Predicting hormesis in mixtures of herbicidal compounds – where are we and how far can we go?

Predicting the occurrence and expression of stimulatory effects of subtoxic doses of phytotoxins or herbicides (hormesis) in mixtures is a challenging and needed task, considering that herbicide exposures in practice often occur in mixtures at low doses due to drift deposition, errors in application, protection by mulch, herbicide resistance, small-scale dose heterogeneity, and other causes. While joint effects in toxin mixtures can be straightforwardly modelled and predicted at toxic doses, the evaluation at stimulatory doses lacks a common statistical approach. Prediction of effective hormetic doses can be adequately facilitated by adopting jointaction models that have been developed for monotonic responses. In contrast, prediction of the magnitude of hormesis as one of the key quantitative features of hormesis is not so easy. Currently, there are no mechanistic models available that could be adopted to predict the hormetic magnitude in mixtures nor is there a generally accepted model available. Nevertheless, some promising attempts were made to predict the hormetic magnitude in herbicidal mixtures demonstrating the fundamental possibility of modelling hormesis in mixtures and providing valuable insights into the phenomenon. The success of these attempts is summarized and future research needs and limits are discussed.