Two-year field study on the emission of N2O from coarse and middle-textured Belgian soils with different land use

In the following study N2O emissions from 3 different grasslands and from 3 different arable lands, representing major agriculture areas with differen t soil textures and normal agricultural practices in Belgium, have been mon itored for 1 to 2 years. One undisturbed soil under deciduous forest was al so included in the study. Nitrous oxide emission was measured directly in t he field from vented closed chambers through photo-acoustic infrared detect ion. Annual N2O emissions from the arable lands ranged from 0.3 to 1.5 kg N ha(-1) y(-1) and represent 0.3 to 1.0% of the fertilizer N applied. Annual N2O emissions from the intensively managed grasslands and an arable land s own with grass were significantly larger than those from the cropped arable lands. Emissions ranged from 14 to 32 kg N ha(-1) y(-1), representing fert ilizer N losses between 3 and 11%. At the forest soil a net N2O uptake of 1 .3 kg N2O-N ha(-1) was recorded over a 2-year period. It seems that the N2O -N loss per unit of fertilizer N applied is larger for intensively managed and heavily fertilized (up to 500 kg N ha(-1)) grasslands than for arable l ands and is substantially larger than the 1.25% figure used for the global emission inventory. Comparison of the annual emission fluxes from the diffe rent soils also indicated that land use rather than soil properties influen ced the N2O emission. Our results also show once again the importance of ye ar-round measurements for a correct estimate of N2O losses from agricultura l soils: 7 to 76% of the total annual N2O was emitted during the winter per iod (October-February). Disregarding the emission during the off-season per iod can lead to serious underestimation of the actual annual N2O flux.