Crude oils are natural substances derived from the degraded remains of organic matter and are typically complex and highly variable in nature. Toxicological assessment following an oil spill is challenging not only due to the variable chemical characteristics as a slick spreads and weathers, but also to the vulnerability of the affected ecosystem. Rapid decisions must be made to determine how to respond to the spill to protect the environment. Hence, a rapid assessment toolkit specific to the ecotoxicological assessment of crude oil and its components has been recommended. This thesis aimed to develop such a rapid assessment toolkit based on state of the art, environmentally relevant, diagnostic chemical and biological tools. Challenges to be addressed included the choice of test species, the dosing method for ensuring a consistent exposure concentration and the extrapolation of the data to predict species sensitivity and ecosystem relevance using computational modelling tools. To determine the applicability and comparative sensitivity of cyst-based zooplankton survival assays (using brine shrimp Artemia franciscana and marine rotifer Brachionus plicatilis) a novel, miniature passive dosing method was developed. Using three exemplar aromatic hydrocarbons, brine shrimp and rotifer were found to be within 51- 75% and 76–100% of most resistant species respectively when compared to 79 other species. When these species sensitivities were extrapolated using computational modelling (PETROTOX) to predict the relative toxicity of crude oil, they provided a conservative estimate when compared to empirical data derived from laboratory tests. In addition, comparison of novel, passively dosed water accommodated fraction preparation methods and traditional low energy, and chemically enhanced, water accommodated fraction methods showed both methods to be applicable to laboratory testing. Finally, investigation of cardiotoxicity as a non-invasive, sublethal biomarker was investigated as an endpoint potentially more sensitive than the whole organism cyst based tests. The sensitivity of three life stages (pediveliger eyed larvae, spat and adults) of Pacific oyster (Crassostrea gigas) was assessed using a novel combination of passive dosing and non-invasive video graphic and infrared methods. Illustrating both the utility of the methods and relative insensitivity of the organisms to phenanthrene as a model compound. Owing to the lack of sensitivity observed in these studies, it was not possible to recommend the biological test methods used above as a complete rapid assessment toolkit. However, the combination of chemical and computational tools showed effectiveness in reaching and maintaining exposure concentrations and providing a conservative estimate of toxicity. A synthesis of findings concludes with suggestions for future directions for the development of oil spill monitoring tools; the consideration of multi-stressor impacts; recommendations of a screening matrix for the future identification of sensitive bioindicators; and alternative endpoints of interest (e.g. indicators of membrane damage, the aryl hydrocarbon receptor and indicators of oxidative stress).

Assessing oil spill toxicity in real time is challenging due to dynamic field exposures and lack of simple, rapid, and sensitive tests. We investigated the relative sensitivity of two commercially available marine toxicity tests to aromatic hydrocarbons using the target lipid model (TLM). State of the art passive dosing in sealed vials was used to assess the sensitivity of brine shrimp (Artemia franciscana) and rotifer (Brachionus plicatilis). Organisms were exposed to toluene, 1-methylnaphthalene and phenanthrene for 24 h. Toxicity results were analysed using the TLM to estimate the critical target lipid body burden and support comparison to empirical data for 79 other aquatic organisms. Our findings demonstrate the applicability of passive dosing to test small volumes and indicate that the two rapid cyst-based assays are insensitive in detecting hydrocarbon exposures compared to other aquatic species. Our results highlight the limitations of applying these tests for oil pollution monitoring and decision-making.

Oil spills of varying magnitude occur every year, each presenting a unique challenge to the local ecosystem. The complex, changeable nature of oil makes standardised risk assessment difficult. Our review of the state of science regarding oil's unique complexity; biological impact of oil spills and use of rapid assessment tools, including commercial toxicity kits and bioassays, allows us to explore the current issues preventing effective, rapid risk assessment of oils. We found that despite the advantages to monitoring programmes of using well validated standardised tests, which investigate impacts across trophic levels at environmentally relevant concentrations, only a small percentage of the available tests are specialised for use within the marine environment, or validated for the assessment of crude oil toxicity. We discuss the use of rapid tests at low trophic levels in addition to relevant sublethal toxicity assays to allow the characterisation of oil, dispersant and oil and dispersant mixture toxicity. We identify novel, passive dosing techniques as a practical and reproducible means of improving the accuracy and maintenance of nominal concentrations. Future work should explore the possibility of linking this tiered testing system with ecosystem models to allow the prediction and risk assessment of the entire ecosystem.