Geostatistical analysis of soil organic carbon in the farming-pastoral ecotone of Northeast China

In this paper, a total of 191 topsoil samples were taken in Tongyu County, a typical area of farming-pastoral ecotone in the Northeast China, and soil organic carbon (SOC) concentrations were investigated using statistics, geostatistics and GIS techniques. Mean concentration of SOC in surface soil of Tongyu County was 0.76%, which was a very low level. The coefficient of variation (C-v) (0.23) indicated the moderate variability of SOC. Significant positive correlations existed between SOC and total N, total P, available N, silt, clay, respectively; negative correlations between SOC and sand, SOC and elevation were observed. The linear regression model of SOC was built based on other soil properties in order to comparing with interpolation results. To obtain an unbiased assessment on the spatial structure of SOC, the spatial outliers were detected using local Moran's I index. The parameters of experimental model fitted for the dataset excluded spatial outliers were better than those for all samples, but the difference was not significant at the regional scale. Based on Kriging interpolation, the spatial distribution of SOC showed a broad regional pattern, with higher values in the eastern part, and lower values in the middle and western part. This spatial pattern was mainly controlled by structural factors, such as climate, parent material and topography.