Split nitrogen application strategies in seed production of two contrasting cultivars of timothy (Phleum pratense L.)

Abstract

Split nitrogen (N) application to seed crops of timothy (Phleum pratense L.) was evaluated in 20 field trials, 11 in the south-Norwegian cultivar ‘Grindstad’ (origin 59°N) and nine in the north Norwegian cultivar ‘Vega’ (origin 67°N) in southeast (58–61°N) and central (63–64°N) Norway during 1998–2000. Three N rates (25, 50 or 75 kg N ha−1) at the start of growth were combined with four N rates (0, 25, 50 or 75 kg N ha−1) at the start of tiller elongation (Z 31). On average for all trials, the highest seed yield was obtained when a total rate of 75 kg N ha−1 was split into a main application of 50 kg N ha−1 at the start of growth and a supplemental application of 25 kg N ha−1 at Z 31. In both cultivars, the optimal N rate at Z 31 was uncorrelated with soil mineral N (NO3+NH4) in early spring and only weakly correlated (r 2=0.16–0.17, p% = 2–5) with the N concentration of aboveground plant material at Z 31. In ‘Grindstad’, a better correlation (r 2=0.46, p% < 0.1) was obtained between the optimal N rate at Z 31 and measurements by the hand-held chlorophyll meter Hydro N-Tester™ on the last fully developed leaf. The equation (HTN model) Optimal N input (kg N ha −1 ) at Z 31 = − 0.37×measured HNT value at Z 31+156 was further evaluated in ten on-farm trials during 2001–2002. Compared to the fixed N levels 0, 25 or 50 kg N ha−1, the HNT model predicted optimal N rates at Z 31 in 2002, but not in 2001, when less rainfall caused less lodging than in 1998–2000 and 2002. Nevertheless, we conclude: 1) that the spring N rates to timothy seed crops should be split into an early and a late application, and 2) that the HTN model will help in adjusting the second N rate at Z 31 in ‘Grindstad’.