Abstract :
[en] Photosynthesis plays a key role in the carbon cycle, being the primary mechanism through which carbon dioxide (CO2) is converted into organic compounds by plants, algae, and certain bacteria. On average, terrestrial ecosystems uptake around one-third of anthropogenic CO2 emissions and participate to the mitigation of climate change. Temperature, solar irradiance or soil water availability are important factors mediating the strength of terrestrial ecosystems. Therefore, drought poses a major threat on the atmospheric CO2 balance due to its profound implications for plant survival, ecosystem dynamics and agricultural productivity.
Water stress induces a cascade of physiological responses in plant functioning. During hot and dry weather, plants often close their stomata to save water and prevent dehydration at the expanse of a decrease in CO2 supply to the chloroplasts, which results in a decrease in both carbon assimilation and transpiration. In addition, photosynthesis and stomatal closure can be impacted by non-stomatal factors, which complexifies the mechanisms regulating plant responses to water stress. Models of photosynthesis and stomatal conductance still lack from a detailed characterization of the intertwined physiological processes under soil water-limiting conditions. Understanding, disentangling and quantifying the importance of photosynthesis limiting factors is crucial for selecting drought-tolerant varieties and for improving model predictions.
The gold standard method for estimating photosynthesis is the measurement of the CO2 net assimilation rate by gas exchange techniques. While providing key information on the factors influencing the temporal variability of photosynthesis, gas exchanges alone do not allow to fully characterize the limitations on carbon assimilation under water stress. Additional measurements of chlorophyll fluorescence by active methods are needed to improve the knowledge about these constraints and to quantify the importance of each limiting factors. In particular, the response of mesophyll conductance to water stress remains a pivotal uncertainty in models of photosynthesis.
The interpretation of actively-induced chlorophyll fluorescence measurements is however limited to the leaf scale and does not allow to elucidate the impact of water stress on large patches of vegetation. The recent emergence of passive techniques has allowed the monitoring of fluorescence emission induced by solar irradiance (SIF) at different temporal scales and spatial resolutions. These measurements provide a promising indicator of vegetation physiological processes which can be used to calibrate empirical models to estimate carbon assimilation at local and large scales. However, such relationships typically fail in reproducing photosynthesis temporal evolution at short timescales or during climate extremes such as droughts or heatwaves. More mechanistic approaches are needed to fully exploit the physiological message carried by SIF measurements.
In this thesis, we first aimed to unravel the origins of photosynthesis limitations under water stress. The first chapters are dedicated to providing a calibration of the response of stomatal and non-stomatal factors to the decrease in soil water availability. This method, first applied on forest ecosystems (chapter 3) and potato (chapter 4) with eddy covariance (EC) data, emphasizes the key role played by non-stomatal factors in regulating canopy-scale photosynthesis for these two very different plant functional types. By performing measurements of leaf-level chlorophyll fluorescence by active methods, and by using a new partitioning method, we also showed (chapter 5) that mesophyll conductance is a pivotal trait which regulates both carbon assimilation and stomatal conductance of potato. These chapters provide functions of soil water availability for stomatal and non-stomatal factors which can be used to reduce uncertainties in climate models. Finally, we evaluated the capability of a new mechanistic approach to model CO2 assimilation and transpiration of a winter wheat crop from proximal sensing of SIF (chapter 6). A very strong correlation between model estimates and EC measurements was found across a wide range of environmental conditions including edaphic drought. A sensitivity analysis revealed that the fraction of open photosystem II centers is a key parameter affecting model robustness. This last chapter paves the way towards an improvement of the understanding of the interactions between the water and carbon cycles by using the physiological information carried by SIF.
