Biological N2 fixation and residual N benefit of pre-rice leguminous crops and green manures

A pre-rice crop experiment, including groundnut, mung bean (Vigna radiata), Sesbania rostrata, and a mixture of Sesbania and multipurpose cowpea (Vigna unguiculata) was conducted on a characteristic sandy soil of NE Thailand. The Sesbania-cowpea intercrop gave a similar total plant biomass as the Sesbania green manure alone (7 t ha-1) but with the advantage of yielding an edible product. The direct economic yield of cowpea was 1.3 t ha-1 green beans and greater than that achieved with groundnut or mung bean. The Sesbania-cowpea combination also enhanced rice yields by 0.8 t ha-1. The benefits in rice production were similar to the Sesbania green manure alone but surpassed the yields with the other grain crops or urea fertilizer of 30-60 kg N ha-1. Sesbania dry matter production increased with increasing planting density. The resulting variation in plant quality, e.g. lignin, however, was low. Rice responses to treatments were more related to the total residue N yields than to changes in plant quality. Apart from mung bean (25%) the pre-rice leguminous crops were able to obtain a considerable (>39%) proportion of their N from N2 fixation. The green manure Sesbania however fixed a larger proportion (79-89%) of its N than the grain crops (25-62%). This led not only to high amounts of N2 fixed by Sesbania but together with a N harvest index of zero yielded a large systems N benefit. With grain legumes this benefit was moderated by the N export in harvestable products. In the case of mung bean this may even result in effective soil N mining. Residue N use efficiency varied between 19-29% and was similar to that obtained from a single application of chemical N fertilizer (17-28%). For the farmer the Sesbania-cowpea intercrop option seems thus the most promising one not only regarding rice yield benefits but also in terms of soil fertility enhancement and generation of edible products