The previous study suggested that climate change would increase the rainfall in the arid region in Northwest China. The purpose of this study was to examine the adaptive strategies and changes of the typical vegetation (Nitraria tangutorum Bobr and Nitraria sphaerocarpa Maxim) in the arid region due to increasing rainfall in the future. We had established three open-simulated rainfall platforms in Dengkou (average annual rainfall 145mm), Minqin (average annual rainfall 115mm), Dunhuang (average annual rainfall 40mm) in 2008. Based on the average annual rainfall of three research regions, five water gradients (increasing water 0%, 25%, 50%, 75% and 100%, respectively) had been set up. Each gradient has 4 replicates (Total 20 plots in each research region). The simulation rainfall was operated per 15 days from 10 May to 10 September in 2008. In 2009, the simulation rainfall interval was adjusted to 30 days, and the rainfall of Dunhuang were adjusted to 0% , 50%, 100%, 200% and 300% based on the average annual rainfall of Dunhuang. This paper had used the platform of simulation rainfall to study: (1) the relationships between rainfall and soil moisture; (2) the relationships among rainfall,δ13C value and leaf traits (including leaf shape: leaf length, width, perimeter, area, ratio of length to width and ratio of perimeter to area; leaf nutrient: area- and mass- based leaf nitrogen content; specific leaf area); (3) The relationships among the leaf traits of common species (110 samples)that collected from the three research regions. The major conclusions as follows:(1) When single simulation rainfall less than 10mm, the soil moisture of 10cm under surface was affected, and the effect usually disappeared within 7 days. When single simulation rainfall was from 10 to 20mm, soil moisture of 20cm under surface had been able to receive weak rainfall signals within 3 days after simulation rainfall. When single simulation rainfall exceeded 20mm, soil moisture of 20cm under surface could be increased significantly. However, even if single simulation rainfall or natural rainfall was up to 29mm or 30mm during the growing season, soil moisture of 50cm under surface could not be affected at all.(2) Leaf shape of Nitraria tangutorum Bobr and Nitraria sphaerocarpa Maxim was close to oval or rectangular, respectively. Leaf length and width were the major factors to determine the leaf morphological changes of. Nitraria tangutorum Bobr and Nitraria sphaerocarpa Maxim With increasing rainfall, leaf length of Nitraria tangutorum Bobr and Nitraria sphaerocarpa Maxim was prior to increase. Compared the plots with 300% increasing water to 0% in Dunhuang in 2009, the leaf length increased 34% (P <0.01), and the leaf width increased 23% (P <0.01); Compared the plots with 100% increasing water to 0% in Minqin in 2008 and 2009, the leaf length increased 11% and 10%, respectively (P <0.01), and the leaf width did not change significantly (P>0.05). Compared the plots with 100% increasing water to that of 0% in Dengkou in 2008, the leaf length increased 9% (P <0.01), and the leaf width did not change significantly (P>0.05). Compared the plots with 100% increasing water to that of 0% in Dengkou in 2009, the leaf length increased 10% (P <0.01), and the leaf width increased 6% (P <0.05). The relationships among leaf length, width and the total annual rainfall in different plots were that leaf length and width increased with the increasing of total annual rainfall. Moreover, the leaf length and width increased 0.52cm and 0.2cm per 100mm, respectively.(3) There were close correlation among leaf shape (leaf length, width, perimeter, area, ratio of length to width and ratio of perimeter to area), leaf nutrients (based area and mass leaf nitrogen content), specific leaf area andδ13C value. With the increasing of total rainfall, leaf area, specific leaf area increased, however, ratio of length to width, ratio of perimeter to area andδ13C value decreased, and the variations of leaf nutrient had no consistent trend. On the other hand, by compared the plots of different rainfall gradient in the three research regions, there were not an obviously trend of the variations of specific leaf area, leaf nutrient content andδ13C value.(4) There was a positive correlation betweenδ13C value and ratios of perimeter to area. This would be caused by that the leaf with high ratio of perimeter to area always have high leaf boundary layer conductance. In the hot growing season, the increasing of leaf boundary layer conductance would help to decrease leaf surface temperature, which could protect photosynthetic machinery and leaf cells against excessive leaf temperatures when stomata conductance decreasing or close in response to water stress, and thus increase water use efficiency. In fact, leaf length of Nitraria increased priority with increasing rainfall, which could slow down the decreasing rate of ratio of perimeter to area, and preventδ13C value (leaf water use efficiency) from rapid declining with increasing rainfall.(5) Compared the leaf traits of common species in Dengkou, Minqin, Dunhuang, the shrub leaf shapes of Dunhuang, which located at the edge of the wetland, had significant differences with Minqin and Dengkou shrubs. Such as, the shrub leaf of Dunhuang had higher leaf length, width and perimeter.Based on the area-based leaf nitrogen content(Narea), all the samples were divided into 3 groups: group 1; Narea <2, group 2: 23. Comparing specific leaf area and mass-based leaf nitrogen content, the results showed that mass-based leaf nitrogen content increased with the increasing of specific leaf area. Given the specific leaf area, mass-based leaf nitrogen content decreased successively in group 3, group 2 and group 1, while given the mass-based leaf nitrogen content, specific leaf area increased successively in group 3, group 2 and group 1. |