RESUMO
This study aimed at evaluating soil nitrous oxide (N2O) and methane (CH4) emissions from integrated farming systems. Soil N2O and CH4 fluxes were assessed in a subtropical Cambisol in southern Brazil, using manual static chambers, over two years, in five farming systems (cropland, livestock, integrated crop-livestock, integrated livestock-forestry, and integrated crop-livestock-forestry). The study was conducted in four growing seasons: summer-1, winter-1, summer-2, winter-2. Integrated farming systems had lower soil N2O emissions than livestock. The observed reduction was possibly due to lower water-filled pore space (WFPS) in soils under integrated systems (average 59.5-64.7%, vs 70.4% in livestock) as indicated by correlation (r = 0.74). Cropland, including cover-crops and maize, also had lower N2O emission (by 40%) relative to livestock, of levels similar to those observed in integrated systems. Methane was consumed in soil, but it was not affected by farming systems, and offset only ~1.4% of the N2O emissions. In the rainiest season of summer-2, the soil had the highest WFPS (on average 71.4%) and thus the highest N2O emission (on average 9.79 kg N2O-N ha-1 season-1) and the lowest CH4 consumption (on average - 0.40 kg CH4-C ha-1 season-1); while the opposite trend occurred in the driest season of winter-2 (on average 57.3% WFPS; 0.64 kg N2O-N ha-1 season-1 and -0.90 kg CH4-C ha-1 season-1). Integrated farming systems including crop-livestock, livestock-forestry and crop-livestock-forestry reduced soil N2O emissions relative to sole livestock by 27-40%, but did not affect CH4 emissions. Seasonal variations of precipitation, and therefore WFPS were driving factors of the N2O and CH4 emissions. Overall, integrated farming systems show the potential to mitigate soil N2O emission compared to livestock system.