Link to home

Development of a Disease Risk Prediction Model for Downy Mildew (Peronospora sparsa) in Boysenberry

January 2014 , Volume 104 , Number  1
Pages  50 - 56

Kwang Soo Kim, Robert M. Beresford, and Monika Walter

First author: Department of Plant Science, Seoul National University, Seoul, 151-921, Korea, and Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 151-921, Korea; second author: Pathology and Applied Mycology, Mount Albert Research Centre, The New Zealand Institute for Plant & Food Research Ltd., Private bag 92 169, Mt. Albert, New Zealand; and third author: Pathology and Applied Mycology, The New Zealand Institute for Plant & Food Research Limited, 55 Old Mill Road, RD3, Motueka, New Zealand.

Go to article:
Accepted for publication 9 July 2013.

Downy mildew caused by Peronospora sparsa has resulted in serious production losses in boysenberry (Rubus hybrid), blackberry (Rubus fruticosus), and rose (Rosa sp.) in New Zealand, Mexico, and the United States and the United Kingdom, respectively. Development of a model to predict downy mildew risk would facilitate development and implementation of a disease warning system for efficient fungicide spray application in the crops affected by this disease. Because detailed disease observation data were not available, a two-step approach was applied to develop an empirical risk prediction model for P. sparsa. To identify the weather patterns associated with a high incidence of downy mildew berry infections (dryberry disease) and derive parameters for the empirical model, classification and regression tree (CART) analysis was performed. Then, fuzzy sets were applied to develop a simple model to predict the disease risk based on the parameters derived from the CART analysis. High-risk seasons with a boysenberry downy mildew incidence >10% coincided with months when the number of hours per day with temperature of 15 to 20°C averaged >9.8 over the month and the number of days with rainfall in the month was >38.7%. The Fuzzy Peronospora Sparsa (FPS) model, developed using fuzzy sets, defined relationships among high-risk events, temperature, and rainfall conditions. In a validation study, the FPS model provided correct identification of both seasons with high downy mildew risk for boysenberry, blackberry, and rose and low risk in seasons when no disease was observed. As a result, the FPS model had a significant degree of agreement between predicted and observed risks of downy mildew for those crops (P = 0.002).

© 2014 The American Phytopathological Society