The Study of Gaining More Detailed Variability Information of Soil Organic Carbon in Surface Soils and Its Significance to Enriching the Existing Soil Database

Abstract

To meet the increasing demands of precision agricultural and environmental management, more abundant and accurate information is needed to describe soil organic carbon (SOC) vertical variation. Based on 923 soil profiles (collected at the depths of 0–15, 15–30, 30–60, 60–90, 90–120, and 120–150 cm) in the central area of Changhua County, Taiwan, the distribution curve of the SOC content of each profile was fitted by the equal-area spline model, and it was possible to obtain the SOC content at all depths. Taking the 0–5 cm (L1), 5–10 cm (L2), and 10–15 cm (L3) sub-layers as examples, their SOC contents and stocks were compared to the mean values of the average 5-cm-thick sub-layers (Lm) derived from the value of the 0–15 cm layer. The results indicated that the SOC contents and stocks both reduced with increasing soil depths. The mean SOC contents of L1, L2, and L3 were 22.1, 21.0, and 18.7 g·kg−1, respectively, with significant variation, and the values of L2 and L3 were 5.0% and 15.4% lower than that of L1. Similarly, the mean SOC stocks were 1.29, 1.25, and 1.16 kg·m−2 of the L1, L2, and L3 layers, also with significant variation, and the values of L2 and L3 were 4.0% and 10.1% lower than that of L1. Meanwhile, it was found that the SOC content and stock of Lm were both close to the corresponding values in L2, but were significantly different to that of L1 and L3. Furthermore, the interpolation contours of the SOC contents and stocks in L1, L2, and L3 by digital soil mapping also presented regular variation with increasing soil depths, while the contours of Lm had nearly identical patterns to that of L2. The results demonstrate that the typically used mean SOC contents with certain thicknesses calculated from the sampling layer can only approximately inflect the SOC situation at intermediate depths, but the SOC content in the upper and lower parts within the sampling layer varies greatly. Therefore, the actual distribution of SOC varies gradually depending on the soil depth. This study indicates that the combination of the equal-area spline model and digital soil mapping can greatly enrich the current soil SOC database and provide more abundant and accurate SOC content and stock information for precision agricultural and environmental management based on legacy soil database.