Adaptation of the New Zealand Psa risk model for forecasting kiwifruit bacterial canker in Korea

Abstract

Bacterial canker of kiwifruit, caused by Pseudomonas syringae pv. actinidiae (Psa), is a severe threat to kiwifruit production in Korea. An existing infection risk model from New Zealand was adopted to respond to this epidemic. Disease incidence (proportion of diseased leaves on each vine) and weather (hourly temperature, rainfall and relative humidity) data required to develop the model were collected and analysed in the study. Disease incidence data were used to modify and validate the existing model. Because the Psa risk model was originally developed in a region where the characteristic climatic conditions are completely different from those in Korea, the temperature and rainfall functions of the existing model were modified. Analyses using statistical correlation and prediction–realization tables revealed that the modified model is valid with high agreement (a correlation coefficient of 0.85 and an accuracy of 85.7%, respectively) between the observed disease incidence and simulated disease risk from the model. The model was also found to be more highly sensitive to the presence or absence of rainfall than any other weather variable inputs. Uncertainty in simulated disease risk was measured based on the level of uncertainty in temperature input from weather forecasts. Overall, these results indicate that the modified Psa risk model can be used to provide practical and applicable information for timely disease control to the kiwifruit growers in Korea.