Increase in atmospheric carbon dioxide (CO2) concentration and soil nitrogen availability affects growths and hydraulic traits of trees. Because stomata are the common pathways in a plant through which diffusion of CO2 and water vapor take place, understanding hydraulic responses of trees to elevated CO2 is essential for estimating productivity and health of future forest ecosystems. We evaluated water uses of loblolly pine (Pinus taeda) and sweetgum (Liquidambar styraciflua), which were exposed to long-term CO2 enrichment (+200 µmol mol-1) under native and nitrogen-fertilized soil conditions, using a demographic terrestrial biosphere model, Ecosystem Demography model 2 (ED2). Model results were compared to sap flux density measured with Granier-type heat dissipation sensors from 1999 to 2010 at the Duke Forest free-air CO2 enrichment site.
Results/Conclusions
After incorporating treatment effects on tree size and stand structure, we found that diurnal patterns of treatment-level sap flux from the model generally agreed with measured sap flux. Comparisons between observed and predicted sap flux with varying parameters for marginal water use efficiency showed a greater reference (no water stress condition) marginal water use efficiency with the lowest error in elevated CO2 under both native and nitrogen fertilized conditions. The model predicted occasional mid-day depressions of stomatal conductance, which led to underestimations of sap flux. The depressions were not related to changes in turgor loss point or vulnerability to cavitation (xylem P50). The model made better predictions of summer daytime average sap flux in the later years of the experiment, and indicated smaller interannual variations than observations under all treatment conditions. The relationships between environmental factors and summer daytime average sap fluxes showed that the predicted sap fluxes were less sensitive to air temperature, vapor pressure deficit and soil moisture than the observed values. Further work is needed to determine other hydraulic traits that can improve the predictions of tree water use.