Abstract: Vegetation net primary productivity (NPP) is one of the key factors determining the ecosystem functions in a basin. In this paper, the NPP of Guangxi Xijiang River basin was estimated by applying a model of CarnegieAmesStanford approach (CASA) with the time series data sets of MODIS, vegetation and meteorological data collected for the period from 2003 to 2012. Then the spatial and temporal patterns of NPP and its influencing factors were analyzed. The results indicated that: (1) the average annual NPP value from 2003 to 2012 in this basin was 524.67 g C·m^-2·a^-1. The high NPP values were mainly concentrated in the south and east while the low values in the central part. From the perspective of terrain, the areas in the valley plain had lower values of NPP but those areas in the hilly region had higher values. The NPP exhibited a significant variation among vegetation types. The evergreen broadleaf forest was the most productive with a NPP value of 788 g C·m^-2·a^-1, and the planting crop the least productive with a NPP value of 386 g C·m^-2·a^-1; (2) The average annual NPP value in this study area was in the range of 430.05-602.48 g C·m^-2·a^-1 from 2003 to 2012. Generally, NPP value exhibited a fluctuating and descending trend. In 88.89% of total study area, the NPP showed a decreasing trend. The NPP value was high from July to October and low from January to March. (3) The NPP was negatively correlated with annual average precipitation but positively correlated with annual average temperature. The correlation between NPP and meteorological elements (precipitation and temperature) was significant, with a multiple correlation coefficient value of 0.67.

Key words: Shangqiu City, ecosystem service value, land use and land cover change, impact assessment