Sustainable agriculture relies greatly on renewable resources like biologically fixed nitrogen. Biological nitrogen fixation plays an important role in maintaining soil fertility. However, as BNF is dependent upon physical, environmental, nutritional and biological factors, mere inclusion of any Nrfixing plant system does not guarantee increased contributions to the soil N pool. In the SAT where plant stover is also removed to feed animals, most legumes might be expected to deplete soil N. Yet beneficial legume effects in terms ofincreased yields in succeeding cereal crops have been reported. Such benefits are partly due to N contribution from legumes through BNF and soil N saving effect. In addition, other non-N rotational benefits, for example. improved nutrient availability. improved soil structure. reduced pests and diseases. hormonal effects are also responsible. In this paper we have reviewed the research on the contribution of grain legumes in cropping systems and the factors affecting BNF. Based on the information available, we have suggested ways for exploiting BNF for developing sustainable agriculture in the semi-arid tropics (SAT). A holistic approach involving host-plant, bacteria. environment and proper management practices including need based inoculation for enhancing BNF in the cropping systems in the SAT is suggested.

Responses to inoculation with Nl-fixing bacteria were studied" in relation to genotypic differences in pea~1 millet, effect of nitrogen levels, and FYM additions in India. In some experiments, inoculation increased mean grain yield up to 33% over the uninoculated control, whereas in the remaining 11 experiments'there was no significant increase. Increased grain yields~ > 10% over the uninoculated controls were observed in' 46% of the experiments with AzospiriJ/um lipojerum (18.7% average increase) and with Azotobacter chroococcum (13.6% average increase). Yield increases were nil or reduced in three experiments with Azos. /ipojerum and four experiments with A'ztb. chroococcum. I,n two experiments continued inoculation fot two