3.1.14
FORECASTING SCLEROTINIA STEM ROT

E TWENGSTROM, R SIGVALD, C SVENSSON and J YUEN

Swedish University of Agricultural Sciences, PO Box 7044, SE-750 07 Uppsala, Sweden

Background and objectives
Sclerotinia stem rot, caused by Sclerotinia sclerotiorum, is a major disease on spring-sown oilseed rape in Sweden. The disease can be efficiently controlled by a fungicide application during flowering. The treatment is expensive and has to be applied before any symptoms are visible. The severity of Sclerotinia attacks varies between years, regions and fields, and to minimize the use of fungicides with maintained profitability for the farmer, a reliable forecasting method is needed.

Materials and methods
During the years 1984-94, approximately 80 fields have been monitored annually for signs of Sclerotinia attacks. Field-specific data such as crop rotation, level of previous Sclerotinia attacks in the field, crop density, fertilization and flowering period have been collected. In addition, weather data from the field or the nearest meteorological station have been registered. The regional risk of apothecia development was estimated from apothecia production from buried scierotia in five to ten fields per district. Naturally produced sclerotia were placed in rapeseed fields immediately after sowing and apothecia development was monitored weekly during spring and summer. The relative importance of different field and weather parameters on disease level has been evaluated by logistic regression with the SAS Procedure GENMOD [1]. To validate the forecasting model, field data and weather parameters were collected from 78 fields during 1996 and 1997. The risk of Sclerotinia infection in these fields was estimated during full flowering and compared to the final disease levels.

Results and conclusions
The most important factors affecting disease incidence were level of previous Sclerotinia attacks in the field, regional risk, crop density, crop rotation, precipitation and weather forecast during flowering. Based on the statistical analysis, the different factors were given points with regard to the risk for disease incidence exceeding the economic threshold. In the forecasting model, the risk points are added and the sum is compared to predetermined threshold values. The need of fungicide sprays could be accurately predicted for fields with high and low risk points. For fields with intermediate risk points predictions were more uncertain.

By using the new forecasting model, the farmers can better predict the need of spraying against Sclerotinia stem rot in individual fields. Use of fungicides can be optimised by spraying only when necessary and any adverse impact on the environment can be limited while maintaining yield level and quality.

References
1. Yuen J, Twengstrom E, Sigvald R, 1996. European Journal of Plant Pathology 102, 847-854.