Evolutionary stomatal adaptations to CO2 impact carbon fluxes and stocks under future climate

Current vegetation models represent photosynthetic and stomatal responses to environmental conditions, including atmospheric CO2 concentrations, as nearly instantaneous, perfectly plastic, and reversible. However, the fossil record shows that plants can exhibit slower, developmental- or evolutionary-scale adaptations through changes in leaf anatomy, including reduced stomatal density under higher atmospheric CO2 concentrations. If such responses were to occur on the timescale of current anthropogenic global change that could have implications for our predictions of terrestrial carbon storage as well as plant response to water stress.

We implement a representation of long-term adaptation to CO2 concentrations in the QUINCY land surface model (LSM) by imposing a maximum stomatal conductance calculated as a function of long-term average CO2 concentration, following the Medlyn stomatal model. This is a functional representation of changes in leaf anatomy that integrates with current representations of leaf level processes in LSMs. We show that even for long response times (100 years), the model predicts changes in carbon and water fluxes, with more pronounced responses for shorter response times (<40 years). While the introduction of this long-term response leads to increased water use efficiency across the globe, the direction of the response in plant growth differs between plant functional types (PFT). Broadleaf deciduous forests show decreased productivity, while evergreen needleleaf forests show increased productivity This pattern is driven by PFT-specific parameters, including the g1 slope parameter as well as PFT-environment interactions like growing season length. Further, the magnitude of the change in productivity is modulated by nutrient availability, with more nutrient-limited regions showing a smaller change in productivity.

These findings highlight the importance of incorporating long-term plastic plant responses to environmental change to vegetation models and differentiating between such responses, short-term acclimation, and interactions between plant function and environment.