Reducing Associated Resource Constraints in Erosion Risk Evaluation in Nigeria

Abstract

Erosion risk determination is time-consuming, cumbersome, and costly. To ensure food security, methods of estimating erosion risk that substantially reduces associated constrains are needed; therefore, this study determined the soil properties central to providing structural stability and using same to build empirical models to forecast  possible response of soil structural framework to the shattering effects of raindrops (D). Five core and auger surface soil samples from five locations were collected across Central Nigeria. A chemical and physico-structural soil properties correlation matrix was produced; ‘D’ was then fitted to a linear multivariate model. Models with the highest coefficient of determination (R²) and minimal standard error with interpretations applicable to real situations were selected for validation on 10 other test soils. Results indicate that the Ca content of soils and soil porosity were the single most important soil chemical and physical property respectively, determining ‘D’, whereas Na (-0.49) and bulk density (-0.73) where the most negatively correlated chemical and physical property to ‘D’. Models 2, 11 and 12 best predicted ‘D’ with ‘r’ values between measured and predicted ‘D’ as 0.97, 0.94 and 0.95, and Model 2 predicted ‘D’ in 80 % of the test soils, whereas Models 11 and 12 did so in 70 % of test soils. However, the cost associated with model 2 was six and four folds higher compared to model 11 and 12 respectively. Based on the related cost, model 11 is the choice, whereas in terms of versatility model 2 is. All models developed were cheap and high in predictive accuracy for ‘D’. The models (2, 11 and 12) with few entries (soil properties) are simpler than existing models.