Influence of soil type, crop management and weather on the recovery of ¹⁵N-labelled fertilizer applied to winter wheat in spring

SUMMARY 15 N-labelled fertilizer was applied, in spring, to winter wheat crops in nine experiments in eastern England over a period of 4 years. Five were on Batcombe Series silty clay loam, two on Beccles Series sandy clay loam (with a mole-drained clay subsoil) and two on Cottenham Series sandy loam. In three of the experiments, different rates of fertilizer N were applied (up to 234 kg N/ha); in the others, a single rate (between 140 and 230 kg/ha) was used. Recovery of fertilizer N in the above-ground crop (grain, chaff, straw and stubble) ranged from 46 to 87% (mean 68%). The quantity of fertilizer N retained in the soil at harvest was remarkably constant between different experiments, averaging 18% where labelled N was applied as 15 NH 4 15 NO 3 , but less (7–14%) where K 16 NO 3 was applied. Of the labelled N present in soil to a depth of 70 cm, 84–88% was within the cultivated layer (0–23 cm). L70 = 5(± 1 63) + 0·264(±00352) R 3 accounted for 73% of the variation in the data where: L 70 = percentage loss of fertilizer N from the crop: soil system, defined as percentage of labelled N not recovered in crop or in soil to a depth of 70 cm at the time of harvest; R 3 = rainfall (in mm) in the 3 weeks following application of N fertilizer. There was a tendency for percentage loss of fertilizer N to be greater when a quantity of N in excess of that required for maximum grain yield was applied. However, a multiple regression relating loss both to rainfall and to quantity of N applied accounted for no more variance than the regression involving rainfall alone. In one experiment, early and late sowing were compared on the first wheat crop that followed oats. The loss of N from the early-sown crop, given fertilizer N late in spring, was only 4% compared with 26 % from the later-sown crop given N at the same time, so that sowing date had a marked effect on the loss of spring-applied fertilizer N. Uptake of unlabelled N, derived from mineralization of organic N in soil, autumn-applied N (where given) and from atmospheric inputs, was < 30 kg/ha on a low organic matter (0·08% total N) sandy soil but > 130 kg/ha when wheat followed potatoes or beans on soil containing c. 0·15 % total N. Unlabelled N accounted for 20–50% of the total N content of fertilized crops at harvest. About 50% of this unlabelled N had already been taken up by the time of fertilizer application in spring and the final quantity was closely correlated with the amount present in the crop at this time. Applications of labelled fertilizer N tended to increase uptake of unlabelled N by 10–20 kg/ha, compared to controls receiving no N fertilizer. This was probably due to pool substitution, i.e. labelled inorganic N standing proxy for unlabelled inorganic N that would otherwise have been immobilized or denitrified.