Abstract
Mesophyll conductance (g(m) ) is known to affect plant photosynthesis. However, g(m) is rarely explicitly considered in land surface models (LSMs), with the consequence that its role in ecosystem and large-scale carbon and water fluxes is poorly understood. In particular, the different magnitudes of g(m) across plant functional types (PFTs) are expected to cause spatially divergent vegetation responses to elevated CO(2) concentrations. Here, an extensive literature compilation of g(m) across major vegetation types is used to parameterize an empirical model of g(m) in the LSM JSBACH and to adjust photosynthetic parameters based on simulated A(n) - C(i) curves. We demonstrate that an explicit representation of g(m) changes the response of photosynthesis to environmental factors, which cannot be entirely compensated by adjusting photosynthetic parameters. These altered responses lead to changes in the photosynthetic sensitivity to atmospheric CO(2) concentrations which depend both on the magnitude of g(m) and the climatic conditions, particularly temperature. We then conducted simulations under ambient and elevated (ambient + 200 μmol/mol) CO(2) concentrations for contrasting ecosystems and for historical and anticipated future climate conditions (representative concentration pathways; RCPs) globally. The g(m) -explicit simulations using the RCP8.5 scenario resulted in significantly higher increases in gross primary productivity (GPP) in high latitudes (+10% to + 25%), intermediate increases in temperate regions (+5% to + 15%), and slightly lower to moderately higher responses in tropical regions (-2% to +5%), which summed up to moderate GPP increases globally. Similar patterns were found for transpiration, but with a lower magnitude. Our results suggest that the effect of an explicit representation of g(m) is most important for simulated carbon and water fluxes in the boreal zone, where a cold climate coincides with evergreen vegetation.