| dc.description.abstract | Terrestrial carbon and water cycling are predicted to undergo large changes over the coming decades due to anthropogenic climate change. These cycles are dependent on the plant hydraulic system, because of its close link with leaf photosynthesis, however these dependencies remain poorly understood. Plant hydraulic traits play a pivotal role in regulating transpiration, water supply, carbon uptake and ecosystem resilience, but how these traits control these ecosystem scale variables is yet to be properly quantified. This is, in part, because the controls on the spatial variations in hydraulic traits themselves remains poorly understood. Eco-evolutionary optimality (EEO) principles have been successfully applied to trait predictions globally, this makes it a promising approach to enhance our ability to predict global hydraulic trait distributions. Within this thesis I progress through a series of steps to develop and test a global hydraulic EEO model.
Close coupling of water loss and carbon uptake requires close coordination between hydraulic and photosynthetic traits. Therefore, firstly I develop a hydraulic trait model incorporating photosynthetic traits based on EEO principles. The model predicts that optimal carbon allocation to leaves and stems generates equilibrium between stem water supply and water demand for photosynthesis, allowing the prediction of the sapwood to leaf area ratio (vH). This model is the first step in quantifying trait-trait and trait-climate relationships in a model that incorporates hydraulic traits. In the second chapter, I use empirical data from 18 sites along an elevation gradient in the Gongga Mountains in China to reveal a trait coordination network with vH at its nexus. This trait network connects hydraulic, photosynthetic, and resource acquisition strategies. vH exhibits a positive correlation with carboxylation capacity, is negatively related to hydraulic efficiency, aligning well with my model predictions. The comparisons between observed and predicted photosynthetic traits suggest the plants are well acclimated to short-term midday climate conditions, supporting them having a water supply-demand balance, as predicted by the model. Finally, I develop this analysis at a global scale using global hydraulic trait databases to evaluate the model and to show that plants in cold, dry, and high-irradiance environments have low hydraulic efficiency and exhibit high vH, consistent with predicted patterns. The EEO-based hydraulic trait model explains 56% of variation in vH on a global scale with only one parameter. These findings provide insights to predict how plants adjust coordinated traits to adapt to climate change.
The hydraulic trait model developed in this thesis offers an alternative way to offer significant improvements to the carbon allocation scheme in dynamic vegetation models, without the use of many additional parameters, which increase model uncertainty. Consequently, this modelling approach offers the opportunity to considerably enhance future capacity to predict vegetation responses to climate change, particularly extreme conditions, such as drought. | en_GB |
