An assessment of P speciation and P:Ca proxy calibration in coral cores from Singapore and Bali

Phosphorus (P) in corals has shown potential as a proxy for dissolved inorganic phosphate (DIP) in seawater, but additional investigations are needed to understand its incorporation mechanism, down-core applicability, and sensitivity, especially in oligotrophic environments with low DIP variability. In this study, we used a new method to distinguish between inorganic and organic P in the skeletons of two Porites sp. cores from Singapore and Bali. We found that around 50% of the total P in the corals is organic P, and both inorganic and organic P in the skeleton can correlate with seawater DIP variability. The Bali core was collected offshore of a major agricultural area in which artificial fertilizer use began in the mid-1970s. Total P/Ca in this core shows a large increase in 1974, and is thereafter strongly related to precipitation. This suggests that P/Ca in this coral records the history of agricultural fertilizer run-off. We used the Singapore coral to directly relate skeletal P/Ca to a contemporaneous seawater DIP record. Despite the overall low DIP concentrations and modest seasonal variability at this site, we found a significant correlation between total P/Ca and seawater DIP (r(2) = 0.42, P = 0.04, N = 10) after excluding highly oligotrophic periods (DIP < 0.050 mu mol/L, N = 4). Based on the global P/Ca-DIP calibration from multiple coral cores, seawater DIP reconstructions probably have an uncertainty of around +/- 0.115 mu mol/L, which is likely to decrease if more studies are undertaken. Considering just the propagated errors from analytical uncertainty and skeletal heterogeneity suggests that the precision within a coral core could be as low as +/- 0.056 mu mol/L, supporting the downcore application of the P/Ca proxy to track relative changes in DIP. We also estimate that the proxy is valid down to a lower boundary of coral P/Ca of roughly 6.5 mu mol/mol, indicating limitations to this proxy only under consistently oligotrophic conditions. Finally, we speculate that organic P in coral skeletons may derive partly from the passive inclusion of dissolved organic P molecules from seawater and partly from the coral holobiont, providing a possible explanation for why total P/Ca can correlate with dissolved phosphate concentration. (C) 2019 Elsevier Ltd. All rights reserved.