Understanding Stress-induced Dynamics of Solar-induced Chlorophyll Fluorescence with Leaf-level Hyperspectral Imaging

Proximal sensing of solar-induced chlorophyll fluorescence (SIF) is widely used as a non-invasive indicator of photosynthetic activity. Together with photochemical quenching (PQ) and non-photochemical quenching (NPQ), SIF represents a third pathway of energy dissipation in excited chlorophyll molecules and provides direct insight into the redox state of the photosystem, particularly under stress conditions, often with shorter response times than vegetation indices (VIs). Because SIF does not require artificial illumination, it can be applied across spatial scales from leaf to global satellite observations. However, two major challenges limit the direct interpretation of SIF in terms of photosynthetic efficiency: (1) accurate estimation of SIF yield due to confounding effects of absorbed PAR (APAR) and fluorescence escape probability, both influenced by illumination geometry, plant structural, and physiological properties; and (2) incomplete understanding of the relationships among SIF yield, NPQ, and PQ under stress. To address part of these challenges, this thesis first introduced a ground-based hyperspectral imaging system (HyScreen) for high-resolution retrieval of red and far-red SIF. The proposed measurement protocol and processing chain achieved high accuracy, with relative errors of 0.4–1.0% for red SIF and 0.2–0.5% for far-red SIF due to noise. Second, a controlled cold-stress experiment was conducted to evaluate the combined use of SIF and VIs as indicators of cold stress tolerance in <em>Arabidopsis thaliana</em>. Specifically, the study examined whether SIF and VIs can (1) detect cold stress duration, (2) distinguish genotypes with contrasting NPQ capacity, and (3) resolve differences in photoinhibition between young and mature leaves. The results show that (1) acute cold stress initially induced high SIF yield with high near-infrared (NIR) reflectance, followed by decreasing SIF, reduced NIR reflectance, and reduced red-edge normalized difference vegetation index (NDVIre) under photoinhibition. Low values of the effective quantum yield of PSII together with high NPQ indicated that NPQ dominated SIF quenching during cold stress, while changes in NIR reflectance and NDVIre indicated cold-induced alterations in leaf structure and reduced chlorophyll content, respectively. (2) Four genotypes differing in NPQ capacity were distinguishable by combining SIF yield, photochemical reflectance index (PRI), and NDVIre. Early SIF responses reflected the fast NPQ component qE, whereas the change and temporal profile of PRI captured both qZ and overall NPQ during prolonged stress. Importantly, higher qE or qZ capacity did not necessarily correspond to reduced photodamage, as chlorophyll degradation (indicated by reduced NDVIre) contributed strongly to photoinhibition. (3) Young leaves exhibited greater cold tolerance than mature leaves, as evidenced by higher Fv/Fm after cold exposure, characterized by higher qE capacity, retention of qZ activity, and better chlorophyll maintenance, as indicated by lower initial SIF yield, higher PRI, and higher NDVIre. In summary, HyScreen enables leaf-level SIF imaging while minimizing background effects and provides new opportunities to investigate spatial heterogeneity in photoprotective responses from leaf to plant scales. This study suggests that the combined use of SIF and VIs can capture NPQ dynamics, chlorophyll degradation, and leaf structural responses linked to cold stress tolerance across genotypes and leaf ages. This work also provides insights relevant to future validation activities for the ESA FLEX mission and to high-throughput screening of stress-tolerant plants in breeding programs.