Estimating the stomatal slope (g1) parameter from the Medlyn model across Australian vegetation using multiple observational datasets

Stomatal conductance plays a key role in the exchange of carbon and water between vegetation and the atmosphere, by controlling photosynthesis and transpiration in plants. However, its representation in land surface models (LSMs) is still considered a major source of uncertainty. The Medlyn model is widely used in LSMs to describe how strongly stomata and the model parameter g1 respond to carbon uptake and atmospheric dryness. This parameter is often introduced as a fixed value in global LSMs for broad vegetation types despite evidence that stomatal behaviour varies with plant water availability, particularly in water limited ecosystems such as those found in large parts of Australia.

We estimated g1 for Australian vegetation using observational data obtained from three intrinsic water use efficiency techniques: leaf gas exchange, stable carbon isotope discrimination, and eddy covariance. These approaches together can provide a comprehensive knowledge on the estimation of g1 across a range of spatial and temporal scales, from leaf to ecosystem and from short to long term responses. To account for water stress, we relate g1 to soil moisture for both leaf-scale gas exchange and eddy covariance datasets, where direct plant water status measurements are rarely available. For the stable isotope dataset, water stress is represented using an aridity index that reflects longer-term water limitation experienced by plants over the period of carbon assimilation.

We compared g1 estimates from these datasets along with soil moisture data to observe the shifts in the sensitivity of stomata under dry conditions and to determine consistency between scales. We found that g1 varied systematically across Australian plant functional types (PFTs), with lower values in xeric shrubs and C4 grasses and higher values in savanna and rainforest trees. Relative differences among PFTs were consistent across methods, but isotope-derived g1 values were generally higher than leaf gas exchange estimates. Eddy covariance data from Australian flux-tower sites showed a clear increasing trend in g1 with increasing soil moisture, and isotope-derived g1 decreased with increasing aridity, indicating more conservative stomatal behaviour under dry conditions.

These findings will be used to generate representative values of g1 for Australian PFTs that can be implemented in land surface models (e.g., The Joint UK Land Environment Simulator JULES) for evaluation at flux-tower sites and the continental scale.