The effect of a range of one-year set-aside treatments on soil mineral nitrogen (SMN), during the set-aside period and in a following wheat crop were studied in a phased experiment at five sites from 1987 to 1991. Ground cover options permitted under the UK government's ‘set-aside’ scheme, including natural regeneration, autumn sown Italian ryegrass (Lolium multiflorum), spring-sown legumes and cultivated fallow, were compared with a control treatment of continuous cereals managed with fertilizer inputs. In the first of three phases in this experiment, an uncultivated fallow (kept weed-free) and autumn-sown forage rape (Brassica napus) were included as extra treatments. There were large differences in total SMN (0·0–0·9 m) between sites, ranging from 16 to 205 kg N/ha, reflecting differences in soil type, which ranged from clays to sands, and previous cropping husbandry. Differences in SMN between set-aside treatments during the first winter of the set-aside year were small, but increased during the following summer. Amounts of SMN were greatest after cultivated fallow (46–178 kg N/ha) and least after ryegrass (26–111 kg N/ha). Natural regeneration and spring sown legumes were more variable in their effect on SMN. Compared to continuous cereals, there was a build up in SMN during bare fallow, but a reduction under ryegrass, prior to returning to wheat cropping in the autumn after set-aside. SMN results suggest there was an increased nitrate leaching risk for bare fallow and natural regeneration set-aside, compared to sown ryegrass covers, in the winter following ploughing out of set-aside. This risk could be minimized by earlier sowing of winter cereals following set-aside or sowing with winter oilseed rape rather than cereals to maximize crop nitrogen (N) uptake, during the autumn growth period.

Averaged across five sites, residual SMN supply in the spring of first test year cereal crops for all set-aside treatments was similar to that for continuous cereals, suggesting over-winter losses by N leaching or immobilization. The low residual N fertility after rotational set-aside suggested that following crop N recommendations should be the same as for continuous cereals. Amounts of SMN were less each year in spring than in the preceding autumn in both the set-aside and first test cereal crops. The results suggested that a ryegrass cover appeared to be the most environmentally favourable option for rotational set-aside management, as it minimized the amount of readily leachable N both during and immediately after the set-aside period.