Abstract
Wheat [Triticum aestivum] grown in pots was subjected to cycles of infection with the take-all fungus (Gaeumannomyces graminis var. tritici) under glasshouse conditions with soil collected from a virgin bushland in the Western Australian wheat belt. Wheat was inoculated with millet-seed-based inoculum of G. g. var. tritici, or with infected roots from plants one to three generations removed from the initial inoculation with millet-seed-based G. g. var. tritici, to determine whether disease suppression could be induced in simulation of talk-all decline. Three runs of such inoculations were made with four, three, and two cycles of infections in the 1st, 2nd, and 3rd runs, respectively. Severity of root disease produced by the root-based inoculum decreased significantly with each generation. Diseased roots from the first three cycles of infection supported larger populations of all bacteria and higher proportions of fluorescent pseudomonads than did healthy roots. When isolates of fluorescent pseudomonads were grouped into sets of isolates based on biochemical tests, most fell within Pseudomonas fluorescens biovar V. Strains of P. putida were more abundant on healthy roots than on roots inoculated with colonized millet seed or second-generation infected root inocula. Although there were differences in the frequency of occurrence of certain sets of isolates on infected roots with generations of the inoculum, most of these sets occurred in both generations studied. A significant portion (33%) of the fluorescent pseudomonad population from roots necrotized by second-generation inoculum was capable of inhibiting the pathogen in vitro.