Beyond Thresholds: A Holistic Approach to Impact Assessment Is Needed to Enable Accurate Predictions of Environmental Risk from Oil Spills.

The risk assessment for the environmental impact of oil spills in Australia is often conducted in part using a combination of spill mapping and toxicological thresholds derived from laboratory studies. While this process is useful in planning operational responses, such as where to position equipment stockpiles and whether to disperse oil, and can be used to identify areas near the spill site where impacts are likely to occur, it cannot accurately predict the environmental consequences of an oil spill or the ecosystem recovery times. Evidence of this disconnect between model predictions and observed impacts is the lack of a profound effect of the Deepwater Horizon wellhead blowout on recruitment to fisheries in the northern Gulf of Mexico, contrary to the predictions made in the Natural Resources Damage Assessment and despite the occurrence of impacts of the spill on marine mammals, marshes, and deep water ecosystems. The incongruity between predictions made with the current approach using threshold monitoring and impacts measured in the field results from some of the assumptions included in the oil spill models. The incorrect assumptions include that toxicity is acute, results from dissolved phase exposure, and would be readily reversible. The toxicity tests from which threshold models are derived use members of the ecosystem that are easily studied in the lab but may not represent the ecosystem as a whole. The test species are typically highly abundant plankton or planktonic life stages, and they have life histories that account for rapid changes in environmental conditions. As a consequence, these organisms recover quickly from an oil spill. The interdependence of ecosystem components, including the reliance of organisms on their microbiomes, is often overlooked. Additional research to assess these data gaps conducted using economically and ecologically relevant species, especially in Australia and other understudied areas of the world, and the use of population dynamic models, will improve the accuracy of environmental risk assessment for oil spills. Integr Environ Assess Manag 2020;16:813-830. © 2020 SETAC.