Development and Validation of an Infection Risk Model for Bacterial Canker of Kiwifruit, Using a Multiplication and Dispersal Concept for Forecasting Bacterial Diseases.

A weather-based disease prediction model for bacterial canker of kiwifruit (known worldwide as Psa; Pseudomonas syringae pv. actinidiae biovar 3) was developed using a new mechanistic scheme for bacterial disease forecasters, the multiplication and dispersal concept. Bacterial multiplication is estimated from a temperature function, the M index, accumulated from hourly air temperature over 3 days for hours when the leaf canopy is wet. Rainfall provides free water to move inoculum to infection sites, and the daily risk indicator, the R index, is the 3-day accumulation of the M index output on days with total rainfall >1 mm; otherwise, R is zero. The model was field-tested using potted kiwifruit trap plants exposed for discrete periods in infected kiwifruit orchards to identify when leaf infection occurred. In a 9-week study during spring, the R index predicted leaf-spot intensity with high accuracy (R2 = 93%) and, in an 82-week seasonal accuracy study, prediction of infection incidence was most accurate from spring to late summer and lower during other times. To implement the risk model for the New Zealand kiwifruit industry, a modified risk index, R', used relative humidity (RH) >81% instead of wetness, so that 2- and 6-day weather forecasts of RH could be used. Risk index values were affected by the shape of the temperature function and an alternative 'low temperature' function for the M index was identified that could be used in climates in which high temperatures are known to limit Psa development during some parts of the year. This study has shown how infection risk for bacterial diseases can be conceptualized as separate processes for temperature-dependent bacterial multiplication and rain-dependent dispersal and infection. This concept has potentially wide application for bacterial disease prediction in the same way that the infection monocycle concept has had for fungal disease prediction.