Rapid Spectroscopy-based Fingerprinting Combined with Data Fusion Chemometrics of Amaranthus tricolor L. for Growing Environment Discrimination

Abstract

Amaranthus tricolor L. is a nutrient-rich plant with significant health-promoting properties, making it a valuable candidate for functional foods and therapeutic applications. This study investigated the effect of the cultivation environment on the metabolite composition of A. tricolor L. by comparing soil-and hydroponically grown samples. Spectral analysis using ultravioletvisible (UV-Vis) and Fourier transform infrared (FTIR) spectroscopy, combined with chemometric data fusion, was conducted to assess the metabolic profiles. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were performed to evaluate the data, facilitating robust classification and accurate identification of the key metabolites. Volcano plot analysis indicated that the soil-grown plants accumulated higher levels of phenolics and flavonoids, whereas hydroponic cultivation favored the synthesis of photosynthetic pigments and membrane lipids. The PCA revealed distinct separation between the groups, with PC1 explaining 84.8% and 69.2% of the FTIR and UV-Vis variance, respectively. OPLS-DA models exhibited the highest classification accuracy (100%) with R²Y > 0.99 and Q² > 0.98. Notably, low-level data fusion further enhanced discrimination, increasing predictive variance to 76.0% and capturing subtle metabolic differences that single techniques cannot detect. These results suggested that soil cultivation promotes a defensive metabolic strategy that enhances the antioxidant capacity, whereas hydroponic systems support growth-oriented metabolism. The integrative chemometric and data fusion approach provided a comprehensive metabolomic fingerprint, offering valuable insights into the optimization of cultivation practices to enhance the nutritional and functional values of A. tricolor L.