Individual sensitivity distribution evaluation from survival data using a mechanistic model : implications for ecotoxicological risk assessment
Résumé
In the literature, there are two main alternatives to model mechanistically dose-survival relationship in ecotoxicity tests. Effects are related to a concentration of concern, for instance body concentration, and either individuals have different concentration thresholds for death, or individuals have equal probability to show an effect, with dying organisms just being the 'unlucky'ones. A general framework to unify both approaches has recently been proposed but only special cases could be confronted to actual data. We used such a unified model to analyse four datasets. We showed the possibility to estimate properly the toxicity parameters together with inter-organisms differences of sensitivity for at least one of these parameters (here the threshold for effect). For all datasets, the model we propose outperformed the standard approach without accounting for differences of sensitivity. More accurate parameters' estimations would be obtained through the incorporation of prior knowledge, in particular relative to background mortality. We also pointed that the standard threshold estimate is close to the medium value of the distribution (which would induce the death of half of the organisms), and that more relevant parameters; such as the concentration protecting 95 % of the population would be 2 to 4 times inferior to this standard threshold. We believe that our approach is easy to understand and to apply by non-experts in modeling. Moreover, it permits to relate two fundamental concepts in understanding the response in ecotoxicity tests : the link between internal concentration and effects and the inter-individual variability regarding sensitivity.