Effects of increased atmospheric carbon dioxide concentrations on water and carbon relations of four co-occurring tree species (Pinus taeda, Liquidambar styraciflua, Cornus florida, Ulmus alata)

Increasing atmospheric CO₂ concentration decreases stomatal conductance (G_S) in many species with potentially direct affects on both water and carbon (C) cycles and secondary effects on ecosystem processes such as competition. Some of these direct effects were investigated in the Duke Forest free-air enrichment experiment ( FACE), concentrating on sap-flux scaled mean canopy G _S, water use, forest canopy C uptake and hydraulic properties of two canopy species, Pinus taeda L. and Liquidambar styraciflua L., and two sub-canopy species, Cornus florida L. and Ulmus alata Michx., subjected to ambient (CO ₂a) and elevated atmospheric CO₂ concentrations (CO₂e) over three and a half years.; The response of G_Sref (G _S at vapor pressure deficit of 1 kPa) to CO₂e changed over the four growing seasons, highlighting the value of long-term studies. Over the study period, G_Sref of the P. taeda for optimal light, temperature and soil moisture conditions increased progressively under CO₂e to 126% of G_Sref under CO₂a, perhaps due to increased hydraulic conductivity of xylem produced under CO₂ e without an increased in xylem vulnerability to cavitation. In contrast, G_Sref of Liquidambar styraciflua increased under CO₂a, such that G_Sref under CO₂e ultimately reached 68% of G_Sref under CO₂a, seemingly because of increased vulnerability to cavitation of xylem produced by this species under CO₂e. The G _Sref of Cornus florida and Ulmus alata did not show clear responses to CO₂e. The resulting stand-level water use was not different between the CO₂ treatments or any of the components of the forest hydrologic balance. Sap-flux scaled canopy stomatal conductance was linked to stomatal-internal CO₂ concentration into a new Canopy Conductance-Constrained Carbon Assimilation (4C-A) model. Under CO₂a annual estimates of canopy net assimilation (A_nC) were within 5% of a component C balance (biomass production and respiration). However, under CO₂ e carbon balanced only 80% of annual A_nC. Of the extra C, 81% are explainable by enhanced forest floor CO₂ flux, indicating that turnover rate of roots was underestimated or that mycorhizea and rhizodeposition became an important component of the carbon balance under CO₂e.; The broader implication of this work is that responses of similar forests to elevated atmospheric CO₂ concentration will be through indirect effects on the water and carbon cycles reflected in stomatal response. These responses reflect changes in xylem hydraulic characteristics and species differences.