Soil-extractable sulfur and pasture response to applied sulfur

Abstract

Summary. A glasshouse experiment was conducted to study the correlation between various soil sulfur (S) extractants and pasture response to applied S grown under controlled environmental conditions (glasshouse) as influenced by variations in levels of extractable S. Intact paired soil cores were collected at 2-month intervals from an S x P factorial experiment at 4 field sites on the Northern Tablelands of New South Wales on 11 occasions. Soil samples were collected from each core and the cores placed in the glasshouse. Sulfur was applied to one core of each pair and basal nutrients applied to all cores which were then oversown with white clover (Trifolium repens L. cv. Haifa). The extractable S in the soil was analysed using the following techniques: 0.1 mol Ca(H2PO4)2/L (MCPt); 0.1 mol Ca(H2PO4)2/L, treated with activated charcoal (MCPi); water (H2O); 0.25 mol KCl/L, heated at 40°C for 3 h (KCl-40); 0.25 mol KCl/L, heated at 100°C for 4 h (KCl-100); 0.5 mol NaHCO3/L (NaHCO3); and an acid digestion of the soil (total). Dry matter yield and S content of the pasture shoots were measured for each 2-month growth period. The relationship between the various soil S extraction techniques and the response of pasture to applied S varied over time and was influenced by the level of inorganic S present in the soil and the amount of organic S extracted. When the level of MCPi-extractable S ranged between 2 and 22 µg S/g soil due to the application of fertiliser (summer 1987–88 and autumn 1988), or after a period of high mineralisation rates (spring 1988), the MCPi, MCPt, H2O and KCl-40 techniques had the highest correlation with the increase in S content of the pasture when S was applied. The higher correlation for the MCPt, H2O and KCl-40 techniques than the MCPi technique indicates that plants obtained some S from the organic S pool. In contrast, when the levels of MCPi-extractable S ranged between 2 and 10 µg S/g soil and the rate of mineralisation increased in response to an increase in soil temperature (June 1988 and June 1989) or moisture (September 1989) upon transfer of samples from the field to the glasshouse, the MCPt, H2O and KCl-40 techniques underestimated the available S pool and had lower r2 values than either the KCl-100 or NaHCO3 techniques.