Analysis of bioassays data from experiments with mixtures require use of special statistical techniques, especially when responses are recorded over time. Often, response curves from bioassays with mixtures do not fit a probit or logit curve. In addition, observations recorded over time are not independent. As a result, standard statistical packages for probit or logit analysis are not applicable. In these bioassays, the shape of the response curve depends on the type of interactions between mixture components. Here we present models that describe response (e.g., mortality) curves over time when mixtures are applied to insects. The shapes of the response curves depend on the modes of both action and interaction between the mixture compounds. We describe a procedure to assess whether the assumed modes of interaction (e.g., independent action or similar action additive models) are appropriate. We demonstrate use of the statistical methods with data obtained when a baculovirus and a synthetic organic insecticide were mixed.
All Science Journal Classification (ASJC) codes
- Insect Science