The HPLC phenolic pro®les of 52 selected uni¯oral honey samples produced in Europe were analysed to detect possible markers for the¯oral origin of the different honeys. Lime-tree (®ve markers), chestnut (®ve markers), rapeseed (one marker), eucalyptus (six markers) and heather (three markers) honeys had speci®c markers with characteristic UV spectra. In addition, the¯avanone hesperetin was con®rmed as a marker for citrus honey, as well as kaempferol for rosemary honey and quercetin for sun¯ower honey. Abscisic acid, which had been reported to be a possible marker for heather honey, was also detected in rapeseed, lime-tree and acacia honeys. Ellagic acid in heather honey and the hydroxycinnamates caffeic, p-coumaric and ferulic acids in chestnut, sun¯ower, lavender and acacia honeys were also detected. The characteristic propolis-derived¯avonoids pinocembrin, pinobanksin and chrysin were present in most samples in variable amounts.

No abstract

European Eucalyptus honeys showed a common and characteristic HPLC profile in which the flavonoids myricetin (3,5,7,3',4', 5'-hexahydroxyflavone), tricetin (5,7,3',4',5'-pentahydroxyflavone), quercetin (3,5,7,3',4'-pentahydroxyflavone), luteolin (5,7,3', 4'-tetrahydroxyflavone), and kaempferol (3,5,7, 4'-tetrahydroxyflavone) were identified. Their contents, and relative amounts, in the analyzed honey samples were quite constant and supported their floral origin. In addition, ellagic acid and the propolis-derived flavonoids pinobanksin, pinocembrin, and chrysin were detected in most samples. The contents of these nonfloral phenolics were much more variable as could be expected for their propolis origin. Myricetin, tricetin, and luteolin had not been identified as floral markers in any other honey sample previously analyzed in our laboratory (chestnut, citrus, rosemary, lavender, acacia, rapeseed, sunflower, heather, lime tree, etc.) or reported in the literature, suggesting that these could be useful markers. Only in some individual heather samples produced in Portugal has tricetin previously been detected in minor amounts. These samples, however, were contaminated with Eucalyptus as revealed by their pollen analysis and the lack of tricetin or their glycosides in heather floral nectar. It remains to be established if myricetin, tricetin, and luteolin originate from Eucalyptus floral nectar where the corresponding glycosides should be present.

Flavonoids and other phenolics of Tunisian honey samples and propolis were analyzed to find correlations between botanical and geographical origin and chemical composition. Flavonoid content of honey was very variable (20−2400 μg/100 g). The richest samples contained the characteristic propolis flavonoids, while those containing fewer flavonoids were devoid of these substances. This is a difference with honeys from temperate areas, in which the poplar-derived flavonoids are present in all honey samples. In Tunisian honeys, and propolis, a new flavonoid, myricetin 3,7,4‘,5‘-tetramethyl ether, was detected. This is characteristic of Cistus spp. leaf exudates. Another minor compound was identified as quercetin 3,7,3‘-trimethyl ether. They were present in high amounts in propolis but were only detected in small amounts in honey. These results show that in border areas, such as Tunisia, where poplars are not always available for propolis collection, other plant sources can be used and their constituents detected in honey. Keywords: Phenolics; honey; propolis; flavonoids; characterization; geographical origin; botanical origin