Environmental control of canopy stomatal conductance in a tropical deciduous forest in northern Thailand

• We revealed the gs of a tropical deciduous forest over 6 years.
• A simple two-layer model ET model was used for estimating gs.
• The seasonality of two ecophysiological parameters, gsref and m, are shown.
• We obtained a simple relationship between soil moisture and gsref.

Tropical seasonal forests in Southeast Asia are among the most important biomes in terms of global and local hydrologic and carbon fluxes, and their vulnerability to climate change. We conducted eddy flux measurements in a teak (Tectona grandis Linn. f.) plantation in northern Thailand over a 6-year period; this forest undergoes a drastic seasonal change in foliage with somewhat constant incident radiative energy. We used a combination of actual evapotranspiration (ET) flux data and an inversed version of a simple two-layer ET model for estimating the mean canopy stomatal conductance (gs). The main novelty of this analysis is that canopy conductance can be extracted from total surface conductance (including the canopy and forest floor effects), and thus environmental and biological controls of gs are explicitly compared among seasons and years. The relationship between seasonal variations in the leaf area index (LAI) and gs revealed an apparent effect of leaf age on leaf gas exchange capacity: within a year, gs peaked earlier than full-leaf expansion and abruptly declined after the peak of LAI. We used this result to classify three leaf age stages: leaf-out, mid-growing, and leaf-senescence seasons. Then, two ecophysiological parameters, the reference value of gs (gsref), and the sensitivity of gs to atmospheric demand (m), as well as their proportion (m/gsref), were derived from the logarithmic response curve of gs to vapor pressure deficit (VPD) for each season. We showed seasonal variation in gsref as follows: leaf-out season ≈ mid-growing season > leaf-senescence season. m demonstrated little seasonality and little interannual variation was observed in either parameter. This resulted in a value of almost 0.6 for m/gsref during the leaf-senescence season and of less than 0.6 in the leaf-out and mid-growing seasons, which suggests that the teak trees had strict stomatal regulation to prevent excessive xylem cavitation during the leaf-senescence season (i.e., under drought conditions) and less strict stomatal regulation during the leaf-out and mid-growing seasons (i.e., under moist conditions) when little risk of water stress-induced hydraulic failure would occur. In addition, we obtained a simple linear relationship between soil moisture and gsref, which can be a powerful tool for further research of land–atmosphere interactions using global climate and vegetation dynamics models.