2.3.1. Statistical analysis
We used SPSS 19.0 (SPSS Inc. software products, Chicago, USA) to distinguish differences in SWC among the tested landcover types. Before the semivariance analysis, all data were tested for normal distribution and distribution consistency across landcover types using One-Sample Kolmogorov-Smirnov Test in SPSS; details and statistical parameters are given in Table 2. The variability in SWC for each study plot or landcover was described by the mean, standard deviation, minimum, maximum, skewness and kurtosis. Regression analyses and graphs conducted using OriginPro 9.0 (OriginLab, USA), and data preprocessing was conducted in Microsoft Excel 2013.