Abstract
Knowledge of soil chemical properties is indispensable to conduct sustainable land use management in alluvial areas. In this study we developed specific pedotransfer functions for cation exchange capacity (CEC-PTFs) in alluvial soils based on soil depth intervals. A soil data set (n = 1094 samples) at different depths from three different Nile River terraces (lower, middle, and upper) and the lower and upper Blue Nile terraces in Sudan was randomly collected and divided into a training data set (n1 = 900 samples) and a testing data set (n2 = 194 samples) for validation. Soil pH, texture, and organic matter were used as predictor variables to estimate CEC. PTF performance was evaluated with the coefficient of determination (R-2), root mean square error (RMSE), and standard error for the estimate (SEE) between the observed and predicted values. Fourteen predictive equations were developed. Results revealed that the CEC of topsoil layers of the lower Nile River terrace were the most difficult to predict (r(2) = 0.29 for training) while the deep soil layers (60-120 cm) of the Blue Nile terraces were predicted well (r(2) = 0.99 for training). Sixty to 76% of CEC variation of the subsoil of the Nile River terraces could be explained by clay alone. From 85 to 90% of the CEC variation in the deep soils could be explained by organic matter, total silt, and total clay. Validated results indicate that the predictive models based on total clay were less reliable at predicting CEC in the top soil layers. Overall, the CEC-PTFs generated by multiple linear regression models (MLR) provided a reasonable estimate of CEC for most soils investigated.