The response of swards which have been previously grazed to N fertilizer applied in early February was studied in two experiments in Northern Ireland. The effect of N fertilizer applied at a range of dates in autumn and spring on swards for out‐of‐season utilization was studied in a further experiment. Deep soil coring was also undertaken, subsequent to grazing with dairy cows, in grazed and protected areas in November and March to investigate the effect of out‐of‐season grazing on soil mineral N levels. Dry‐matter (DM) yield response to early spring N application in previously grazed swards was low, with no effect on DM yield in February or March. Progressively delaying N application (and commencement of herbage accumulation) in autumn from 8 September until 18 October reduced herbage availability in late autumn and early spring but increased leaf lamina content. The greater the amount of herbage accumulated to 1 December, the lower the tiller density in the following April. N fertilizer had a greater impact on soil mineral N in spring than in late autumn/early winter, suggesting that fertilizer N was more prone to loss in the latter. Soil mineral N was not significantly affected by out‐of‐season grazing. It is concluded that in well‐fertilized, previously grazed swards response to N for out‐of‐season herbage is low and the probability for N loss is increased. Herbage quality will decline and the sward may be damaged if about 2 t DM ha−1 or more of harvestable herbage accumulates for use in winter or in early spring.

The pool of nitrate‐N (NO3–‐N) in the soil is more prone to losses than that of ammonium‐N (NH4+‐N) so any shift towards NO3–‐N dominance in the soil pools, caused by management intensity, could have environmental implications. The change in the balance of soil NH4+‐N and NO3–‐N content with time was studied using grazed grass swards receiving different fertilizer N inputs. In addition, the effect of past management on net nitrification of 400 μg NH4+‐N g–1 was investigated in a soil incubation study. Mineral N was determined at frequent intervals (at least every 2 weeks) throughout the year in the top 5 or 7·5 cm of a sandy clay‐loam soil at the Agricultural Research Institute of Northern Ireland at Hillsborough, County Down, for a 7‐year period (1989–90 to 1995–96). The treatments were a perennial ryegrass–white clover sward receiving no fertilizer N, together with perennial ryegrass swards receiving 100, 200, 300, 400 or 500 kg N ha–1 year–1 as calcium ammonium nitrate. The plots were continuously grazed by beef steers from April to October to maintain a constant sward height of 7 cm. There was little or no change in average soil NO3–‐N and NH4+‐N content from 1989–90 to 1995–96 on the grass–clover sward and plots receiving 100 and 200 kg N ha–1 year–1. However, with the plots receiving 300, 400 and 500 kg N ha–1 year–1 NO3–‐N became progressively more dominant with time. The incubation study confirmed that this was due to an increase in net nitrification rate. There was evidence that rapid microbial assimilation of NO3–‐N occurred during the soil incubations. Past management history can play an important role in determining soil NO3–‐N content and hence potential losses of N to the environment.