Soil N contributes to the oscillations of the white clover content in mixed swards of perennial ryegrass under conditions that simulate grazing over five years

The effect of the initial N-supplying capacity of soils (SoilN, 90–230 kg N ha^–1 year^–1) was tested on the dry-matter and N yields of pure or mixed white clover and perennial ryegrass swards, managed under simulated grazing over a 5-year period. The cumulated N harvested in the mixed swards was similar, both for white clover and perennial ryegrass, but the proportion of white clover showed oscillations over a 2-year period. In the first year, the SoilN effect was similar to that of fertilizer N. During the course of the experiment, the effect was always positive on the pure perennial ryegrass sward, alternately negative and nil for the white clover in the mixed sward and alternately positive and nil for the perennial ryegrass in the mixed sward; the period of these oscillations was 2 years. From the third regrowth period after sowing, the ratio between the actual N concentration and the concentration non-limiting to growth for the perennial ryegrass in the mixed sward, increased above that of the pure perennial ryegrass sward. It was in turn greater in the soils that were initially poor and then greater in those that were initially rich in soil N. The periodic oscillation of the initial SoilN effects implies that the initial SoilN gradient was alternately compensated and restored. It was concluded that N fluxes are partly responsible for the temporal oscillations in the proportion of white clover in mixed swards.