Variations in water sources used by winter wheat across distinct rainfall years in the North China Plain

Quantifying water availability in different soil layers to crop root water uptake (RWU) in different rainfall years helps deliver strategies to combat water scarcity and improve water productivity. To investigate the responses of RWU to soil water availability across distinct rainfall years, a field experiment of winter wheat was conducted in 2017–2018 (wet year), 2018–2019 (dry year), and 2019–2020 (ordinary year) in the North China Plain. Stable isotopes (δ18O and δD) in precipitation, irrigation water, xylem water, and soil water, as well as soil water content (SWC), were investigated during the three wheat growth seasons and were used to determine the soil water evaporation line (SWEL: the linear relationship between δD and δ18O in soil water), the line-conditioned excess (lc-excess: the deviation between δD and the local meteoric water line in the dual-isotope space) and the proportion of RWU from different soil layers with the MixSIAR model. The results indicated that there were significant differences in δ18O and δD in soil water and xylem water across different rainfall years. The slopes of SWEL and lc-excess of the 0–100 cm soil layer in the dry year were significantly smaller than those in the ordinary and wet years, which suggests that more substantial soil evaporation occurred in the dry year. The MixSIAR results revealed that the three-year wheat RWU exhibited different patterns, which were related to soil water availability. Wheat mainly utilized soil water from the 0–20 cm and 20–60 cm layers during the greening to filling stage under the dry year, while it mainly utilized the 0–20 cm soil layer under the wet and ordinary years. However, at the harvest stage, wheat under all precipitation years could transfer its main RWU zone to the 20–60 cm soil layer. There were significantly different proportions of RWU from 0 to 60 cm under different rainfall years, shown as 75.3 %, 67.5 %, and 69.5 % in the wet year, dry year, and ordinary year, respectively. These findings indicated that the appropriate irrigation wetting depth should be set as<60 cm for all kinds of rainfall years.