Nitrogen fixation by groundnut and soyabean and residual nitrogen benefits to rice in farmers' fields in Northeast Thailand

Abstract

Nitrogen fixation in groundnut and soyabean and the residual benefits of incoporated legume stover to subsequent rice crops were estimated in farmers' fields using¹⁵N-isotope methods. Three field experiments were conducted, two which examined N₂-fixation in groundnut by¹⁵N-isotope dilution using a non-nodulating groundnut as a reference crop and one in which N₂-fixation in two soyabean genotypes was compared using maize as the non-fixing reference crop. Groundnut fixed 72–77% of its N amounting to 150–200 kg N ha^-1 in 106–119 days and soyabean derived 66–68% of its N from N₂-fixation which amounted to 108–152 kg N ha^-1 under similar conditions. When legume stover was returned to the soil, there was a net contribution of N from N₂-fixing varieties of groundnut in all cases ranging from 13–100 kg N ha^-1, whilst due to the high % N harvest index in soyabean (87–88%) there was a net removal of N of 37–46 kg N ha^-1. In all cases if the legume stover was removed there was a net removal of N in the legume crop which ranged between 54 and 74 kg N ha^-1 in N₂-fixing varieties of groundnut and from 58 to 73 kg N ha^-1 in soyabean, whilst maize removed 66 kg N ha^-1 if its stover was returned and 101 kg N ha^-1 when the stover was removed.

Growth of rice was improved in all cases where groundnut stover was returned resulting in increases in grain yield of 12–26% and increases in total dry matter production of 26–31%. Soyabean residues gave no increases in rice grain yield but increased total dry matter production by 12–20%. Rice accumulated more N in all cases where legume stover was returned to the soil, and N yields were larger in all cases after the N₂-fixing legumes than after the non-fixing reference crops. N difference estimates of the total residual N benefits from the N₂-fixing legumes ranged from 11–19 kg N ha^-1 after groundnut and 15–16 kg N ha^-1 after soyabean. The amounts of N estimated directly by application of¹⁵N-labelled stover amounted to 7.2–20.5 kg N ha^-1 with groundnut which represented recovery of 8–22% of the N added in the stover. In soyabean only 3.0–5.8 kg N ha^-1 was estimated to be recovered by¹⁵N-labelling which was 15–23% of the added N, whilst only 1.3 kg N ha^-1 (4% of the N added) was recovered by rice from the maize stover. An indirect¹⁵N-method based on addition of unlabelled stover to microplots where the soil had previously been labelled with¹⁵N gave extremely variable and often negative estimates of residual N benefits. Estimates of residual N from the added stover made by N difference calculations did not correspond with the estimates by direct¹⁵N-labelling in all cases and possible reasons for this are discussed.