Estimating the crop response to fertilizer nitrogen residues in long-continued field experiments

Knowledge of the cumulated effect of long-continued nitrogen (N) inputs is important for both agronomic and environmental reasons. However, only little attention has been paid to estimate the crop response to mineral fertilizer N residues. Before interpreting estimates for the crop response to previous N input rates, the experimental design for testing needs to be examined. Experimental designs that suspend the customary N inputs, leaving the test crop unfertilized, ignore any interaction between the rate of N applied in the past and the rate applied in the test year. We estimated the interaction using data from five long-continued field experiments on mineral and organic N input rates where the main plots were subdivided for incremental rates of mineral N fertilizer in the test year. The interaction between N applied in the past and in the test year significantly affected grain yield and N offtake when the residual effect originated from organic applications, but the interaction was not significant when mineral N fertilizer had been used in the past, making the residual effect of N applied in the past additive to the effect of N applied in the test year. The dry matter (DM) grain yield of spring barley decreased by an average of 5 kg DM/ha and the grain N offtake by 46 g N/ha for a decrease in the annual mineral N rate of 1 kg N/ha applied for more than three decades. Although statistically significant, the crop response to mineral fertilizer N residues was of minor importance when compared with the residual effect of organic inputs.