Abstract: Field experiments were carried out from June 2003 to May 2004. Three crops of maize, soybean and rice were planted in summer season of 2003 followed by winter wheat. A static opaque chamber-GC technique was used to detect the emissions of N2O and CO2 in situ. Soil temperature, moisture and relevant crop parameters were measured. Results showed that there was a significant difference in the annual N2O emissions from different rotation plots. Higher N2O emission was observed in the plot of maize-wheat rotation with an annual amount of (18.5±0.7) kgN·ha-1. Lower N2O emissions were found in the plots of rice-wheat and soybean-wheat rotation with the annual amounts of (11.7±0.7) kgN·ha-1 and (13.2±0.4) kgN·ha-1, respectively. The soybean-growing season did not receive any additional fertilizer in the soybean-wheat rotation. In the rice-wheat plot, no-plowing practice before wheat sowing enhanced N2O emission significantly (P<0.001) in wheat-growing season from sowing to the winter as compared with that from the plowing plot, while no significant increase was observed over the wheat-growing season. A further investigation suggests that a stepwise regression function of y=a·W+b·T+c·L+d can be employed to quantify the influence of soil moisture (W), temperature (T) and leave area index (L) on N2O emission. The y is expressed as ln (FLUXN2O). The function did not show significant correlation between y and L for the maize plot, and thus c=0. The y did not correlate with W and T for the soybean plot, and thus a=0 and b=0. In the irrigated rice paddy, the y did not correlate with L. In addition, N2O emission was generally correlated to ecosystem respiratory rate for the plots of soybean, rice and wheat, while no significant correlation was found for the maize plot.

Key words: Nitrous oxide, Rotation method, Soil temperature, Soil moisture, LAI