Species-Dependent Antibacterial Responses and Bioactivity Patterns Associated with Declared Botanical Type in Honey:...

Honey antibacterial activity reflects both the properties of the honey matrix and the susceptibility of the target microorganism, but the consistency with which these sources of variation are reflected across antimicrobial endpoints remains insufficiently characterised. This study investigated whether species-dependent differences in antibacterial responses were detected more consistently than differences among declared botanical honey types and whether the integration of antimicrobial and biochemical traits could reveal botanical-type-associated bioactivity patterns not apparent from individual endpoints. Thirty-seven honey samples representing seven declared botanical types from Western Pomerania, Poland, were analysed for total phenolic content (TPC), ferric-reducing antioxidant power (FRAP), operational minimum inhibitory concentration producing ≥90% growth inhibition (MIC90), and minimum bactericidal concentration (MBC) against Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Pseudomonas aeruginosa, and Proteus vulgaris. Buckwheat honeys showed the highest observed TPC and FRAP values among the analysed samples. Species-dependent differences in MIC90 and MBC were detected more consistently than differences among declared botanical types, whereas significant botanical-type effects for individual antimicrobial endpoints were observed only for S. aureus. P. vulgaris was generally the most susceptible species, although susceptibility patterns differed between inhibitory and bactericidal endpoints. Associations between biochemical parameters and antimicrobial activity were limited and directionally inconsistent. In contrast, the integration of five MIC90 variables, five MBC variables, FRAP activity, and TPC revealed significant botanical-type-associated structuring of the overall bioactivity profile. This association remained significant after the exclusion of botanical types represented by single samples (pseudo-F = 2.66, R² = 0.283, p = 0.0001), while multivariate dispersion no longer differed significantly among groups. The resulting bioactivity fingerprint should be regarded as a descriptive multidimensional representation of the investigated samples rather than as a validated classification or prediction tool. Within the present dataset, combined species-resolved microbiological and biochemical profiling provided a more comprehensive description of honey bioactivity than individual antimicrobial or biochemical endpoints considered separately.