Tomato Late Blight Forecasts: True Forecasting with Adaptation to Disease Management Practices 1998
Project Leaders: Robert Seem, Helene Dillard
Implement and evaluate BLITECAST / E-Weather disease forecast system for tomato late blight.
a. Establish transitional criteria for shifting from TOMCAST recommendations to BLITECAST recommendations, with and without the confounding influence of bacterial diseases.
b. Compare accuracy of E-Weather forecasts to the weather conditions monitored at each cooperating grower site, and assess the user friendliness of the BLITECAST delivery.
c. Evaluate fungicide program efficiency for tomato late blight control, with special reference to new P. infestans genotypes.
d. Monitor tomato crops canopy to parameterize moisture balance model for improved surface wetness forecasts.
No late blight was observed at any of the four project locations. However, at two of the locations (Chemung and Ontario Counties) the switchover from early blight control was initiated when late blight was confirmed in the area. The SkyBit late blight forecast was not satisfactory for estimating infection periods on tomato under our test conditions. The forecast system made consistent warnings of infection from the beginning of May for all locations. This was partly due to the fact that the E-weather forecast appeared to overestimate the weather conditions that are crucial to late blight forecasts (relative humidity and surface wetness duration), thus creating estimates for infection more frequently than it actually occurred. It is also possible that the biology incorporated into the modification of BLITECAST did not hold up under field conditions. Finally, the model assumes that inoculum is always present, a situation that did not occur in at least two of the four test locations. Additional work on the models (weather and disease forecast) must be carried out before this system of disease management can be considered for commercial implementation.