Photosynthetic parameter estimations by considering interactive effects of light, temperature and CO2 concentration

Biochemical leaf photosynthesis models are evaluated by laboratory results and have been widely used at field scale for quantification of plant production, biochemical cycles and land surface processes. It is a key issue to search for appropriate model structure and parameterization, which determine model uncertainty. A leaf photosynthesis model that couples the Farquhar-von Caemmerer-Berry (FvCB) formulation to four different leaf temperature models is used to investigate the photosynthetic characteristics across a range of temperature gradients using both light (A-Q) and CO2 response curves (A-Ci). We used the Bayesian approach to fit the model to trial data of C3 crop plants (soybean, wheat) in the North China Plain and estimated key photosynthetic parameters, such as the maximum carboxylation rate of Rubisco (Vcmax25), the potential electron transport rate (Jmax25), leaf dark respiration in the light (Rd25), mesophyll conductance (gm25) and the kinetic parameter of Rubisco (Г*25) at a reference temperature of 25 °C. The results showed that 1) the model with moderate complexity showed the best goodness of fit, while conversely the simpler and more complex models were under and over fitting with their corresponding data, respectively; 2) the nonpeaked Arrhenius temperature response, which including both light and CO2 responses data gave the best estimates for the key parameters among the four models; and 3) the temperature gradient used to verify the model has greatly improved the estimation of six key parameters (Jmax25, Vcmax25, Rd25, Г*25, Kc25, g 25) with relatively more narrow confidence intervals (CIs) and showing regular variation on temperature gradient. Overall, this method offers an accurate basis for estimating leaf photosynthesis parameters and may enhance the accuracy of canopy, ecosystem and even global vegetation models.