Denitrification, leaching and immobilisation of applied 15N following legume and grass pastures in a semi-arid climate in Australia

Abstract

Four consecutive 15N mass balance experiments lasting 18 months from February 1993 to August 1994 were carried out to assess the fate of applied 15N at 3 sites after 4 years of lucerne or snail medic and 20 years of Mitchell grass/naturalised medic pastures respectively in the Roma district of Queensland, Australia. 15N loss via denitrification was estimated from the difference between the recovery of applied Br(100kg Br/ha) and that of applied 15N(40kg N/ha) in the top 250mm at the end of each mass balance experiment. From February to August 1993, denitrification losses were 12–38% of applied 15N. N losses increased to 36–51% during August to November 1993, responding to the higher rainfall during this period. With even more rainfall during the period between November 1993 and March 1994, N losses were estimated to be 16–23%while displacement of 15N below 250 mm was 74–81%. When rainfall was much less between March 1994 and August 1994, N losses of only 15–19% of the applied 15N occurred at the 3 sites. It was found that although rainfall was the dominant factor controlling denitrification of the applied 15N, soil available carbon (C) (measured as water-soluble C) and the quantity of nitrate available were also important for soils already containing a considerable quantity of organic matter and N from residues of pasture legumes.