Abstract

Reliable quantification of nitrous oxide emission is a key to assessing efficiency of use and environmental impacts of N fertilizers in crop production. In this study, N₂O emission and yield were quantified with a database of 853 field measurements in 104 reported studies and a regression model was fitted to the associated environmental attributes and management practices from China’s croplands. The fitted emission model explained 48% of the variance in N₂O emissions as a function of fertilizer rate, crop type, temperature, soil clay content, and the interaction between N rate and fertilizer type. With all other variables fixed, N₂O emissions were lower with rice than with legumes and then other upland crops, lower with organic fertilizers than with mineral fertilizers. We used the subset of the dataset for rice – covering a full range of different typical water regimes, and estimated emissions from China’s rice cultivation to be 31.1 Gg N₂O-N per year. The fitted yield model explained 35% of the variance in crop yield as a function of fertilizer rate, temperature, crop type, and soil clay content. Finally, the empirical models for N₂O emission and crop yield were coupled to explore the optimum N rates (N rate with minimum N₂O emission per unit yield) for combinations of crop and fertilizer types. Consequently, the optimum N application rate ranged between 100 kg N ha−1 and 190 kg N ha−1 respectively with organic and mineral fertilizers, and different crop types. This study therefore improved on existing empirical methods to estimate N₂O emissions from China’s croplands and suggests how N rate may be optimized for different crops, fertilizers and site conditions.