Abstract
Stomatal conductance of plants exposed to elevated CO 2 is often reduced. Whether this leads to water savings in tall forest‐trees under future CO 2 concentrations is largely unknown but could have significant implications for climate and hydrology. We used three different sets of measurements (sap flow, soil moisture and canopy temperature) to quantify potential water savings under elevated CO 2 in a ca. 35 m tall, ca. 100 years old mixed deciduous forest. Part of the forest canopy was exposed to 540 ppm CO 2 during daylight hours using free air CO 2 enrichment (FACE) and the Swiss Canopy Crane (SCC). Across species and a wide range of weather conditions, sap flow was reduced by 14% in trees subjected to elevated CO 2 , yielding ca. 10% reduction in evapotranspiration. This signal is likely to diminish as atmospheric feedback through reduced moistening of the air comes into play at landscape scale. Vapour pressure deficit (VPD)‐sap flow response curves show that the CO 2 effect is greatest at low VPD, and that sap flow saturation tends to occur at lower VPD in CO 2 ‐treated trees. Matching stomatal response data, the CO 2 effect was largely produced by Carpinus and Fagus , with Quercus contributing little. In line with these findings, soil moisture at 10 cm depth decreased at a slower rate under high‐CO 2 trees than under control trees during rainless periods, with a reversal of this trend during prolonged drought when CO 2 ‐treated trees take advantage from initial water savings. High‐resolution thermal images taken at different heights above the forest canopy did detect reduced water loss through altered energy balance only at 2 ‐treated Fagus trees). Short discontinuations of CO 2 supply during morning hours had no measurable canopy temperature effects, most likely because the stomatal effects were small compared with the aerodynamic constraints in these dense, broad‐leaved canopies. Hence, on a seasonal basis, these data suggest a 2 .