Seasonal SIMS δ18O record in Astarte borealis from the Baltic Sea tracks a modern regime shift in the NAO

• Published in: Frontiers in Marine Science
• Main Authors: Hunter P. Hughes, Donna Surge, Ian J. Orland, Michael L. Zettler, David K. Moss
• Format: Article in Journal/Newspaper
• Language: English
• Published: Frontiers Media S.A. 2023
• Subjects: Bivalvia; secondary ion mass spectrometry; oxygen isotope ratios; North Atlantic oscillation; paleoclimate; Science; Q; General. Including nature conservation; geographical distribution; QH1-199.5; North Atlantic
• Online Access: https://doi.org/10.3389/fmars.2023.1293823; https://doaj.org/article/cf33d1a17c884ce68c0ffde7a3f2f605

IntroductionAstarte borealis holds great potential as an archive of seasonal paleoclimate, especially due to its long lifespan (several decades to more than a century) and ubiquitous distribution across high northern latitudes. Furthermore, recent work demonstrates that the isotope geochemistry of the aragonite shell is a faithful proxy of environmental conditions. However, the exceedingly slow growth rates of A. borealis in some locations (<0.2mm/year) make it difficult to achieve seasonal resolution using standard micromilling techniques for conventional stable isotope analysis. Moreover, oxygen isotope (δ18O) records from species inhabiting brackish environments are notoriously difficult to use as paleoclimate archives because of the simultaneous variation in temperature and δ18Owater values.MethodsHere we use secondary ion mass spectrometry (SIMS) to microsample an A. borealis specimen from the southern Baltic Sea, yielding 451 SIMS δ18Oshell values at sub-monthly resolution.ResultsSIMS δ18Oshell values exhibit a quasi-sinusoidal pattern with 24 local maxima and minima coinciding with 24 annual growth increments between March 1977 and the month before specimen collection in May 2001.DiscussionAge-modeled SIMS δ18Oshell values correlate significantly with both in situ temperature measured from shipborne CTD casts (r2 = 0.52, p<0.001) and sea surface temperature from the ORAS5-SST global reanalysis product for the Baltic Sea region (r2 = 0.42, p<0.001). We observe the strongest correlation between SIMS δ18Oshell values and salinity when both datasets are run through a 36-month LOWESS function (r2 = 0.71, p < 0.001). Similarly, we find that LOWESS-smoothed SIMS δ18Oshell values exhibit a moderate correlation with the LOWESS-smoothed North Atlantic Oscillation (NAO) Index (r2 = 0.46, p<0.001). Change point analysis supports that SIMS δ18Oshell values capture a well-documented regime shift in the NAO circa 1989. We hypothesize that the correlation between the SIMS δ18Oshell time series and the NAO ...