Nitrogen and sulfur dynamics of contrasting grazed pastures

The experimental area was located at the Big Ridge 2 site, CSIRO, Chiswick (30°31′S, 151°39′E), 20 km south of Armidale, New South Wales, Australia. The site was established in 1955. In March 1966, phalaris and white clover were sown and pastures were fertilised annually with superphosphate until 1993. There were 3 pasture treatments, each with 2 replicates: degraded pasture (low phalaris content), phalaris dominant, and phalaris–white clover. Each of 6 experimental plots was divided into 3 strata. Two representative areas 1 m by 0.5 m were selected in each stratum of each treatment. The selected areas were labelled with 34S-enriched (90%) elemental sulfur and 15N-enriched (99%) NH4Cl solution. All plots were grazed continuously by sheep. No effect of pasture type on N leaching was apparent in this experiment. Seasonal variation of total soil mineral N in different soil layers, low 15N recovery down to 60 cm soil depth, and low nitrate-N concentrations in drainage water obtained in this experiment suggest that synchronisation of pasture growth with mineralisation and nitrification, together with ammonium domination of the soil N system, is the key ecological feature in preventing N leaching in this environment. Unlike N, potential S leaching was found with evidence of a large amount of sulfate stored deeper in the soil profile and high S concentrations in drainage water. High KCl-40 extractable S concentration in the top 20 cm soil layers was associated with the long history of superphosphate application. Long-term applications of superphosphate (1967–93), together with an increase in sulfate sorption capacity at lower soil depths, resulted in a large amount of sulfate stored at greater depth. However, retention of the 34S applied in 1995 in the top 10 cm soils suggests that sulfate-S movement down the soil profile is slow.