Semi-arid lands represent one fifth of the global land area but our understanding of greenhouse gas fluxes from these regions is poor. We investigated if inclusion of a grain legume and/or lime in a crop rotation altered greenhouse gas emissions from an acidic soil. Nitrous oxide (N 2O) and methane (CH 4) fluxes were measured from a rain-fed, cropped soil in a semi-arid region of Australia for two years on a sub-daily basis. The randomised-block design included two cropping rotations (lupin-wheat, wheat-wheat) by two liming treatments (0, 3.5 t ha -1) by three replicates. The lupin-wheat rotation only received N fertilizer during the wheat phase (20 kg N ha -1), while the wheat-wheat received 125 kg N ha -1 during the two year study. Fluxes were measured using soil chambers connected to a fully automated system that measured N 2O and CH 4 by gas chromatography. Nitrous oxide fluxes were low (-1.4 to 9.2 g N 2O-N ha -1 day -1), and less than those reported for arable soils in temperate climates. Including a grain legume in the cropping rotation did not enhance soil N 2O; total N 2O losses were approximately 0.1 kg N 2O-N ha -1 after two years for both lupin-wheat and wheat-wheat rotations when averaged across liming treatment. Liming decreased cumulative N 2O emissions from the wheat-wheat rotation by 30% by lowering the contribution of N 2O emissions following summer-autumn rainfall events, but had no effect on N 2O emissions from the lupin-wheat rotation. Daily CH 4 fluxes ranged from -14 to 5 g CH 4-C ha -1 day -1. Methane uptake after two years was lower from the wheat-wheat rotation (601 g CH 4-C ha -1) than from either lupin-wheat rotations (967 g CH 4-C ha -1), however liming the wheat-wheat rotation increased CH 4 uptake (1078 g CH 4-C ha -1) to a value similar to the lupin-wheat rotation. Liming provides a strategy for lowering on-farm greenhouse gas emissions from N fertilised soils in semi-arid environments via decreased N 2O fluxes and increased CH 4 uptake.
