3.1.11
FIELD EVALUATION OF A BROWN SPOT DISEASE PREDICTOR FOR SCHEDULING FUNGICIDE SPRAYS AGAINST STEMPHYLIUM VESICARIUM ON PEAR

I LLORENTE1, E MONTESINOS1, P VILARDELL1, R BUGIANI2, I GHERARDI3 and P GOVONI2

1Institute of Food and Agricultural Technology-CeRTA, University of Girona, Avda Lluis Santalo s/n, 17071 Girona, Spain; 2Servizio Fitosanitario - Regione Emilia Romagna, Via Corticelia 133, Bologna, Italy; 3DI.PRO.VAL. University of Bologna, Via Filippo Re 8, 40126 Bologna, Italy

Background and objectives
Brown spot of pear is an important fungal disease caused by Stemphylium vesicarium, affecting fruit-growing areas of Europe, mainly Girona (Spain), Emilia-Romagna (Italy), Bouches du Rhone (France) and Netherlands. Infections and necrosis occur on leaves and fruits. If the environmental conditions are favourable for the disease, the loss of production will be very high. The control of pear brown spot is based on protectant sprays of fungicide applied at 7- to 15-day intervals depending on the type of fungicide. The high number of fungicide applications needed to maintain acceptable levels of disease may produce non-target effects and increased costs of pear production. An infection model for S. vesicarium was developed (STREP) [1] which quantifies the effect of wetness duration and temperature in the severity of disease and can be used to eliminate unnecessary sprays. The objective of our study was to evaluate the model STREP for scheduling fungicide sprays in orchard plots naturally affected by brown spot disease.

Materials and methods
The pear brown spot forecasting system is based on an empirical model which determines periods of environmental orchard conditions favourable to disease development. The model was derived and synthesized from previous research [1]. Every 24 h the model generates a daily risk (S) and a 3-day cumulative risk index (SA) calculated by totalling S values for the past 3 days. Values of SA were used as thresholds for spraying fungicides. The STREP model was evaluated for 3 years (1995, 1996 and 1997) in five commercial orchards located in Girona (Spain) and Emilia-Romagna (Italy). The cultivars used in the study were Passe Crassane, Abate Fetel and Conference which are susceptible to the disease. The fungicide most frequently tested was tiram, but in some cases procymidone and kresoxim-methyl were used. Field evaluation of the forecasting system consisted of comparing disease control levels for different treatments. The treatments used were the following: (i) fungicide sprays applied according to the thresholds SA:0.4, SA:0.5 or SA:0.6; (ii) fixed sprays every week (tiram) or every 2 weeks (procymidone or kresoxim-methyl); and (iii) a non-treated control.

Results and conclusions
A relationship between periods with high values of SA and increases of disease levels through the time in non-treated plots was observed. In all trials the level of disease incidence at harvest time in non-treated plots was greater than 30%. Disease control on fruits with the fungicide tiram and SA:0.4 or SA:0.5 thresholds was similar to the fixed spray schedule, but saving 25-50% of fungicide sprays. The use of a threshold SA:0.6 showed a lower efficacy than fixed sprays but it saved applications about 50-70%. The efficacy using procymidone or kresoxim-methyl with SA:0.4 was similar to fixed sprays but saving 70-75 and 20% of procymidone or kresoximmethyl fixed sprays, respectively. More trials using kresoxim-methyl or procymidone sprayed according to the STREP-model should be done to generalize the results for these fungicides. However, existing results indicate that the model STREP is a good predictor of pear brown spot infection and will be used in future mesoscale pilot trials as a forecaster for scheduling sprays. If the results are acceptable, the model will be integrated on Grower's Warning Stations.

References
1. Montesinos E, Moragrega C, Liorente I et al., 1995. Phytopathology 85, 586-592.