Simultaneous modeling of multiple end points in life-cycle toxicity tests.

Abstract

Standard toxicity tests do not allow extrapolations to the population level, mainly because these tests apply a short, fixed exposure time and focus on a single end point only. These limitations can be overcome by (partial) life-cycle toxicity testing, although these test results are harder to analyze. DEBtox is an existing software tool for the process-based analysis of standardized bioassays, and this paper presents two extensions of this method, making it applicable to life-cycle tests: the simultaneous assessment of end points and the description of aging (senescence) of the animals. We demonstrate these adaptations by describing life-cycle tests with the springtail Folsomia candida, exposed to cadmium and triphenyltin in their food. The extended model is able to describe the data for all end points simultaneously overtime with few, physiologically relevant parameters. Furthermore, the analysis reveals these chemicals to have distinctly different modes of action: cadmium apparently decreases the assimilation of energy from the food whereas triphenyltin increases the maintenance costs. The model fit allows calculation of the intrinsic rate of population increase, integrating effects on survival and reproduction. As the analysis is process based, population responses under food limitation can be explored, which depends critically on the selected mode of action.