Biological Control Agents for Fire Blight of Apple Compared Under Conditions Limiting Natural Dispersal

Fire blight, which is a major constraint to pome fruit production in many areas of the world, is most commonly initiated by epiphytic populations of Erwinia amylovora that develop on blossoms (1,26). Under relatively dry climatic conditions, bacteria colonize flower stigmas, and subsequent rain or heavy dew facilitates movement to the floral cup (hypanthium), where infection generally occurs (24). Management programs in the United States have focused on suppressing E. amylovora on floral parts using antibiotics, and disease control has been especially dependent on the use of streptomycin; however, pathogen resistance to this antibiotic has developed in many production areas (13,14). Suppression of the blossom blight phase of fire blight with microbial antagonists has been widely demonstrated as an alternative (4,8) or complementary measure (12,22) to the use of antibiotics. Pseudomonas fluorescens strain A506, the active ingredient in BlightBan A506 (J. R. Simplot Co., Boise, ID), was originally selected based on its inhibition of an icenucleating strain of Pseudomonas syringae on corn leaves, then later found to reduce fire blight incidence (12). Another bacterial antagonist of commercial interest, Pantoea agglomerans (syn. Erwinia herbicola) strain C9-1 (8), was selected based on in vitro production of antibiotics inhibitory to E. amylovora (7). Other microorganisms reported as potential biological control agents for fire blight, including many strains of P. agglomerans, were first discovered in screening assays on artificial media (2,26) or immature pear fruit (3,5,29). Laboratory screening assays on detached crabapple flowers led to our recent discovery of P. agglomerans strain E325, which was significantly more effective than other test organisms in suppressing populations of E. amylovora on flower stigmas (17,19,20). The effects of bacterial antagonists on blossom blight in field trials often have been inconsistent (4,27), perhaps because of environmental factors and the instability or small size of established populations. Another factor that complicates field experiments is the natural dispersal of antagonists following treatment applications. Bacterial antagonists have been shown to spread from blossom to blossom and tree to tree (8,16), and honeybees have been implicated as the primary vector (8,9,25,27). The capacity of biocontrol agents to be vectored in this way and to colonize plant surfaces away from treated sites is one advantage they have over antibiotics, which protect only the contact area. However, we can reasonably assume that, to some extent, such spread interferes with evaluating and comparing different biocontrol treatments in the same experimental orchard. It may distort the comparison among treatments and the comparison between each treatment and the untreated control. A field study was conducted to compare the efficacy of P. agglomerans strain E325 with that of other antagonistic strains in reducing fire blight incidence and to simultaneously assess the importance of bacterial dispersal as a complicating factor. This was done under two sets of conditions, one that allowed for honeybee dispersal of bacteria and one that minimized it. Preliminary results of this study were previously reported (20).