Plant disease models are a key part in the implementation of IPM (Integrated Pest Management), as they provide useful information for deciding whether and when crop protection actions must be implemented.
Prediction models can either be empirical or mechanistic. Empirical models are based on the identification of mathematical or statistical relationship in field-collected data, but these relationship do not necessarily express a cause-effect linkage. In order to be used in contexts different from the ones they were developed in, this kind of models require accurate validation and adaptation. In recent years, several analysis techniques based on big data and artificial intelligence have been developed, but this does not overcome the intrinsic weaknesses of empirical models, which are mainly the lack of knowledge, accuracy, and robustness.
On the other hand, mechanistic models are based on the study and modellisation of biological processes of plants and pathogens, linking them to external weather variables by mean of mathematical equations. Mechanistic models are dynamic, as they analyze changes in the components of an epidemic over time, characterizing the state of a pathosystem in every moment quantitatively. Mechanistic models are characterized by a greater accuracy and robustness if compared to empirical models. Mathematical model are a useful tool to improve pest and disease management in crops, as they can serve for a better timing of treatment application, leading to an increase efficacy and reduced number of interventions.