Importance of mesophyll diffusion conductance in estimation of plant photosynthesis in the field.

Mesophyll diffusion conductance to CO(2) (g(m)) is an important leaf characteristic determining the drawdown of CO(2) from substomatal cavities (C(i)) to chloroplasts (C(C)). Finite g(m) results in modifications in the shape of the net assimilation (A) versus C(i) response curves, with the final outcome of reduced maximal carboxylase activity of Rubisco (V(cmax)), and a greater ratio of the capacity for photosynthetic electron transport to V(cmax) (J(max)/V(cmax)) and alterations in mitochondrial respiration rate (R(d)) when estimated from A/C(i) responses without considering g(m). The influence of different Farquhar et al. model parameterizations on daily photosynthesis under non-stressed (C(i) kept constant throughout the day) and stressed conditions (mid-day reduction in C(i)) was compared. The model was parameterized on the basis of A/C(C) curves and A/C(i) curves using both the conventional fitting procedure (V(cmax) and R(d) fitted separately to the linear part of the response curve and J(max) to the saturating part) and a procedure that fitted all parameters simultaneously. The analyses demonstrated that A/C(i) parameterizations overestimated daily assimilation by 6-8% for high g(m) values, while they underestimated if by up to 70% for low g(m) values. Qualitative differences between the A/C(i) and A/C(C) parameterizations were observed under stressed conditions, when underestimated V(cmax) and overestimated R(d) of A/C(i) parameterizations resulted in excessive mid-day depression of photosynthesis. Comparison with measured diurnal assimilation rates in the Mediterranean sclerophyll species Quercus ilex under drought further supported this bias of A/C(i) parameterizations. While A/C(i) parameterization predicted negative carbon balance at mid-day, actual measurements and simulations with the A/C(C) approach yielded positive carbon gain under these conditions. In addition, overall variation captured by the best A/C(i) parameterization was poor compared with the A/C(C) approach. This analysis strongly suggests that for correct parameterization of daily time-courses of photosynthesis under realistic field conditions, g(m) must be included in photosynthesis models.