- 1Department of Geography, University of Cambridge, Cambridge, UK (adf10@cam.ac.uk)
- 2Conservation Research Institute, University of Cambridge, Cambridge, UK (adf10@cam.ac.uk)
- 3College of Ecology and Environment, Nanjing Forestry University, Nanjing, China (chenyzvest@gmail.com)
- 4Department of Physical Geography and Ecosystem Science, Lund University, Lund, Sweden (annemarie.eckes-shephard@nateko.lu.se)
- 5Research Unit Forest Dynamics, Swiss Federal Institute for Forest Snow and Landscape Research WSL, Birmensdorf, Switzerland (patrick.fonti@wsl.ch)
- 6Sainsbury Laboratory University of Cambridge, University of Cambridge, Cambridge, UK (eva.hellmann@slcu.cam.ac.uk)
- 7Institute of Temperate Forest Sciences, University of Quebec in Outaouais, Outaouais, Canada (Tim.Rademacher@uvm.edu)
- 8Center for Ecosystem Science and Society, Northern Arizona University, Flagstaff, Arizona, USA (Andrew.Richardson@nau.edu)
- 9School of Informatics, Computing, and Cyber Systems, Northern Arizona University, Flagstaff, Arizona, USA (Andrew.Richardson@nau.edu)
- 10Department of Pure Mathematics and Mathematical Statistics, University of Cambridge, Cambridge, UK (prt38@cam.ac.uk)
Dynamic global vegetation models (DGVMs) are used to attribute historical and forecast future atmosphere-land carbon (C) exchange. While mean long-term behaviour across DGVMs is compatible with observational constraints on the global C cycle over recent decades, differences between models are high both in annual fluxes and attribution of the long-term net carbon uptake to drivers such as atmospheric CO2, indicating significant uncertainty regarding process understanding. These models are largely C source-driven, with behaviour primarily determined by the environmental responses of photosynthesis. However, real plants are integrated wholes, with feedbacks between sources (e.g. photosynthesis) and sinks (e.g. growth) resulting in homeostatic concentrations of metabolites such as sugars. An approach to implementing such behaviour in a plant growth model is presented and its implications for responses to environmental factors assessed. The approach uses Hill functions to represent inhibition of C sources (net photosynthesis) and activation of sinks (structural growth) based on sugar concentrations. The model is parameterised for a mature tropical rainforest site and its qualitative behaviour is found to be consistent with experimental observations. Key findings are that sinks and sources strongly regulate each other. For example, doubling potential net photosynthesis (i.e. the rate that would occur without feedback) results in growth increasing by only 1/3 at equilibrium, with increased sugar concentration causing feedback-inhibition of photosynthesis. A C source-only driven response, as in current DGVMs, would result in close to a doubling of growth. Hence, in this approach, environmental factors that affect potential net photosynthesis, such as atmospheric CO2, have greatly reduced effects on growth and net C uptake when homeostatic behaviour of sugars is considered. Implications for understanding and modelling the global carbon cycle are discussed.
How to cite: Friend, A., Chen, Y., Eckes-Shephard, A., Fonti, P., Hellmann, E., Rademacher, T., Richardson, A., and Thomas, P.: Implications of plant metabolic source-sink feedbacks for modelling the terrestrial carbon balance, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20116, https://doi.org/10.5194/egusphere-egu25-20116, 2025.
