Organic carbon, organic matter and bulk density regression models for forest soils in Lanlingxi watershed, Three Gorges Reservoir area

Objective To establish regression models of soil organic carbon, organic matter and bulk density for forest soils and improve the regional soil attribute database in Lanlingxi watershed, Three Gorges Reservoir area.

Method Using forest soil survey data of this watershed, the conversion factor for soil organic matter (SOM) to soil organic carbon (SOC) was established, and the regression models linking soil bulk density (BD) and SOM (SOC) content were built. The whole evaluation consisted of determining the coefficient of determination (R²), Nash-Sutcliffe coefficient of efficiency (E), and the percentage error (P_e).

Result The Van Bemmelen conversion coefficient (0.58) could not be directly applied in this watershed. The proper SOC-SOM conversion coefficient was 0.455, as SOC-SOD conversion coefficients varied from different depth of soil, declining quickly with the increase of depth. The BD-SOM (SOC) regression models built in other regions could not be directly applied to this region. When parameters of the BD-SOM models were optimized, the logarithm polynomial model could be used for this region.

Conclusion Overall, the simulated values of BD-SOM regression models are better than those of BD-SOC models, and it is recommended to use BD-SOM regression models to improve the soil database. Among the optimized BD-SOM models, the recommended model for this study is Federer organic density model with the highest efficiency (E=0.81) and the lowest error (P_e = 5.4%).