Simulation of long-term changes in soil organic carbon and nitrous oxide emissions from permanent grass silage using the DNDC model

Quantification and reporting of soil organic carbon density (SOCρ) changes and greenhouse gases (GHGs), particularly nitrous oxide (N₂O), emissions from agricultural soils using higher tiers remain a key challenge. Modelling approaches can provide largescale land use and management coverage whilst minimizing spatial and temporal variability. Identification of an advanced tool to simulate the net balance of SOC and GHG for mitigation, offsetting and policy formulation is a global concern. We tested the widely used latest version of Denitrification-Decomposition (DNDC95) model, a process-based one, to simulate both SOCρ and N₂O emissions and their annual changes over 45 years. The moist temperate grass silage was managed with inorganic fertilizer as urea and organic ones as cattle and pig slurry applied at low, medium and high rates. The model performed well for urea, cattle slurry and pig slurry to predict both SOCρ and N₂O emissions. The measured data for SOCρ at a 0-15 cm depth for unfertilized and urea-fertilized fields (73-77 t C ha^-1) were significantly higher than the simulated ones (54-55). However, the model-estimates showed good agreement with the measured values (R² = 0.66) and revealed increased C sequestration with increasing added-C (0.46±0.06 vs. 0.37±0.01 t C ha^-1 yr^-1). The model simulated N₂O emissions well and the resulted emission factors (EFs) estimated on average to be 0.35 ± 0.02, 1.80 ± 0.28 and 1.53 ± 0.41%, respectively, which are close to national and IPCC estimates. Variations in the simulated-SOCρ and derived-EFs could be explained mainly by differences in nitrogen inputs (49%) and added-C (62%), respectively, where the impact of rainfall (15-16%) and temperature (10-11%) was identical. Generally, SOCρ and N₂O EFs were sensitive to soil texture, pH, bulk density and organic carbon (R² = 0.77-0.99) but annual changes in SOCρ decreased with the latter two (R² = -0.99). Application of animal slurry during autumn demonstrated more C being sequestered in the clay loam soil (Dystric Gleysol) and strategic replacement of slurry either after the second or third silage cuts by urea decreased N₂O EFs significantly. Results  imply that the updated DNDC95 could provide an accurate representation of the key drivers influencing both SOCρ and N₂O fluxes in temperate grass silage.