Modelling nitrous oxide emissions from dairy-grazed pastures

Soil N2O emissions were measured during four seasons from two highly productive grass-clover dairy pastures to assess the influences of soil moisture, temperature, availability of N (NH4(+) and NO3(-)) and soluble C on N2O emissions, and to use the emission data to validate and refine a simulation model (DNDC). The soils at these pasture sites (Karapoti fine sandy loam, and Tokomaru silt loam) differed in texture and drainage characteristics. Emission peaks for N2O coincided with rainfall events and high soil moisture content. Large inherent variations in N2O fluxes were observed throughout the year in both the ungrazed (control) and grazed pastures. Fluxes averaged 4.3 and 5.0 g N2O/ha/day for the two ungrazed sites. The N2O fluxes from the grazed sites were much higher than for the ungrazed sites, averaging 26.4 g N2O/ha/day for the fine sandy loam soil, and 32.0 g N2O/ha/day for the silt loam soil. Our results showed that excretal and fertiliser-N input, and water-filled pore space (WFPS) were the variables that most strongly regulated N2O fluxes. The DNDC model was modified to include the effects of day length on pasture growth, and of excretal-N inputs from grazing animals; the value of the WFPS threshold was also modified. The modified model 'NZ-DNDC' simulated effectively most of the WFPS and N2O emission pulses and trends from both the ungrazed and grazed pastures. The modified model fairly reproduced the real variability in underlying processes regulating N2O emissions and could be suitable for simulating N2O emissions from a range of New Zealand grazed pastures. The NZ-DNDC estimates of total yearly emissions of N2O from the grazed and ungrazed sites of both farms were within the uncertainty range of the measured emissions. The measured emissions changed with changes in soil moisture resulting from rainfall and were about 20% higher in the poorly drained silt loam soil than in the well-drained sandy loam soil. The model accounts for these climatic variations in rainfall, and was also able to pick up differences in emissions resulting from differences in soil texture.