Evaluation of local and imported fire blight warning systems in Israel.

ABSTRACT The possibility of using local and imported warning systems for the management of fire blight (caused by the bacterium Erwinia amylovora) in pears was tested in Israel from 1997 to 2000. Three imported systems (MARYBLYT 4.3, BIS95, and Cougarblight 98C) and one local system (Fire Blight Control Advisory [FBCA]) were used. All systems were tested in simulation experiments; MARYBLYT 4.3 and FBCA were also tested in orchard experiments under natural infections. Simulation experiments included 193 orchard-plots in which the time of disease onset enabled us to determine the date of infection. Thirty-five experiments were conducted in commercial orchards; in 10 of these, fire blight developed naturally. The performance of the imported warning systems was too variable to be accurately used under Israeli conditions. In the simulation experiments, the success rate (i.e., the capacity of the systems to predict the exact date of the occurrence of infection episodes) of the imported systems was low (3 to 55%) with considerably large variability among years (CV = 30 to 67%). Similar results were obtained in the orchard experiments for MARYBLYT 4.3: in only two of five experiments where plots were managed according to that system was disease severity significantly lower than that recorded in untreated control plots. In comparison, the local system, FBCA, predicted most infection episodes in the simulation experiments with low variability (99%, CV = 1.0%). In the orchard experiments, adequate disease suppression was achieved in all eight experiments in which FBCA recommendations were followed. We concluded that it was not possible to import and successfully implement fire blight warning systems in Israel that have been developed in regions with dissimilar environmental conditions.