COUNTING RIFAMPICIN-RESISTANT RHIZOBIA WHEN A MINOR COMPONENT OF THE SOIL RHIZOBIAL FLORA

This paper describes the development of a method for studying the popu lation dynamics of rifampicin-resistant rhizobia even when they are fe w in number compared with naturalised strains in soil. Spontaneous mut ants resistant to 200,ug rifampicin ml-1 were selected from two strain s of Rhizobium leguminosarum bv. trifolii and a strain of Bradyrhizobi um japonicum. These mutants were used in determining whether incorpora tion of rifampicin into the rooting medium of seedlings used in plant infection counts (MPN) would discriminate between the sensitive and re sistant phenotypes. Tolerance of seedlings of Trifolium subterraneum t o rifampicin and streptomycin and small-seeded Glycine soja to rifampi cin was tested within the range 40-200 mug ml-1. Rifampicin reduced pr imary root growth and increased lateral branching. At 100 mug ml-1 the seedlings were judged acceptable for MPN estimates. Streptomycin at 1 00 mug ml-1 reduced shoot dry weight by 54% and was deemed unsuitable. Nodulation of plants in rifampicin seedling agar was a function of bo th inoculum potential and rifampicin concentration. When pure cultures of a sensitive strain were counted, providing they did not exceed log 10 6.43 cfu ml-1, 100 mug rifampicin ml-1 prevented them nodulating T. subterraneum. A rifampicin-sensitive strain when grown in finely-mill ed peat remained sensitive to rifampicin in the seedling agar even whe n log10 8.02 cfu were added to the tube. Resistant mutants were unaffe cted at this concentration. When mixtures of sensitive and a resistant strain in soil were counted, rifampicin prevented nodulation by the s ensitive strain even when it outnumbered the resistant strain 200-fold . The technique provides an ecological tool for studying population dy namics of a strain against a background population of naturalized stra ins avoiding the problems associated with direct plate counts.