Predicting aggregate stability index in ferralsols

Pedotransfer functions are advantageous for predicting attributes in which the analysis or transport is laborious or expensive, such as the aggregate stability index (ASI), which is directly associated with soil use and ecosystem services. The aims of this study were (i) to predict the ASI in Ferralsols based on four soil attributes: clay concentration (Clay), bulk density (Bd), organic matter (OM) and soil pH; and (ii) to evaluate the contribution of these attributes to the ASI in soils with Clay 600 g/kg. Samples (n = 405) were collected at the 0.0–0.2 m layer in Ferralsols with different clay concentration and agricultural uses in Paraná, Brazil and used for predicting ASI through multiple linear regression using the least squares method with stepwise to test the persistence of the explanatory attributes in the models. The models were trained using 2/3 of the samples (n = 270), and the remaining samples (n = 135) were used to test the model. The ASI could be predicted with good accuracy (Radj2 > 0.77, P