Spectroscopic Methods for Botanical and Geographical Authentication of Honey from Mediterranean Area

Identifying the geographical and botanical origin of honey is critical to ensure authenticity. A wide range of spectroscopic techniques has been applied for honey authentication, including UV-Vis spectroscopy, Fourier-transform infrared spectroscopy, near-infrared spectroscopy, Raman spectroscopy, Nuclear magnetic resonance, isotope ratio mass spectrometry, and gas chromatography–mass spectrometry. Some of them require expensive instrumentation, complex sample preparation, and specialized expertise (e.g. GC-MS or IRMS). In this study, visible (VIS), near-infrared (NIR) spectroscopy, and energy-dispersive X-ray fluorescence (ED-XRF) were applied as methods potentially providing a rapid, non-destructive, and more accessible alternative for honey authentication. This study evaluated the effectiveness of diverse spectroscopic techniques in predicting the geographical and botanical origin of honeys of the Mediterranean area (Italy n = 20, Spain n = 6, France n = 8, Greece n = 11, and Croatia n = 9). Fifty-six honey samples were sourced from commercial retailers, including black locust (n = 14), chestnut (n = 15), linden (n = 13), and lavender (n = 12). Spectra were acquired using two NIR devices (benchtop and portable), one VIS colorimeter, and a benchtop ED-XRF instrument. Prediction models were developed using partial least square discriminant analysis (PLS-DA). Spectral data were divided into a calibration set (including 70% of the samples) and a validation set (including 30% of the samples). Model performance was assessed using sensitivity, specificity, positive and negative predictive values, and balanced accuracy for both the calibration and validation sets. In external validation, the VIS colorimeter achieved the highest performance for botanical origin (accuracy = 1). Conversely, the geographical origin showed limited predictability, with accuracy values ≤ 0.57 across all the spectroscopy techniques. Findings of the present study highlight the considerable potential of spectrometry-based approaches for discriminating botanical origins, while emphasizing the need for further methodological refinement to improve the robustness of geographical origin classification.