|
(069) Acute to chronic extrapolation of LCx by a mechanistic model of toxicity. Duboudin, Cédric1, Bonnomet, Vincent1,2, Magaud, Hélène2, 1 2 ABSTRACT- To understand and predict the temporal evolution of the toxicity, a mechanistic model was developed and fitted to real data. Derived from Bedaux and Kooijman's approach (Bedaux and Kooijman, The analysis of aquatic toxicty data, VU Univ. Pr. Amsterdam, 1997), it considers the hazard rate as a function of the toxic concentration in the organism. By linear approximations, LCx (concentration that induces x% of lethality), and in particular the LC50, can be expressed analytically as a function of time. Only three parameters (expressing an asymptotical effect concentration, a time constant and an effect velocity) are necessary to express this model. It can be demonstrated that the linear relation between LCx and 1/t empirically proposed by Mayer et al. (Mayer et al., Environ. toxicol. chem. 13(4), 671-678, 1994) is a good approximation of our model for many cases. In addition, with a more sophisticated (but still analytic) model considering a time dependency on the asymptotical effect concentration, some complex toxicity patterns can be interpreted mechanistically and taken into account such as, for example, toxic induced aging or adaptation. Some cases were fitted from LC50 for various freshwater species and diverse chemicals (organics and inorganics). The fitted parameters differ according to chemicals and species. This obviously suggests differences in sensibilities to toxicants. The interpretation of the parameters values could help to explain the toxicity process. The parameters of the models also reflect the influence of the water quality on the toxicity of some substances to organisms. Different toxicity processes can be characterized by such models and this can lead to a better extrapolation from acute to chronic toxicity. Key words: toxicity, modelling, acute to chronic |
