The nitrogen fertilizer conundrum: why is yield a poor determinant of crops’ nitrogen fertilizer requirements?

The application of nitrogen (N) fertilizer both underpins high productivity of agricultural systems and contributes to multiple environmental harms. The search for ways that farmers can optimize the N fertilizer applications to their crops is of global significance. A common concept in developing recommendations for N fertilizer applications is the “mass balance paradigm” – that is, bigger crops need more N, and smaller less – despite several studies showing that the crop yield at the optimum N rate (N_opt) is poorly related to N_opt. In this study we simulated two contrasting field experiments where crops were grown for 5 and 16 consecutive years under uniform management, but in which yield at N_opt was poorly correlated to N_opt. We found that N lost to the environment relative to yields (i.e., kg N t^-1) varied +/- 124 and 164 % of the mean in the simulations of the experiments. Conversely, N exported in harvested produce (kg N t^-1) was +/- 11 and 48 % of the mean. Given the experiments were uniformly managed across time, the variations result from crop-to-crop climatic differences. These results provide, for the first time, a quantitative example of the importance of climatic causes of the poor correlation between yield at N_opt and N_opt. An implication of this result is that, even if yield of the coming crop could be accurately predicted it would be of little use in determining the amount of N fertilizer farmers need to apply because of the variability in environmental N losses and/or crop N uptake. These results, in addition to previous empirical evidence that yield at N_opt and N_opt are poorly correlated, may help industry and farmers move to more credible systems of N fertilizer management.