Pharmacokinetic modelling of multi-decadal luminescence time series in coral skeletons

Abstract

As corals grow, they incorporate chemical indicators of seawater conditions into their aragonite skeleton after they have traversed an outer living tissue layer. Long-lived, massive coral skeletons can record decade- and century-long time series of seawater status. One such environmental clue is luminescence intensity which can correspond to river flow patterns and has been attributed to humic acid incorporation. Seawater humic acid levels are linked to river flow as rainfall extracts them from catchment soils to then flow into rivers and coastal seas. However, discrepancies exist when validating coral luminescence records against river flow data with intense luminescence sometimes occurring in the absence of increased flows. This contributes to uncertainty when reconstructing pre-instrumental river flows and rainfall from coral luminescence. Here we demonstrate that a major portion of coral core luminescence time series can be explained using a single-compartment, pharmacokinetic model that incorporates river flow measurements as the equivalent of drug dose. The model was robust for luminescence series in corals from near-shore reefs regularly influenced by river flow. The model implies that after floods, a proportion of subsequent luminescence peaks can be derived from the initial flood. This explains why some luminescence peaks after floods often do not correspond to additional significant river flows. This provides the first mechanism-based explanation for temporal changes in coral skeleton luminescence that incorporates a mathematical link between two independent time series making this proxy even more robust for reconstructing river flow and rainfall.