Abstract:

To reveal the spatial distribution characteristics of forest soil factors in Shanxi Province, provide reference for the evaluation of forest soil fertility, technical support for sustainable forest management, and a basis for the establishment of forestry sustainable development policy, based on the soil sampling and testing data of 297 sample spots in Shanxi, the study used GIS techniques to do trend analysis and select the optimal semi-variogram model of 7 factors of forest soil, including alkali-hydrolyzed nitrogen, total nitrogen, organic substance, available phosphorus, total potassium, pH and water content. Then, Kriging interpolation was adopted to analyze the spatial variability characteristics of the forest soil factors. The results show that: the optimal semi-variogram model of total nitrogen, alkali-hydrolyzed nitrogen and pH is Gaussian function model; that of total potassium is exponential function; that of the organic matter and water content is trigonometric and spherical function, respectively. Finally, we used Kriging interpolation method to map the spatial distribution of the forest soil factors, to display the distribution characteristics of them in Shanxi directly. The spatial differentiation characteristics of the forest soil factors in Shanxi are obvious. For alkali-hydrolyzed nitrogen, total nitrogen and organic substance, the relatively high contents are mainly concentrated in the mid-high mountainous area along the main mountain range in the province. The content of total potassium is relatively rich. The water content shows obvious difference of more in southeast and less in northwest. As an instantaneous variable, the spatial distribution of available phosphorus is rather disordered. The spatial variability of pH indicates that the whole forest soil in Shanxi is weakly alkaline, particularly in some areas of Shuozhou, the forest soil is strong alkalinity, which could be used as the typical area of saline-alkali land in research.

Key words: forest soil, spatial variability, trend analysis, Kriging interpolation, semi-variogram, optimal model, geostatistics