University of Missouri-Columbia, Columbia, Missouri 65211, USA

Background and objectives
Environmental monitoring has been used to forecast infection periods for plant diseases. Mills [1] monitored air temperature and leaf wetness in order to predict infection periods for the apple scab pathogen Venturia inaequalis, and Jones et al. [2] developed a computerized system to monitor environmental conditions and predict apple scab infection periods. Later, Neogen Corp. developed the EnvirocasterTM to predict infection periods for several plant diseases. However, these instruments cost from US$3000 to 6000 and growers were reluctant to buy them. The objectives of this study were: (i) to develop a moderately priced automated system to monitor environmental conditions and forecast infection periods for many plant diseases; and (ii) to determine the usefulness of this system to accurately time sprays and reduce pesticide use.

Materials and methods
A computerized weather station was developed that monitors air temperature, relative humidity, leaf wetness, rainfall, soil temperature, wind speed and wind direction. Data are transmitted, via radio telemetry, every 2-5 min to an IBM-compatible base computer using a Microsoft Windows operating system. Software has been developed that collects weather data from the remote weather stations and predicts infection periods for many plant diseases. These include apple scab, cedar apple rust, canker fire blight, blossom fire blight, apple blotch, pear scab, Cougar blight, grape black rot, grape Botrytis bunch rot, grape and cherry powdery mildew, wheat scab, Stewart's wilt of corn, alfalfa Sclerotinia rot, turfgrass brown patch, and Pythium blight of turfgrass. A model has also been developed to predict codling moth activity.

Results and conclusions
Predicted post-infection fungicide sprays were compared with standard calendar-timed pesticide sprays, to compare their efficacy. Post-infection fungicide sprays used to control apple scab allowed a 21-50% reduction in the number of applications compared to the standard calendar-timed applications. Control was equally good in both cases. Likewise, sprays for cedar apple rust and streptomycin sprays for blossom fire blight were reduced 33% using the predicted schedule. Recent tests indicated that the use of plant disease forecasting allowed the reduction of fungicide applications for sooty blotch of apple.

1. Mills WD, 1944. NY Agricultural Experiment Station, Ithaca, Bulletin 630, 4.
2. Jones AL, Lillevik SL, Fisher PD, Stebbins TC, 1980. Plant Disease 64, 69-72.