Joint action of chemicals in algal toxicity tests: influence of response level and dose-response regression model.

The joint toxicity of nonylamine and decylamine and of atrazine and decylamine was evaluated in assays with the green alga Selenastrum capricornutum based on an isobologram method. In this method, curves of constant response, isoboles, are plotted versus concentrations of two toxicants. The response parameter was growth rate based on biomass, and several response levels were used. Dose-response curves were developed for dilution series using fixed ratios between concentrations in toxic units of the compounds. Probit and Weibull dose-response curves were then determined by nonlinear regression. A model for isoboles for partially similar action was used when applicable. The no-effect concentration (NEC or EC0) was estimated based on a newly proposed model containing median effective concentration (EC50) and EC0 as explicit variables. Results show that nonylamine and decylamine are nearly concentration additive at EC50 and EC10 (similarity parameter lambda = 0.70-0.76) and to a lesser extent at EC0. By contrast, the mixtures of atrazine and decylamine show antagonism in that atrazine acts as an antidote to decylamine. The shapes of these isoboles are independent of response level. The EC50 values (mg/L) for chemicals acting singly were 0.090 (nonylamine), 0.039 to 0.044 (decylamine), and 0.225 (atrazine). In order to determine NEC effectively, the level of inhibition must be fairly low, with observed growth rates between 0.6 and 1.0 times the average growth rate of the controls.