Simulation models of potato late blight

Abstract

Plant disease epidemiology has its roots in the study of infectious diseases in man. The oldest mathematical model of a human disease stems from the 18th century (Bernoulli 1760), but it was not till two centuries later that mathematical analysis of plant diseases was initiated (Van der Plank 1963). Potato late blight featured prominently in Van der Plank’s analyses, and by now there is a rich literature of blight models. Before the mid 1980s blight models focused on the life cycle of the fungus, and on effects of the environment on the various stages in that life cycle. In recent years more emphasis has been given to the role of host growth in epidemic development and loss of yield. Sophisticated pathogen-crop combination models are now available.

Potato late blight modelling has served two purposes mainly. Firstly, blight models were used to evaluate strategies for disease control, especially the scheduling of fungicide application. Secondly, blight models have been used to explain differences in loss of yield among cultivars, and among various temporal and spatial patterns of disease development.

Many models have been made and their behaviour differs strongly. The sensitivity of the models to changes in parameters or inputs depends largely on the structure of the model. The models may be put to better use if more attention is paid to correct initialization and parameterization, and if comprehensive sensitivity analyses are carried out.