How to model N2O emissions of grazed pastures with DayCent?

Nitrous oxide (N₂O) is one of the main greenhouse gases (GHG), and it occurs particularly in agricultural soils due to fertilizer applications and livestock grazing. Cattle, for instance, excrete 75-95% of their nitrogen (N) intake. National N₂O emission estimates for grazing cattle excreta are highly uncertain as they are typically calculated using global emission factors. A few countries including the U.S. and Australia use biogeochemical models to report N₂O emissions from agricultural soils. In Switzerland, the biogeochemical model DayCent was successfully applied for simulations of N₂O emissions from cropland with diverse crop rotations (dos Reis Martins et al. 2022; Wang et al. 2025) and mown grasslands with different mangement intensities (dos Reis Martins et al. 2024). However, it has not yet been validated for grazing pasture systems in Switzerland.

In this study, we aim to test DayCent for pasture systems in Switzerland, by a) examining how the grazing activities can be represented appropriately in the model, and b) to test the preformence of Daycent  in reproducing observed N₂O emissions.

Datasets from two Swiss field experiments in Posieux (Voglmeier et al. 2019; 2020) and Waldegg (Barczyk et al. 2024) were used. In both experiments, pasture N₂O emissions had been measured by eddy covariance over several years (Posieux: 2013-2017; Waldegg: 2020-2023) and the pasture management like the timing of grazing and fertilising events was precisely documented. First, a sensitivity analysis of the model was performed by varying the main grazing parameter flgrem (fraction of live shoots removed by a grazing event) in DayCent. Secondly, the model was applied in two scenarios: GrazMod (using the specific grazing module of DayCent) and HarvFert (representing cattle grazing intake by harvests and excreta depositions by fertilizer applications).

For both sites, the amount of biomass N consumed by the cattle on the pasture varied between 2-30 g N m^-2 yr^-1 initially increasing in correlation to the flgrem value, however following a saturation curve at higher flgrem values. The amount of N excreted on the pasture was proportional to the amount of N consumed (DayCent default: 80%), which was close to the values estimated by a cattle N budget approach as used in the national GHG inventory. N₂O emissions were higher for the HarvFert scenario, possibly due to a lower aboveground biomass which favors the emission loss of N. DayCent tends to underestimate the observed N₂O emissions of both pastures. Further results of DayCent simulations will be shown and discussed.