Key points are not available for this paper at this time.
Rhizobacteria were isolated from seedlings of seven economically important weeds and characterized for potential phytopathogenicity, effects on seedling growth, and antibiosis to assess the possibility of developing deleterious rhizobacteria as biological control agents. The abundance and composition of rhizobacteria varied among the different weed species. For example, fluorescent pseudomonads represented from 11 to 42% of the total rhizobacterial populations from jimsonweed and lambsquarters, respectively. Other bacteria frequently isolated were nonfluorescent pseudomonads, Erwinia herbicola, Alcaligenes spp., and Flavobacterium spp. Only 18% of all isolates were potentially phytopathogenic, based on an Escherichia coli indicator bioassay. However, the proportion of isolates that inhibited growth in seedling assays ranged from 35 to 65% depending on the weed host. Antibiosis was most prevalent among isolates of fluorescent Pseudomonas spp., the activity of which was due to siderophore production in over 75% of these isolates. Overall, rhizobacterial isolates exhibited a complex array of properties that were inconsistent with accepted definitions for plant growth-promoting and deleterious rhizobacteria. It is suggested that for development of effective biological control agents for weed control, deleterious rhizobacteria must be screened directly on host seedlings and must possess several properties including high colonizing ability, specific phytotoxin production, and resistance or tolerance to antibiotics produced by other rhizosphere microorganisms, and they must either synthesize or utilize other bacterial siderophores.
Kremer et al. (Fri,) studied this question.