Temperature-dependent toxicity: is an assessment factor of 10 appropriate for ecological risk assessment?

Abstract

Laboratory derived toxicity data provide important information for ecological risk assessment (ERA) of pollutants and form the basis for deriving water quality criteria. However, toxicity tests are often conducted following standard methods and/or guidelines (e.g., ‘U.S. EPA Ecological Effects Test Guidelines’ and ‘OECD Guidelines for Testing of Chemicals’), and thus carried out under controlled laboratory conditions. For instance, most of toxicity tests are run at a constant temperature which usually represents the optimum temperature for the test species. Yet, toxicity of a contaminant may vary in a temperature-dependent manner, depending on the physiology of the test organism and the chemical properties of the contaminant. Even though an assessment factor of 10 is often used to account for the variability of toxicity data in ERA, no one has investigated the appropriateness of this value to cover the temperature-dependent variability adequately. Here, we compared median lethal concentration (LC50) data of several trace metals and organic pollutants on a range of aquatic species at different temperatures to assess whether an assessment factor of 10 is under- or over-protective for ERA. Our preliminary results indicated a considerable variability of copper toxicity with temperature across different freshwater species. For each species, we used the temperature at which the LC50 is the highest as a reference temperature (i.e., LC50ref) and the corresponding LC50 value as 1 reference toxic unit (TUref). The relative difference in the copper toxicity to the same species at a different temperature (T) is expressed as a relative TU value (RTU) which is equal to LC50T/LC50ref. A smaller RTU indicates a higher toxicity. For copper, we detected an average decrease of 0.139 TU/ºC increase in temperature and a decrease of 0.076 TU/ºC decrease in temperature. The results imply that an assessment factor of 10 will cover up to 6.1ºC increase and 11.8ºC decrease in temperature for copper, translating into a maximum protective range of 17.9ºC temperature variation. Global surface freshwater temperature was found to have an average variation of 11.7±1.4oC. Thus, application of a factor of 10 in this case might be adequately protective to freshwater organisms. In this presentation, we will present more results on other chemicals before drawing a solid conclusion.