Methods for modeling spatial variability of soil organic carbon under different land use in middle reaches of the Heihe river basin, northwestern China

The aim of this paper is to identify the influence of land uses on soil organic carbon (SOC) distribution in the middle of Heihe, China. Geostatistical methods including ordinary kriging (OK), spline (SPLIN), inverse distance weight (IDW) and local polynomial interpolation (LP) were compared. It is noteworthy that OK engenders smaller prediction errors than SPLIN, IDW, and LP. We selected the OK method to estimate SOC distribution. Results show that spatial distribution of SOC has an obvious gradual decreasing trend from high to low sections at different layers in the depth of the 100 cm soil profile. Most notable is that SOC content of cultivated land is higher than that of desert, sandlot, salinealkali and naked fields. From a spatial perspective, paddy fields with a long history of cultivation are distributed on the two sides of Heihe river, while land with a shorter history of cultivation is located at a greater distance from the river. Land use change will increase or decrease SOC content. The extension of cultivated land with desert, naked land, sandlot, saline-alkali fields and middle density grassland with low SOC content will enhance the content of SOC in soil profile by a carbon fixation process, although sandy desertification is a reverse process. The results show that cultivation is an important process of increasing SOC content.