Photosynthesis responses to intrinsic water use efficiency depend on atmospheric feedbacks and modify the magnitude of response to elevated CO2

Abstract

Photosynthesis and transpiration affect the Earth system by altering carbon, water, and energy fluxes. Plants have valves on leaves called stomata that regulate gas exchange needed for these processes, thus tightly linking the flux of water and carbon. Water use efficiency represents plant water loss per carbon gain and can vary based on plant type and environmental conditions. It has observational uncertainty but in land models is typically simplified to one value per plant type. Here we examined how perturbations of water use efficiency influence photosynthesis using coupled Earth System model experiments. We found that while low water use efficiency perturbations led to a monotonic reduction in photosynthesis nearly everywhere, high water use efficiency perturbations have a photosynthetic response that is regionally dependent and sensitive to the inclusion of a dynamic atmosphere and leaf area. Notably, photosynthesis in the Amazon and central North America increases with an increase in water use efficiency, while photosynthesis in boreal regions decreases with increasing water use efficiency. The photosynthetic response to an increase in water use efficiency in these regions reversed with the inclusion of a dynamic atmosphere and dynamic leaf area, largely due to sensitivity to temperature and vapor pressure deficit increases. In contrast, low water use efficiency perturbations consistently resulted in a global decline in photosynthesis, with the inclusion of a dynamic atmosphere and leaf area modifying only the magnitude of the response. We also found that water use efficiency influenced the plant photosynthetic response to elevated atmospheric carbon dioxide (CO2), with both high and low perturbations modifying the total photosynthetic response on average by 89% and 37%, respectively. Assumptions about water use efficiency in land models and what components of the model we allow to dynamically respond significantly affect the climate outcome. By evaluating different assumptions of water use efficiency, our study contributes to the understanding of uncertainty in plant photosynthesis under future climate projections and how these projections may differ from real-world scenarios.