Abstract
Two experiments were conducted in the greenhouse to study the interaction between alfalfa cultivars (Medicago sativa L. and M. falcata L.) and strains of Rhizobium meliloti Dang. for acetylene reduction rate, plant height and dry weights of shoot, root and whole plant. Fifteen alfalfa cultivars were inoculated with 10 strains of Rhizobium in Experiment I. Variance component analysis revealed that more than 30% of the total variance was due to alfalfa cultivars for acetylene reduction rate and 26% was accounted for by Rhizobium strains. More than 36% of the total variation was attributed to the interaction between alfalfa cultivars and Rhizobium strains for this character. Twenty-five host cultivars and 11 Rhizobium strains were included in Experiment II. The results also showed that the interaction of alfalfa cultivars and Rhizobium strains contributed the largest portion of the total variation for dry weights of shoot, root and whole plant and acetylene reduction rate. The results clearly demonstrated that the non-additive effects were the major component of variation for these characters associated with nitrogen fixation in alfalfa. Therefore, an effective way of improving nitrogen fixation in alfalfa is to select for a favourable combination of specific Rhizobium strains and alfalfa cultivars.