Significant Influence of EC₅₀ Estimation by Model Choice and EC₅₀ Type

Abstract

The effective control to 50% growth inhibition (EC₅₀) is a standard statistic for evaluating dose-response relationships. Many statistical software packages are available to estimate dose-response relationships but, recently, an open source package ("drc") in R has been utilized. This package is highly adaptable, having many models to describe dose-response relationships and flexibility to describe both hormetic relationships and absolute and relative EC₅₀. These models and definitions are generally left out of phytopathology literature. Here, we demonstrate that model choice and type of EC₅₀ (relative versus absolute) can matter for EC₅₀ estimation using data from Pythium oopapillum and Fusarium virguliforme. For some P. oopapillum isolates, the difference between absolute and relative EC₅₀ was significant. Hormetic effects changed F. virguliforme EC₅₀ distributions, leading to higher estimates than when using four- or three-parameter log-logistic models. Future studies should pay careful attention to model selection and interpretation in EC₅₀ estimation and clearly indicate which model and EC₅₀ measure (relative versus absolute) was used. We provide guidelines for model choice and interpretation for those wishing to set up experiments for accurate EC₅₀ estimation.