Scale-dependent responses of water use efficiency to precipitation variability in a rain-fed maize field of Northeast China

Precipitation changes alter cropland water use efficiency (WUE), yet their effects across ecosystem, canopy, and leaf scales remain poorly understood. Here, we used the eddy covariance technique to measure CO 2 and H 2 O fluxes in a rain-fed maize field of Northeast China during a wet (2020–2021), dry (2021–2022), and normal (2022–2023) year. On this basis, we quantified ecosystem (EWUE), transpiration (TWUE), and intrinsic (IWUE) WUE, and examined their temporal dynamics and responses to precipitation anomalies. Increased precipitation reduced EWUE because gross primary productivity (GPP) declined more than evapotranspiration (ET), while yielding a marginal decrease in TWUE due to proportional reductions in GPP and transpiration (T). Decreased precipitation suppressed both EWUE and TWUE, as the decline in GPP far exceeded that in ET and T. IWUE exhibited contrasting responses to precipitation changes: increased precipitation lowered IWUE by inducing a larger reduction in GPP than in canopy conductance, whereas the opposite pattern occurred under decreased precipitation. Over the diel cycle, increased precipitation dampened variations of EWUE, TWUE, and IWUE, but decreased precipitation amplified them. On the seasonal scale, altered precipitation steepened the positive EWUE-GPP relationship. Higher precipitation strengthened the temperature dependence of EWUE, whereas lower precipitation weakened it. Both increased and decreased precipitation attenuated TWUE responses to GPP and vapor pressure deficit (VPD), while intensifying IWUE sensitivity to VPD and radiation. These findings reveal multi-scale WUE responses to precipitation variability and highlight that intensified precipitation anomalies under climate change may increase water consumption for crop production, potentially threatening the resilience of rain-fed agroecosystems.