A dispersibility index for soils and its dependence on other soil properties tested with a group of soils from the Lockyer Valley Uplands, Qld, and the Lower Namoi Valley, NSW

Abstract

The ratio between the weight of a sample of the suspension of a soil given minimal dispersion treatment and a sample of a similar suspension where the particles were completely dispersed provides an index of dispersibility. The weight is proportional to the component elements. Values for K, Mg, Fe and A1 were obtained by HF/HClO4 digestion of a small portion of the suspension, and the ratio of the amount of any of these elements estimated before and after dispersion was found to provide a reproducible index of dispersibility. The relationships between this dispersibility index and exchangeable plus soluble cations replaced by Ag or Cu thiourea, pH and specific conductivity of a 1 : 5 soil/water suspension, and the K, Mg, Fe and A1 content of the suspension of dispersed soil sampled to provide the index were investigated for a set of 105 soil samples that included soloths and grey clays from the Lockyer Valley and grey clays from the Namoi region. It was found that, while the B horizons of soloths from the Lockyer Valley uplands were non-dispersive, the deep subsoils were often highly dispersible. With the grey clays of the Namoi region, surface soils assessed as being poorly structured were often non-dispersive, but such soils showed a rapid increase in dispersibility with increasing depth. A principal component analysis of the data showed that soils from the two sites were different in terms of the cations extractable with copper thiourea, clay content, pH and conductivity. The Mg content of the soil suspension was related to the dispersibility index, i.e. smectitic clays were more dispersible. However, the K, Fe and A1 contents were not related to dispersibility. A simple linear regression with soluble plus exchangeable Na, Ca and Mg, pH, clay content and conductivity accounted for 78-81% of the variation in the dispersibility index. When Na, Ca and Mg were expressed as a percentage of the cation exchange capacity, 76-84% of the variability in the dispersibility index was accounted for. A log-transformation of the data was not as successful overall, with 50% of the variability being accounted for, but with the Namoi soils, which were higher in Ca, 84% was accounted for. To explain the dispersibility observed in soils from a landslip in the southern Lockyer Valley, a separate regression equation was able to account for 88% of the variation. When the sequence of profiles from the landslip site was examined individually, there was a catenary trend in the attributes related to dispersibility. Similarly, regression equations on sub-groups of closely related soils showed that the control over dispersibility was site-specific.