W. W. Shane and P. S. Teng
Department of Plant Pathology
University of Minnesota
St. Paul, MN 55108
Inocula were obtained from 8 to 10 day old V-8 juice agar Cercospora beticola cultures grown at 25 C under a 14 hr photoperiod. Cultures were started from a water suspension of freeze-dried spores and mycelia. The isolate used was collected in Chippewa county during 1982.
Plants were inoculated with spore suspensions (10,000/ml) using an air brush sprayer such that both sides of each leaf were lightly wetted. Inoculated plants were then moved to an Environmental Growth Chamber set at a constant temperature and a 14 hr photoperiod. A misting system kept leaf surfaces wet with deionized water. Plants were removed at specified intervals, sprayed with protectant fungicides to kill non-established fungi, placed on greenhouse benches, and evaluated for disease 3 and 5 weeks later. Each temperature X time combination was replicated 3 to 5 times.
The range of daily infection condition values (DICV) is on an arbitrary scale of 0 to 7. A value of 7 indicates that conditions sufficient for infection have occurred in that day. DICV values from adjacent days are combined to indicate the potential for infection according to the following scheme:
|
If the sum is: |
then the potential for infection is: |
| <6 6 >6 |
Unfavorable Marginal Favorable |
The structure for the daily infection value model is based on a model developed for Cercospora arachidicola on peanut (Parvin et al. Phytopathology 64:385-388, 1974) The performance of both the peanut model and the beet model were evaluated with weather data collected during 1983 at Rosemount in Dakota Co, Clara City in Chippewa Co., and Bird Island in Renville Co. Two versions of the beet model were used--beet model 1 used a 95 % relative humidity threshold value, model 2 used a 90 % relative humidity threshold. Both the peanut and beet models were written in Pascal to speed calculations. Weather data were recorded on hygrothermographs set in weather shelters at a height of 1 foot. Temperature and relative humidity were read from the charts with the use of a digitizer program from Cornell modified in our laboratory to work on an Apple II+ computer.
Output comparing the three models for two locations are shown in Figure 1 and Figure 2. It appears that the peanut model is not as sensitive as beet models 1 and 2. Only four and two days had infection values greater than zero for Clara City and Rosemount, respectively. The inadequacy of the peanut model was apparent because disease progress curves and fungicide efficacy data for Rosemount (see report on yield loss) indicated considerable infection must have taken place during the end of July and the beginning of August. The poor performance of the peanut model is not unexpected because it was developed for the warmer peanut-growing regions of the United States.
Beet model 1 and 2 appear to give more realistic pictures of conditions occurring at the two locations during 1983. With the 95% threshold for beet model 1 only a few days were indicated as actually favorable for infection although the model suggested that many days had marginal conditions. Beet model 2 indicated 14 favorable days for the Clara City region but only 3 days for Rosemount. Thus it appears that even model 2 is not sensitive enough, based on our observations at Rosemount. Part of the problem may be that ambient relative humidity adjacent to the canopy can be less than 95 or 90 % even though leaves in the canopy are still wet. Switching to measurements of leaf wetness instead of relative humidity may help with this problem. Another unknown area is the effects of interrupted wet periods on infections by Cercospora beticola. The ability of the conidia to survive under dry conditions is not well understood.
One of the benefits of the model is that locations can be compared on the basis of daily infection values (Fig. 3). Thus, it appears that conditions were more favorable for infection for Bird Island and perhaps Clara City than for Rosemount during the month of August. More experience is needed interpret the significance of the differences. We are in the process of analyzing weather data from additional locations and years. One important question is whether the daily infection value approach can anticipate severe leafspot outbreaks such as the one that occurred in 1981. We plan to test the significance of the daily infection values further in 1984.
It should be stressed that the daily infection values indicate only if weather conditions are adequate for infection. An equally important consideration is whether inoculum is available for infection. At Rosemount where inoculum pressure was heavy it was safe to assume that the favorable conditions indicated for the middle of July (Fig. 3) were going to result in appreciable infection by Cercospora beticola.
An important set of problems outside the task of building a workable infection prediction model deal with the difficulties of collecting weather data, plugging the data into the model, and making the information readily and immediately available. The daily infection values can be calculated without the use of a computer simply by using Table 1. We plan to make the information available on the Minpest computer news file system for a limited number of sites for southwestern Minnesota in 1984. The ideal system would be to have such information collected by microcomputers and automatically dumped via telephone line to a computer newsfile system. Such on-line weather services are available in other states such as Michigan and California and surely are feasible here.
1983 Sugarbeet Research and Extension Reports. Volume 14, pages 174-179.
