Contrasting responses of water use efficiency to increasing aridity in alpine shrubs: A modelling perspective

Water use efficiency (WUE), which is strongly related to carbon and water cycles, is crucial for maintaining fragile and sensitive alpine ecosystems. An accurate assessment of the spatial and temporal variations in WUE among alpine shrubs under different aridity levels is essential for quantifying the carbon and water balance in alpine environments. We calibrated the Biome-BGC model using the parameter estimation (PEST) approach with one year of data from an eddy covariance tower and validated the model against three years of carbon–water flux and carbon storage data from 80 biomass sampling sites in Qinghai Province, China. We then simulated the carbon and water cycles of alpine shrubs in Qinghai Province from 1980 to 2019. Using meteorological data from the study area, we analyzed the spatiotemporal variations and factors influencing WUE in regions with different aridity levels. The results showed that after optimization using the PEST approach, the mean absolute error (MAE) and root mean square error (RMSE) of gross primary productivity (GPP) decreased by 0.58 and 1.05 g C m⁻² d⁻¹, respectively, and those of evapotranspiration (ET) decreased by 0.41 and 0.77 mm d⁻¹, respectively. Spatial distribution analysis revealed that the annual mean GPP and ET generally decreased from southeast to northwest in the order of humid, subhumid, semi-arid, arid, and hyper-arid climate regions. In other regions, WUE exhibited a biphasic trend with the aridity index, decreasing under severe dryness but increasing as aridity increased. The primary controlling factor in humid and sub-humid regions is the mean annual temperature, whereas in arid and semi-arid regions it is the mean annual precipitation. These findings are critical for improving the prediction of carbon sequestration and water-holding capacity of alpine shrublands under drought conditions.