Stomata are the common pathways through which carbon uptake and water vapor loss take place in a plant. Therefore, the response of stomatal conductance (GS) to environmental factors, including the rise of atmospheric CO2 concentration (eCO2), affect tree and forest water use and photosynthesis. eCO2 may affect GS by increasing intercellular CO2 concentration (direct) and by adjusting allometrics or hydraulics (indirect). Direct and indirect processes may take different exposure time to play out, inducing differing responses among species. Separating direct from indirect responses of different functional types (e.g., coniferous v. broadleaved deciduous species), is necessary for proper rendering in models from leaf to regional scales.
We investigated the GS responses of Pinus taeda and Liquidambar styraciflua during and following free-air CO2 enrichment (FACE). The study was conducted at the Duke FACE experiment site, NC, USA from 2009 to 2012. Four plots in both ambient CO2 (aCO2) and eCO2 were established in 1996, and the CO2 enrichment terminated in 2010. Half of each plot received nitrogen (N) fertilization since 2005. Using the hierarchical Bayesian state-space framework, we modeled GS of each species as a multiplicative function of reference GS, vapor pressure deficit, photosynthetic photon flux density and soil moisture.
Total canopy conductance and transpiration in eCO2 was not significantly different from that in aCO2 both before and after the termination of FACE regardless of N fertility. Ratio of monthly daytime mean GS of P. taeda (eCO2/aCO2) before termination of FACE (0.69 and 0.74 under native soil and N-fertilized condition, respectively) increased to 1.12 and 0.97 after the termination of FACE. The mean ratio of L. styraciflua was slightly higher after termination than before (0.94 and 0.84, respectively) under native soil condition; the ratio increased from 0.64 to 1.11 under N-fertilized condition. Leaf area index (L) of P. taeda was greater in eCO2, but the enhancement of L declined quickly after FACE termination. L enhancement of L. styraciflua was only significant under N-fertilized condition under eCO2, but the enhancement disappeared after termination of FACE. Because direct response of GS to short-term manipulation of CO2 was not observed, the decline of L likely has led to the increase in GS indirectly. Because ci/ca was unaffected by eCO2, we conclude that, at the canopy scale, trees operating under eCO2 for 17 years, assimilated similar amount of carbohydrates as those in aCO2 soon after enrichment termination.