Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed, Xinjiang, China

Transcription

1 Science Press Journal of Arid Land 2012, 4(4): doi: /SP.J jal.xjegi.com; Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed, Xinjiang, China Jilili ABUDUWAILI 1, Yang TANG 1, Mireban ABULIMITI 2, DongWei LIU 3, Long MA 1 1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi , China; 2 College of Resources and Environmental Science, Xinjiang University, Urumqi , China; 3 Inner Mongolia University, Hohhot , China Abstract: With the classical statistical and geostatistical methods, the study of the spatial distribution and its influence factors of soil water, salinity and organic matter was carried out for 0 70 cm soil layers in Manas River watershed. The results showed that the soil moisture data from all soil layers exhibited a normal distribution, with average values of 14.08% 21.55%. Geostatistical analysis revealed that the content of soil moisture had a moderate spatial autocorrelation with the ratios of nugget/sill ranging from to 0.718, which implies that the spatial pattern of soil moisture is influenced by the combined effects of structural factors and random factors. Remarkable spatial distributions with stripped and mottled features were found for soil moisture in all different soil layers. The landform and crop planting had a relatively big influence on the spatial distribution of soil moisture; total soil salinity was high in east but low in west, and non-salinized soil and lightly salinized soil appeared at the northwest and southwest of the study area. Under the effect of reservoir leakage, the heavily salinized soils are widely distributed in the middle of the study area. The areas of the non-salinized and lightly salinized soils decreased gradually with soil depth increment, which is contrary to the case for saline soils that reached a maximum of km 2 at the layer of cm. The types of soil salinization in Manas River watershed were classified into four classes: the sulfate, chloride-sulfate, sulfate-chloride and chloride. The sulfate salinized soil is most widely distributed in the surface layer. The areas of chloride-sulfate, sulfate-chloride, and chloride salinized soils increased gradually along with the increment of soil depth; the variation range of the average values of soil organic matter content was between 7.48% 11.33%. The ratios of nugget/sill reduced gradually from to with soil depth increment, which shows that the content of soil organic matter has a moderate spatial autocorrelation. The soil organic matter in all soil layers met normal distribution after logarithmic transformation. The spatial distribution patterns of soil organic matter and soil moisture were similar; the areas with high organic matter contents were mainly distributed in the south of the study area, with the lowest contents in the middle. Keywords: soil water and salt; organic matter; spatial distribution; Manas River watershed The soil moisture, salinity and organic matter are important soil properties, and they are important objects for the study of soil spatial variation. The soil moisture is an important part of the terrestrial water resources (Chen et al., 2005), which links soil, land cover and atmosphere systems, influences the atmospheric precipitation and the redistribution of energy between the latent heat and the sensible heat, and then determines the distribution of vegetation cover and the land productivity level (Western et al., 1998; Petrone et al., 2004; Pan et al., 2007). The study of the spatial variation of soil moisture has an important role in analyzing and simulating the fluid movement and the material migration process of soil dissolved matters (Williams et al., 2009), simulating hydrological processes and optimizing agricultural production activities. The salinization of soils is harmful to the production potential of soils, causing sharp reductions in crop yields Received ; accepted Corresponding author: Jilili ABUDUWAILI ( jilil@ms.xjb.ac.cn)

2 442 JOURNAL OF ARID LAND Vol. 4 and changing the adaptability of land cropping; furthermore, the accumulation of salt will also change the environment of plant growing and cause the vegetation degradation. Understanding the soil salinization degree and characteristics will help to improve and recover the ecological environment. The soil organic matter is also an important part of soils. It is not only the source of plant nutritive elements, but also has a big influence on the physical and chemical properties of soils (Chen et al., 2011). Especially in the arid desert areas, the soil organic matter content determines the direction of ecological evolution. Understanding the soil organic matter variation has a very important role in improving the salinized soils (Reeves 1997; Dalal et al., 2011). The soil moisture, salinity and organic matter in Manas River watershed were studied by using the classic statistical and geostatistical methods, which will help us to discover the spatial variation regularity at large scales. The influential factors of spatial variations were also discussed, which will provide a theoretical base for the protection of water and soil resources and the management of oases. 1 Materials and methods 1.1 Spatial sampling With the 1: relief map of the Manas River watershed, 68 sampling sites were selected along the Manas River according to geomorphic features, soil type, and land use conditions. The sampling was conducted in early April, 2010 (before summer irrigation). In this period, the air temperature went up, and the soil water had strong evaporation when the soil unfreezed and melted, resulting in the upward migration of salt. It is a special period for the redistribution of soil water and salt. The study of spatial distributions of soil moisture, salt, and organic matter in this period has a profound influence on the soil improvement. There was no irrigation or fertilizing in this period, and human disturbances were rare. For sampling, both representativeness and uniformity were considered. Sampling points were chosen at 5-km intervals. The sampling depth was 70 cm, which was divided into 5 layers (0 10, 10 20, 20 30, 30 50, and cm). The samples were obtained through drilling, and soil samples obtained from 3 different adjacent drilling locations were mixed to form the samples for testing. The soil samples for soil moisture determination were placed in aluminum boxes after being labeled with serial numbers. The soil samples for determination of soil salt and organic matter were sealed in air-proof bags. All sampling points were localized with GPS, and the conditions of surrounding vegetation cover and plant growth were recorded. Finally, the spatial data and attribute data of the sampling points were established, and projection conversion was performed with ArcGIS to generate the sampling point distribution diagram of the study area (Fig. 1). 1.2 Sample analysis The soil samples for salt determination were brought back to the laboratory for natural air drying, grinding, and 2-mm sieving, and then to prepare leaching solution of 1:5 water and soil ratio. The soil moisture contents were determined by oven-drying method, and the soil ph by the acidimeter method. The soil samples for the determination of soil organic matter were sieved after air drying. The salt was measured as Ca 2+ and Mg 2+ using EDTA complexometric titration; Na + and K +, flame photometer method; CO 2 3 and HCO 3, double indicator titration; Cl and AgNO 3, titration; SO 2 4, EDTA indirect complexometric titration; organic matter, potassium dichromate external heating method. 1.3 Data analysis The Digital Elevation Model (DEM, resolution is 30 m) of the study area was used to extract the two terrain factors of slope and aspect, and then the generated aspect distribution diagram, slope distribution diagram, and the soil sampling point distribution diagram were superposed to obtain correspondent attribute database of the terrain factors and the soil sampling points. The software SPSS was used to calculate the statistical characteristic values of the soil moisture, salt, and organic matter. The normality of the data was used in the Kolmogorov-Smirnov (K-S) method for checking, and logarithmic transformation was carried out on the data failing to meet normal distribution. Both the correlation analysis and the principal component analysis for the soil moisture, salinity, organic matter and ter-

3 No.4 Jilili ABUDUWAILI et al.: Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed 443 Fig. 1 Location of the soil sampling sites rain factors, and land use types were completed in the SPSS. The geostatistical analysis was carried out by software GS+9.0 to determine the theoretical model curve and select the best fitting curve. The spatial interpolation was completed by the software ArcGIS 9.3 to cut the generated GRID diagram through the boundary of the study area, and then to generate the diagrams of soil moisture, salt, and organic matter contents. 2 Results and discussion 2.1 Spatial variation analysis for soil moisture It is observed from Table 1 and Fig. 2 that the spatial variations of soil moisture contents in all layers are similar. The soil moisture semivariance models for the layers of 0 10 cm, cm, cm, and cm are exponential, while that for the layer cm is gaussian. Both types of models are well fitted, but the nugget values are relatively big. The nugget value generally represents the variation caused by the experimental error and the sampling scale smaller than the experimental sampling scale. The nugget values of soil moisture in the study area have no apparent variation with the increment of soil depths with positive values between and 41.1, indicating that there exist different kinds of positive base effects caused by the randomicity of sampling errors, sampling scale, and inherent variation. The nugget values for all soil layers are above 10, of which, that of the layer cm is the most apparent, being The relatively big nugget value also indicates that on a relatively small scale the ecological process may be non-negligible, but whether or not the relatively big value is generated by structure variation on a small scale still needs to be verified by increasing the sampling point density at an even smaller scale (Trangmar et al., 1986). The ratio of the nugget value to the sill value indicates the proportion of spatial heterogeneity brought about by random factors and the whole variance within

4 444 JOURNAL OF ARID LAND Vol. 4 Table 1 Characteristic parameters of semivariance of the spatial variation in soil moisture Depth (cm) Model Nugget Sill (C 0 ) (C 0 +C) Nugget/Sill (C 0 /(C 0 +C)) Range (km) R Exponential Exponential Exponential Exponential Gaussian Fig. 2 Spatial distribution of soil moisture the system; it reflects the spatial dependence of the soil attribution (Chien et al., 1997). According to the dividing criteria of the Cambardella, it represents strong spatial dependence when C 0 /(C 0 +C) 25%, moderate spatial dependence when 25%<C 0 /(C 0 +C)<75%, and relatively weak spatial dependence when C 0 /(C 0 +C) 75%. Table 1 reveals that the structure variance ratios of soil moisture at all soil layers of the study area are between and 0.718, which belong to moderate correlation. The results indicate that for the soil moisture spatial variations of all layers, some external factors such as the cropping system, level of management, and planting system also have non-negligible effects besides the effects of the internal factors of soil parent material, climate, landform, etc. These artificial factors made the spatial correlation of soil moisture reduce and gradually develop to homogenization, which implies that the spatial variation of soil moisture was caused by the combined effects of structural factors and random factors. Except for the layer 0 10 cm, the ratio of the nugget value to the sill value droped with the increment of soil depth, indicating that the influence of the external factors kept on weakening with the increment of soil depth. Accordingly, the influence of the internal factors increased, and the spatial correlation of soil moisture increased. The range represents the area influenced by the environmental factors at different scales, and its values are different under different study areas and sampling scales. Within the range, a variable takes on spatial autocorrelation, otherwise it has no spatial autocorrelation. The range of soil moisture for all soil layers of the study area is between 40.2 and 489 km. The layer cm has the largest range, being 489 km. The study indicates that the sample interval is the main cause resulting in the range difference. Therefore, the range works as guidance to the validity of the sampling design (Yang et al., 2010). The average distance between the selected sampling points is 5 km, while the ranges for all the soil moisture sampling points exceed 5 km, indicating that it is valid to carry out spatial interpolation. 2.2 Spatial distribution structure of soil moisture The soil moisture of all soil layers takes on stripped and mottled structures, and the difference is apparent. The highest values of soil moisture contents for all layers mainly appear in the southwest and northeast corners of the study area, and the lowest values in the middle part (Fig. 2). The spatial distributions of soil moisture for the layers of cm, cm, and cm show similar structures in the middle area. They all have big

5 No.4 Jilili ABUDUWAILI et al.: Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed 445 mottled areas. The distribution areas of the yellow low-value mottled parts of all layers are relatively big (Fig. 2). The reasons may lie in two aspects: (1) the soil had strong evaporation during the sampling period when the air went up in the arid study area; (2) the relatively developed agriculture in the Manas River basin brought about soil loosening which favored evaporation and thus generated low soil moisture content. The purple areas indicating high soil moisture values are mainly concentrated in the piedmont alluvial-proluvial fan areas, which implies the influence of landform on the soil moisture. The relatively high terrain of the alluvia-proluvial fan areas is inducive to abundant precipitation and big soil particles, which results in good infiltration replenishment and water permeability. In addition, the heights of the areas above sea level lead to relatively small evaporation, which is helpful for maintaining the surface soil moisture. Furthermore, at the northeast corner of the study area the soil moisture contents in the layers of cm,20 30 cm, and cm are also relatively high, which is owing to the little soil moisture utilization by the sparse desert vegetation at the edge of the oasis and the surface sandy soils that inhibit the capillary evaporation of the moisture in the under layer (Zhang et al., 2007). 2.3 Spatial distribution characteristics of soil salt Classification of soil salinization High degree of soil salinity harms the growth of crops. Under the same conditions, the higher the soil salt content, the heavier the harm to the growth of crops. In general, the degrees of the soil salinization can be divided into non-salinization, light salinization, moderate salinization, and heavy salinization. The grades of soil salinity can be classified according to the percentage of saline matter to total soil mass Spatial distribution of soil salt The total soil salt content of the study area is high in east and low in west; the non-salinized soil and lightly salinized soil are distributed in the northwest and southwest of the study area; the heavily salinized soil and saline soil are distributed in the southeast and the middle of the study area (Fig. 3). The saline soils in 0 10 cm, cm, and cm soil layers are mainly distributed in the alluvial plains which are close to the desert and has low terrains. The water from the midstream and upstream enters into the channels to discharge salt, and then flows to the downstream area, resulting in high soil salt concentration. In the middle of the study area, saline soils appear in the layers of cm and cm, which is because, on the one hand, the unsmoothed irrigation and drainage areas resulted in years of soil salt accumulation, and on the other hand, a great deal of saline matter did not migrate downward along with water though the air temperature rose to drive the melting of the frozen surface soil after the freezing season of winter. Under different soil depths, the heavily salinized soils are widely distributed in the middle of the study area, and with the increment of soil depth, the distribution area increases gradually due to the infiltration effect of the reservoirs in the Manas Oasis. The discharge of underground water may reduce the accumulation of ground surface soil salt. As the precipitation in spring is relatively rich in the Manas Oasis (Qin et al., 2006), infiltration carries the topsoil salt downward into deeper layers to generate Fig. 3 Spatial distribution of soil salinity

6 446 JOURNAL OF ARID LAND Vol. 4 a larger area of salinized soil. The distribution of non-salinized soil and lightly salinized soil imply the influence of landform and human activities on soil salinity. The south of the study area is mainly characterized by alluvial-proluvial fans with relative deep underground water and rich precipitation, so the soil salinization does not easily occur. While the degree of salinity in the middle of the study area is relatively high for two main reasons: (1) the intense agriculture leached away part of the soil salt; (2) the groundwater exploitation lowers the groundwater table of the area to induce desalinization. But with the increment of soil depth, the areas of the non-salinized and lightly salinized soils reduce gradually, which is mainly because of the less interference of human agricultural activities to the soils and the increase in the influence of the underground water on the soil salinity. The areas of the non-salinized and lightly salinized soils reduce gradually with the increment of soil depth, but that of the saline soil increases, reaching a maximum of km 2 (8.88% of the total) at the layer of cm; The layers of 0 10 cm, cm, cm, and cm have the highest proportion of moderately salinized soils in the whole area, especially for the soil layer 0 10 cm, where the area of moderately salinized soil reaches 1, km 2 (36.18% of the total), indicating that salt usually accumulates in the surface layer. In the layer cm, the area of heavily salinized soil is km 2, which is 29.35% of the total. In the layers of cm and cm, the areas of lightly salinized soil, moderately salinized soil, and heavily salinized soil are about the same, indicating the total salinity distributions of the two layers are basically uniformed (Table 2). To sum up, the topsoil has a relatively high proportion of moderately salinized soil, indicating that the threatening of salinization to the study area is relatively serious, and the growth of crops will be inhibited in a certain degree. Further work is therefore needed with regard to the balance between irrigation and discharge, as well as the establishment of monitoring and early warning systems. 2.4 Chemical types of soil salt Soil water soluble ions have an important influence on the growth of crops. Through spatial analysis of soil water soluble ions of the study area, and plotting the spatial distribution diagram, it is possible to grasp the rules for the spatial distribution of soil salt. The classification of the chemical types of soil salt is helpful to improving saline-alkali soils, preventing and controling secondary salinization, and better developing agricultural production (Wang et al., 2006). The harm of different kinds of lyotropic salt ions to the crops is different. Under the same conditions, the harm of the chloride salts is the greatest, while that of the sulfate type is relatively light. Due to the little precipitation in the Manas Oasis, the seasonal cycle of soil salt is very weak, and the negative ions are mainly Cl and SO 4 2. So Cl and SO 4 2, rather than CO 3 2 and HCO 3, are considered for the study of major negative ion components in the area. Table 2 The area and proportion of soil salinity for different soil depths Depth (cm) Category Total salinity (g/kg) > Proportion (%) Proportion (%) Proportion (%) Proportion (%) Proportion (%)

7 No.4 Jilili ABUDUWAILI et al.: Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed Spatial variation of soil organic matter To obtain the semivariance model and its parameter values, semivariance function analogue to soil organic matter was carried out with the software GS+9.0 (Table 3). The results indicate that the models for the soil organic matter in the soil layers 0 10 cm and cm are both exponential; that for the soil layer cm spherical; and that for the soil layer cm linear. The determination coefficient R 2 of all models is more than 0.5, indicating that all layers can well fit their theoretical models. The models can basically reflect the spatial variation characteristics of the organic matter. The range of nugget values of soil organic matter is between , which is relatively small; it shows that when the scale is smaller than the scale of the current study, the effect influencing the spatial variation of soil organic matter is relatively strong. The soil organic matter of the soil layer cm presents a linear model, the nugget value being the biggest and the nugget effects apparent. It indicates the difference of spatial distribution of soil organic matter at this layer is not big. At the whole study scale, the distributions of soil organic matter have constant variations, and the spatial variation caused by the structural factors at the minimum sampling interval is not apparent. The ranges of sill values for the soil organic matter in the layers of 0 10 cm, cm, cm, and cm are between , with the sill value for the layer 0 10 cm being relatively high, reaching 0.821, and the sill values for the layers of cm, cm, and cm increasing with the increment of soil depth, reaching a maximum of The nugget/sill values of soil organic matter have ranges between , indicating a moderate spatial autocorrelation. With the increment of soil depth, the nugget/sill values reduce gradually, indicating the effects of the structural factors of the climate, parent material, and landform on the gradual increase of the organic matter contents. The nugget/sill value of the layer cm is the biggest, indicating that the spatial distribution of the soil organic matter in this layer is the result of the effect of the structural factors and the influence of artificial factors such as fertilization is very limited. As for the topsoil of 0 10 cm and cm layers, the anthropogenic factors have a bigger influence on the soil organic matter, so the nugget/sill values are relatively big. The range of the soil organic matter is between km, and the difference is relatively big, indicating the variational scales of spatial autocorrelation of soil organic matter contents for all soil layers are different. The ranges of organic matter all exceed the average sampling interval of 5 km arranged in this study, indicating that adopting this sampling interval can meet the requirements for the analysis on the spatial variability of organic matter. 2.6 Spatial distribution of soil organic matter To reflect the spatial distribution of soil organic matter more intuitively, the study input the parameters of the semivariance functional model, carried out Kriging interpolation, and plotted the spatial distribution diagram of soil organic matter (Fig. 4). Soil organic matter contents at all layers take on stripped and mottled horizontal structures. The areas with high organic matter contents are basically distributed in the south of the study area, and the higher organic matter contents can also occur in the soil layers of 0 10 cm, cm, and cm in the north, while the organic matter contents of all layers in the middle are relatively low. The distribution structures of the soil organic matter and the soil moisture are similar. Table 3 Characteristic parameters of semi-variance of the spatial variation of soil organic matter Nugget Sill Nugget/Sill Depth (cm) Model Range (km) R 2 (C o ) (C 0 +C) (C 0 /(C 0 +C)) 0 10 Exponential Exponential Epherical Liner

8 448 JOURNAL OF ARID LAND Vol. 4 Fig. 4 Spatial distribution of soil organic matter With the vertical increment of soil depth, the areas of low soil organic matter contents increased gradually. The areas with high soil organic matter contents for the layers of 0 10 cm, cm, cm, and cm are mainly concentrated in the south of the study area, which is close to the piedmont alluvial-proluvial fans and has a low air temperature due to high altitude, weakening the mineralization speed and promoting the accumulation of organic matter. While in the middle of the study area, several factors function together to speed up the decomposition of organic matter, including frequent human activities, good soil permeability, strong mineralization, and ascending air temperature during the sampling period. The areas near canals and reservoirs, where the threatening of secondary salinization is present, have relatively low underground water levels, resulting in low fertile soil concentrations and organic matter contents. The soil organic matter content in the northern interlacing zones of desert and oasis is also relatively high as a result of the low terrain, gentle slope, and salt and alkali drainage. 3 Conclusions The distribution of soil moisture in the study area shows stripped and mottled inlay structures, and the difference is apparent. The areas with high soil moisture contents are mainly concentrated in the south with relatively high terrains, indicating that the landform has important influences on the soil moisture distribution structure. The areas with low soil moisture contents in all soil layers are widely distributed, but basically in the developed agricultural zones of the oasis. The total soil salt contents are high in east and low in west, and the non-salinized soil and lightly salinized soil both appear at the northwest and southwest of the study area. Under the leakage effect of the reservoirs, the heavily salinized soils are widely distributed in the middle of the study area. Under the effect of underground water discharge and precipitation, the area of heavily salinized soil is less in the surface layer than in deeper layers. The salinized soils in 0 10 cm, cm, and cm layers are mainly distributed in the alluvial plains which are close to the desert; and those in the layers of cm and cm were found in the middle of the study area. The soil organic matter distribution in the soil layers of 0 10 cm and cm follows the exponential model; that of the soil layer cm meets the spherical model; and that of the soil layer cm meets the linear model. The nugget/sill values of soil organic matter range between , which indicates a moderate spatial autocorrelation. With the increment of soil depth, the nugget/sill values reduce gradually. The ranges of the soil organic matter are between and km, and the difference is relatively big. The distributions of soil organic matter take on stripped and mottled structures. Under the influence of topographic factors, the areas with high organic matter contents are widely distributed in the south of the study area. Acknowledgments This research was funded by the National Basic Research Program of China (2009CB825101) and the National Natural Science Foundation of China ( ).

9 No.4 Jilili ABUDUWAILI et al.: Spatial distribution of soil moisture, salinity and organic matter in Manas River watershed 449 References Chen H S, Shao M A, Wang K L Desiccation of deep soil layer and soil water cycle characteristics on the Loess Plateau. Acta Ecologica Sinica, 25(10): Chen L D, Qi X, Zhang X Y, et al Effect of agricultural land use changes on soil nutrient use efficiency in an agricultural area, Beijing, China. Chinese Geographical Science, 21(4): Chien Y J, Lee D Y, Guo H Y, et al Geostatistical analysis of soil properties of mid-west Taiwan soils. Soil Science, 162(4): Dalal R C, Allen D E, Chan K Y, et al Soil organic matter, soil health and climate change. Soil Health and Climate Change, 29(2): Pan Y X, Wang X P, Su Y G, et al Variability characteristic of surface soil moisture content in sand areas covered by different vegetation types. Journal of Soil and Water Conservation, 21(5): Petrone R M, Price J S, Carey S K, et al Statistical characterization of the spatial variability of soil moisture in a cutover peatland. Hydrological Processes, 18(1): Qin R, Cui C X, Luo J, et al Analysis on secular variation of air humidity and precipitation in the Shihezi Oasis. Bimonthly of Xinjiang Meteorology, 29(3): 6 8. Reeves D W The role of soil organic matter in maintaining soil quality in continuous cropping systems. Soil and Tillage Research, 43(1 2): Trangmar B B, Yost R S, Uehara G Application of geostatistics to spatial studies of soil properties. Advances in Agronomy, 38: Wang H, Gong P, Liu G H Multi-scale spatial variations in soil salt in the Yellow River Delta. Geographical Research, 25(4): Western A W, Bloschl G, Grayson R B Geostatistical characterisation of soil moisture patterns in the Tarrawarra catchment. Journal of Hydrology, 205: Williams C J, McNamara J P, Chandler D G Controls on the temporal and spatial variability of soil moisture in a mountainous landscape: the signature of snow and complex terrain. Hydrology and Earth System Sciences, 13(7): Yang Z P, Yang H, Xu X L, et al Spatial heterogeneity of soil moisture and vegetation coverage of alpine grassland in permafrost area of the Qinghai-Tibet Plateau. Journal of Natural Resources, 25(3): Zhang B, Zhang H, Zhang K, et al Study on spatial diversification of soil moisture content of oasis and oasis-desert ecotone in the middle reaches of the Heihe River. Geographical Research, 26(2):