Growth CO2 concentration modifies the transpiration response of Populus deltoides to drought and vapor pressure deficit.

Cottonwood (Populus deltoides Bartr. ex Marsh.) trees grown for 9 months in elevated carbon dioxide concentration ([CO2]) showed significant increases in height, leaf area and basal diameter relative to trees in a near-ambient [CO2] control treatment. Sample trees in the CO2 treatments were subjected to high and low atmospheric vapor pressure deficits (VPD) over a 5-week period at both high and low soil water contents (SWC). During these periods, transpiration rates at both the leaf and canopy levels were calculated based on sap flow measurements and leaf-to-sapwood area ratios. Leaf-level transpiration rates were approximately equivalent across [CO2] treatments when soil water was not limiting. In contrast, during drought stress, canopy-level transpiration rates were approximately equivalent across [CO2] treatments, indicating that leaf-level fluxes during drought stress were reduced in elevated [CO2] by a factor equal to the leaf area ratio of the two canopies. The shift from equivalent leaf-level transpiration to equivalent canopy-level transpiration with increasing drought stress suggests maximum water use rates were controlled primarily by atmospheric demand at high SWC and by soil water availability at low SWC. Changes in VPD had less effect on transpiration than changes in SWC for trees in both CO2 treatments. Transpiration rates of trees in both CO2 treatments reached maximum values at a VPD of about 2.0 kPa at high SWC, but leveled off and decreased slightly in both canopies as VPD increased above this value. At low SWC, increasing VPD from approximately 1.4 to 2.5 kPa caused transpiration rates to decline slightly in the canopies of trees in both treatments, with significant (P = 0.004) decreases occurring in trees in the near-ambient [CO2] treatment. The transpiration responses at high VPD in the presence of high SWC and throughout the low SWC treatment suggest some hydraulic limitations to water use occurred. Comparisons of midday leaf water potentials of trees in both CO2 treatments support this conclusion.