SPATIAL COLONIZATION PATTERNS AND INTERACTION OF BACTERIA ON INOCULATED SUGAR-BEET SEED

Development and spatial distribution of microcolonies of Pseudomonas s pp. and Bacillus subtilis GB03 inoculated singly and in combination on sugar beet (Beta vulgaris) seed were observed with a scanning electro n microscope (SEM). SEM examination of seed directly after inoculation with Pseudomonas strain 33-2 or ML5 at population densities of approx imately 10(4) cfu per seed revealed a random distribution of individua l cells. By 24 h, when population densities had reached the stationary phase (approximately 10(6) cfu per seed), microcolonies had developed in a random pattern over the seed surface. However, even at these pop ulations, only 10-40% of the seed surface was colonized. Most microcol onies developed as separate entities on the indented surface of cells of the perianth and the operculum. The colonization patterns at 48 h w ere similar to those at 24 h, except that the colonies were larger. Si nce the number of cfu measured by dilution plating (detectable populat ion) was similar at both time periods, it was assumed that many cells were dead or dormant in the larger microcolonies. The spatial coloniza tion patterns were entirely different, depending on the density of the initial inoculum. The entire seed surface was covered when sufficient inoculum was applied to attain a detectable population size of approx imately 10(7) cfu per seed. Yet, even when the detectable population s ize increased to 10(7) cfu per seed following growth from an initial i noculum density of 10(4) cfu per seed, only 40-50% of the seed surface was colonized. This indicates the need for differentiating among live , dormant, and dead cells. The spatial colonization pattern of strain GB03 differed greatly from Pseudomonas strains. At temperatures favori ng its growth, microcolonies of GB03 were located primarily near the b asal pore of the seed, whether inoculated singly or coinoculated with Pseudomonas putida 33-2. In coinoculations, few microcolonies of 33-2 developed near the basal pore. However, this localized interaction cou ld not be detected by dilution plating of bacteria from the spermosphe re, demonstrating the difficulty of determining microbial interactions without visual examination. The overall conclusion from the study was that the spatial distribution patterns of developing and established microbial colonies of pseudomonads are such that little direct interac tion occurs in the spermosphere unless massive amounts of inocula are present, and that availability of nutrients is the limiting factor in population size. An interaction was detected with GB03 only at 37 C.