N2O emissions of low input cropping systems as affected by legume and cover crops use
Résumé
Reducing environmental impacts of agriculture is a key issue for preserving water, soil and air quality. Among available options, the incorporation of legumes allows to reduce synthetic N fertilizer use, the use of cover crops is an efficient way to decrease nitrate leaching and recycle nutrients, while mechanical weeding together with crop rotation help to limit the use of pesticides. However, how these options affect N2O emissions remains uncertain. Here we studied N2O emissions of four cropping systems which were designed as low input alternatives to the conventional wheat-sunflower rotation in south-west France and make use of the aforementioned options. The used cropping systems differ by synthetic N fertilizer input: 96 kg N ha(-1) year(-1) for the Low Input (LI) cropping system and 33 kg N ha(-1) year for the Very Low Input (VLI) cropping system which incorporates a grain legume in the rotation. For each of the two systems (LI and VLI), both a variation with cover crop (+CC) and without cover crop ( CC) were studied. Daily N2O fluxes were measured almost continuously (at least 87% of the 3-year rotation duration) using automatic chambers. Measurements of soil water content, soil temperature, mineral nitrogen (N) content were also carried out. Cumulative N2O emissions over the whole 3-year rotation were in the low range of literature values for all systems but they were significantly higher for the VLI cropping system (1.12 kg N2O-N ha(-1) year(-1)) than for the LI cropping system (0.78 kg N2O-N ha(-1) year(-1)) on average over +CC and CC. Patterns of emissions were also different between the two systems, with emissions taking place in the-form of short-lived peaks only after N fertilization for LI, while peaks occurred more frequently over the whole rotation for VLI. No difference between the +CC and CC cropping systems could be detected at the scale of the rotation (P = 0.47). Although N2O fluxes increased significantly fora few days after cover crop destruction (P = 0.002), the contribution of such events to cumulative N2O emissions remained negligible (<10 g N2O-N ha(-1)). The effect of tillage events, especially when they were not associated with crop residues return to soil, was also found to be small. As expected, both temperature and soil water content had a strong influence on N2O emissions, although the underlying relationships often only appear clearly during specific periods, e.g. during summer for the soil moisture effect. Because agronomical practices such as increased use of legume crops and use of cover crops only have a limited effect on N2O emissions, their adoption can be considered on the basis of the benefits they have to address other important agronomic or environmental issues.