Derivation and validation of a pedotransfer function for estimating the bulk density of tropical forest soils

Bulk density (ρ_b), a soil physical property critical for estimating soil carbon storing potentials, is often under-reported in many tropical soil databases because of the difficulty and tedious nature of its measurement in the field. In this study, a pedotransfer function was developed to estimate the bulk density of topsoils (0–30 cm) of Nimbia Forest Reserve, Nigeria. Easily measured soil variables including sand content, total organic carbon and moisture contents were used as predictor variables. A pedotransfer function (PTF) was derived based on multiple linear regression model. Bulk density values for the forest soils varied from 0.78 to 1.28 g/cm³ with a mean value of 1.03 g/cm³. All the metrics used for the validation of the pedotransfer function show the model is statistically significant confirming its usefulness. The RMSE, R-squared and MAE of the pedotransfer function are 0.07498, 0.42231 and 0.058934, respectively. Given that this is the first PTF developed for bulk density estimation in the study area, we recommend a comparison of this new model with others that can be developed using machine learning algorithms and other statistical techniques. The PTF developed in this study will find application in the calculation of the soil organic carbon stocks in the Nimbia Forest Reserve to provide baseline estimates—a decision support tool for sustainable soil management and possible climate change mitigation.