SUN-INDUCED CHLOROPHYLL FLUORESCENCE SPECTRA: A POTENTIAL REMOTE SENSING SIGNAL FOR LEAF PHOTOSYNTHETIC PIGMENT ASSESSMENT IN RICE CROPS

Monitoring in situ photosynthetic pigment contents is of great significance for assessing photosynthetic capacity. Reported studies have focused on reflectance measurements for evaluating photosynthetic pigment contents, although the performance could vary among different cultivars. The capability of sun-induced chlorophyll fluorescence (SIF) is rarely explored and is currently considered a proxy for the photosynthesis of vegetation. This study aims to investigate the feasibility of evaluating leaf photosynthetic pigment contents via SIF yield spectra combined with empirical models and to compare their performance in different rice cultivars. The SIF signal of rice leaves was acquired by a FluoWat clip, and the reflectance spectra at the same leaf position were collected for comparison. The leaf chlorophyll and carotenoid contents were measured by chemical methods. The "lambda-by-lambda"band-optimization (LLBO) algorithm, classical partial least squares regression (PLSR), and potential Gaussian process regression (GPR) were used to develop models for evaluating photosynthetic pigments. The results showed that both reflectance spectra and SIF yield can effectively evaluate the photosynthetic pigments of the total dataset, although the reflectance-based GPR model for chlorophyll had the best results (R2 = 0.68, RMSE = 7.63 μg cm-2). In addition, the R2 of the reflectance-based GPR model for carotenoids was 0.58 with an RMSE of 1.17 μg cm-2. However, the SIF yield-based PLSR models constructed by using a single cultivar/material predicted other cultivars/materials better than the reflectance-based PLSR models. Whether based on the total dataset or the single cultivar/material, the generalization of SIF yield-based spectral indices to assess photosynthetic pigments was better than that of reflectance-based spectral indices, which is promising for developing a portable instrument based on SIF yieldbased spectral indices. © 2024 American Society of Agricultural and Biological Engineers.