Several new and existing short-term forage legumes could be used to provide nitrogen (N) inputs for grain crops in subtropical farming systems. The fixed-N inputs from summer-growing forage legumes lablab (Lablab purpureus), burgundy bean (Macroptilium bracteatum) and lucerne (Medicago sativa) and winter-growing legume species snail medic (Medicago scutellata), sulla (Hedysarum coronarium) and purple vetch (Vicia benghalensis) were compared over several growing seasons at four locations in southern Queensland, Australia. Available soil mineral N and grain yield of a following cereal crop were compared among summer-growing legumes and forage sorghum (Sorghum spp. hybrid) and among winter-growing legumes and forage oats (Avena sativa). In the first year at all sites, legumes utilised the high initial soil mineral N, with <30% of the legume N estimated to have been derived from atmospheric N2 (%Ndfa) and legume-fixed N <30 kg/ha. In subsequent years, once soil mineral N had been depleted, %Ndfa increased to 50–70% in the summer-growing legumes and to 60–80% in winter-growing legumes. However, because forage shoot N was removed, rarely did fixed N provide a positive N balance. Both lablab and burgundy bean fixed up to 150 kg N/ha, which was more than lucerne in all seasons. Prior to sowing cereal grain crops, soil nitrate was 30–50 kg/ha higher after summer legumes than after forage sorghum. At one site, lablab and lucerne increased the growth and yield of a subsequent grain sorghum crop by 1.4 t/ha compared with growth after forage sorghum or burgundy bean. Of the winter-growing legumes, sulla had the highest total N2 fixation (up to 150 kg N/ha.year) and inputs of fixed N (up to 75 kg N/ha), and resulted in the highest concentrations of soil N (80–100 kg N/ha more than oats) before sowing of the following crop. Wheat protein was increased after winter legumes, but there was no observed yield benefit for wheat or grain sorghum crops.
New forage legume options, lablab, burgundy bean and sulla, showed potential to increase N supply in crop rotations in subtropical farming systems, contributing significant fixed N (75–150 kg/ha) and increasing available soil N for subsequent crops compared to non-legume forage crops. However, high soil mineral N (>50 kg N/ha) greatly reduced N2 fixation by forage legumes, and significant N2 fixation only occurred once legume shoot N uptake exceeded soil mineral N at the start of the growing season. Further work is required to explore the impact of different management strategies, such as livestock grazing rather than harvesting for hay, on the long-term implications for nutrient supply for subsequent crops.