Improved prediction of sediment toxicity using a combination of sediment and overlying water contaminant exposures

Abstract

The choice of sediment quality assessment methodologies can strongly influence assessment outcomes and management decisions for contaminated sites. While in situ (field) methods may potentially provide greater realism, high costs and/or complex logistics often prevent their use and assessment must rely on laboratory-based methods. In this study, we utilised static-renewal and flow-through ecotoxicology tests in parallel on sediments with a wide range of properties and varying types and concentrations of contaminants. The prediction of chronic effects to amphipod reproduction was explored using multiple linear regression (MLR). The study confirmed the considerable over-estimation of the risk of toxicity of contaminated sediments in field locations when assessments rely on the results of laboratory-based static and static-renewal tests. Improved prediction of toxicity risks was achieved using a combination of contaminant exposure measures from sediment and overlying water. Existing sediment and water quality guideline values (GVs) were effective for predicting risks posed by sediments containing mixtures of common metal and organic contaminants. For 17 sediments with paired data sets from static-renewal and flow-through tests, the best prediction of toxicity to reproduction was achieved using a 2-parameter MLR that included hazard quotients for sediment contaminants and toxic units for dissolved metals (r² = 0.892). The inclusion of particle size, organic carbon and acid-volatile sulfide did not improve toxicity predictions, despite these parameters being recognised as modifying contaminant bioavailability. The use of dilute-acid-extractable metal concentrations in place total recoverable metal concentrations did not improve the predictions. The study also confirmed that sediments existing within the estuarine and marine bays of Sydney Harbour pose significant risks of adverse effects to benthic organisms.

Keywords: Benthic invertebrates; Chronic toxicity prediction; Field-based assessments; Multiple linear regression; Risk assessment; Sediment quality.