Abstract: By using GLOPEMCEVSA model, the net primary productivity (NPP) of terrestrial vegetations in Jiangxi Province in 2000-2006 was simulated, and the spatial patterns of the NPP in relation to climate factors were analyzed. The simulated NPP had a significant linear correlation with the surveyed NPP, and the multiple correlation coefficient was 0.85 (P< 0.001). Among the main vegetations in the Province, evergreen needleleaf forest had the highest NPP (1091.38 g C·m^-2·a^-1), followed by evergreen broadleaf forest (846.09 g C·m^-2·a^-1), shrub (596.62 g C·m^-2·a^-1), and grass (325.50 g C·m^-2·a^-1). The analysis of the NPP along the climate gradient showed that in the areas with the precipitation less than 1900 mm, the NPP increased with increasing precipitation, but the increment was small and had a greater fluctuation. In the areas with the precipitation being 1900-1950 mm, the NPP increased significantly with increasing precipitation; while in the areas with the precipitation >1950 mm, the NPP decreased with increasing precipitation. In the areas with the air temperature lower than 17 ℃, the NPP increased with increasing temperature; while in the areas with air temperature higher than 17 ℃, the NPP decreased with increasing temperature. The further analysis on the relationships between the NPP and precipitation in the areas with lower air temperature (< 17.25 ℃), medium air temperature (17.25-18.55  ℃), and higher air temperature (>18.55 ℃) showed that in the areas with lower and medium air temperature, the dominant vegetation was evergreen needle-leaf forest, which had higher NPP, while in the areas with higher air temperature, crop and shrub were the main vegetations, whose NPP was lower and had greater variability.

Key words: late sowing winter wheat, straw returning, nitrogen application rate, soil nitrogen surplus.