Can element chemical impurities in aragonitic shells of marine bivalves serve as proxies for environmental variability?

Abstract

In many biogenic and geogenic materials, ion impurities can provide serviceable proxies for environmental conditions. However, the element/Ca ratios of bivalve shells are notoriously challenging to interpret. Due to strong vital effects, nonclassical nucleation and growth mechanisms, and/or kinetic processes, the concentration of trace and minor elements in marine shells typically remains below values observed in inorganic CaCO3 precipitated from a solution resembling seawater chemistry but above those expected for thermodynamic equilibrium. The interpretation is further complicated by non-lattice bound and microstructure-specific element content. If environmental conditions were still encoded in the shells, they should result in statistically significantly reproducible element/Ca chronologies between contemporaneous specimens from the same site. Here, we tested this hypothesis and exemplarily studied seven elements in twelve modern specimens of Arctica islandica collected from four different localities in the North Atlantic (Faroe Islands, NE Iceland, Isle of Man, Gulf of Maine). Age-detrended chronologies of weighted annual B, Mg, Sr and Ba/Ca ratios (Al, Zn and Pb largely remained below detection limit) measured in the shells were reproducible between most specimens from the same site, supporting the hypothesis that the incorporation of these elements was at least partly controlled by environmental forcings. Notably, the agreement (explored with linear regression analyses and sign tests) between shell element/Ca ratios and environmental quantities was weaker than the agreement of respective element/Ca ratios between specimens suggesting that the available information on temperature, food and water chemistry did not properly reflect the in-situ conditions to which the bivalves were exposed or other extrinsic factors were at work. As in inorganic aragonite – but in contrast to thermodynamic expectations –, annual Sr/Ca, Mg/Ca and B/Ca ratios were negatively correlated to water temperature (up to 40% explained variability). The link between Ba/Ca and bulk phytoplankton often remained below the significance threshold, but was otherwise positive. Quantitative environmental reconstructions based on ion impurities in bivalve shells will remain challenging or impossible unless the chemistry of the parent solution (= extrapallial fluid) from which the shell actually formed is known, including temporal changes thereof. This information is crucial to compute representative partition coefficients required to calibrate transfer functions.