Causal Inference in Plant Disease Contexts

Many questions in plant pathology are of a causal nature. However, the formal framework of causal inference remains underutilized in the discipline. This review provides an accessible, higher-level introduction to causal concepts for agricultural researchers. The potential outcomes framework is introduced to define counterfactuals and directed acyclic graphs to visualize causal assumptions. We discuss the "gold standard" randomized controlled trial for estimating causal effects, with attention to agriculturally relevant contexts such as interplot interference and the estimation of heterogeneous treatment effects. The relevance and application of causal inference methods to observational data are also covered, in recognition of the growing role of this type of data. The limited but pioneering studies already applying these methods in plant pathology are highlighted.