Information on N transfer from field‐grown forage legumes to associated grass would contribute to elucidation of N budgets and improved N management of pasture mixtures. This research evaluated N transfer by white clover (Trifolium repens L.) (WC), red clover (Trifolium prutense L.) (RC) or birdsfoot trefoil (Lotus corniculatus L.) (BT), grown with tall fescue (Festuca arundinacea Schreb.) (TF) at various legume proportions. Two identical experiments were established on a Typic Argiudoll soil in Uruguay, one in 1983 (Exp. 1) and one in 1984 (Exp. 2), and each was evaluated for approximately 2 yr. The soil of the area was a fine, montmorillonitic, mesic, Typic Argiudoll. Transfer of N to TF was estimated by 15N isotope dilution, using TF grown alone as the reference crop. Mean values for the proportion of grass N derived from legumes (%NT) were 20% for the 60% harvest after seeding, and later increased up to 60%. The increase on %NT over time was greater for WC‐TF and RC‐TF in Exp. 1 and for WC‐TF in Exp. 2. The annual uptake of transferred N (TNY) averaged 18 and 34 kg ha−1 in 1st and 2nd yr, respectively. The WC‐TF mixture usually showed greater annual TNY than RC‐TF on BT‐TF. Transfer by WC in summer often was greater than for RC or BT. Increasing legume proportion generally resulted in linear increases of %NT for the three mixtures, but TNY either did not change or decreased. We conclude that these legumes transferred small amounts of N shortly after seeding, but transfer increased to significant values with time. The concentration of legume‐derived N in TF herbage increased when legume proportion increased, but the amount transferred was not affected or reduced.

Optimal utilization of the N2‐fixation capability of legumes improves forage productivity and quality and reduces the need for N fertilizer. This research assessed N2 fixation in binary mixtures of white clover (Trifolium repens L.) (WC), red clover (Trifolium pratense L.) (RC), or bitdsfoot trefoil (Lotus corniculatus L.) (BT) with tall fescue (Festuca arundinacea Schreb.) (TF) at four legume proportions. Two identical experiments were established in Uruguay, one in 1983 (Exp. 1) and one in 1984 (Exp. 2) that were evaluated for 2 yr. The soil of the area was a fine, montmorillonitic, mesic, Typic Argiudoll. Symbiotically fixed N was estimated by 15N isotope dilution by using TF pure stands as nonfixing reference. The proportion of legume N derived from air (%Ndfa) was larger in winter and spring harvests (82‐95%) than in other harvests, and differences among legumes were minimal. In first harvests after seeding, %Ndfa was 58% for BT and 74% for WC or RC, whereas in summer, WC showed the least %Ndfa. Total fixed‐N yield over 2 yr was greatest for RC‐TF (390 and 330 kg ha−1 for Exp. 1 and Exp. 2) because of both large %Ndfa and legume yield. As legume proportion increased %Ndfa decreased linearly, usually with similar slopes for all legumes. Fixed‐N yield increased linearly as legume proportion increased for RC‐TF and BT‐TF mixtures. For WC‐TF, there were optimal WC proportions between 50 and 70% in seeding years. We conclude that %Ndfa was negatively and linearly related to legume proportion for the three mixtures and that legume‐dominant swards were required to maximize fixed N yields for RC‐TF and BT‐TF but not for WC‐TF. Red clover fixed the greatest amount of N2 under the conditions of this study.