The forestry industry has integrated somatic embryogenesis into its clonal programs due to the generation of a high number of plants from selected genotypes at low cost. Somatic embryos are generated in a stressful environment after multiplication of the proembryogenic masses; thus, it is critical to determine the degree of stability of the embryogenic cultures and their potential for mass propagation. Maturation ability in cultures of different ages was evaluated in conjunction with the integrity of the proembryogenic masses, germination rate, hypocotyl and root length, plant conversion, and ex vitro survival. To identify differences in embryogenic tissue from different subcultures, their DNA was analyzed using FT-IR spectroscopy. A significant decrease in the production of somatic embryos was detected from week 15, and some lines even stopped producing embryos. Germination rate, hypocotyl length, and plant conversion were negatively affected by long-term cultivation, while root length and ex vitro survival were not significantly affected. The results obtained from the FT-IR spectroscopy analysis indicate that it is feasible to use mid-infrared spectroscopy to differentiate between embryogenic tissues with different cumulative subculture times based on the spectra obtained from their DNA, which is directly related to maturation ability.
Molasses, a byproduct of sugarcane and sugar beet processing, is widely utilized in the food, fermentation, and animal feed industries. However, authenticating its botanical origin remains challenging, often relying on costly, time-consuming, chemically intensive, and environmentally unsustainable methods. In response to increasing demands for sustainable analytical alternatives, this study aimed to develop and compare infrared spectroscopic methods to classify cane and beet molasses, focusing on sustainability of techniques while maintaining analytical performance. Data of portable and benchtop Fourier Transform Near-Infrared (FT-NIR) and Fourier Transform Mid-Infrared (FT-IR) spectrometers were evaluated using chemometric approaches, such as Principal Component Analysis (PCA) and classification models like Partial Least Squares Discriminant Analysis (PLS-DA) and k-Nearest Neighbors (k-NN). The Analytical GREEnness (AGREE) metric was employed to assess the sustainability of each technique, while analytical accuracy was evaluated using figures of merit derived from confusion matrices. FT-IR spectroscopy achieved the highest classification accuracy (0 % error) and revealed that beet molasses exhibiting higher protein content, whereas cane molasses contained more fructose. However, FT-IR scored the lowest in terms of greenness due to higher energy demands and sample handling in comparison with the other techniques. In contrast, portable FT-NIR was the most sustainable technique (AGREE score = 0.86, scale from 0 to 1), albeit with a slightly higher classification error (8.3 %). These findings demonstrate the potential of infrared spectroscopy as a reliable and sustainable solution for molasses authentication and show that sustainability–accuracy trade-offs can be quantitatively assessed to support informed decision-making in the analytical process of sugar industry.
The monitoring of storage-induced changes in plant-based products is essential to ensure quality and reduce postharvest losses. Spectroscopic techniques such as visible–near infrared (VIS–NIR) spectroscopy and hyperspectral imaging (HSI), combined with chemometric analysis, have shown strong potential for rapid and non-destructive assessment of compositional changes. However, their integrated application for modeling temporal changes during storage, particularly in sugar beet, remains insufficiently explored. This study evaluates the potential of VIS–NIR spectroscopy and HSI coupled with chemometric methods to monitor and predict storage-induced changes in sugar beet. Spectral data acquired using probe-based and hyperspectral approaches were analyzed using principal component analysis (PCA), k-nearest neighbors (k-NN), and partial least squares discriminant analysis (PLS-DA). PCA revealed clear discrimination between early (t0–t30) and advanced (t45–t60) storage stages, highlighting spectral features associated with chlorophyll degradation and phenolic oxidation. Supervised models showed high classification performance, with the HSI-Avg model, defined as the average of all pixel spectra within each hyperspectral image, providing the best results. The k-NN model achieved 93.35% classification accuracy (external validation), while PLS-DA reached 100% accuracy with 0% error. In contrast, models based solely on HPLC sugar concentrations failed to adequately discriminate storage stages. Overall, this integrated approach provides a rapid and reliable strategy for predicting storage time and assessing quality changes in sugar beet, supporting the development of predictive tools for postharvest management and quality control in the agro-food industry.