Increasing transparency in the implications of variability in contaminant partitioning

Abstract

Porewater concentrations of contaminants in aquatic sediment (Cfree) are often measured with passive sampling methods (PSM) to quantify the true contaminant activity more accurately. This Cfree correlates more closely with uptake and toxicity in benthic organisms than solid-phase contaminant concentrations (Ctotal) or model predictions of Cfree from Ctotal, and thus represent a better indicator of bioavailability and risk. While sediment managers rely on accurate measures of Cfree to estimate risk, they also rely on Ctotal as the basis for defining and monitoring clean-up goals for restoration and remediation. In real environments, the partitioning of contaminants between Cfree and Ctotal is highly variable among samples collected, even those in close proximity. The variability is due to intractable differences in adsorption capacity among different carbon phases; however, this variability can provide a quantitative basis for converting between Ctotal and Cfree in a stochastic approach. The stochastic approach can be used to estimate the likelihood that Ctotal would exceed a Cfree-based biological threshold in the case in which Cfree was not measured (or vice versa). When the stochastic approach to bioavailability is implemented at the beginning of the risk assessment process, screening-level evaluations can be refined and possibly reduce the number of sediments in which additional testing (e.g. toxicity testing) is required to elucidate risk. This dissertation uses field-collected measurements and statistical modeling to illustrate how PSM-measured Cfree and a stochastic view of contaminant partitioning can provide a more nuanced way of understanding the implications of variability in contaminant partitioning. This view can support more transparency in decision making at contaminated sediment sites.