Potassium in soil and pasture and leaching of cations on an allophanic soil in New Zealand

Potassium chloride (KCl) was applied at 0, 150, 450, and 750 kg ha(-1) in May 2000 and April 2001 to an allophanic soil (Egmont brown loam) in South Taranaki to generate a range of soil and pasture potassium (K) levels typical of grazed dairy pastures. Initial mean soil quick test (QT) K levels of 7.0 increased to 7.3, 9.2, 13.6, and 18.5 respectively for each rate of K at the March 2001 sampling. In the March 2002 sampling, soil QT K decreased to 5.5, 7.8, 8.1, and 8.5 for each rate of K. There was only a small effect of rate of K on initial soil magnesium (Mg), calcium (Ca), and sodium (Na) soil QT levels of 30.8, 8.6, and 8.4. Pasture K content was significantly higher (P < 0.05) at the highest rate of K compared with control from July to December in Year I (July 2000-May 2001) and January-March in Year 2 (July 2001-May 2002). Increasing rate of K did not significantly change pasture Mg and Ca in most months but significantly decreased pasture Na. Leachate collected at I m soil depth showed that \ with increasing rate of KCl there was a significant increase (P < 0.10) in the amount of leached K in Year 1 ((July-August 2001) (7.0 kg ha(-1) for control to 17.5 kg ha(-1) for highest rate of K)), chloride (64.6-146.6 kg ha(-1) in Year I (P < 0.01), 104.1-202.5 kg ha(-1) in Year 2 (June-August 2002) (P < 0.05)), Ca (30.4-60.1 kg ha(-1) in Year 1 (P < 0.05), 34.1-74.3 kg ha(-1) in Year 2 (P < 0.10)), and Na (59.4-74.1 kg ha(-1) in Year 2 (P < 0.10)). There was no measured significant K effect (P > 0.10) on the amount of Mg, nitrate-nitrogen and sulphate-sulphur leached in either year. Therefore, application of excess K fertiliser on allophanic soils to elevate soil QT K > 7 will reduce pasture uptake and the availability of Na to animals. If the leached Na and Ca is not replaced, pasture Na and Ca could become deficient for dairy cow requirements.