Modeling the support factor (P) as a function of socio-economic factors for improved erosion prediction on the hillslopes of Lake Victoria Basin of Uganda

Abstract

A major challenge to erosion prediction using the Universal Soil Loss Equation (USLE) is the uncertainty in parametrizing the support factor (P). This P factor is usually regarded as 1 in areas with no structural management practices. However, in agrarian landscapes which are dominated with agronomic management practices, the P factor is difficult to parameterize. Moreover, the agronomic practices are usually the most simplest and affordable soil and water conservation technologies for mitigating runoff and soil losses in many developing countries. Our objective was to model the support factor (P) as a function of socio-economic factors for adoption of management practices in order to improve erosion prediction. Our methodology involved four (4) steps; namely, (a) estimating potential erosion using RUSLE; (b) establishing the socio-economic for adoption of management practices using Probit regression analysis; (c) integrating socio-economic factors with biophysical parameters to form a Systems Dynamic (SD) model for soil erosion; and (d) validating the Systems Dynamic (SD) model at watershed level using empirical data and RUSLE as the baseline model.  Validation results showed that on Acric Ferralsols at slope gradient 10-15% the potential erosion as predicted by RUSLE model ranged between 120-140 t ha^-1yr^-1.  On the other hand, soil loss as predicted from the Systems Dynamic (SD) model, based on the same slope gradient and soil condition as the case in RUSLE, ranged between 11-50 t ha^-1yr^-1. This accounted for about 67-90% decrease in soil loss. Model outputs were calibrated and validated by field data measured using Un-bound runoff plots (Gerlach Troughs). The results showed that in sole banana soil loss increased step-wise with increasing gradient in the measured and predicted data (P < 0.05); while in sole coffee contradicting results were achieved. We concluded that modelling the support factor (P) as a function of socio-economic factors provides a pragmatic solution to the uncertainty in its parameterization. Generalizing the support factor (P) as one (1) even in areas with agronomic management technology tends to over-estimate the risk of soil erosion. Thus, it can potentially stand out as a dis-incentive that undermines farmers’ efforts to mitigate runoff and soil loss in degraded watersheds.

Key words: 1. Erosion, 2. Geo information science, 3. System Dynamics, 4. Support factor, 5. Uganda

Funding was by SIDA SAREC Project 377 under Makerere University for the period 2016-2022.

DOI: 10.7176/JEES/13-3-06

Publication date: May 30^th 2023