Based on the daily rainfall data of ten rain gauging stations from 1971 to 2010 in the XiangxiRiver watershed of the Three Gorges Reservoir region, we studied the spatio-temporal distribution characteristics of rainfall erosivity estimated with our focus on the annual and inter-annual trends of therainfall erosivity (R-factor) for the watershed through Mann Kendall nonparametric tests and Kriging space interpolation of Arcgis software. The results are shown as follows. 1) From 1971 to 2010, theannual rainfall erosivity of the watershed changed from 2 465.26 to 7 419.29 MJ·mm/ (hm²·h), with an，average annual rainfall erosivity of 4 535.63 MJ·mm/ (hm²·h). 2) There was a great variation of inter-annual R-factor with the maximum three times greater than the minimum. The intra-annual distribution of rainfall,erosive rainfall and rainfall erosivity were highly concentrated with a single peak, mainly during the period from April to October, accounting for 85.4%, 92.4% and 94% of whole year values, respectively. 3) For the whole Xiangxi River watershed, annual precipitation and rainfall erosivity had no significant tendencies of change over 40 years. 4) The results also showed that the spatial distribution of R in the Xiangxi River Watershed decreased rapidly from the west to the east.

We investigated the soil erodibility (K) of red soil in Guangdong Province and established its predication model based on field artificial rainfall events and indoor measurement of soil physical and chemical indexes. There was a significant correlation between K and clay, sand and organic matter content (P < 0.01), with Pearson’s correlation coefficient of - 0.920 4, 0.925 9 and - 0.6424,respectively. The stepwise regression analysis (SRA) filtrated the single factors and interacted terms,which significantly affected K. The interacted influence between the silt and organic content was the largest. The SRA prediction model for K had a good estimation, and the relative error between predicted and observed values was less than 25%. Using the single factors and interacted terms filtrated by SRA as the input parameter, and K data as the output parameter, we established the BP predication model for K. The grey relational analysis indicated that the optimal network structure was 11-11-1 and the best training algorithm was the Levenberg-Marquardt. The relative error of over 90% data was less than 10% by the BP prediction model. The precision of BP model was obviously better than that of SRA model, and the former can more accurately reflect the inherent law between K and factors.

The soil crust on slope farmland is a sort of compact slab structure formed by rainfall on the soil surface, and it has the characteristics of being smooth, compact, and easy to become ramous. Previous researchers have done a multitude of experiments regarding development mechanism, process of formation and factors affecting them, and their findings indicated that the soil crust has effects in decreasing infiltration, improving soil surface runoff amount, influencing the process of sediment yield, and reducing the biomass and yield of crop and so forth. Protecting slope farmland against soil and water loss has become an important issue of constructing bio-environment and realizing sustainable development in China. It is evident that the soil crust is a special underlying surface and it has significant effect on runoff amount and sediment yield on the slope. Planting is the main activity on slope farmland, and it has certain influence on the soil crust. The objectives of this study were to investigate the effects of soil crust on runoff and sediment yield under soybean cover and to provide theoretical basis for soil and water conservation and rational utilization of slope lands. Our study was carried out in the southern fringe of the Loess Plateau in 2010. Soil in the study area is Eum-Orthic Anthrosols. The intensity of the precipitation in this research was 80 mm/ h, with each rainfall lasting for 60 min. We took slopes planted with soybean as object, and by using simulated rainfalls we investigated the existence of soil crust before rain and influences of its thickness and coverage on runoff and sediment yield. Soil crust before rain had considerable effects on runoff and sediment yield. Analysis of the effects of soil crust on runoff and sediment yield on slope land under soybean cover showed that, the volume of runoff was higher on soil crusted slopes and amount of soil loss was much lower compared to that on slopes without soil crust. The soil crust made the beginning time of runoff yield earlier. With the growth of the soybean, the beginning time of runoff yield on the slope without crust was postponed much. The runoff and sediment yield became lower on slope land with soybean growing, and the soil crust on slope made the surface smoother. Besides, the structure of the soil crust is compact, which speeds up the runoff, lower the infiltration of rainfall and runoff, and keeps the runoff amount on slope with prophase soil crust at a relatively high level. On bare soil, the runoff and sediment yield on soil crusted slopes was 5.27% higher and 27.66% lower than that on slopes without soil crust, respectively. On slopes with soybean cover, the runoff and sediment yield on soil crusted slopes was 24.81% higher and 14.26% lower than that on slopes without soil crust, respectively. The crust has great influence on the processes of the runoff and sediment yield, which have significant difference between the slope with prophase soil crust and without soil crust. On the whole, the fluctuation of the process of runoff and sediment yield was smooth on slope with prophase soil crust, while the volatility of slope without soil crust fluctuates greatly. The presence of soybean cover increased the difference in runoff yield between slopes with and without soil crust, but reduced the difference in sediment yield. The effect of soybean on the differences of runoff and sediment yield between slopes with and without soil crust became stronger with the growth of soybean plants. Effects of thickness and coverage of soil crust on runoff and sediment yield were not significant. The influence of soil crust on the soil erosion varied greatly due to different soil types, terrain, rainfall, crops and other factors, which makes the problem become more complex. The research on the principle of the influence of soil crust on soil erosion is helpful for illustrating the progress of such effects, which provides reference for agricultural sectors to weigh the pros and cons in the real production.

Under global warming, desertification has brought threats to the ecological environment of arid and semi-arid regions like the Loess Plateau. It is helpful to understand the spatial-temporal trend of desertification for ecological environment projects and desertification management in these regions. The vegetation cover and its spatial distribution can reflect the degree of desertification effectively. It is anefficient way using the remote sensing techniques to observe the trend of vegetation cover and subsequently the development of desertification. The data of GIMMS AVHRR NDVI from 1986 to 2006 on the Loess Plateau were used, and the vegetation cover index was then calculated based on the principle of the dimidiate pixel model. The correlation between degree of desertification and vegetation cover was determined based on researches of wind and water erosion in the study area. The results showed that about 64% of the plateau presented moderate and severe degrees of desertification with 10% - 50% vegetation cover. A third of total plateau belonged to the slight and non-desertification degrees with vegetation cover greater than 50%, and only 3% areas were subjected to extreme desertification with vegetation cover less than 10%. From 1986 to 2006, desertification generally showed a decreasing trend on the Loess Plateau. The areas with moderate and severe degrees of desertification decreased significantly, while the areas without desertification had an apparent increase. However, the area with extreme degree of desertification expanded. The degree of desertification showed a changing rule with a decadal interval. The moderately and severely desertified areas shifted reciprocally and frequently. The areas transferred between slight and non-desertification degrees were sparsely scattered. The extremely desertified areas appeared mainly along the Yellow River of Ningxia and Gansu provinces.

The purpose of this study was to understand the characteristics of variations in soil surface micro-topography and soil erosion, providing evidence for the effective prevention and control of soil erosion in sloping farmland of purple soil region. Based on the indoor simulated rainfall experiments, the spatial variation of soil surface micro-topography was studied in sloping farmland with cross ridge in purple soil region under the conditions of increasing and decreasing rainfall series. Soil surface roughness was used to indicate the dynamic changes in the soil surface micro-topography. On this basis, soil surface roughness was calculated and DEM was established, and the effect of soil surface micro-topography on runoff and sediment yield was explored. The results showed that the change of soil surface elevation was characterized primarily by the reduction in increasing and decreasing rainfall series, concentrated mainly in the range of -2 to 0 cm. Under the increasing rainfall series, soil surface micro-topography changed obviously in the middle course of rainfall and the rill erosion kept intensifying. However, in the decreasing rainfall series, soil surface micro-topography began to collapse in earlier rainfall course. In the late rainfall course, rills tended to broaden and deepen, but the starting point of rill formation was relatively stable, which did not extend and develop along the ridge direction. In the increasing rainfall series, the variation of soil surface roughness was not obvious in the first two rainfalls, whereas the soil surface roughness showed a significant increase in the third rainfall. The soil surface roughness increased first and then decreased under the decreasing rainfall series. The difference of sediment yield was not significant between the increasing and decreasing rainfall series. The sediment yield in the increasing rainfall series was 3.9% higher than that in the decreasing rainfall series, while the volume of runoff under the condition of the increasing rainfall series was 14.8% lower than that in the decreasing rainfall series. When cross ridge-breaking happened, surface water movement dominated in surface runoff on the ridge-breaking slope. The sediment yield depended on the soil loss of ridge breaking and runoff washing under high-intensity rainfalls in slope farmland of purple soil region. Compared with the decreasing rainfall series, the initial time of runoff and sediment yield was later in the increasing rainfall series. Effect of rainfall conditions on runoff was greater than the effect of erosion yield. Cross ridge could be used to effectively reduce soil erosion in light-intensity rainfalls in slope farmland of purple soil region. However, ridge breaking was likely to happen for the cross ridges of sloping farmland in purple soil region, and it might increase soil erosion when the rainfall intensity increased to 1.5 mm/ min.

In order to investigate the effect of litters on the water storage in the upper reaches of Chishui River, we measured water holding capacity and water loss of 27 major tree species in the study area. The results showed that the amount of litters was 0.15 -4.50 t/ hm², natural moisture content was 10.23% - 137.66%, the maximum water holding rate was 3 122.83 - 9 555.80 g/ (kg·h), the maximum water holding depth was 0.04 - 1.70 mm, effective water interception depth was 0.02 - 1.27 mm, the maximum rate of water loss was 209.52 - 2 423.21 g/ ( kg·h), moisture content was 85.02% - 256.18% after water loss for 24 h, and the maximum water loss depth was 0.01 -0.43 mm. Moreover, the rates of water holding and water loss of litters were higher in the earlier stage, declined quickly and became stable eventually. Several major factors contributing to water storage capacity of litters were determined as the amount of litters, the maximum water holding depth, effective water interception depth, the maximum water loss depth, leaf life stage and leaf texture. At last, litters were divided into three main functional groups according to their capacities of water holding, water loss and the leaf characteristics. Leaf life phase could be an indicator to evaluate the water-storing capacity of litters.

We investigated the rainfall redistribution of canopy on Pinus tabulaeformis plantations in the semi-arid areas of the Loess Plateau based on the data collected from 2012 to 2013 using the permanent plot method. The canopy interception was simulated separately by using Wang's and Cui's interception models, and the applicability of the two models were compared. The results showed that: 1) There was a linear relationship between rainfall and throughfall (R² = 0.978 9), but an exponential relationship between rainfall and canopy interception (R² = 0.867 3), and the average of interception rate was 25.5%. 2) The determination coefficient of Wang's interception model was 0.724 5 and it was 0.755 4 for Cui's. In general, Cui's interception model was slightly better than Wang's.

Preferential flow is a major infiltration process in unsaturated soils. We conducted a dye-staining experiment in Beijing Jiufeng National Forest Park, in order to analyze the differences of soil porosity and soil moisture in preferential flow area and matrix flow area, as well as the effect of antecedent soil moisture (ASM) on preferential flow characteristics (i. e. , maximum dyed soil depth, 50% dyed soil depth, dyed soil percentage). Results indicated that: 1) the non-capillary porosity and the total porosity of the soil in preferential flow area were statistically higher than those in the matrix area, and no statistical difference in capillary porosity was found between the two areas; 2) soil moisture in the preferential flow area decreased by 21% after the experiment, while there was no difference in soil moisture in the matrix flow area;3) ASM was positively correlated with the dyed soil percentage (R² =0.679) and 50% dyed soil depth (R² = 0.527); in contrast, no apparent correlations were found between ASM and the maximum dyed soil depth (R² =0.236).

Desertification and salinization of the Minqin Oasis in western China are getting more and more serious in China. Ephedra przewalskii is a desert salt bush widely distributed in Minqin. In this study we took soil surrounding E. przewalskii as research object, and investigated the main characteristic factors of salinization of soil around E. przewalskii by using correlation analysis and main component analysis, aimed to provide a theoretical basis and technical support for the improvement and utilization of soil salinization in the Minqin Oasis. The results showed that E. przewalskii could reduce the pH of its soil around it. The soil pH value varied from 8.21 to 8.69 in the horizontal direction from 0 to 30 cm away from the plant; in the vertical profile, the soil pH first dropped with the soil depth increasing and then ascended, showing a “double S" pattern. The soil pH value changed from 7.89 to 8.95 at a distance of 30 to 60 cm away from the plant, while it first decreased and then increased with the soil depth increasing in the vertical profile. The soil pH value ranged from 8.50 to 8.96 at a distance of 60 to 90 cm away from the plant, while it decreased with the soil depth increasing in the vertical profile. The degree of soil salinization was mainly influenced by chloride, sulfate and bicarbonate, and profile distribution of soil salinity was in bottom accumulation. Ca²⁺ , SO^2-₄ , HCO^-₃ , Mg²⁺ and K⁺ were the main characteristic factors for the salinization of soil 30 cm away from E. przewalskii, where soil salinity existed mainly as sulfate and bicarbonate, and chloride followed. Total salt content, Cl^- , Na⁺ , Mg²⁺ and pH were the main characteristic factors for the salinization of soil 60 cm away from E. przewalskii, where soil salinity existed mainly as chloride, and sulfate and bicarbonate followed. Total salt content, Na⁺ , Cl^- and Ca²⁺ were the main characteristic factors for the salinization of soil 90 cm away from E. przewalskii, where soil salinity existed mainly as chloride, and bicarbonate and sulfate followed. Our study suggests that E. przewalskii can be used to alleviate soil salinization in the Minqin Oasis.

In order to specify the optimal scope of salinity of irrigation water which could be used for freezing irrigation with saline water, the effect of freezing irrigation with saline water on coastal saline land soil was studied by simulated experiment on soil columns under different salinities of irrigation water. Compared with the pre-treatment samples, the top soil layer (0 -50 cm) of the soil columns treated by the irrigation water of low salinity (8 g/ L, 16 g/ L) was desalinized, while that treated by the irrigation water of high salinity (24 g/ L, 32 g/ L) was salinized. When the frozen layer of soil thawed completely next spring, the top soil layer (0 - 20 cm) of the soil columns treated by the irrigation water of high salinity (24 g/ L, 32 g/ L) showed drastic salinization, while that of treated by the irrigation water of low salinity (8 g/ L, 16 g/ L) did not. Therefore, the top soil layer (0 - 50 cm) showed desalinization by freezing irrigation with low saline water when frozen soil layer thawed next spring. The lower the salinity of irrigation water was, the better the effect of desalinization was. However, in the same situation, the top soil layer (0 -50 cm) showed salinization by freezing irrigation with highly saline water; therefore, the salinization of coastal saline land soil may be worsened if the soil was treated by freezing irrigation with high saline water. The results of the experiment suggest that the irrigation water with a salinity lower than 16 g/ L is optimum for freezing irrigation under the conditions of this experiment.

A pot experiment was conducted to explore the effects of soil conditioners on soil properties and growth of Juglans regia seedlings, through the determination of soil physical properties and the indexes of growth and physiological on J. regia seedlings with different treatments. The effects of three self-made multifunctional soil conditioners ( I，II ，III) on the growth of J. regia seedlings were compared with an ordinary soil conditioner by using variance analysis and factor analysis on soil physical properties and the physiological indicators. The results showed that the application of soil conditioner could significantly improve soil porosity and soil water holding capacity, decrease soil bulk density, and promote the growth of J. regia seedlings. Soil conditioners had significant influences on soil properties and growth of J.regia seedlings. The effects of self-made multifunctional soil conditioners were better than the ordinary soil conditioner. The length of shoots treated with self-made multifunctional soil conditioner I was 44.1% higher than that of the control group and 36.3% higher than that of the ordinary soil conditioner. Compared with the control group and the ordinary soil conditioner, the ground diameter of self-made multifunctional soil conditioners I increased by 20.4% and 11.6%, and the biomass of self-made multifunctional soil conditioners I increased by 32.3% and 8.2% ,and the photosynthetic rate of self-made multifunctional soil conditioner I increased by 35.6% and 15.1%,while the transpiration rate of self-made multifunctional soil conditioner I decreased by 26.7 % and 15.6%, respectively. Hence, the self-made multifunctional soil conditioner I can be used as the optimal formula. This study had a certain guiding function for exploring soil conditioner of environmental protection composite.

We analyzed the content and density of soil organic carbon along a chronosequence of Pinus tabulaeformis plantations (9, 23, 33 and 47 years old) in the Loess Hilly Region, and explored the relationships between soil organic carbon content and main factors affecting it, i. e. , total nitrogen, soil particle density, root biomass and litter biomass. The results showed that the average contents of soil organic carbon were 4.9, 5.9, 9.2 and 6.5 g/ kg,and the densities of soil organic carbon were 63.0, 66.8, 100.7 and 72.5 mg/ hm², for 9-, 23-, 33-and 47-year-old plantations, respectively. Both the content and density of soil organic carbon increased first (9 - 33 years) and then decreased (33 - 47 years) over stand age. In addition, the content and density of soil organic carbon decreased gradually as soil depth increased. The density of organic carbon in surface soil layer (0 - 30 cm) accounted for 48.5% - 57.9% of that of the entire observed soil layer. 2) The soil organic carbon content had a significantly positive correlation with total soil nitrogen, root biomass and litter biomass. However, there was a significantly negative relation between soil organic carbon and soil particle density. The correlation coefficients between soil organic carbon content and root biomass and litter biomass decreased gradually with soil depth increasing. The result of stepwise regression indicated that total soil nitrogen and litter biomass were the main factors affecting soil organic carbon content.

The spatial heterogeneity of soil organic carbon (SOC) levels is important in researches of both global carbon - nitrogen cycle and climate change. By employing geostatistics methods and GIS technology, we studied the spatially heterogeneous distribution of SOC in the Yaoxiang small watershed in the hilly area of central and southern Shandong Province. Results showed that: 1) SOC, total nitrogen and total phosphorus contents showed a trend of declining with the increase of soil depth, and there was an extremely significant difference in SOC content between 0 - 10 cm and 10 - 30 cm soil layers, and significant differences in total nitrogen and total phosphorus contents were found between 0 -10 cm and 10 -30 cm soil layers. 2) SOC content in 0 - 10 cm soil layer had an extremely significantly positive correlation with total nitrogen content and elevation, and had a significantly positive correlation with total phosphorus content; SOC content in 10 -30 cm soil layer had a significantly positive correlation with total nitrogen and elevation. 3) Semivariogram model of SOC content belongs to the Gaussian model, and SOC content in 0 -10 cm soil layer had a strong regional pattern, and that in 10 -30 cm soil layer showed a moderate spatial correlation. 4) SOC content was higher in the eastern and northwestern parts and lower in the central and southern parts of the small watershed; the distribution of SOC diminished gradually from the eastern and northwestern parts to central and southern parts, which was basically consistent with the trend of DEM in the Yaoxiang small watershed. 5) Fractal dimension of SOC content in 0 -10 cm soil layer was smaller than that in 10 -30 cm soil layer, and the spatial pattern of SOC was relatively simple; in contrast, the spatial pattern of SOC in 10 -30 cm soil layer was more complex. The fractal dimensions of SOC content in different directions showed that the east-west (E-W) direction was the advantageous pattern of SOC distribution in the Yaoxiang small watershed.

Based on the data from 32 meteorological stations in extremely arid region in northwestern China from 1960 to 2013, we analyzed the characteristics of temperature and precipitation with the methods of the linear regression, cumulative anomaly and Mann-Kendall test. The results indicated that the climate of 30 out of 32 stations in the study region showed a significant trend of warming by 0.034℃/a (P<0.01) in recent 54 years with an average annual temperature of 9.3℃. The increases of temperature in winter and spring seasons were higher than those in summer and autumn. Spatially, the highest temperature was found in Tarim Basin, and the lowest in the southeast of the study region. The average annual precipitation (55.35 mm) also showed a significant trend of increasing by 0.46 mm per 10 years (P<0.01), and 29 stations presented such trend, of which 11 passed the significant test. The highest rate of increase was found in summer, and the lowest in spring. The northwest and east part had higher precipitation than the south part of the study region. In general, the climate in the recent 54 years tended to be warmer and wetter.

In order to explore the change law of plant diversity in Pinus tabulaeformis and Picea meyeri stands at different stand ages and figure out the driving factors for plant diversity, we took Margalef index, Simpson index, Shannon-Wiener index and Pielou index as criteria for measuring biodiversity. We analyzed and compared the characteristics of biodiversity and soil physico-chemical properties of Pinus tabulaeformis stands and Picea meyeri stands at different stand ages in Youyi Reservoir, Erdaoying Basin, Xinghe County of Inner Mongolia. The results indicated that: 1) The biodiversity index of Pinus tabulaeformis artificial forest was greater than that of Picea meyeri artificial forest. The biodiversity of herbaceous community presented a nonlinear rising tendency with the increase of forest stand age. Five-year-old Pinus tabulaeformis forest had the highest biodiversity index of 2.09, whereas the biodiversity index of the thirteen-year-old forest was the lowest, only 1.63. The Margalef index of Picea meyeri artificial forest was higher than that of Pinus tabulaeformis forest, 3.79 in thirteen-year-old Picea meyeri and 2.16 in two-year-old Pinus tabulaeformis artificial forest. 2) During the process of vegetation restoration succession, perennial herbs replaced annual herbs gradually, ongoing towards a positive direction. 3) Among various edaphic factors, the content of soil water, soil anti-scouring ability and SOC had direct effects on biodiversity index. The results of our study could provide a theoretical support for vegetation restoration and rehabilitation in the forest-steppe ecotone.

The objective of this study was to explore the characteristics of diameter at breast height(DBH) classes and tree height classes of 358 gap border trees (GBTs), and analyze the relationships among DBH class,tree height class, distribution of rate of crown inclination (RCI) and forest gap size.The gap border trees were selected from 32 gaps. We fit the distributions of DBH and tree height classes with Weibull and normal distribution models, and then verified their fitness by Kolmogorov and Chi-square tests. The results showed that the distribution of DBH classes of GBTs followed Weibull distribution rather than normal distribution,but the distribution of tree height classes complied with bothWeibull and normal distributions. The RCI of 65.36% of GBTs ranged from 0.5 to 0.7 while 15.08% of GBTs had regular crowns. There was a highly significant difference in RCIs between different tree species. No significantly linear correlation could be found between mean RCI in a single gap and its size of area. The proportion of gap with 10 GBTs was the highest,reaching 21.88%,and each gap had averagely 11.19 GBTs.

In order to accurately carry out the ecological service evaluation of the other forests of the same forest types in the same ecological unit that lacked ecological monitoring data, we propose the forest ecological function correction coefficient based on the biomass in this paper in order to adjust the ecological monitoring data for better application in the rest of the sites of the same forest types in the same ecological unit through the conversion of biomass. As an example, the ecological service of the project of conversion of cropland to forest in Liaoning Province was evaluated. The evaluation showed that the total ecological service value of the project of conversion of cropland to forest in Liaoning Province was 49.194 billion yuan in 2013. In six ecosystem service evaluation indicators such as water conservation, soil conservation, carbon fixation and oxygen releasing, nutrient accumulation, atmosphere purification, and biodiversity protection, the proportions of the value of water conservation and biodiversity protection to the total value were the largest, 26.84% and 26.30%, respectively. Our study would provide reference for the revision of Specifications for assessment of forest ecosystem services and the accurate evaluation of the forest ecosystem service at regional scale.

The objective of this study was to explore the responses and acclimations of wild Spiraea trilobata and Rosa xanthina to soil water and nutrient conditions at different slope positions. We selected five wild adult plants for each of the two shrub species grown at three different slope positions (upper slope, mid-slope and lower slope) on a south-facing slope as experiment materials. In growing seasons of 2013 and 2014, we monitored soil water contents at the three slope positions once a week from June to September. Soil nutrient contents were measured in July 2014. In August 2014, we measured leaf morphological traits and nutrient contents of the plants. The results showed that the average soil water contents decreased significantly with ascending slope positions. Although soil pH, contents of total phosphorus and potassium were generally low and similar at all three positions, soil nitrogen and contents of organic matters were significantly lower at upper slope than at mid-and lower slope sites. Meanwhile, leaf traits of the two shrub species grown at different slope positions differed significantly. With slope position rising, both leaf area and mass of the adult plants of the two shrub species decreased, but the specific leaf area kept unchanged. Furthermore, contents of soluble sugar, non-structural carbohydrate and total nitrogen and total phosphorous in plant leaves of both species at upper slope position were significantly lower than those of plants grown at mid-and lower slopes, while the mass ratio of total carbon to total nitrogen was significantly higher at upper slope position. However, contents of organic carbon, total phosphorous and soluble protein in leaves did not change with slope positions for both species. Although there were significant differences in most leaf traits of S.trilobata and R. xanthina, their responses to changing slope positions were similar, both in direction and in magnitude, well coupled with the soil water gradient created by the slope. The above results suggest a similar long-term response and acclimation of leaves of wild S. trilobata and R. xanthina plants to soil water gradient in the natural environment.

In order to study the effect of water and nutrient utilization in the Triticum aestivum-Zea mays cropping system, we conducted the experiment of integrated water and fertilizer management in Yuanyang Experimental Station of Crop Water Use,Ministry of Agriculture. The treatments of nitrogen fertilizer were set as base fertilizer, topdressing once and twice. The water conditions were set as irrigation once, twice and three times and the volume of each irrigation was 450 m³ / hm². The results showed that the combination of supplementary irrigation and fertilizer played a positive role in the growth and development of T. aestivum and Z. mays, which increased the spike length of T. aestivum by 0.2 - 0.8 cm, the number of grains per spike by 2 -10, the thousand-grain-weight by 2 -9 g, and the number of panicles by 90 000 -570000 per hectare. For comparison, the same treatment increased the spike length of Z. mays by 0.48 -1.82 cm, the number of grains per row by 2.0 -8.4, the weight of five panicles by 15 - 374 g, and the hundred-kernel-weight by 2.0 -13.0 g. At the same time, the yield of T. aestivum was increased by 9.85% - 37.93% under supplementary irrigation and nitrogen fertilizer topdressing compared with the control. The yield of T. aestivum under supplementary irrigation combined with topdressing treatments was increased by 9.92% -25.66% compared to that only with application of only basal fertilizer, in which the treatment with irrigation for three times and topdressing twice was the optimum. Compared with the treatment of topdressing once, topdressing twice increased the yield by 3.95% -6.11%, in which the treatment with irrigation twice and topdressing twice was the optimum. The yield of Z. mays was increased by 11.05% -46.62%, 18.71% -32.03% and 2.80% -5.42%, respectively, compared with the control, application of basal fertilizer and topdressing once treatment. The total yield of T. aestivum and Z. mays was increased by 10.46% -42.36%, 14.37% -28.87% and 3.34% - 5.75%, respectively, compared with the control, application of basal fertilizer and topdressing once treatment, showing the same tendency with the increase of T. aestivum. The comprehensive water use efficiency of T. aestivum and Z. mays was increased by 0.95 -5.41, 0.56 - 3.81 and 0.35 - 0.67 kg/ m3, respectively, compared with the control, application basal fertilizer and topdressing once treatment. The treatment with irrigation once and topdressing twice was the best. According to our results, the optimum water-saving and yield-increasing treatment was irrigation twice with topdressing twice or irrigation once with topdressing twice.

In order to reveal the spatial variation of grain crop water requirement (W_R ) in the North China Plain under climate change, we downscaled the meteorological elements of RCP4.5 scenarios through the methods of SDSM, calculated the crop water requirement by Penman equation, and evaluated the water deficiency rates of piedmont plain, central alluvial plain and littoral plain through statistical methods. The results indicated that: 1) There was a strong relationship between crop W_R  and annual average maximum temperature. Under present climate conditions, a rise of 1℃ would increase the annual crop W_R  by 38.8 mm in Baoding, 44.8 mm in Dezhou, and 50.6 mm in Cangzhou. In RCP4.5 scenarios, the annual crop W_R  would increase by 29.1 mm in Baoding, 44.2 mm in Dezhou, and 39.6 mm in Cangzhou as the annual average maximum temperature elevates by 1℃. 2) From the current climate conditions to the RCP4.5 scenarios, there is an increase of crop irrigation water requirement (I_R)to different degrees, with the largest increase of 5.4% in Cangzhou and the lowest 4.8% in Baoding. Under the present climate conditions, the I_R is 717.14 mm in Baoding, 729.52 mm in Dezhou, and686.32 mm in Cangzhou. In RCP4.5 scenarios, the I_R is 751.50 mm in Baoding, 729.52 mm in Dezhou, and 723.64 mm Cangzhou. 3) From the present climate status to the RCP4.5 scenarios, thewater deficiency rates (W_a) of the piedmont plain and central alluvial plain will increase, while that of the littoral plain has the tendency of declining. 4) Under the present climate conditions, the annualaverage Wa(2011—2070) is 53% in Baoding, 47% in Dezhou, and 48% in Cangzhou. In RCP4.5 scenarios, the annual average W_a is 55% in Baoding, 49% in Dezhou, and 47% in Cangzhou. 5) Through building the evaluation index system of natural water deficiency rates, we can find that the highest level of W_a is greater than 50%, the medium level 30% -50%, and the lowest level less than 30%. 6) The probability density function (PDF) of W_a fits Weibull distribution well. Under the present climates, the highest level of maximum probability of W_a is 0.634 9 in Baoding, followed by Dezhou, 0.490 2, and the lowest in Cangzhou, 0.476 0. 7) In RCP4.5 scenarios, the maximum probability of W_a is at a high level in Baoding (0.698 2), and the next is at a medium level. In Dezhou, the maximum probability of Wa is also at a high level, reaching 0.515 2, and the next is at a medium level. In Cangzhou, the Wa of maximum probability is at a medium level (0.506 2), and the next is at a high level.