Prediction, mapping, and implication for better soil organic carbon management in Ethiopia

Abstract

Abstract A precise soil organic carbon (SOC) content estimate is crucial soil quality parameter for agricultural produce and ecological safety. Moreover, geospatial modeling of SOC is critical when there are limited laboratory equipment and chemical reagents for soil analysis. This study used geostatistics—ordinary kriging (OK) and inverse distance weighting (IDW)—to map SOC in Libokemkem area, Northwest Ethiopia, for improved SOC management. About 107 soil samples were obtained from the plow layer at a 20‐cm depth and SOC was determined. Statistical Package for Social Sciences version 24.0 was used to generate descriptive statistics, and geostatistical analysis was also performed on the data using ArcGIS platform. The coefficient of determination ( R 2 ) and root mean square error (RMSE) derived from the validation of the predicted maps were used to assess the models. The results revealed homogeneity (coefficient of variation < 10%), low (0.12%–1.74%), and optimal (1.74%–4.06%) mean levels of SOC in study area. The OK showed an R 2 of 0.74 and an RMSE of 13%, and the IDW revealed an R 2 of 0.69 and an RMSE of 14%. The semivariogram results indicate a moderate dependence for SOC with stable, circular, spherical, exponential, and Gaussian models. We conclude that the sustainable monitoring of SOC is significant in enhancing soil quality. However, further study considering all drivers of spatial variability for SOC in the study and other soil sampling approaches improving performance of the prediction models is needed.