Considerable progress has been made with the quantification of inputs and losses of nitrogen (N) for a number of legume-based dryland rotations, enabling the fate of legume-derived N to be determined with greater accuracy than previously. Analyses of nitrate (NO3-) in soil profiles to a depth of at least 0.6 m, during and after legume phases, together with measurements of net N mineralisation are providing a much clearer insight of the capacity of legume phases to supply inorganic N. Advances in procedures used to determine drainage have improved estimates of NO3- leaching for a range of soils and rainfall conditions. The loss of N from urine patches and rates of ammonia volatilisation from grazed fields are fairly well characterised. In contrast, the amounts of N lost from legume-based rotations through denitrification are largely unknown. The ingestion of 60-70% of the legume N by animals in intensively managed pastures highlights the pivotal roles grazing animals can play in the transformations of N in pastures. Most of the ingested N is excreted, with the proportion returned in urine dependent on the N content of feed consumed. The tendency of sheep and cattle to defecate close to camping areas when set-stocked can cause large transfers of N within pasture paddocks. Transfer of N from pastures to laneways and to milking sheds (about 55 kg N/ha. year), and export of N in milk (80 kg N/ha. year), are major loss processes in intensively managed dairy pastures. Export of N in meat and wool are insignificant in respect to N-2 fixation in improved pastures. Gaseous losses, specifically ammonia (NH3) volatilisation, can account for between 30 to 50% of urine voided to dead pasture or dry soil in summer and autumn. Lower proportions of urine N (10-25% of N applied) are lost after application to green pasture, with gaseous losses further reduced where rainfall occurs soon after urination. Although these losses of N are significant in the context of urine patches, micrometeorological techniques that measure NH3 volatilisation over an area of several hectares of grazed green pasture indicate that NH3 losses chiefly fall in the range 1-7% of urine N excreted. Annual rates of leaching of the order 15-35 kg NO3- N/ha have often been obtained under grazed legume pastures for a range of soil and climatic conditions. Uptake of NO3- by non-leguminous species in mixed pastures appears to be the main reason for the smaller quantities of leached NO3- than might be anticipated from the high rates of N addition in urine patches. The maintenance of low NO3- concentrations in field soils, together with low temperatures during periods of excess soil water, also appear to restrict denitrification in soil under mixed pasture swards, even though measurements undertaken in controlled soil environments suggest that denitrification could potentially account for up to 25-30% of urine N. The magnitude and timing of N release from legume residues remaining after grazing, and subsequent immobilisation of mineralised N, is affected by the efficiency of C use by the decomposer population, the demand for N, the chemical nature of the plant residues, and a range of soil factors. Green residues decompose rapidly with up to 40% of residue mineralised within 12 months. A slower rate of decomposition occurs in mature residues that possess a wider C: N ratio, and greater lignin: N ratio and/or polyphenol: N ratios. Where legume phases are followed by a crop phase, 10-20% of previously green legume residue N is typically used by the first succeeding crop, while less than 10% of N in mature pasture residue is normally in the first following crop. Loss of mineralised residue N from soil, by either NO3- leaching or denitrification, are small in Mediterranean-type climates, but can be large in wet temperate or tropical regions. Nevertheless, the soil organic matter pool is the main sink for N in legume residues. Mineralisation of soil organic matter after legume phases can result in the accumulation of 70-150 kg N/ha, chiefly as NO3-, in many soils during either winter or summer fallows. Between 40 and 100 kg NO3- N can be leached from the rooting zone of the first succeeding crop in soils that possess large hydraulic conductivities, highlighting that the greatest risk of N loss from legume-based rotations exists at the onset of subsequent cropping phases when the crop demand for mineral N is low. Few studies have evaluated the loss of legume- or soil-derived NO3- by denitrification in crops that follow legumes, making it difficult to assess the importance of this N loss process. Australian cereal crops can use as little as 21-36% of available soil-derived mineral N after legume phases on sandy soils with low water holding capacity, and up to 45-50% in the case of finer-textured red or red brown earths. Better synchronisation of N supply from legume phases with subsequent demand for mineral N would further enhance the efficiency of recovery of legume N. The review outlines ways in which this might be achieved, and it also discusses options that could be used to reduce N loss in grazed pastures.
