Description

Abstract: Weather based models help time fungicide applications to the periods when the diseases are most likely to occur. The first objectives of this work was to compare and adapt weather-based advisories developed for the control of Cercospora arachidicola on peanuts for resistant cultivars. It was achieved by comparing the disease progress curves of the 1997-1999 growing seasons in Lewiston NC, to spray schedules simulated by the Virginia Advisory, the Parvin, Smith and Crosby Advisory (PSC), NC Advisory, and AU-Pnuts Advisory and their adaptations for resistance. Field trials were conducted in 1997, 1998 and 1999 to test adaptations for resistant genotypes based on the NC Advisory. In all three years the leaf spot epidemics started late in the season (September). There was no yield difference due to leaf spot control except in 1999 in Lewiston for the susceptible genotypes (NC 7 and NC 11). All the advisories had a tendency to overspray at the beginning of the season, this might be due to a lack of inoculum at this time. The resistant genotype used for the study, GP-NC 343, did not lose any yield due to leaf spot in any of the tests and therefore did not need to be sprayed. The model that had the best fit to the disease progress curve of the susceptible genotypes

was the AU-Pnuts 12/4. The AU-Pnuts advisory 7/3, currently used in the Southeastern US, started spraying to early in the season for NC. The Virginia advisories also oversprayed. The NC advisory and the PSC were considered almost equivalent, and the adaptations for the PSC did not differ from the PSC itself. The second objectives was to develop a simulation model to predict epidemics of Stagonospora nodorum on winter wheat. The CERES-Wheat model was used to simulated leaf area indexes (LAI) for the wheat plant throughout the season. The disease model developed in this work simulated the spread of spores onto the plant leaves and heads, infection, the latent period and, lesion extension. The model equations were inferred from the literature and were calibrated with disease assessments made on Coker 9904 during the spring of 1998 in Plymouth NC. For 1998 and 1999, disease increase in the lower leaves took place 20 days after the disease increase was simulated by the model both years. The most effective spray timing corresponded to a period when disease was first observed in the lower leaves, no disease was seen on the flag leaf, and simulated onset of disease on the flag leaf had occurred. A sharp simulated disease increase in the flag leaf compartment may be a very good indicator for a spray recommendation. Combining a disease model to an already existing crop growth model facilitated modeling disease progress. Further work will be needed to fully validate both the CERES-wheat and the S. nodorum models in North Carolina Coastal Plains.