Survival and cell culturability of biocontrol Pseudomonas fluorescens CHA0 in lysimeter effluent water and utilization of a deleterious genetic modification to study the impact of the strain on numbers of resident culturable bacteria

Little is known on the behavior of soil-inoculated biocontrol pseudomonads once they are transported to deeper soil layers and/or groundwater levels after a heavy rain. This issue was investigated in inoculated microcosms containing lysimeter effluent water, and experimental conditions mimicking a worse-case scenario for potential bacterial dissemination were chosen. First, the survival of the polyketide-producing biocontrol strain Pseudomonas fluorescens CHA0-Rif was studied for 175 days at two inoculation levels in unamended and nutrient-amended lysimeter effluent water, and its impact on numbers of resident culturable bacteria was determined. Cell numbers of CHA0-Rif declined to 3-4 log cells ml−1 (at high inoculum level) or reached the detection limit or below (at low inoculum level) by day 175, without generating significant numbers of non-culturable cells. At high inoculum level, strain CHA0-Rif resulted durably (from day 50 to 175) in higher numbers of the total resident culturable bacteria when compared with the uninoculated control. This effect, which did not take place at low inoculum level or when nutrients had been added, contrasts with the transient ecological impact of the strain on rhizosphere bacterial populations in previous studies. Neither 2,4-diacetylphloroglucinol nor pyoluteorin were found in the water using HPLC, and inoculation with CHA0-Rif had no effect on the percentages of the total culturable aerobic bacteria sensitive to either antimicrobial polyketide on day 20. Second, the impact of CHA0-Rif on numbers of resident culturable bacteria was compared with that of CHA0-Rif(pME3424). Plasmid pME3424 carries an extra copy of the strain's rpoD gene (encoding sigma factor σ70). CHA0-Rif(pME3424) disappeared within 50 days in the water, but had the same impact as CHA0-Rif on the total number of resident culturable bacteria. This suggests that the impact of CHA0-Rif took place at the early stages of the experiment and was probably linked to the release of nutrients by introduced cells during inoculant declin