A multivariate framework for interpreting the effects of soil properties, soil management and landuse on water repellency

This paper reviews recent progress in relating the incidence of water repellency to other soil attributes, relating the severity of water repellency to different soil management and landuses and determining the value of soil survey in predicting the risk of water repellency. Data sets of soils from south-western Australia are used for this analysis. The relationship between water repellency and other soil attributes such as clay and organic matter contents has been found to be multivariate in nature, with a general form: Water repellency approximately equal to Organic Matter(b)/Clay(c) where a, b and c are constants. Multiple regressions, which use a variety of attributes related to soil organic matter and soil surface area (e.g. clay, silt, amorphous iron contents), can explain up to 63% of the variation in water repellency. Water repellency increases in severity with increasing organic carbon content or decreasing soil surface area and vice versa. This relationship explains the poor or non-existent bivariate relationships between various soil attributes and water repellency reported in several studies and allows the re-interpretation of studies into the effects of liming, zero-till and different rotation lengths on water repellency as each affect soil organic carbon (OC) content. It is also consistent with observations that water repellency is more prevalent on sandy soils and reported reductions in water repellency following applications of clay and fine inorganic materials. It similarly explains the occurrence of water repellency on soils with >5% clay; accumulation of sufficient amounts of OC can induce water repellency in any soil. Soil management and landuse affect water repellency. Water repellency is less severe in soils under crops compared to pastures and this has been considered as being caused by differences in organic matter composition. This is re-interpreted as being due to differences in organic matter amount, as cultivation mineralises and dilutes soil organic matter in the topsoil. Water repellency is also a feature of soils under natural vegetation (Eucalyptus, Banksia spp.) and may be more severe than that for agricultural soils. Increments of organic matter from natural vegetation can induce water repellency to a greater extent than equivalent amounts of organic matter from agricultural species. Thus, water repellency can be considered as a natural feature of soils, rather than a form of land degradation due to farming. Comparisons between the water repellency associated with different soil management or landuses are problematic due to differences in not only OC content and composition but also other soil properties such as surface area, due to the spatial separation of the areas compared. Given the underlying relationship between clay content and water repellency, field texture data from soil surveys can be used to predict the risk of water repellency developing, this providing a basis for optimum management practice. The degree to which this water repellency develops will be dependent on management practices and their effects on soil organic matter content. Quite costly, but profitable, ameliorative management (e.g. clay application) can thus be applied to sites where required, rather than to large areas on a non-specific basis.