祁连山区土壤砂粒含量剖面分布模式及其影响因素

Mode of Profile Distribution of Soil Sand Content and Its Affecting Factors in the Qilian Mountains

祁连山区是西北黑河流域的水源,理解祁连山区土壤颗粒组成的剖面分布模式及其与环境要素之间的关系,有助于进一步定量理解土壤中水、肥、气、热的运移和开展颗粒组成/质地的三维制图。以2012—2013年获取的深度≥1 m的69个代表性土壤剖面为对象,分析了含量变异最大的砂粒的剖面分布模式及其与海拔、坡度、坡向、年均降水、平面曲率、剖面曲率、地形湿度指数、NDVI、土地利用、成土母质等环境变量之间的关系。结果表明:祁连山区砂粒含量的剖面分布模式可分为均一型、递增型、递减型、先增后减型、先减后增型和不规则型六类,但其很难用单一的环境变量来定量预测,其根源在于祁连山区成土母质复杂多样,加上复杂的地貌地形、土地利用等导致母质易发生运移而产生不连续性;对应复杂山区的土壤颗粒剖面分布模式的定量预测,还需从复合环境变量的角度进行深入的探索研究。

[Objective] The Qilian Mountains is the main water head of the Heihe River Valley in Northwest China. To balance the need of water between socio-economic development and natural ecology protection, it is very important to understand the hydrology and other scientific basis of water management. Situated at the headwater area of the basin, soil properties especially soil physical properties such as particle size distribution in the Qilian Mountains have significant impacts on hydrological processes. Mode of profile distribution （ PDM ） of soil particle size composition and its relationships with various environmental factors were discussed in an attempt to further understand quantitatively the transport of water, nutrient, gas and heat in soil and to develop 3D predication and mapping of soil particle size distribution. [Method] A total of sixty-nine soil profiles - lm in depth typical of the Qilian Mountains were investigated and soil samples collected from layers of the profiles for fractionation of clay, silt and sand fractions with the laser diffraction method. Sand fraction, the most variable one, was selected to study mode of its vertical distribution in the profile and its relationships with environmental factors, including elevation, slope, aspect, annual mean precipitation, plan curvature, profile curvature, terrain wetness index, land use and parent materials. [Result] Results show that the sand fraction in that region had 6 PDMs, i.e., uniform, increasing, decreasing, increasing-decreasing, decreasing-increasing and irregular, which were little related to any quantifiable environmental factors.We predicted the PDM with random forest, the accuracy was low. It is difficult to predict the PDM with quantifiable environmental information. [Conclusion] Because of the diversity of soil forming parent materials and complex landform and land relief in the Qilian Mountains, migration and discontinuity of parent materials often occurs as a result of land use. So it is difficult to predict PDM quantitatively only with single environmental variables, and therefore, its essential to explore in depth the issue from the angle of composite environmental variables.