Grain yield and crop N offtake in response to residual fertilizer N in long‐term field experiments

Organic inputs [e.g. animal manure (AM) and plant residues] contribute directly to the soil organic N pool, whereas mineral N fertilizer contributes indirectly by increasing the return of the crop residues and by microbial immobilization. To evaluate the residual effect of N treatments established in four long‐term (>35 yr) field experiments, we measured the response of barley (grain yield and N offtake at crop maturity) to six rates (0, 30, 60, 90, 120 and 150 kg N/ha) of mineral fertilizer N (N_new) applied in subplots replacing the customary long‐term plot treatments of fertilizer inputs (N_prev). Rates of N_prev above 50–100 kg N/ha had no consistent effect on the soil N content, but this was up to 20% greater than that in unfertilized treatments. Long‐term unfertilized plots should not be used as control to test the residual value of N in modern agriculture with large production potentials. Although the effect of mineral N_prev on grain yield and N offtake could be substituted by N_new within a range of previous inputs, the value of N_prev was not eliminated irrespective of N_new rate. Provided a sufficient supply of plant nutrients other than N, the use‐efficiency of N_new did not change significantly with previous mineral N fertilizer rate. The residual effect of mineral N fertilizer was negligible compared with the residual effect of N from AM and catch crop residues.