Modeling the concentration–response function of the herbicide dinoseb on `Daphnia magna' (survival time, reproduction) and `Pseudokirchneriella subcapitata' (growth rate)

Models describing dose–response relationships are becoming increasingly popular in ecotoxicology. They allow simple and thorough evaluations of toxicity test results, including inter- and extrapolations to concentrations or exposure times other than those tested. Simple parametric regression models are of particular interest because their parameters may be attributed mechanistic meanings and they can be applied without sophisticated mathematical and computational support. We recently proposed a four-parameter logistic regression model to fit the survival data of `Daphnia magna' under dinoseb stress. The model parameters are the maximum survival time, the minimum time required for an individual to die, effect concentration, EC50, and a curve shape parameter. This model has now been applied to compare the lethality and reproduction toxicity of `D. magna' and the growth inhibition of `Pseudokirchneriella subcapitata' under dinoseb stress. It can be fitted adequately to all the measured data and the parameters can be attributed biological meanings in any of the three endpoints. A comparison of the modeled concentration-response functions of all three endpoints for dinoseb toxicity shows that the range of ECs with respect to both `D. magna' and algae is steep (a decrease of between 0.1 and 0.6 mg/L). The survival and reproduction of `D. magna' exhibit similar characteristic concentration–response functions and toxicities. The statistical no-effect concentration (SNEC) is 0.14 (survival) and 0.11 (reproduction) mg/L, respectively. On the other hand, algae seem to be less sensitive to dinoseb than `D. magna', (SNEC: 0.48 mg/L). However, further investigations of individual algae may lead to a more suitable comparison. We speculate that the four parameters of the model function can be related to specific properties of chemicals and organisms. Characterization of these properties would allow simple and appropriate estimation of the toxic effects of these chemicals.