This work characterises volatile floral source descriptors for a variety of Australian unifloral honeys. The chemical analysis of naturally occurring honey volatiles is useful for authenticating the floral source of Australian honeys for domestic and overseas markets. Further, the floral and geographic origin of honey will need to be supplied by Australian honey exporters competing in the lucrative European market. With the European market, honey labelling now has to include information that shows the floral and geographical origin of the product, particularly if it is labelled as floral. Chemical fingerprinting of these volatiles has potential application to the authentication of the botanical source, which is of increasing commercial importance to the Australian honey industry in quality assurance programs.
This methodology was successfully applied to both Eucalyptus and non-Eucalyptus honey samples. Also, the components responsible for the strong, distinctive aroma and flavour of Australian honeys were identified. These, together with taste flavour compounds such as organic acids, make up the fiiU flavour of honey. In addition, high boiling semivolatiles are included in the flavour profile of honey. Floral source marker compounds for high quality unifloral (or 'straightline') samples of 11 floral types of Australian high quality species-specific or 'straightline' samples of floral honey have been identified using the volatile fraction. This flavour profile of volatile compounds represents a distinctive 'fingerprint' of the honey that can authenticate the floral source of the honey.
The three-step Floral Certification Test which has been developed involves firstly a solvent extraction of volatiles from honey, followed by their detection and quantification using gas chromatography (GC) and gas chromatography-mass spectrometry (GC-MS). Finally, multivariate statistical analysis of the chemical data is performed. This methodology is applicable to any honey throughout the world once an initial survey is done on the composition of its volatile fraction. These procedures have permitted the identification of numerous compounds, including norisoprenoids, monoterpenes, benzene derivatives, aliphatic compounds, and Maillard reaction products. The characterisation of these volatiles is, therefore, of interest in the search for possible floral source descriptors for 11 floral types of Australian honey. The honeys studied were the eucalypt floral types, red gum, yapunyah, pilliga box, Caley's ironbark, spotted gum, and bloodwood; and the non-eucalypt floral types, crows ash, jelly bush, heath, tea tree and brush box.
The six eucalypt floral types of honey examined [red gum (E. camaldulensis), yapunyah (E. ochrophloia), pilliga box (E. pilligaenis), Caley's ironbark (E. caleyi), spotted gum (E. maculata) and bloodwood (E. intermedia)] were subjected to principal components analysis (PCA). A total of 21 variables (volatiles/groups of volatiles) were selected for these floral types of honey and analysed with corresponding published data on two eucalypt floral types of honey [yellow box (Eucalyptus melliodora) and blue gum (E. leucoxylon)]. These variables represent a diverse range of compound categories. In conclusion, it will be possible in the fiiture to authenticate these eight eucalypt floral types of honey using this methodology and the accimiulated chemical data summarised above, which will support quality assurance programs in the honey industry.
A total of five non-eucalypt floral types of honey were analysed. Each was considered separately with respect to the floral source descriptor volatiles and PCA of collected GC data. Crow ash (Guioa semiglaucd) honey has some similarities to leatherwood honey, particularly related to monoterpene levels and variety.
Jelly bush (Leptospermum sp.) honey has some similarities to the New Zealand Leptospermum honeys, manuka and kanuka honeys. For jelly bush honey, GC analysis was able to distinguish between non-peroxide biologically active and inactive samples. The main distinguishing feature in the GC fingerprint of unmethylated extracts of active samples was a very large broad peak that elutes early and is absent in unmethylated extracts of inactive samples. When quantification of this compound(s) was combined with that for the other floral source descriptors, o-anisic acid (2-methoxybenzoic acid), dehydrovomifoliol, methyl syringate, and combined meso- and levo-butane-2,3-diols, and PCA of the data was undertaken, it is possible to authenticate non-peroxide active jelly bush honey samples from non-peroxide inactive ones.
For heath (Banksia ericifolia) honey and tea tee (Melaleuca quinguenervia) honey, there were similarities in the composition of volatiles. Floral source descriptors for these two floral types of honey were identified. Benzene derivatives were well represented amongst the volatiles floral source descriptors for these two honey types.
The final non-eucalypt honey studied in detail was brush box (Lophostemon conferta) honey. The composition of the volatile fraction of samples of this floral type was variable. Some unusual Cg and Cio norisoprenoids characterise this variety from other types.
In conclusion, this project has developed a rapid chemical procedure (Floral Certification Test based on 'chemical fingerprinting') for authenticating the floral origin of samples of species-specific floral types of Australian honey. The large amount of accumulated chemical data will support authenticity testing of floral Australian honeys in the future, as part of quality assurance programs of the Australian honey industry.