Variation of soil organic carbon estimates in mountain regions: A case study from Southwest China

Abstract Soil organic carbon (SOC) is an important soil component of farming systems and plays a key role in terrestrial ecosystems. However, there is a large variation in SOC estimates at both regional and global scales. The widely used soil type method is usually projected using a planimetric approach, and hence SOC estimates vary notably compared to those generated from more rigorous 3-D surfaces describing rugged terrain. In order to improve the accuracy of SOC storage estimates for regions with complex landforms, this paper examined the causes of variability in estimated SOC storage and SOC density in the upper 1 m soil depth based on 798 soil profiles from Southwest China. The study area is a region with rugged terrain, including the Guangxi Zhuang Autonomous Region, and the Guizhou and Yunnan Provinces. Three methods, the soil profile statistics (SPS), the GIS-based planar soil type (GST-2D), and the GIS-based three-dimensional soil type (GST-3D), were applied to estimate SOC storage. Results demonstrate that the GST-3D, which used soil surface area data, was more accurate than the other two methods. The SOC storages estimated by the SPS and the GST-2D methods were lower than the GST-3D mainly due to the underestimation of soil acreage. Of the four geomorphologic units represented in the study area, the complex landforms with slopes greater than 18.2° covered more than 30%. There is a relatively big difference (> 6%) between planimetric projection area and surface area in this region, making the effect of landform on the estimate of SOC an important factor to be considered. However, such thresholds (30% and 18.2°) as terrain descriptor boundaries need to be further verified in other mountainous regions.