Development of a Downy Mildew Advisory Model Based on Downcast

Abstract

A downy mildew advisory model for use in practical onion growing was constructed according to DOWNCAST, an existing downy mildew forecasting model. The model was empirically improved after comparison between observed and calculated disease development. Onion plants grown in pots were placed as trap plants in artificially inoculated onion fields (one in 1995, three in 1996 and one in 1997) for 1–7 days, then removed and incubated outdoors in a sheltered place at 1 km distance from the onion field. Symptom development was observed. Hourly measurements of leaf wetness, temperature and relative humidity were taken in the crop and hourly data on rainfall were obtained from a nearby automatic weather station. Sporulation of the fungus was visually observed in the onion crop in the morning on several days in all three years. Sporulation-infection periods and sporulation as predicted by the model were compared with symptom development on the trap plants and observations on sporulation in the field. The initial model corresponded to the trap plant observations in only 18 out of 42 and to visual sporulation in 25 out of 40 occasions. Criteria in both the sporulation and infection submodels of the advisory model were subsequently calibrated to obtain a best fit between the model and observations. In 30 out of 42 occasions the improved model corresponded with trap plant observations while 30 out of 40 sporulation observations were now calculated correctly. The improved model needs to be evaluated with additional and independent data.