EXPERIENCE WITH DIFFERENT PREDICTION SYSTEMS FOR CONTROL OF FIRE BLIGHT IN SOUTHWEST GERMANY

Due to a severe outbreak of fire blight in 1993 in many orchards of southwest Germany, Plantomycin (a.i. streptomycin sulfate) was allowed to be applied during bloom according to prediction systems in 1994 and 1995. The Billing Revised System (Billing 1992) and the Maryblyt model (Steiner 1990) transformed into a computer program were used by the plant protection service for timing of the applications.
Spraying was recommended, when both systems calculated infections risk days.
In 1994 during bloom of apple and in 1995 during bloom of pear and apple infection risk days were indicated.
Applications on, before, or in a few cases after an infection risk day were recommended.
They had been effective, since sprayed orchards were protected from heavy blossom infections, while untreated orchards were heavily damaged.
In experiments, the efficiency of the treatments was high (>90 %). In both years, first symptoms on apple appeared, when predicted by the Maryblyt model, whereas symptoms on pear were observed earlier than predicted.
In 1994 and 1995, both systems apparently indicated more days with an infection risk than infections actually took place, as concluded by comparing the efficiency of applications carried out on different infection risk days.
However, in no case did the systems fail to indicate a real infection day.
The significance of dry and warm days as infection risk days is discussed.

VII International Workshop on Fire Blight

E. Moltmann

https://doi.org/10.17660/ActaHortic.1996.411.28