Simulating N2O emissions under different tillage systems of irrigated corn using RZ-SHAW model

Abstract

Nitrous oxide (N2O) is a potent greenhouse gas (GHG), and agriculture is a global source of N2O emissions from soil fertility management. Yet emissions vary by agronomic practices and environmental factors that govern soil moisture and temperature. Ecosystem models are important tools to estimate N2O emissions by accounting for such variables, and models can strengthen field research. The objective of this study was to test RZ-SHAW predictions of crop production and N2O emissions from conventional till (CT) and no-till (NT) systems at high (HN) rate and low nitrogen input (LN) treatments in an irrigated corn (Zea mays L.) field in Colorado from 2003 to 2006 growing seasons. The model was calibrated using the HN-CT, and other treatments were used as validations. Additionally, the SHAW model was run in conjunction with RZWQM2 to account for differences in soil surface temperatures. Simulated crop yields were within 0.7 and 0.9% of measured yield for HN-NT and CT treatments, and 32 and 3% of measured yield for LN-NT and CT treatments, respectively. Spring soil temperatures were cooler by 2°C in NT compared to CT, and were correctly simulated using RZ-SHAW coupled model. RZ-SHAW simulated N2O emissions were slightly under predicted by 0.10 (1.5%) and 0.56 (7.1%) kgNha−1 for HN-NT and HN-CT treatments, respectively. Results for LN treatments showed larger differences in simulated N2O emissions and were over predicted by 0.11 (16%) kgNha−1 in NT and under predicted by 0.29 (29%) gNha−1day−1 in CT. Annual emissions were in close agreement, with observed and simulated showing 12 and 10% lower N2O emissions from HN-NT than HN-CT, respectively. Cooler surface soil temperature and higher soil water content in the HN-NT treatment caused slower breakdown of crop residue and slightly more denitrification than HN-CT, resulting in lower N2O emissions in HN-NT. This is the first test of the newly added GHG component in RZ-SHAW under no-till management, and results suggest with some improvements the model could be applied to quantify N2O emissions from different management practices.