Chemometric Analysis of FTIR Spectra to Classify Red Spinach Cultivation Systems

Abstract

Environmental stress from different cultivation systems, such as hydroponic and conventional methods, can significantly influence the metabolic composition of plants, affecting both functional compounds and overall plant quality. Rapid characterization of these differences is essential for monitoring crop responses and improving agricultural strategies. This study aims to classify red spinach (Amaranthus tricolor L.) parts—leaves, stems, and roots—cultivated under hydroponic and conventional systems using Fourier Transform Infrared (FTIR) spectroscopy combined with chemometric analysis. A total of 30 samples (hydroponic: HL, HS, HR; conventional: CL, CS, CR) were analyzed by using FTIR spectroscopy. Data were processed using Principal Component Analysis (PCA), Heatmap clustering, Partial Least Squares Discriminant Analysis (PLS-DA), and Orthogonal Projections to Latent Structures Discriminant Analysis (OPLS-DA). PCA captured 88.9% of total variance in the first two components, clearly separating sample groups. CL formed a distinct cluster, while HS also showed unique spectral characteristics. Overlaps among HR, HL, and CS indicated shared chemical profiles. Heatmap analysis supported this separation based on intensities at key functional group regions, including O–H, C–H, and C=O stretches. PLS-DA demonstrated strong predictive performance (R²Y = 0.998, Q² = 0.995), while OPLS-DA provided a more moderate yet robust classification (R²Y = 0.465, Q² = 0.446). In conclusion, FTIR spectroscopy combined with multivariate analysis is an effective tool to distinguish red spinach samples by cultivation method and plant part. This approach offers a rapid, non-destructive means for assessing plant responses to environmental conditions.