How does crop rotation influence soil moisture, mineral nitrogen and nitrogen-use efficiency?

Rice–wheat (RW) cropping systems are integral to global food security. Despite being practiced for decades, Chinese RW cropping systems often suffer from low productivity and poor nitrogen use efficiency (NUE), reflecting management approaches that are not well-contextualized to region and season. Here, we develop the best management guides for N fertilizer in RW systems that are designed to help raise the productivity, NUE, and environmental sustainability of winter wheat over the long term. 2-year field experiments were conducted with four N fertilizer rates (0, 135, 180, and 225 kg N ha–1), allowing contrasts of yields, soil moisture, and NUE of wheat in RW in the humid climates zones on the Jianghan Plain. We compared RW systems with soybean/maize dryland wheat (DW) systems that are similarly endemic to China: after soybean/maize is harvested, soils are often drier compared with moisture content following rice harvest. With high seasonal N application rates (180–225 kg N ha–1), wheat crop yields increased by 24% in RW which were greater than comparable yields of wheat in DW, mainly due to greater kernels per spike in the former. Across treatments and years, N accumulation in plant tissue and kernel dry matter of DW was higher than that in RW, although mean agronomic efficiency of nitrogen (AEN) and physiological efficiency of nitrogen (PEN) of RW systems were greater. As N application rates increased from 135 to 225 kg ha–1, AEN and PEN of DW decreased but changed little for RW. Soil ammonium N was much lower than that of nitrate N; changes in NH4+ and NO3– as a consequence of increasing N fertilization were similar for RW and DW. We recommend that tactical application of N fertilizer continue seasonally until midgrain filling for both the DW and RW systems. At fertilization rates above 180 kg N ha–1, yield responses disappeared but nitrate leaching increased significantly, suggesting declining environmental sustainability above this N ceiling threshold. Collectively, this study elicits many functional and agronomic trade-offs between yields, NUE, and environmental sustainability as a function of N fertilization. Our results show that yield and NUE responses measured as part of crop rotations are both more robust and more variable when derived over multiple seasons, management conditions, and sites.