The effects of clay and sand additions on the strength of sandy topsoils

The clay contents of sandy soils in south-western Australia are often modified, either intentionally or inadvertently, as a result of management practices and erosion. Although the strength of sandy surfaced soils has previously been shown to be related to clay content, in natural soils the effects of induced changes in clay content on soil strength have not been assessed. Increasing amounts of subsoil clay were added to their respective topsoils in increments ranging from 5 to 20% by weight, and these systematically increased soil strength. A strong log–log relationship between clay content and soil strength explained 69% of the variation, with soil strength further affected by sodicity. This enhancement of soil strength has implications for the practice of claying water-repellent soils, particularly where non-uniform application or poor incorporation results in high concentrations of clay, where very high rates (e.g. 300 t/ha) of application are used, or where clayey subsoils are brought to the surface by deep cultivation or the removal of topsoils by erosion. Drift sand, with a clay content of around 1% and negligible strength, was added in increasing increments to typical topsoils, over the range of 0–100% by weight to replicate the effects of wind-induced deposition and winnowing of clay particles. Increasing additions of drift sand systematically decreased soil strength, with a log–log relationship between clay content and strength of the mixtures explaining 81% of the variation. This suggests that wind erosion, and the winnowing of clay or deposition of drift sand, permanently destabilises soil surfaces by reducing soil strength. It is feasible that strategic applications of sand on the surfaces of soils affected by hardsetting may reduce soil strength and encourage soil structure development and seedling emergence. Keywords: claying, soil strength, water repellency, wind erosion, sandy soils, desertification. Australian Journal of Soil Research 42(1) 39 - 44 Full text doi:10.1071/SR03063 © CSIRO 2004