Agricultural management and environment controls long-term soil nitrous oxide fluxes

Nitrous oxide (N₂O) is an important greenhouse gas with a longatmospheric half-life. Understanding the controls on soil nitrous oxide fluxes is vital for the development of mitigation opportunities and for under standing their climatic impact. The spatial and temporal variability of soil nitrous oxide fluxes, however, makes it difficult to predict such fluxes. We examined the longest available dataset, 22 years of continues measurements, which contains biweekly measurements of soil nitrous oxide emissions together with measurements of an array of environmental and management parameters from eleven ecosystems, including four corn-soybean-wheat rotations under different management(conventional, no-till, biological, and reduced input), one perenni alalfalfa system, two tree plantations, three successional systems, and one deciduous forest. This dataset was used to assess the effect ofdifferent agricultural and land management practices on soil N₂O emissions. Using statistical and correlation analyses, we found that, in general, annual crops emitted 2-3 times more N₂O annually than didperennial crops. Among the annual crops, there were no differences in the annual emissions among the cropping systems; the conventional, no-till, reduced input, and biologically managed systems emitted similar amounts of N₂O with very different emission patterns. Among the perennial crops, alfalfa emitted 2 times more N2O than did poplar, approximately 1.6 times more than did the coniferous plantation, and ~3times more than did the unmanaged successional communities and the deciduous forest, which emitted similar amounts. Within the annual crop rotation phases, the wheat phase of the conventionally and no-till-managed rotations emitted approximately twice as much N₂O than did the reduced input- and biologically managed systems, largely due to the length of the bare soil fallow. The corn and soybean phases of the conventionally managed rotation emitted between 70 and 100% less N₂O than did the other row crops. The annual cumulative soil N₂O emissions were correlated with the growing season average soil nitrate concentration and the soil nitrate production potential. The daily soilN2O emissions were poorly predicted by the soil temperature, water content, and extractable soil N. The low predictability of the daily N2Ofluxes was likely due to the highly variable individual fluxes and the non-normal flux distribution, which was greatly skewed toward high extremes.