A comparison between measured and modeled N2O emissions from Inner Mongolian semi-arid grassland

The objectives of this study were (1) to determine the effect of land use on N2O emissions from Inner Mongolian semi-arid grasslands of China and (2) to evaluate the process-based DNDC model to extrapolate our field measurements from a limited number of sites to a larger temporal and spatial scale. The results suggest the following. Rainfall event was the dominant controlling factor for the seasonal variations of the N2O fluxes. The seven selected sites exhibited a similar seasonal trend in N2O emission, despite their different vegetation, land use and textures. In the typical steppe, N2O fluxes generally decrease with decreasing soil organic C (SOC) and total N content, indicating that soil C and N pools are very important in determining the spatial magnitude of the N2O flux. N2O emissions were very small during the entire growing season, averaging only 0.76 g N2O-N ha^-1 day^-1 for the five typical steppe sites, 0.35 g N2O-N ha^-1 day^-1 for the mown meadow steppe site, and 0.83 g N2O-N ha^-1 day^-1 from the cropped meadow steppe site. No enhanced effect due to overgrazing was observed for the N2O emission from the semi-arid grasslands. This was mainly results from the decreased SOC content due to overgrazing, which may have reduced the promoting effect of increased soil bulk density by trampling and animal excreta. Except for the mown steppe site, the model predictions of the N2O flux for the six different sites agree well with the observed values (r ^2 ranging from 0.35 to 0.68). It would be concluded that the DNDC model captured the key driving process for N2O emission. Nitrification was the predominant process, contributing 64-88% to the N2O emission. However, in terms of the magnitude of the N2O emission, further modifications should focus on the underestimated N2O flux during the spring and autumn periods (nitrification, freeze/thaw cycles) and the effect of topography and the mowing on N2O emission.