N2 fixation and its potential contribution to increasing soil total N were estimated in field-grown crops of lupin and pea in 21 trials at 10 locations in New South Wales and Victoria, during 1984 to 1987. Chickpea, faba bean and annual medic were included at some sites. Across experiments there were differences in annual rainfall (267 to 646 mm), soil N (0.02 to 0.20%), soil pH (CaCl2,4.3 to 8.0) and sowing date (24 April to 16 June). Most experiments were conducted on acidic (pH < 4.8) red-earth, the others on grey-cracking clay or sandy soil, both of higher pH The differing sites, seasons, and sowing time contributed to variation in legume biomass (2.02 to 14.33 t/ha) and total N (45 to 297 kg N/ha), and the amount of N harvested with grain (8 to 153 kg N/ha), which were related.Lupin fixed an average of 65% of total crop N, and pea 61%, but there was considerable variation about these averages (20 to 97%). Significant differences in % N2 fixation between legumes within sites were few. The amount of N2 fixed averaged 98.5 kg N/ha by lupin and 80.5 kg N/ha by pea, varying 26 to 288 kg N/ha and 16 to 177 kg N/ha, respectively. Variation in proportional and total N2 fixation was associated with biomass, soil mineral N, and sowing date. N2 fixation increased with more biomass and declined with higher soil mineral N, and later sowing (lupin). Each additional tonne of dry matter increased fixed N by c. 20 kg N/ha. Differences in amounts of fixed N between legumes within sites were due primarily to biomass differences.N2fixed by lupin contributed an average of 38.2 kg N/ha to soil N, and by pea, 17.9 kg N/ha. The contribution was variable, -41 to 135 kg N/ha (lupin) and -32 to 96 kg N/ha (pea), and correlated with proportional and total N2 fixation. Positive increase to soil total N occurred when lupin fixed at least 50% of its crop N, and pea 65%. This occurred in most crops. Legumes frequently used less of the available soil N than cereals.

At 15 sites in the cereal belt of New South Wales and Victoria, wheat after lupin or pea produced more biomass and had a greater nitrogen (N) content than wheat after wheat or barley; on average these crops assimilated 36 kg N/ha more. The improved wheat yield after lupin averaged 0 . 9 t/ha and after pea 0.7 t/ha, increases of 44 and 32% respectively. The responses were variable with site, year and legume. Soil available N was increased by both lupin and pea and the levels of surface inorganic N measured at the maturity of first year crops was often related to N in wheat grown in the following year. Of two possible sources of additional N for wheat after legumes, namely mineral N conserved in soil by lupin or pea (up to 60 kg N/ha) and the total N added in the residues of these legumes (up to 152 kg N/ha), both were considered significant to the growth of a following wheat crop. Their relative contribution to explaining variance in wheat N is analysed, and it is suggested wheat may acquire up to 40 kg N/ha from legume stubbles. Non-legume break crops also increased subsequent wheat yield but this effect was not as great as the combined effect of added N and disease break attained with crop legumes.

Using an acid sand with differing proportions of added ground limestone, a range of soil pH levels was achieved from 5.4 to 9.2. Four winter grain legumes (field peas, faba beans, lupins and chickpeas) were grown for 8 weeks before being sampled for nodulation, dry matter and nitrogen production. Chickpeas appeared the least affected by the range of soil pH values and calcium levels in terms of root and shoot growth; all the legumes produced the greatest number of root nodules at pH 7 or 8. Top dry weight production in lupins was maximized at pH 6, and declined rapidly above pH 7; field peas, chickpeas and faba beans all showed maximum above-ground dry weights between pH 7 and 8.

SUMMARYTwo field experiments examined the influence of soil nitrogen on grain yield, wateruse efficiency and residual nitrate accumulation of four winter grain legumes in a legume–fallow–wheat rotation. In the first experiment, conducted on a clay soil with a high total soil nitrogen content (0·194%), the three legumes Cicer arietinum, Vicia faba and Pisum sativum gave similar yields, with Lupinus angustifolius showing poor yields. Water-use efficiencies followed similar trends to the grain yields. Wheat yields and concentration of nitrogen in the grain, following a fallow after the legume crops, showed no differences from a full 2-year fallow. In the second experiment on a low fertility soil (0·081 % total soil nitrogen), Cicer arietinum gave the highest yield and the greatest water-use efficiency. All legumes and long fallow resulted in a similar increase in wheat yield when compared with a wheat-fallow-wheat rotation. The legume crops generally increased soil nitrate status between their sowing and the sowing of a wheat crop 2 years later; such increases were greatest with Pisum sativum at the high fertility site, with a suggestion of the reverse occurring at the low fertility site.

No abstract