DAO 63:61-68 (2005)  -  doi:10.3354/dao063061

ABSTRACT: Use of Pseudomonas sp. strain MT5 to prevent and treat Flavobacterium columnare infection was studied in 2 experiments with fingerling rainbow trout Oncorhynchus mykiss. In the first experiment, length heterogeneity analysis of PCR-amplified DNA fragments (LH-PCR) was used to assess the effect of antagonistic baths on the microbial diversity of healthy and experimentally infected fish. In the 148 samples studied, no difference was found between bathed and unbathed fish, and 3 fragment lengths were detected most frequently: 500 (in 75.7% of the samples), 523 (62.2%) and 517 bp (40.5%). The species contributing to these fragment sizes were Pseudomonas sp., Rhodococcus sp. and F. columnare, respectively. A specific PCR for detection of Pseudomonas sp. MT5 was designed, but none of the tissue samples were found to be positive, most likely indicating poor adhesion of the strain during bathing. LH-PCR was found to be a more powerful tool for detecting F. columnare in fish tissue than traditional culture methods (χ² = 3.9, df = 1, p < 0.05). Antagonistic baths had no effect on the outbreak of infection or on fish mortality. F. columnare was also detected in healthy fish prior to and after experimental infection, indicating that these fish were carriers of the disease. In the second experiment, intensive Pseudomonas sp. MT5 antagonistic baths were given daily to rainbow trout suffering from a natural columnaris infection. Again, the antagonistic bacteria had no effect on fish mortality, which reached 95% in both control and antagonist-treated groups in 7 d.

KEY WORDS: Flavobacterium columnare · Antagonism · LH-PCR · Microbial diversity