Factors regulating the contributions of fixed nitrogen by pasture and crop legumes to different farming systems of eastern Australia

On-farm and experimental measures of the proportion (% Ndfa) and amounts of N₂ fixed were undertaken for 158 pastures either based on annual legume species (annual medics, clovers or vetch), or lucerne (alfalfa), and 170 winter pulse crops (chickpea, faba bean, field pea, lentil, lupin) over a 1200 km north-south transect of eastern Australia. The average annual amounts of N₂ fixed ranged from 30 to 160 kg shoot N fixed ha⁻¹ yr⁻¹ for annual pasture species, 37-128 kg N ha⁻¹ yr⁻¹ for lucerne, and 14 to 160 kg N ha⁻¹ yr⁻¹ by pulses. These data have provided new insights into differences in factors controlling N₂ fixation in the main agricultural systems. Mean levels of %Ndfa were uniformly high (65-94%) for legumes growing at different locations under dryland (rainfed) conditions in the winter-dominant rainfall areas of the cereal-livestock belt of Victoria and southern New South Wales, and under irrigation in the main cotton-growing areas of northern New South Wales. Consequently N₂ fixation was primarily regulated by biomass production in these areas and both pasture and crop legumes fixed between 20 and 25 kg shoot N for every tonne of shoot dry matter (DM) produced. Nitrogen fixation by legumes in the dryland systems of the summer-dominant rainfall regions of central and northern New South Wales on the other hand was greatly influenced by large variations in %Ndfa (0-81%) caused by yearly fluctuations in growing season (April-October) rainfall and common farmer practice which resulted in a build up of soil mineral-N prior to sowing. The net result was a lower average reliance of legumes upon N₂ fixation for growth (19-74%) and more variable relationships between N₂ fixation and DM accumulation (9-16 kg shoot N fixed/t legume DM). Although pulses often fixed more N than pastures, legume-dominant pastures provided greater net inputs of fixed N, since a much larger fraction of the total plant N was removed when pulses were harvested for grain than was estimated to be removed or lost from grazed pastures. Conclusions about the relative size of the contributions of fixed N to the N-economies of the different farming systems depended upon the inclusion or omission of an estimate of fixed N associated with the nodulated roots. The net amounts of fixed N remaining after each year of either legumebased pasture or pulse crop were calculated to be sufficient to balance the N removed by at least one subsequent non-legume crop only when below-ground N components were