Authentication of a wine’s origin has been an issue in the European Community for many years as the wine industry tries to minimise the potential for fraud and/or adulteration. Extending Bordeaux wine with wine from northern Africa was one such case. Much of the early work was based on element distribution and isotopic composition. The team of Professor Gérard-Jean Martin at the University of Nantes was at the forefront of geographical origin identification. For example, working with wines from the 1990 vintage, carbon, hydrogen and oxygen isotope analysis coupled with the analysis of eleven metals allowed the discrimination of wines from four regions in France as well as four subregions within Burgundy (J. Sci. Food Agric. 1995, vol. 67, pp. 113–23).
When I first started as Director of the National Wine and Grape Industry Centre (NWGIC), I raised the issue of establishing a basis for authentication of origin. At the time, however, there was little interest in the concept, especially as the success of the Australian wine industry in the export market was based around style and consistency of style from year to year. Blending across regions was necessary to achieve this consistency. Jacob’s Creek Chardonnay was a classic example of this strategy. One interesting side effect of this approach was the difficulty in communicating to overseas markets that there were wine regions in Australia. A survey of UK wine consumers on Australian wine regions gave the top three named regions as Hunter Valley, Barossa Valley and Chardonnay!
Times change, perhaps driven more by the realisation that some top-bracket Australian wines were the subject of fraud in overseas markets. Dr Martin Day at the Australian Wine Research Institute, previously with the Nantes research group, is leading a project on origin authentication. Martin makes the significant point that stable isotopes selected for origin analysis need to reflect the geology and water source in the vineyard, but are not influenced by winemaking. This rules out many metals found in wine – see Eric Wilkes’ article in the November/December issue (p. 37). Successful (98%) discrimination between 60 Australian and overseas wines could be achieved with this strategy (bit.ly/39YZzir), making the approach useful in detecting fraud.
The establishment of geographical indications (GIs) has been a positive step not just to authentication of origin but to defining regionality. The main purpose of a GI is ‘to protect the use of the regional name under international law, limiting its use to describe wines produced from wine grape fruit grown within that GI’. There is much more about GIs and their legal impact, as well as a list of GIs, on the Wine Australia website (bit.ly/3da7jQo).
A GI is in reality a physical description of a zone, region or sub-region. For example, the Hunter Valley (NSW) zone includes the Hunter region within which there are the Broke Fordwich, Pokolbin and Upper Hunter subregions. At the other extreme, the South Eastern Australia zone extends from North Queensland through Ceduna to the Great Australian Bight and includes all wine grapes grown to the south-east of that line. The legal definition and its use require 85% of the grapes to be grown within the claimed GI.
While the GIs are well defined in terms of place, little is known about the chemical composition of wines in a selected region and how this may differ between regions. That is, can compositional profiles, possibly coupled with sensory analysis, allow a region to be clearly identified in a blind tasting? This is now a major research endeavour, both here and in Europe.
A recent publication from the NWGIC described the use of 2D gas chromatography coupled to time-of-flight mass spectrometry to discriminate the Shiraz wine volatome between two regions in NSW (J. Agric. Food Chem. 2019, vol. 67, pp. 10 273−84). ‘Volatome’ was a new ‘ome’ word to me: it simply refers to the composition of volatile components. Wines were made in triplicate at two harvest dates according to a standard winemaking protocol, to remove winemaking influence of the profiles. The cooler region (Orange GI) showed a higher concentration of grape-derived volatile compounds than the warmer (Riverina) region. Fermentation-derived compounds were not related to the difference in climate. Harvest date also turned out to be an effective regional discriminator.
The next step in this regional differentiation is to try and establish the typicality of a varietal wine from a region and use this to discriminate between regions. Dr Wendy Parr has commented that ‘typicality estimates the degree to which a wine reflects geographical origin and varietal purity’ (bit.ly/390iuIj), whereas Dr Yves Cadot has proposed that typicality is the ‘perceived representativeness’ of the region (Anal. Chim. Acta 2010, vol. 660, pp. 53–62). The intriguing aspect of Cadot’s approach is that it requires a group of ‘wine experts’ to come up with descriptors of the region’s characters, often leading to heated debate.
Wine Australia is continuing to fund work on the regionality of Shiraz in the Barossa, Canberra, Heathcote, Hunter Valley, Yarra Valley and McLaren Vale regions (bit.ly/3bgkZI7). The results to date indicate that all six regions can be separated by their volatile composition, but that some compounds showed ‘large within region variations’, possibly reflecting ‘the different viticultural and oenological practices’ used by the individual wineries. Significant advances in understanding regionality and typicality are clearly being made.