Optimizing nitrogen input by balancing winter wheat yield and residual nitrate-N in soil in a long-term dryland field experiment in the Loess Plateau of China

Abstract Nitrate-N residue (NR) in soil is attracting increasing attention because of its environmental risks. However, research information to regulate NR in soil by optimizing chemical nitrogen (N) fertilizer input and at the same time ensuring a satisfactory high grain yield is still lacking, particularly in the drylands of northwestern China. A 9-yr field experiment was initiated in 2004 in the Loess Plateau of China to investigate the dominant factors affecting NR in soil at harvest and optimize the N input for winter wheat by balancing the grain yield and NR in soil. Obtained results showed that varied grain yields and N application rates resulted in different NR in soil in dryland. The NR increased quadratically with the N rates, and the average NR was 59.0, 98.7, 175.1, 326.0, and 475.5 kg N ha −1 in 0–300 cm soil, respectively, at N rates of 0, 80, 160, 240, and 320 kg N ha −1 , of which 35.1, 54.5, 94.5, 186.5, and 222.0 kg N ha −1 was in the 0–100 cm soil. Seasonal N residue (SNR) in soil also increased quadratically with N rate, with the yearly increases correspondingly being 0, 4.4, 8.8, 13.2, and 17.7 kg N ha −1 . The fertilizer N residue (FNR) increased linearly with N rate, with the yearly FNR increases, respectively, to be 0.02, 7.1, 14.1, and 21.1 kg N ha −1 at N rates of 80, 160, 240, and 320 kg N ha −1 . The seasonal nitrate-N residual depth (NRD) was mainly within the 0–100 cm soil layer. The NR, SNR, FNR and NRD were all affected by but not correlated with precipitation. In the experimental site, the NR was able to be slowly leached downward to deeper soil by 13.3–33.3 cm each year driven by precipitation. The findings suggest that if the NR is to be decreased to 55–67 kg N ha −1 in the 0–100 cm soil while maintaining a relatively high grain yield (4487–5000 kg ha −1 ), an N application rate of 66–92 kg N ha −1 should be recommended.