Prediction and mapping of erodibility factors (USLE and WEPP) by magnetic susceptibility in basalt-derived soils in northeastern SAo Paulo state, Brazil

Abstract

Spatial assessment of soil erosion is essential for the adaptation of agricultural practices and monitoring of soil losses. In this sense, this study aims to assess the efficiency of magnetic susceptibility (MS) as a predictor of soil erodibility factors (K for USLE model; K-i and K-r for WEPP model) fora detailed mapping of Oxisols with different iron contents in northeastern SAo Paulo State, Brazil. This study was carried out in an area of 380 hectares under sugarcane cultivation in SAo Paulo State. Soil samples were collected in a sampling grid (150) and in a transect (86) and physical and chemical analyses and calculations of the erodibility factors/parameters K, K-i, and K-r were performed. Pedotransfer functions (PTFs) were calibrated using simple linear regression analysis to predict the factors/parameters K and K-i using MS as a predictor variable. The observed values of MS and the predicted values of the factors/parameters K, K-i, and K-r were submitted to geostatistical analysis for constructing maps. Magnetic susceptibility can be used as a predictor of erodibility factors (K for USLE model; K-i and K-r for WEPP model) for Oxisols with total iron content ranging from 1 to 20% Fe2O3, with a precision of up to 60% and an accuracy of up to 85%. The results can guide future studies on water erosion in a tropical environment using magnetic soil data as an environmental covariate in the modeling process for large areas.