Volume 47 Number 1 Spring 2000
Peanut Disease Advisory Now Contains Rainfall Estimates
Austin Hagan, Kira Bowen, Ellen Bauske, Roger Getz, and Steve Adams
The days of spraying on a regular schedule for early and late leafspot may not be totally gone, but the common use of highly efficient, reduced-rate fungicides has made these blanket treatments obsolete. Researchers have long known that leafspot strikes a peanut field only under hot, wet conditions. Now, AAES researchers have refined a system that takes into consideration daily rainfall in a peanut field, making timely application of the more expensive, new fungicides more economical and efficient.
Control of early and late leaf spot as well as white mold is critical for producing high peanut yields. Currently, fungicides and application costs can easily account for 20% or more of the variable costs needed to grow an acre of peanuts. Weather-based expert systems have been developed to improve the timing of fungicide sprays. Research has shown that by adopting weather-based spray advisories such as AU-Pnut, fungicide spray numbers and disease control program costs can be reduced without jeopardizing peanut yield or crop quality. With the possible exception of irrigated peanuts, fewer than the seven standard fungicide sprays, applied every 14 days, are often needed to maintain effective disease control, particularly during extended periods of hot, dry weather.
The AU-Pnut leaf spot advisory was developed nearly a decade ago by AAES researchers to improve the timing of fungicide sprays needed to control early and late leaf spot on peanut. Typically, use of this expert system saves one to three leaf spot fungicide sprays per year. Also, Folicur 3.6F applications, which were scheduled using the AU-Pnut advisory, controlled white mold. However, adoption of AU-Pnut has been limited in part by the need to record total daily rainfall in widely scattered peanut fields.
Doppler radar (WSR-88D) offers a near real-time method of identifying daily rain events of 0.1 inch, which are needed to implement the AU-Pnut leaf spot advisory. Doppler radar data can be compiled to generate rainfall estimates on a 1.2 x 1.2-mile grid. Individual fields may be located within the appropriate grid square using latitude and longitude coordinates generated by a GPS unit.
AAES researchers have investigated the integration of Doppler radar rainfall estimates into the AU-Pnut leaf spot advisory. In addition, private sector cooperators on this project have developed an Internet-accessible version of AU-Pnut, along with supporting data management sofware.
To verify the occurrence of rain events by Doppler Radar, automated weather stations, which monitor temperature and rainfall, were installed in March 1998 near three peanut fields in Henry County. Two of the fields (the Williams site and the Murphy site) had a history of frequent peanut production while at the Trawick Farm site peanuts followed bahiagrass. The Williams field, which was planted with the ViruGard peanut, was irrigated as needed. During the first week of May, Georgia Green peanut was planted in the two non-irrigated fields. On the Trawick Farm, Georgia Green was planted on twin rows. Peanuts were planted on single 36-inch rows at the other sites. At all sites, seeding rates, tillage practices, soil fertility, and weed control were followed as recommended by the Alabama Cooperative Extension System.
Each morning, the archived Doppler Radar images and rain gauge data were reviewed to determine whether a rain event had occurred within the preceding 24 hours. Also, the daily and five-day precipitation forecast, which was provided by private sector cooperator AWIS Inc., was compared with the total number of rain events tallied to determine whether to issue a spray advisory based on the Doppler Radar or rain gauge data for any of the test sites. If possible, the advisory plots were then treated with a fungicide either that afternoon or the next morning.
Fungicides were applied at a rate of approximately 15 gallons of spray volume (water + fungicide) per acre. Applications were scheduled using either the AU-Pnut advisories generated from Doppler Radar and automated rain gauge data or the standard 14-day calendar spray program, beginning approximately 30 days after seedlings emerged.
For the calendar program, two applications of Echo 720 at a rate of 1.5 pints per acre were made, followed by four applications of Folicur 3.6F at a rate of 7.2 fluid ounces per acre and typically a final application of Echo 720 at a rate of 1.5 pints per acre.
For both AU-Pnut advisoriesthe Doppler radar and the automated rain gaugeone or two applications of Echo 720 at a rate of 1.5 pints per acre were made after the first and usually the second spray advisory. These applications were followed by three or four applications of Folicur 3.6F at a rate of 7.2 fluid ounces per acre.
Leaf spot severity was rated just before the plots were turned. Ratings were based on the Florida 1 to 10 leaf spot scoring system. Counts of southern stem rot hits (one hit was defined as one foot of consecutive white mold damaged plant[s] per row) were made immediately after the plots were dug. Yields were reported at 10% moisture.
With the exception of late July and early August 1998, Doppler Radar and the automated rain gauge generally identified rain events during the same 24-hour period at all three farm sites (Table 1). At the Williams and Murphy farms, the occurrence of rain events identified by Doppler Radar and the automated rain gauge prior to the third and/or fourth advisories were not recorded on the same day. The number of rain events recorded by the Doppler Radar and automated rain gauge at the Trawick farm also differed for the third spray advisory. Differences in the timing of rain events during the above time period were attributed to equipment failure and subsequent down time of the Doppler Radar unit near Echo, Alabama. During this same period, data from the much more distant Tallahassee, Florida, Doppler Radar site was used to identify rain events. Despite the discrepancies between Doppler Radar and automated rain gauge, an equal number of AU-Pnut spray advisories based on the radar and automated rain gauge data were issued for all three sites.
Leaf spot and southern stem rot severity in the plots sprayed according to both the AU-Pnut advisories were similar at all three sites (Table 2). Also, no differences in the severity of leaf spot and white mold were seen between the standard calendar spray program and the Doppler Radar and automated rain gauge-generated advisories. According to both the Doppler Radar and automated rain gauge derived AU-Pnut advisories, yield of peanut did not significantly differ at any of the on-farm sites. Depending on the test site, both advisories saved from one to three fungicide applications as compared with the standard calendar spray program.
In 1999, project efforts focused on the development of site-specific, interactive software for the AU-Pnut leaf spot advisory for Alabama peanut producers. The software needed to compile Doppler radar data and generate rainfall estimates was operational before the 1999 growing season. Also, Novartis Crop Protection sponsored the Peanut Weather Web site, which was maintained by the Agricultural Weather Information Service (AWIS). This web site included the five-day precipitation forecast, radar data, and other information required to implement the AU-Pnut leaf spot advisory.
In summary, Doppler Radar proved a valid method for estimating precipitation inputs for a weather-based expert system, particularly one with a specific rainfall threshold, such as the AU-Pnut leaf spot advisory. Rain gauges are no longer needed to implement the AU-Pnut advisory. Software was developed to archive, compile, and access Doppler Radar-based rainfall estimates for any field over the peanut-production region of Alabama. Finally, the concept of customized fungicide spray advisories on a field-by-field basis, which are generated with regional weather data and delivered on-line to farm clientele, has been demonstrated.