s) 2019. Norsminde Fjord has received extensive geoarchaeological investigation, hosting one of the classic Stone Age shell midden sites in Denmark, and one of the best examples of the widespread oyster decline at the Mesolithic–Neolithic transition. Here, intra-shell δ18O (and δ13C) analyses from the common periwinkle Littorina littorea (L.) are used to infer inter-annual environmental changes at the Mesolithic–Neolithic transition (four from each period). This study utilises a modern δ18O L. littorea-salinity training set previously developed for the Limfjord, Denmark to quantify winter salinity. δ18O values range between +1.6% and +4.0% in the late Mesolithic and ‒6.3% to +2.0% in the early Neolithic. Using maximum δ18O values, winter salinity at the known temperature of growth cessation in L. littorea (i.e. +3.7 ± 1°C) for the first annual cycle of each shell ranges between 25.5 and 26.8 psu (standard deviation (SD): 0.56) for the late Mesolithic, with an average salinity of 26.1 psu. Early-Neolithic shells range between 19.4 and 28.2 psu (SD: 4.59) with an average salinity of 23.7 psu. No statistically significant change in salinity occurs between the late Mesolithic and early Neolithic. This result supports recent diatom/mollusc-based inferences that salinity was not the sole cause of the oyster decline, although some evidence is presented here for more variable seasonal salinity conditions in the early Neolithic, which (along sedimentary change and temperature deterioration) might have increased stress on oyster populations in some years. It is recommended here that for robust palaeoenvironmental inferences, where possible, multiple specimens should be used from the same time period in conjunction with multiproxy data