Articles
THE ROLE OF ANTIBIOTICS IN BIOLOGICAL CONTROL OF FIRE BLIGHT BY ERWINIA HERBICOLA STRAIN EH318
Several mechanisms have been proposed; competition for nutrients, chemoattractants or sites on blossoms (Klopmeyer and Ries, 1987; Hattingh et al., 1986; Wilson et al., 1992), acidification of the environment (Riggle and Klos, 1972) and the production of bacteriocins or antibiotics (Ishimaru et al., 1988; Vanneste et al., 1992). To determine the contribution of possible mechanisms to biocontrol ability, it is necessary to separately assess the role of each individual possible mechanism.
We have studied the role of antibiosis in the biocontrol ability of E. herbicola, strain Eh318, which had been isolated from apple shoots and has provided good control in orchard trials (Thomson and Gouk, 1992).
Strain Eh318 produces two distinct antibiotics; termed 717 and 719 that inhibit E. amylovora in vitro (Wright-Dobrzeniecka and Beer, 1993). Marker-exchange mutagenesis of the structural genes in the pathways for synthesis of 717 and/or 719 created mutants of Eh318 that produced either one or no antibiotic.
Any detected difference in biocontrol ability of the Eh318 mutants and the wild-type strain would be due to the presence of antibiotics 717 or 719. The mutants of Eh318 were named Eh421 (717-deficient), Eh439 (719-deficient) and Eh440 (717- and 719-deficient).
The biocontrol ability of the mutants against E. amylovora was tested in the immature pear fruit assay (Beer et al., 1984) and on apple blossoms in a controlled environment chamber.
For the immature pear fruit assay, 50 μl of cell suspensions of 108, 107 and 106 cfu/ml of strains Eh318 and Eh440, or phosphate buffer was added to individual fruits.
For each strain and concentration of inoculum 36 pear halves were used.
A broad range of concentrations was used to assure that the test could distinguish biocontrol ability between strains.
Inoculum of E. amylovora, strain Ea273 was added at 107 cfu/ml, 40 μl per pear half, two hours after E. herbicola had been applied.
The arrangement of the treated pears in sealed plastic boxes on moist towels was completely random.
The pears were maintained at 28°C. Disease was scored every twelve hours over a period of eight days, and the data collected were analyzed with an ANOVA for repeated measures.
On day eight, three apparently healthy pear halves per treatment were macerated in a blender and plated to check for differences in growth between Eh318 and Eh440. There were no significant differences in biocontrol ability detected between E. herbicola Eh318 and Eh440 in the pear test, ie. the ANOVA did not detected a treatment effect.
There was no significant difference in the populations of Eh318 and Eh440 isolated from pear fruits.
The apple blossom assay was repeated thrice with container-grown trees that had been stored in a cold room and then brought to bloom in the greenhouse.
At bloom the trees were taken to a controlled environment chamber at 80% RH, 25°C and a day length of 14 hours.
The blossoms were hand-pollinated, the next day they were sprayed with a compressed-air-powered atomizer to runoff with 2 x 108 cfu/ml of the E. herbicola