Long-term variability in N2O emissions and emission factors for corn and soybeans induced by weather and management at a cold climate site

Nitrous oxide (N2O) emissions are highly variable in space and time due to the complex interplay between soil, management practices and weather conditions. Micrometeorological techniques integrate emissions over large areas at high temporal resolution. This allows identification of causes of intra- and inter-annual variability of N2O emissions and development of robust emission factors (EF). I Here, we investigated factors responsible for variability in N2O emissions during growing and non-growing seasons of corn and soybeans grown in an imperfectly drained silt loam soil, in Ontario, Canada. We used quasi-continuously (at half-hourly to hourly intervals) N2O fluxes measured via the fluxgradient technique over 11 years for corn and 5 years for soybeans and evaluated the uncertainty of default IPCC and Canada-specific EFs. In the growing season, emissions were controlled by soil nitrate content, soil moisture and temperature in the fertilized corn, while moisture and temperature regulated N2O emissions in the unfertilized soybeans. In the non-growing season, nitrogen (N) input from the crop residue did not affect the emissions, pointing to freeze-thaw cycles as mechanisms for enhanced N2O emissions. The non-growing season contribution to annual emissions was 38% in corn and 43% in soybeans. On average, annual emissions were 2.6-fold higher in corn than soybeans. Observed mean N2O EFs were 0.84% (012-2.02%) for growing season and 1.69% (0.29-7.32%) for yearly emissions. The growing season EF derived from long-term N2O emissions was 0.9 +/- 0.14%. The interannual variability in N2O emissions and EFs can be attributed to management practices and annual weather variability. The default IPCC approach based on overall N input had poorer performance in predicting annual N2O emissions compared to the current Canadian methodology, which includes management and environmental factor in addition to N inputs. The observed emissions were further evaluated with a newly developed growing season N2O emission prediction approach Canada. However, performance of the approach was poorer than IPCC or the current national Canadian approach. Additional tests of the new national methodology arc recommended as well as consideration of non-growing season emissions.