As a typical inland river basin in China, the Shiyang River Basin is characterized by its special mountain-basin structure. The ecological health in the basin is related to the sustainable development of the economy and society. At present, there are few studies on net primary productivity ($NPP$) in the Shiyang River Basin, and the existing analysis of the relationship between $NPP$ and climatic factors is lacking. The upper mountainous area and the middle and lower oasis areas in the Shiyang River Basin were selected as the study area. The $NPP$ of the study area was estimated using the Thornthwaite Memorial Model. The spatial and temporal characteristics of $NPP$ were analyzed by Sen’s slope method. Based on the sensitivity analysis, the correlation of the main climate factors to $NPP$ was estimated. According to the aforementioned work, the variation trend of the future $NPP$ was predicted. The results showed that $NPP$ in the study area increased from 1981 to 2015 with the increase in temperature and precipitation. The spatial heterogeneity of the change trend of $NPP$ was not significant, but the spatial heterogeneity of the rangeability was strong. The $NPP$ was highly sensitive to precipitation, relative humidity and net solar radiation. By integrating the changes in climatic elements, the temperature, precipitation and relative humidity contributed the main parts of the change in $NPP$. The $NPP$ is predicted to increase by 4.9–8.1% by 2050 according to the amplitude of climate change over the past 35 yrs.

Reservoir construction has led to the development of numerous wetlands, and these wetlands play an important role in global environmental change. In this paper, we investigate the relationship between reservoir wetlands and hydroclimatic variables. We used the MODIS land cover product to extract the wetland area of the Sanmenxia Reservoir, China. Then, various indices of reservoir wetland landscape patterns were calculated. Principal component analysis was performed to build the Sanmenxia Reservoir wetland comprehensive landscape pattern index (CLPI) to depict the changes in Sanmenxia Reservoir wetlands from 2001 to 2013. Pearson correlation analysis was used to assess their relationship. The following results were obtained. Firstly, the Sanmenxia Reservoir wetland area considerably declined and the landscape heterogeneity decreased from 2001 to 2013, especially in 2004. Secondly, the CLPI is significantly negatively correlated with annual runoff and significantly positively correlated with annual sediment discharge, annual average water level and annual shallow groundwater table in Sanmenxia Reservoir regions. Additionally, due to the decline in the reservoir wetland area, the values of Shannon’s diversity index and Simpson’s diversity index decreased in the study area. Therefore, the study suggests that maintaining a stable and healthy reservoir wetland area should be the focus of ecological reservoir management.