Stable oxygen isotopes reveal different thermal regimes during the Early Pliocene in the southern North Sea Basin: a multi-species approach

ABSTRACT Oxygen stable isotope (δ18O) analysis of marine biogenic carbonates is widely used to study the temperature at which the hard parts precipitated. This technique enables investigation of the environmental conditions experienced by carbonate-shelled organisms throughout their life. We used this approach on four bivalve species from the Coralline Crag Formation (Early Pliocene; UK), where warm- and cold-associated species co-occur in the Ramsholt Member. Our results reveal species-specific differences in reconstructed temperature. Two distinct climatic settings were identified: a cold setting with seafloor temperatures of 7.3 °C in winter and 12.4 °C in summer, and a warm one with seafloor temperatures of 10.4 °C in winter to above 14 °C in summer. These absolute temperatures were calculated using a modelled value of seawater δ18O (+0.1‰) that is open to question, but the disparity in temperatures exists whatever value is used. Interestingly, the reconstructed temperatures for some of the studied species align with those of their modern relatives: Arctica islandica records the coldest temperatures in this dataset, while Centrocardita squamulosa ampla indicates warmer conditions. Aequipecten opercularis, a eurythermal pectinid, exhibited the broadest range of temperatures. The interval during which the Coralline Crag was deposited (c. 4.4‒3.8 Ma) is generally regarded as relatively warm. However, the presence of A. islandica and the absolute seawater temperatures derived from shell δ18O suggest the occurrence of sufficiently long cool periods for the establishment and survival of this long-lived species. Data from specimens of C. squamulosa ampla and A. islandica collected from a single bed point to glacial/interglacial-type climate fluctuation. These results confirm that the presence of A. islandica is a marker for cold temperate environments. These results also confirm that assemblage information is not sufficient to determine palaeoclimatic conditions, and that reconstructed temperatures from shell δ18O can reveal time-averaging in fossil shell beds.