土壤理化性质空间分布及其相关关系分析

使用地理加权相关（GW correlation）与多因素方差分析对张家口地区土壤理化参数空间分异及其相关关系进行分析。结果显示：土壤理化参数具有不同的空间自相关性，各参数之间地理加权相关系数存在正负或大小差异。林地土壤养分含量与饱和导水率显著大于其他土地利用类型（P<0.05）；土壤理化参数之间相关性主要表现为坝上高原>坝下区域、栗钙土>其他土壤类型（P<0.05）。这表明，土地利用是影响张家口地区土壤理化参数空间分异的主要因素，地貌和土壤类型是影响土壤理化参数相关关系空间分异的主要因素，结果可以为认识区域土壤地域分异规律及制定生态脆弱区水土涵养措施提供数据参考。

The Spatial Distribution of Soil Properties and Their GW Correlation: A Case in Zhangjiakou City

The development of soil is decided by many environmental factors, such as climate, lithology, landforms and human activities, etc. The complex of these environmental factors leads to the spatial heterogeneity of soil physic-chemical properties and their interrelations. However, the knowledge of interrelations between soil properties on regional scale is still limited. In this study, the topsoil samples were collected from Zhangjiakou city, and six parameters of soil properties respectively are median grain size, bulk density, field water capacity (FW), saturated hydraulic conductivity (Ks), soil organic carbon (SOC) and total nitrogen (TN) were analyzed using a geographical statistical analysis and nonparametric statistics models of two-way analysis of variance to discuss the spatial distribution of soil property, as well as the spatial distribution of interrelations between theses six parameters. Our results show that, the spatial distribution patterns are different between six parameters of soil property, median grain size of soil exhibits significant spatial cluster patterns (P<0.05). However, other five soil parameters do not represent significant spatial cluster patterns (P<0.05). The spatial distribution difference of six parameters would cause uncertainty of interrelations between these six parameters. GW correlation analysis shows that the interrelations between six parameters represent various spatial distribution patterns. Furthermore, our results found that the spatial distribution of soil properties was only affected by land use which often represents as human activities. The content of soil nutrient (soil organic carbon and total nitrogen) and the saturated hydraulic conductivity (Ks) of soil are larger in forestry land than other land use types (P<0.05). Other soil properties such as grainsize, bulk density, field water capacity of soil do not represent difference in different unites of geomorphic factors, soil types and land use (P<0.05). The spatial distribution of interrelations between soil properties were primarily decided by geomorphic factors and soil types. In most cases of interrelations between soil properties, the GW correlation coefficient showed larger values in Bashang plateau while lower values in other parts of Zhangjiakou city, and the GW correlation also represented larger values in chestnut soil type while lower values in other soil types (P<0.05). In addition, we found the GW correlation coefficients are nearly similar to the Pearson correlation coefficients, suggests that GW correlation can be used to detect the general correlation between variables. Overall, our results suggest that even though the distribution of soil properties are sensitive to human activities such as land use changes or reforestation, however, the interrelations between soil properties are not sensitive to land use. In other words, the influence of human activities on soil properties is comprehensive, not just on single soil property.