Early Blight Forecasting Systems: Evaluation, Modification, and Validation for Use in Fresh-market Tomato Production in Northern New Jersey

Abstract

Research trials, conducted from 1991 to 1998, evaluated early blight forecasting systems for use in fresh-market tomato (Lycopersicon esculentum) production in northern New Jersey. Initial trials focused on determining which of three forecast systems—NJ-FAST, CU-FAST, TOM-CAST—would optimize fungicide use. The TOM-CAST system generated fungicide application schedules that reduced foliar disease rating compared to the untreated check and, in 1 year, controlled diseases as well as a weekly schedule with 3 rather than 14 applications. TOM-CAST was easier to use than NJ-FAST or CU-FAST because it required fewer weather data inputs and simpler forecast calculations. Subsequent trials evaluated and defined thresholds for using TOM-CAST in northern New Jersey and evaluated the efficacy of several fungicides with TOM-CAST. Of the six TOM-CAST modifications evaluated, TOM-CAST beginning fungicide applications at 25 cumulative dew severity values (dew SV) and reapplying fungicide at 15 or 25 cumulative dew SV reduced disease rating as much as a weekly schedule in 1995 and 1996 and with fewer applications. After 5 years of trials, decision thresholds for using TOM-CAST in northern New Jersey were chosen and this new version of the forecast system designated NJ-TOM-CAST. It was verified in 1997 and 1998 and shown to generate fungicide application schedules that reduced foliar disease rating compared to the untreated check in both years and as much as the weekly schedule in one year. From 1995 through 1998, the conservative TOM-CAST schedules, TOM-CAST 25-15 or NJ-TOM-CAST, required on average 6 fungicide applications per year compared to weekly schedules that required on average 15 applications per year. In 1996, marketable yield was increased with TOM-CAST scheduled treatment compared to the untreated check and was the same as or greater than yield with weekly treatment. In the other 3 years, fungicide applications, whether applied on a calendar-based or TOM-CAST-based schedule, did not increase marketable yields compared to the untreated check. Fungicides shown to be effective when used with NJ-TOM-CAST schedules included both low cost and new chemistry materials. Copper fungicides, some of which are allowed in organic crop production, did not consistently control fungal diseases when applied on the NJ-TOM-CAST schedule. Applying fungicides on the NJ-TOM-CAST schedule instead of calendar-based schedules did not increase bacterial disease severity. Powdery mildew damage was more severe with NJ-TOM-CAST-scheduled applications than weekly applications in 1 year, affirming the importance of disease monitoring in the field when using NJ-TOM-CAST. By 2000, through a cooperative effort of Rutgers Cooperative Extension and SkyBit, Inc. (Boalsburg, Pa.), a commercial weather service, NJ-TOM-CAST was available to northern New Jersey tomato growers by subscription.