Abstract

Three surveys of a pastoral–cropping farming system were carried out over a period of 1 year, using an electromagnetic sensor and real-time-kinematic (RTK)-GPS. The maps produced delineated areas of different apparent soil electrical conductivity (EC_a). These delineated areas were compared with soil units of a conventional soil map and results showed the EC_a map related well to soil-particle-size classes. In addition EC_a could be used to predict groupings of soil phases accurately within one soil type.

Soil coring to depths of 1 m, to determine soil physical and chemical properties, showed EC_a values were moderately well correlated (R² = 0.72) to soil clay percentage, weighted for the soil profile. Soil fertility indicators, Olsen P (R² = 0.61), cation exchange capacity (R² = 0.59), and exchangeable magnesium (R² = 0.76) also related well. The linear regression (R² = 0.76) of EC_a with exchangeable magnesium is thought to reflect the dominant clay mineralogy of the study area, i.e. chlorites weathering to illites and releasing magnesium to the soil solution. Discriminant statistical analysis of results showed point EC_a values could be used to predict 2 major groupings of the mapped soil phases with 100% accuracy. More precise prediction of these mapped soil units is constrained by localised management effects. Elevated EC_a values occur at areas of soil compaction, which have been deduced from measurements of soil strength, aggregate size distribution and visual soil assessment.