1Department of Horticulture, SLU, Box 55, SE-230 53 Ainarp, Sweden; 2USDA-ARS, Root Disease and Biocontrol Research Unit, Pullman, Waahington, USA

Background and objectives
Pythium ultimum is an ubiquitous pathogen in closed hydroponic systems with tomato. It may occur as an acute pathogen leading to root damage and damping-off or as a minor pathogen decreasing plant growth and yield. As there are no chemical alternatives for control of Pythium in cultures with continuous harvest in Sweden, control by the use of bacterial isolates is an option. Pseudomonas fluorescens strain 5.014, originating from tomato grown in a closed hydroponic greenhouse system, contributes to the promotion of growth and yield of tomato when applied in a multiple strain mixture, and has shown in vitro and in vivo antagonism of Pythium ultimum. However, at the end of the vegetation period nonbacterized tomato plants and those bacterized with strain 5.014 or a multiple strain mixture showed Pythium on the roots and did not differ with respect to their plant disease index. Different studies have demonstrated the antagonizing effect of 2,4-diacetylphloroglucinol (Phi) on Pythium. Therefore, enhanced biocontrol ability of strain 5.014 by reinforcment of Phi was suggested to be a feasible alternative. The objective of the present study was to chromosomally insert the genes responsible for Phi production and to screen the mutants' effect on Ciavibacter michiganesis ssp. michiganesis (Cmm) and Pythium ultimum var. ultimum (Puu) in vitro and on growth of tomato seedling in vivo.

Materials and methods
Twenty-three mutants were obtained and studies physiologically on the basis of their ability to utilize 95 carbon sources. The test was performed on the Biolog GN-panel. Readings were made spectrophotometrically by the colour change of the redox dye after 4, 18, 24, 48 and 72 ;h. In the bioassays, the mutants were screened for their ability to antagonize Cmm and Puu, using the parental strain P. ;fluorescens 5.014 and P. ;fluorescens Q2-87, which naturally produces Phi, as controls. The Cmm bioassay was conducted as an overlay, spreading Cmm on KMB. Evaluation of growth inhibition was performed after 72 ;h. To test Puu, KMB, PDA and YMA were used as test media. Evaluation of the growth of Puu was performed after 24 and 48 ;h.

The plant bioassay was performed under greenhouse conditions using rockwool as growing medium. The mutants and the parental strain were propagated in King's B broth supplemented with 100 ;l/ml rifampicin for 48 ;h, washed by repeated spinning and resuspended in Ringer solution at a a density of 108. Of the bacterial suspensions, 6 ;ml were inoculated into each rockwool cube before sowing. The cubes were kept at 24C in the dark until emergence and were then grown in the greenhouse at 19C, 75% humidity and a 16-h daylength.

Results and conclusion
In the physiological test, significant differences occurred with respect to adonitol, i-erythtol, xylitol and sebabic acid, which were utilized to a greater extent by the parental strain 5.014. In all tests, the positive control Q2-87 showed the strongest growth inhibition of both Cmm and Puu. All of the mutants had an increased ability to antagonize Cmm but most differed significantly from Q2-87. However, two mutants, 5-214 and 5-215, inhibited growth to an extent that was comparable to Q287. With Puu, Q2-87 was a superior antagonist on all media, compared with both mutants and the parental strain. Most of the mutants did not differ significantly from the parental strain 24 ;h after fungal inoculation. Mutant 5-214 was significantly better at suppressing Puu at an early stage on all media.

In the plant bioassay, the effect of the Phi-positive mutants on the performance of tomato was studied with respect to germination rate, emergence and plant height in rockwool cubes. They resulted in decreased emergence and plant height.