| dc.description.abstract | Honey differentiation based on botanical origin remains challenging due to complex physical and compositional characteristics, particularly in tropical multifloral systems. In this study, an integrated analytical framework combining physical characterization, ATR–FTIR spectroscopy, and GC–MS volatile profiling with chemometric modeling was applied to differentiate Tanzanian honeys from miombo, mangrove, and mixed-flower origins. Floral origin exerted significant influence on honey composition, with Brix and water activity emerging as dominant discriminators (p < 0.001), reflecting differences in sugar concentration and moisture availability. Miombo honey exhibited a distinct profile, characterized by significantly higher viscosity, lower water activity, and enhanced yellowness, indicating a shift toward a more concentrated carbohydrate matrix. Spectroscopic analysis revealed that differentiation was primarily influenced by carbohydrate-related and moisture-associated spectral features, particularly the bands at 979.66, 1635.51, and 3471.87 cm?¹. Chemometric modelling confirmed that supervised classification (PLS-DA) significantly improved discrimination compared with PCA, with VIP analysis (VIP > 1) identifying these spectral markers as the major contributors to floral origin differentiation. Complementary GC-MS profiling demonstrated that relative abundance patterns of phenolic and carbonyl compounds, rather than unique markers, underpin floral differentiation, with miombo honey exhibiting a more distinct and consistent volatile signature.
All analytical approaches aligned to a consistent pattern, demonstrating that honey differentiation is driven by coordinated variations across physical, spectral, and volatile profiles. This study highlights the strength of integrated multi-analytical approaches for robust honey classification and provides a scientifically grounded framework for the authentication of tropical honeys. | |
