Contaminant chemistry and toxicity of sediments in Sydney Harbour, Australia: spatial extent and chemistry–toxicity relationships

Abstract

The spatial distribution of chemical contamination and toxicity of surficial sediments in Sydney Harbour, Australia, was investigated in a 3-tiered, hierarchical approach. An initial chemical investigation throughout the entire estuary (Stage 1) indicated wide ranges and different spatial patterns in sediment chemical concentrations. Sediment quality guidelines (SQGs) were used as a preliminary estimate of possible toxicity in Stage 2 of the investigation. Assessment of chemical mixtures indicated that sediments in a small part (~2%) of the harbour had the highest probability of being toxic (~75%), whereas sediment in almost 25% of the port was estimated to have an inter- mediate (~50%) probability of being toxic. The SQG assessment in Stage 2 enabled careful stratifica- tion of the harbour into areas with different toxicity risks, reducing cost and time commitments in the final tier of assessment. The spatial survey carried out in Stage 3 involved concurrent chemical and ecotoxicological analyses. In this final stage, the degree of response in tests of amphipod survival in whole sediment samples, as well as in tests of microbial metabolism (Microtox©) and sea urchin egg fertilisation and embryo development in pore waters, generally increased with increasing chemical concentrations. However, amphipod response was lower than predicted due to relatively low sensi- tivity of the indigenous amphipod Corophium colo. Areas initially predicted, using SQGs, to be most at risk were highly toxic in the combined chemistry-toxicity investigation, while sediment from areas with the lowest predicted risk were least toxic, but still toxic in at least one ecotoxicological test. The results demonstrate that the empirical approach used for this study, which was originally developed in North America, produced plausible outcomes elsewhere and that observed toxicity, based on SQGs, matched predictions using different species but similar methodologies.