Shell growth of A. islandica reveals depth-dependent temperature variability

Bivalve shells can provide high-resolution proxy records that facilitate our understanding of environmental variability. Conceptually similar to tree rings, indices of A. islandica shell growth have become a valuable proxy for reconstructions of ocean variability in the North Atlantic sector. Past studies showed that there is a site-specific relationship between shell growth and environmental variability. Water temperature is a major albeit not the only driver of shell growth. Yet, it is challenging to establish a relationship with temperature within the well-mixed ocean layer, simply because monthly-resolved observational subsurface temperature data are quite scarce. We present a study case to explore from where exactly do bivalve shells residing at greater depths pick up the temperature signal, and how well it synchronizes locally and at large-scale. We correlate an annually-resolved master shell chronology from Butler et al. (2013), collected at 82 m water depth from the N. Icelandic shelf, with monthly-resolved EN4.2.1 temperature dataset across several depth layers in the North Atlantic region over the common analysis period 1900-2005. Results show that near collection site, the sea surface temperature signal is recorded poorly, whereas the strongest positive in-phase correlations (r>0.4) occur during the summer and autumn months with increasing water depth, corresponding to patterns of thermal expansion of local water bodies. We then use this information to thermally link distant water bodies, revealing a large-scale oceanographic pattern extending into the Arctic region. This pattern allows us to identify potential source regions of growth variability (i.e., a predominance of warm Atlantic waters), and modes of natural variability (e.g., North Atlantic Oscillation features observed during winter). By analyzing the synchronicity with the strongest temperature signal, we find periods of decoupling which could aid in assessing climatic anomalies. Our results strengthen the reliability and application of A. islandica as recorder of temperature variability. To maximize the use of the shell proxy, future correlation analyses should include not only the surface layer, but also various water depths.