Large-scale fuzzy rule-based prediction for suitable chestnut ink disease sites: a case study in north-east Italy.
In the past few decades, economic interest in the cultivation of chestnuts for both timber and nut production has resurfaced in the Mediterranean area. However, chestnut cultivation has suffered in recent years from the spread of exotic pests, such as the gall wasp Dryocosmus kuriphilus, and from the resurgence of previously present diseases, most likely due to anomalous climate dynamics. This is the case with chestnut ink disease, caused by the soilborne pathogens Phytophthora cinnamomi and P. cambivora. Scientific and technical support in monitoring and management, that utilizes new forecasting approaches incorporating related environmental variables, is therefore essential. The main aim of this study was to develop a mathematical model assessing the potential for the establishment of chestnut ink disease at a large scale. Towards this goal, fuzzy rule-based theory was applied to the environmental variables associated with host presence, pathogens' ecological niches and ink disease symptoms expression. The effectiveness of the rule-based modelling outcomes, provided with uncertainty maps to facilitate their correct interpretation, was confirmed by detailed field data collected from a large chestnut-growing area where ink disease has been increasing in recent years. The final model gave consistent predictions for disease presence. For this reason, it represents a flexible and valuable decision-support tool to forecast which sites are at risk from CID.