Partitioning of eddy covariance derived pasture N2O emissions for different sources and respective emission factors

Nitrous oxide (N₂O) is a very potent greenhouse gas, and the majority of the emissions are associated with intensive livestock production. The magnitude of the emissions depends on the nitrogen (N) input to the soil, and on grazed pastures the largest share of the emissions is typically originating from the N applied via fertilization and excreta of the grazing animals. The uneven spatial distribution of the excretion leads to emission hot spots on grazing systems and makes the quantification of the gaseous emissions difficult. Micrometeorological methods like the eddy covariance (EC) that integrate emissions over a larger area method are well suited to quantify total field-scale N₂O emissions of grazed pastures. But the partioning of emissions for different sources and the determination of source-specific emission factors is still a challenge.

We present results of a 5-year field experiment carried out in western Switzerland. The investigated pasture was grazed by dairy cows in an intensive rotational management. The field was additionally fertilized with organic and mineral fertilizer each year, according to the N requirement of the grassland. The field-scale N₂O fluxes were quantified with the EC technique using a fast response Quantum cascade laser spectrometer for N₂O concentration measurements. The experimental setup and the environmental conditions resulted in high temporal and spatial dynamics of the N₂O fluxes with highest values typically occurring after mineral fertilization events in the summer month. Using N₂O background parametrizations retrieved from chamber measurements in one year and subtracting the background emission from the measured N₂O fluxes allowed us to calculate excreta-related emission factors (EFs) according to the IPCC guidelines. EFs for fertilizer N input were calculated using a pre-defined time window after the fertilizer was applied. The subtracted background emissions during the fertilization events were calculated from the EC measurements outside this time window. We attribute the observed emissions to the different N inputs and discuss potential reasons for the supposedly higher emissions after mineral fertilizer applications in comparison to organic fertilizer emissions.