Field measurement of lupin belowground nitrogen accumulation and recovery in the subsequent cereal-soil system in a semi-arid Mediterranean-type climate

In situ ¹⁵N-labelling was used to provide a quantitative assessment of the total contribution of lupin (Lupinus angustifolius) to below-ground (BG) N accumulation during a growing season under field conditions, and to directly trace the fate of the lupin BG N in the next season, including quantifying the N benefit from lupin to a following wheat (Triticum aestivum) crop. The experiments were conducted at two sites, both experiencing a semi-arid Mediterranean-type climate in the wheat-growing region of Western Australia but with differing soil types, a deep sand (Moora) and a sand-over-clay shallow duplex soil (East Beverley, EB). Lupin shoot and root dry matter and total plant N accumulation, proportional dependence on nitrogen fixation and grain yield were greater at the deep sand site than the duplex soil site, although there was a similar proportion of shoot N to estimated total BG N at both sites. The proportion of total plant BG N decreased from the vegetative stage (42–51%) to peak biomass (25–39%) and maturity (23–34%). From 56–67% of BG N on the deep sand and 74–86% on the duplex soil was not recovered in coarse roots (>2 mm) or as soluble N, but was present in the insoluble organic N fraction. There was evidence for cycling of lupin root-derived N into soil microbial biomass and soluble organic N during lupin growth (by the late vegetative stage), but no evidence for leaching of legume derived BG N during the lupin season. Estimates of fixed N input BG were at least four times greater if based on total lupin BG N rather than on N recovered in coarse roots (>2 mm). There were no apparent losses of lupin BG N during the summer fallow period subsequent to lupin harvest at either site. Also, immediately prior to sowing of wheat there were similar proportions of lupin BG N in the inorganic (20–25%) and microbial biomass (6–9%) pools at both sites, with the majority of BG N detected in the <2 mm fraction of the soil column. However, the proportion of residual lupin BG N estimated to benefit the aboveground wheat biomass was relatively low, 10% on the deep sand and only 3% on the shallow duplex. Some (14%) residual lupin BG N was leached as nitrate to 1 m on the deep sand compared to 8% of residual lupin BG N leached to the clay layer (0.3 m) on the shallow duplex. About 27% of the residual lupin BG N on the deep sand at Moora had apparently mineralised by the end of the succeeding wheat season (i.e. recovered either in the wheat shoots, as inorganic N in the soil profile or as leached nitrate) compared to only 12% at EB. There was an unaccounted for large loss of residual lupin BG N (50%) from the duplex soil at EB during the wheat season, postulated to be chiefly via denitrification. At both sites after the wheat season a substantial proportion (32–55%) of legume derived BG N was still present as residual insoluble organic N, considered to be an important contribution to structural and nutritional long-term sustainability of these soils.