Abstract
Context: To revive the practice of planting legume green manure (GM) in the fallow period in rainfed agricultural areas, it is essential to demonstrate the benefits of this practice on the yields and water use efficiency (WUE) of subsequent crops, especially when integrating with optimized water and fertilizer management. Objectives: We conducted a field experiment to determine the positive effects of planting legume GM in the summer fallow on the yield, WUE, and nitrogen uptake efficiency (NupE) of subsequent winter wheat, which was grown with plastic film mulching and integrated fertilization in the Loess Plateau of China. Methods: A split-plot-designed experiment was arranged with two main treatments, namely (1) wheat planting followed by GM planting in the summer fallow (GM) and (2) conventional wheat monoculture followed by bare land summer fallow (BL), and three sub-treatments: (1) control treatment without any chemical fertilizer (Ct), (2) application of chemical N, P, and K as basal fertilizer (B), and (3) application of basal fertilizer plus wheat straw return (BS). Results: In the initial two years, even in a dry year, GM did not decrease the soil water content and storage (0-200 cm layer) during the subsequent winter wheat season, relative to BL. But in the third and fourth years, GM increased the grain yield of winter wheat by 3.2% and 3.8%, respectively. B and BS increased the grain yield of winter wheat by 14.4% and 22.2%, respectively, during the third experimental year, and by 12.7% and 19.4% during the fourth experimental year, primarily through increasing the population density of winter wheat. The increase in the grain yield contributed to a higher WUE of winter wheat. In the third year, GM increased the water consumption (WC) and WUE of wheat by 2.4% and 1.7%, respectively, though they were far lower than B (8.3% and 5.6%) and BS (10.4% and 10.7%). B and BS resulted in a higher yield and N nutrition than GM alone, but GM combined with B and BS resulted in the highest yield and N nutrition, thus greatly decreasing the NupE and increasing N productivity. Conclusions: Planting legume GM in the fallow can further increase the long-term yield, WUE, and N utilization of winter wheat when integrated with chemical fertilization and wheat straw return in rainfed agriculture. Implications: Our study yields new insights into the agronomic benefits of legume GM application in semi-arid or analogous rainfed agroecosystems and underscores the critical role of water conservation in ensuring dryland agricultural production, particularly in regions undergoing optimization of fertilization.