|
(W/EH104) Evaluation of QSAR and models for the screening of potentially bioaccumulative substances. Arnot, Jon1, Gobas, Frank1, Kelly, Barry1, 1 ABSTRACT- Several countries in Europe, the US, Canada and international organizations (e.g. United Nations) are evaluating large numbers of chemical substances on their potential to be Persistent, Bioaccumulative and Toxic in the environment. Current policies use Bioaccumulation factors (BAF), Bioconcentration factors (BCF) and n-octanol/water partition coefficients (logKow) to assess the bioaccumulation potential of a substance. For the majority of chemicals that are to be screened, there is a lack of empirical BAF and BCF data. For these substances, the application of QSARs and bioaccumulation models are necessary tools to assess the bioaccumulation potential. In this paper, we evaluate the behaviour of a series of QSARs and food-chain bioaccumulation models to assess the BAF of organic substances. Predictions of the QSARs and models are compared to a large BAF and BCF database including more than 2,500 observations. The analysis illustrates (i) the role of experimental artifacts in the assessment of the bioaccumulation potential, (ii) the underestimation of BCF-QSARs of bioaccumulation in the field and (iii) a generic BAF-QSAR that can be used to assess the BAF of chemicals if metabolic transformation is absent. We further illustrate that log Kow is not an adequate property for assessing the bioaccumulation potential of substances in terrestrial food-chains. We propose a bioaccumulation QSAR for organic substances in terrestrial food-chains that is based on the octanol-air partition coefficient. It is concluded that (i) a Kow based BAF-QSAR provides a reliable assessment of the bioaccumulation potential of organic substances in aquatic food-chains and (ii) a Koa based bioaccumulation-QSAR may be a useful for assessing the bioaccumulation potential of organic substances in terrestrial food-chains. Key words: qsar, marine, terrestrial, bioaccumulation |
