Effects of mesophyll conductance on vegetation responses to elevated CO ₂ concentrations in a land surface model

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.