The populations of diazotrophic and nondiazotrophic bacteria were estimated in the endorhizosphere and on the rhizoplane of Kallar grass (Leptochloa fusca) and in nonrhizosphere soil. Microaerophilic diazotrophs were counted by the most-probable-number method, using two semisolid malate media, one of them adapted to the saline-sodic Kallar grass soil. Plate counts of aerobic heterotrophic bacteria were done on nutrient agar. The dominating N2-fixing bacteria were differentiated by morphological, serological, and physiological criteria. Isolates, which could not be assigned to a known species, were shown to fix nitrogen unequivocally by 15N2 incorporation. On the rhizoplane we found 2.0 × 107 diazotrophs per g (dry weight) of root, which consisted in equal numbers of Azospirillum lipoferum and Azospirillum-like bacteria showing characteristics different from those of known Azospirillum species. Surface sterilization by NaOCI treatment effectively reduced the rhizoplane population, so that bacteria released by homogenization of roots could be regarded as endorhizosphere bacteria. Azospirillum spp. were not detected in the endorhizosphere, but diazotrophic, motile, straight rods producing a yellow pigment occurred with 7.3 × 107 cells per g (dry weight) of root in the root interior. In nonrhizosphere soil we found 3.1 × 104 nitrogen-fixing bacteria per g. Diazotrophs were preferentially enriched in the Kallar grass rhizosphere. In nonrhizosphere soil they made up 0.2% of the total aerobic heterotrophic microflora, on the rhizoplane they made up 7.1%, and in the endorhizosphere they made up 85%. Owing to high numbers in and on roots and their preferential enrichment, we concluded that diazotrophs are in close association with Kallar grass. They formed entirely different populations on the rhizoplane and in the endorhizosphere.
展开▼
