Abstract
Dinitrogen fixation as a result of Gliricidia sepium/Rhizobium symbiosis (involving five plant genotypes and five bacterial strains) was evaluated in the glasshouse. In particular, the influence of plant genotype, bacterial strain and their interactions on the degree of fixation was determined. In general, no one strain was superior over all the host genotypes nor one genotype over the strains. The contribution of the host genotype × strain interaction to the total phenotypic variation was consistently significant (P < 0.05) and highest (over 50%), with the exception of shoot height where the contribution of the host genotype was high. Thus, the dinitrogen fixation and general performance of a Gliricidia sepium genotype was either low or high depending on the inoculum strain. The need for selecting optimal symbiotic partners for a particular environment is highlighted.