Applicability of optimal stomatal conductance models during a simulated heat wave in grapevine

Optimization models of stomatal conductance (gs) provide a conceptual framework to understand stomatal responses to environmental changes in terms of tradeoffs between the benefits and costs of stomatal opening. Theoretically, stomata maximize the benefits (carbon gain, A) and minimize the costs (water loss through transpiration, E). However, during heat waves, there can be conflicts between the need to maintain a high E to control leaf temperature via evaporative cooling and avoid heat damage, and the decrease in gs in response to increased VPD and reduced soil water availability. Under these conditions, the balance between the gains and the associated costs remains unclear. We measured leaf gas exchange (A, E and gs) in potted grapevines, cv Sauvignon Blanc, before, during and after a simulated six-day heat wave (Tmax = 40 °C), in the morning (10 to 12) and the afternoon (15 to 17), using heated well-watered (HW), heated drought-stressed (HD), non-heated well-watered (CW) and non-heated dry (CD) vines. We also measured plant transpiration (Elys) and leaf temperature (Tleaf) continuously during the heat wave using lysimeters and infrared cameras. We test the hypothesis that under combined stress, in addition to the stomatal limitation to A, there is an additional non-stomatal cost for A caused by the heat damage in the photosystem. We explore whether this potential additional cost is captured by two gs optimization models that include soil-to-leaf hydraulic conductance (model 2) or not (model 1). We also hypothesize that gs in HW vines may not follow the optimal predictions because these models do not include stress-related risks such as heat damage due to higher-than-optimal Tleaf resulting from stomal closure and reduced E. There were no significant differences in gs in HD and CD vines, during the morning or the afternoon measurements. Consequently, there were no differences in measured E and A between HD and CD vines during the peak of the heat wave. This was probably due to the stronger effect of water stress (soil water potential from -400 to -800 kPa) than high VDP (5 kPa) on gs during the peak of the heat wave. Under well-watered conditions, measured gs, E and A in the morning were much higher in HW than in CW vines and the values decreased from the morning to the afternoon. Preliminary results suggest that optimal models including dynamic responses to soil water potential can correctly integrate plant responses to heat and drought stress. The incorporation of stress-related risks (such as heat damage to the photosystem) into these models will be discussed.