15N-labelled fertilizer was applied in spring to winter wheat, winter oilseed rape, potatoes, sugarbeet and spring beans in field experiments done in 1987 and 1988 in SE England on four contrasting soil types – a silty clay loam, a chalky loam, a sandy loam and a heavy clay. The 15N-labelled fertilizers were applied at recommended rates; for oilseed rape, a two-thirds rate was also tested. Whole-crop recoveries of labelled nitrogen averaged 52% for winter wheat, 45% for oilseed rape, 61% for potatoes and 61% for sugarbeet. Spring beans, which received only 2·5 kg ha−1 of labelled N, recovered 26%. Removals of 15N-labelled fertilizer N in the harvested products were rather less, averaging 32, 25, 49, 27 and 13% in wheat grain, rape seed, potato tubers, beet root and bean grain, respectively.

Crop residues were either baled and removed, as with wheat and rape straw, or were flailed or ‘topped’ and left on the soil surface, as was the case with potato tops and sugarbeet tops. Wheat stubble and rape stubble, together with leaf litter and weeds, were incorporated after harvest. The ploughing in of crop residues returned 4–35% of the original nitrogen fertilizer application to the soil, in addition to that which already remained at harvest, which averaged 24, 29 and 25% of that applied to winter wheat, oilseed rape and sugarbeet respectively. Less remained at harvest after potatoes (c. 21%) and more after spring beans (c. 49%). Most of the labelled residue remained in the top-soil (0–23cm) layer.

15N-labelled fertilizer unaccounted for in crop and soil (0–100 cm) at harvest of winter wheat, oilseed rape, potatoes, sugarbeet and spring beans averaged 23, 25, 19, 14 and 26% of that applied, respectively. Gaseous losses of fertilizer N by denitrification were probably greater following applications to winter wheat and oilseed rape, where the N was applied earlier (and the soils were wetter) than with potatoes and sugarbeet. Consequently, it may well be advantageous to delay the application of fertilizer N to winter wheat and oilseed rape if the soil is wet.

Total inorganic N (labelled and unlabelled) in soils (0–100 cm) following harvest of potatoes given 15N-labelled fertilizer in spring averaged 70 kg N ha−1 and was often greater than after the corresponding crops of winter wheat and oilseed rape, which averaged 53 kg N ha−1 and 49 kg N ha−1, respectively. On average, 91 kg ha−1 of inorganic N was found in soil (0–100 cm) following spring beans. Least inorganic N remained in the soil following sugarbeet, averaging only 19 kg N ha−1. The risk of nitrate leaching in the following winter, based on that which remained in the soil at harvest, ranked in decreasing order, was: spring beans=potatoes>oilseed rape=winter wheat>sugarbeet. On average, only 2·9% of the labelled fertilizer applied to winter wheat and oilseed rape remained in the soil (0–100 cm) as inorganic N (NO−3+NH+4) at harvest; with sugarbeet only 1·1% remained. In most cases c. 10% of the mineral N present in the soil at this time was derived from the nitrogen fertilizer applied to arable crops in spring. However, substantially more (c. 21%) was derived from fertilizer following harvest of winter wheat infected with take-all (Gaeumannomyces graminis var. tritici) and after potatoes. With winter wheat and sugarbeet, withholding fertilizer N had little effect on the total quantity of inorganic N present in the soil profile at harvest, but with oilseed rape and potatoes there was a decrease of, on average, 38 and 50%, respectively. A decrease in the amount of nitrogen applied to winter wheat and sugarbeet in spring would therefore not significantly decrease the quantity of nitrate at risk to leaching during the following autumn and winter, but may be more effective with rape and potatoes. However, if wheat growth is severely impaired by take-all, significant amounts of fertilizer-derived nitrate will remain in the soil at harvest, at risk to leaching.