Bacterial associations with the mycorrhizosphere and hyphosphere of the arbuscular mycorrhizal fungus Glomus mosseae

Roots and mycorrhizal fungi may not associate with soil bacteria randomly, but rather in a hierarchical structure of mutual preferences. Elucidation of such structures would facilitate the management of the soil biota to enhance the stability of the plant-soil system. We conducted an experiment utilizing two isolates of soil bacteria to determine their persistence in distinct mycorrhizal regions of the root zone, and their effects on general rhizosphere populations of fluorescent pseudomonads (FP). Split-root sorghum (Sorghum bicolor L.) plants were grown in four-compartment containers, constructed so that the soils in individual compartments held either (1) roots colonized by the arbuscular-mycorrhizal (AM) fungus Glomus mosseae (M), (2) nonAM roots only (R), (3) hyphae of G. mosseae (H), or (4) no mycorrhizal structures (S). The soils were inoculated (10(7) cells g(-1) dry soil) with antibiotic-resistant (rifampicin, rif; streptomycin, sm) strains of the soil bacteria, Alcaligenes eutrophus (rif(r) 50) or Arthrobacter globiformis (sm(r) 250), or were left uninoculated as control. A. eutrophus had been isolated from a specific source (hyphosphere soil of G. mosseae), and A. globiformis from mycorrhizosphere soils of two AM fungi. After 10 wk of growth, the presence of A. eutrophus was barely detectable (<10 cfu g(-1) dry soil) in nonAM (R and S) soils, but persisted well (10(4) cfu g(-1) dry soil) in AM (H and M) soils. Numbers of A. globiformis were more evenly distributed between all soils, but were highest in the presence of AM roots (M soil). There were varied bacterial effects on root and AM-hyphal development: A. eutrophus decreased hyphal length in H soil, while A. globiformis stimulated root length in M soil. The two bacterial inoculants did not affect numbers of FP in H, R, and M soils, but the AM status of the soils did: the numbers of FP increased in the order M>R>H>S. There was a positive correlation of FP numbers with both bacterial inoculants in M and H soils. Numbers of FP changed with root or hyphal lengths, an effect that was related to changes in the numbers of the inoculated bacteria. The results indicate that the hyphosphere-specific A. eutrophus depended on the presence of G. mosseae, but that the nonspecific A. globiformis did not. The mycorrhizal status of soils may selectively influence persistence of bacterial inoculants as well as affecting the numbers of other native bacteria.