Persistent increase in canopy temperature in a Eucalypt woodland exposed to elevated CO2

Background/Question/Methods

One of the most pertinent effects of rising atmospheric CO₂ on vegetation is a reduction in water use, as a direct result of stomatal response to CO₂. Most of the forest free-air CO₂ enrichment (FACE) sites have demonstrated a reduction in leaf-level rates of water use, which in most cases led to a decrease in total water use for the canopy. Arising from energy balance considerations, reductions in stomatal conductance and transpiratory cooling could result in increased leaf temperatures. Measurements of stomatal conductance (g_s) are typically conducted with well-mixed cuvette-based systems, which subjects the leaf to conditions different from those normally experienced in terms of radiation load and wind speed and thus leaf temperatures. Alternatively, measurements of sap flow velocity are used to assess CO₂ effects on water use, but this is difficult to interpret in terms of water use per unit leaf area when estimates of tree leaf area are uncertain.

Here we use a continuous, nearly three-year record of canopy temperature measured at the EucFACE, a native, mature, Eucalyptus woodland exposed to elevated CO₂ (ambient + 150 µmol mol^-1 CO₂). The site has 3 plots of 25m diameter in elevated CO₂, and 3 control plots with ambient CO₂ levels. Four infrared thermometers per plot (24 in total) have monitored canopy temperature at minutely intervals for nearly 3 years. We test the hypothesis that reduced stomatal conductance under elevated CO₂ increases the leaf temperature, as is expected from basic energy balance considerations.

Results/Conclusions

The difference between leaf and air temperature (ΔT=T_leaf-T_air) was strongly correlated with solar radiation, frequently reaching up to 4 °C. Overall for the dataset (over 40 million temperature readings on n = 968 days) there was a significant effect of CO₂ on ΔT (p < 0.01). The average increase in ΔT was only 0.1 °C with elevated CO₂ (evaluated either as daytime average ΔT, or leaf warming at the peak solar input for every day), and so not surprisingly we only found significant differences between CO₂ treatments on ca. 8 out of 32 months of the study. However, the effect of CO₂ was much larger during periods of very high solar radiation load (up to 0.5 °C higher in elevated than ambient CO₂ treatment). Preliminary analyses suggest this leaf temperature increase is roughly consistent with expected differences that would arise from up to a 10% reduction in stomatal conductance.