Distribution and recovery of nitrogen from legume residues decomposing in soils sown to wheat in the field

Abstract

Unground 15N-labelled medic material ( Medicago littoralis) was mixed with topsoils at 3 field sites in South Australia, allowed to decompose for about 8 months before sowing wheat, and then for a further 7 months until crop maturity. The site locations were chosen to permit comparisons of recoveries and distribution of 15N in soils (organic N and inorganic N to 90 cm depth) and wheat (grain, straw and roots to 20 cm depth) in areas where rainfall (and wheat yields) differed greatly. Soils differed also in their texture and organic matter contents. Recoveries of applied 15N in wheat plus soil were 93.1% from a sandy loam (Caliph) and 92.3% from a sandy soil (Roseworthy) despite differences in rainfall and extent of leaching of the 15NO 3 − formed from the decomposing medic residues. From a heavy clay soil (Northfield), which received the highest rainfall, the 15N recovery was 87.7%. The loss of 15N at this site was not due to leaching, as judged by 15NO 3 − distribution in the soil profile at seeding and crop maturity. Wheat plants took up only 10.9–17.3% of the 15N added as legume material. Percentage uptakes of 15N were not related to grain yields. The proportions of wheat N derived from decomposing medic residues were 9.2% at Caliph (input medic, N, 38 kg N ha −1), 10.5% at Roseworthy (input medic N, 57 kg N ha −1), and only 4.6% at Northfield (input medic N, 57 kg N ha −1). Most (51–70%) of the 15N recovered in wheat was accounted for in the grain. Inorganic 15N in the soil profiles was depleted during the cropping phase, and at wheat harvest represented from 0.6 to 3.1% only of 15N inputs. The major 15N pool was soil organic 15N accounting for 71.9–77.7% of 15N inputs. We conclude that, in the context of N supply from decomposing medic tissues to wheat crops, the main value of the legume is long-term, i.e. in maintaining soil organic N concentrations to ensure adequate delivery of N to future cereal crops. The N of the wheat was not uniformly labelled, root N being generally of the highest atom% enrichmensts, and straw N of the lowest. Nevertheless, at the Roseworthy site, the enrichments of wheat N were similar to those of NO 3 − N in the profile at seeding, indicating that the proportions of 14N and 15N in the inorganic N pool did not change appreciably during the cropping period. By assuming equilibrium at this site, we calculate that during 15 months decomposition the soil plus legume delivered about 189 kg N ha −1, of which 93.2 kg ha −1 (49.3%) was taken up by the wheat, 37.2 kg ha −1 (19.7%) was immobilized or remained as fine root residues, and 17.3 kg ha −1 (9.2%) remained as inorganic N in the soil profile; 41.7 kg ha −1 (22.1%) was unaccounted for in the soil-plant system, and was probably lost via inorganic N. Thus about 6.5 kg inorganic N ha −1 was supplied by the soil plus medic residues per 100 kg dry matter ha −1 removed as wheat grain.