Physiological tradeoffs in the parameterization of a model of canopy transpiration

We examined physiological parameter tradeoffs in modeling stomatal control of transpiration from a number of forest species. Measurements of sapflux, micrometeorology, and leaf area index were made in stands representing 85% of the forest ecosystems around the WLEF eddy flux tower in northern Wisconsin. A Jarvis-based canopy conductance model was used to simulate canopy transpiration (EC) for five tree species from these stands. They consisted of conifers and deciduous species in both upland and wetland locations. Parameter estimation was used to assess the tradeoffs between physiological parameters used in the calculation of stomatal conductance. These tradeoffs were then evaluated against current theory on stomatal regulation of leaf water potential. The results show that the best simulations of EC were obtained with values of maximum stomatal conductance (gSmax) and stomatal sensitivity to vapor pressure deficit (δ) that closely followed this hydraulic theory. The model predictions reveal a large variation in the strategies used to regulate water potential among species. Aspen showed the greatest tendency towards efficiency, indicating that it has high EC under low vapor pressure deficit (D) conditions, but is susceptible to rapid EC decline at moderate to high D. Other species showed more conservative water use. The results indicate that inter-specific differences in dynamic response to D can produce large spatial variation in EC under typical environmental conditions.