To quantify the main factors limiting grain yield in a water-repellent sandy gravel soil.

Crop establishment and nutrient supply had a greater effect on crop growth than soil water repellence at this site. Interestingly, grain yield showed a weak, slightly positive relationship with our measurements of water repellence e.g. grain yield was highest at points that had higher levels of water repellence (data not shown here). The positive relationship between soil water repellence reflects the role of soil water repellent organic matter and soil nitrogen supply. Our results suggest nitrogen supply was a major limitation to growth therefore it is reasonable that higher grain yields occurred where soil organic carbon (soil nitrogen supply) was high. Water repellence may have had a minor effect because it was not severe at this site; the MED values ranged from 0.1 (Low water repellence, King 1981) to 2.1 (Moderate water repellence) and the average was 1.0 (Moderate water repellence). However, the samples for MED measurement were taken in winter when the expression of soil water repellence is lowest.

Gamma thorium provided a reasonable prediction of shoot K concentration and gravel content and provides a feasible method for delineating K management zones on sandy gravels in the Badgingarra area. Figure 2 shows that high shoot K concentration is associated with gravel content and vice versa. The critical shoot K concentration for wheat at anthesis is 1.5%; based on the data shown in Figure 2 the areas where gamma Thorium is less than 40 ppm would be responsive to potassium fertilizer application.

The gravel measurements separated into 2 clusters; non-gravelly soil where the fraction of gravel was less than 15% by weight and gravel soil where the fraction was greater than 15%. The mean Colwell K 0 to 10 cm was 17 mg kg which is well below the critical range of 51 to 57 for high yield potential on deep grey sands. Based on current prices and a wheat yield potential of 2t/ha, the DAFWA K calculator shows profit would be maximized with 40 kg K/ha applied. However, the efficacy of topdressed K on this sand may be low. The mean Colwell K for the gravel soil was 24 mg / kg which is also well below critical levels but shoot K concentration was adequate. This suggests that the crop is accessing K from the subsoil and that tissue testing may be an important component of assessing nutrient supply from different parts of the sandy gravel landscape.