Seasonal analysis of photosynthetic limitations in mature Scots pine trees reveals that models based on intracellular CO2 concentration grossly underestimate photosynthetic capacity

Global models of photosynthesis commonly use photosynthetic capacity parameters (V_cmax, J_max) that are estimated based on C_i, the intercellular CO₂ concentration. The underlying assumption of these models is that mesophyll conductance (gm) is infinite and therefore the CO₂ concentration at the site of carboxylation (C_c) is equal to C_i, despite a growing body of literature acknowledging that these assumptions are incorrect. Because relatively few studies on gm have been conducted under natural conditions and with high enough resolution, it is currently unclear how significant the assumption of infinite C_c is for the accuracy of long-term predictions by large-scale photosynthesis models. In this study we investigated this question with data collected in a mature Scots pine stand, one of the dominant species of the boreal region. We conducted high-resolution gas-exchange and online ¹³C discrimination measurements over a whole growing season (May-October), and analysed the relative contribution of diffusional and biochemical limitations to photosynthesis. We hypothesised that diffusional limitation will be significant in this species, as conifers typically have low stomatal and mesophyll conductance (g_s and g_m respectively). Accordingly, we found that diffusional limitations were similar or stronger than biochemical limitation during May-July, and that all limitations reached minima around the end of June when A_net values were highest. However, during August-October biochemical limitation became increasingly dominant, as the diffusional limitations were relatively small and stable. Over the whole period, both g_m and the relative mesophyll limitation were similar in magnitude to g_s and the stomatal limitation, respectively, resulting in a 40-100 ppm reduction in CO2 concentration between C_i and C_c. This meant that V_cmax was under-estimated by 20-40% when calculated from C_icompared to C_c, highlighting the importance of accounting for the finite g_m when determining photosynthetic capacity and modelling photosynthesis under natural conditions.