Rapid and accurate assessment of nutritional quality, particularly crude protein content and essential nutrient concentrations, remains a major challenge in the food and feed industries. In this study, laser-induced breakdown spectroscopy (LIBS) was combined with advanced chemometric modeling to predict the levels of crude protein and key macro- and micronutrients (Ca, Mg, K, Na, Fe, Mn, P, Zn) in 61 barley forage samples composed of whole aerial plant parts ground prior to analysis. LIBS offers a compelling alternative to traditional analytical methods by enabling real-time analysis with minimal sample preparation. To minimize interference from atmospheric nitrogen, nitrogen spectral lines were excluded from the protein calibration model in favor of spectral lines from elements biochemically associated with proteins. We compared the performance of Partial Least Squares (PLSR) regression and Extreme Learning Machine (ELM) using fivefold cross-validation. ELM outperformed PLS in terms of prediction, achieving a coefficient of determination (R²) close to 1 and a ratio of performance to deviation (RPD) exceeding 2.5 for proteins and several nutrients. These results underscore the potential of LIBS-ELM integration as a robust, non-destructive, and in situ tool for rapid forage quality assessment, particularly in complex and heterogeneous plant matrices.

Rapid quantitative characterization of protein-rich agricultural materials remains challenging due to the molecular and elemental complexity of plant-derived matrices. In this study, external-cavity quantum cascade laser-based mid-infrared (QCL-MIR) spectroscopy and laser-induced breakdown spectroscopy (LIBS) were investigated as complementary approaches for predicting protein and elemental composition in pea flour (Pisum sativum L.) samples. QCL-MIR spectroscopy primarily probed molecular vibrational information, whereas LIBS directly monitored elemental composition through plasma-generated atomic and ionic emission signatures. To address the high dimensionality and nonlinear relationships present in both spectroscopic datasets, advanced chemometric strategies combining Uninformative Variable Elimination (UVE), Partial Least Squares (PLS) regression, and Extreme Learning Machine (ELM) modeling were implemented. Given the limited dataset size (N = 20), predictive performance was assessed using a concentration-stratified five-fold prediction strategy. The optimized UVE–PLS–ELM approach improved predictive performance compared with conventional linear PLS models for both techniques. QCL-MIR spectroscopy showed strong predictive performance within the present dataset for total protein content (R² = 0.97) and also provided promising predictions of phosphorus and potassium through indirect correlations with sample composition. LIBS also demonstrated strong calibration and predictive capability for total protein content (R² = 0.93) and good predictive performance for Na, Mg, Ca, Mn, and Fe. However, LIBS performance decreased at concentration extremes, particularly for high-concentration elements (K and P) and trace elements such as Zn and Cu. These results demonstrate the complementary strengths of QCL-MIR and LIBS and highlight the potential of nonlinear chemometric approaches for accurate compositional analysis of protein-rich agricultural materials within the present limited dataset.