|MERCADIER, GUY - European Biological Control Laboratory (EBCL)|
|GUERMACHE, FATIHA - European Biological Control Laboratory (EBCL)|
|BON, MARIE-CLAUDE - European Biological Control Laboratory (EBCL)|
Submitted to: Biological Control
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/9/2011
Publication Date: N/A
Interpretive Summary: A field experiment was conducted concerning a fungus-based biopesticide that we were developing against varroa mites of honey bees. We obtained a result that we did not expect at all. Most of the treated hives died. We then found that the biopesticide was contaminated with bacteria. The bacteria needed to be identified and tested to see if they had anything to do with the failed experiment. We identified the bacteria, and found that they were two strains of a species that are important contaminants of wastewater and stored food, and are also known to interfere with fungal pathogens. We tested the bacteria with the fungus and found that the bacteria do interfere with fungus growth when there are about as many bacteria as fungal spores. This paper showed the possible effect of contamination of biopesticides on experimental results. This is important because biopesticides are being used in many different systems now, such as against plant pests in greenhouses and against mosquitoes. Workers need to know if poor results in a particular experiment might be the result of contamination.
Technical Abstract: A biopesticide, formulated with commercially-prepared conidia of a strain of Beauveria bassiana isolated from Varroa mites, was tested in a field experiment in southern France to compare published results from the same geographic region with the same biopesticide but made with laboratory-prepared conidia. Results were dramatically different from those of previous experiments: bee hives treated with the biopesticide died at a much higher rate than control hives, and surviving treated hives lost significantly more weight than control hives. In addition, phoretic mite densities were significantly higher in treated hives than in control hives. The biopesticide was subsequently found to be contaminated with bacteria. Two strains of bacteria were isolated and identified following the combination of biotyping and sequencing data of two genomic regions. The two strains were found to be distinct although both were Pseudomonas sp. belonging to the P. fluorescens group. In dual cultures of B. bassiana and each bacterial strain, B. bassiana growth was slowed or suppressed when bacterial cfu density was about equal or greater than that of B. bassiana. To reduce or avoid such risk in biopesticide development, the authors encourage a systematic screen for bacterial contamination prior to field application, particularly for pathogens or biopesticides produced on a large scale.