Abstract/Summary

Current, how to use limited water resources efficiently and improve agricultural water use efficiency, has become one of the greatest challenges for global food security. In this study, multiple site-years of carbon and water flux data across the major crops including maize, winter wheat and soybean, were used to quantify the variability in canopy-scale transpiration (T), ecosystem-scale evapotranspiration (ET) as well as the associated water use efficiencies (WUET and WUEET). On the basis of ET partitioning, the results indicated that the transpiration ratio–T/ET as well as T and ET exhibited an obvious single-peak seasonal pattern across the typical croplands. However, at the early and late growing stages, there existed large discrepancies in T and ET owing to low vegetation coverage, while T and ET were very close during the peak period. Among them, maize exhibited the largest T/ET by 0.50 ± 0.12, followed by soybean of 0.43 ± 0.08 and winter wheat of 0.38 ± 0.09, respectively. Furthermore, the coupling relationships between gross primary productivity (GPP) and water fluxes including T and ET changed from linear to nonlinear. The study also found that the variability in WUET and WUEET were not consistent. Specifically, WUEET showed distinct seasonal characteristic whereas WUET kept constant as a plateau almost throughout the growth period, which reflected the inherent physiological property controlled by plant stomata at the canopy scale. Among these crops, maize exhibited the largest WUET and WUEET (5.30 ± 0.89 and 2.48 ± 1.14 g C kg−1 H2O), followed by winter wheat (4.97 ± 1.52 and 2.35 ± 0.64 g C kg−1 H2O) and soybean (4.88 ± 1.59 and 1.89 ± 0.99 g C kg−1 H2O), respectively.