The traditional wood used to make barrels destined for use in the world wide wine industry is oak. However, oak chips and shavings can substitute for barrels to add flavour to wine and are very much more cost effective. As with the heat treatment of barrels, oak chips are toasted before use. This serves to pyrolyse lignin and hemicellulose, generating families of compounds that impart desirable flavours to wine. Other woods are very occasionally used in wine barrel construction, but no chips other than oak chips have been used to flavour wine. This is surprising given that all woods contain lignin and hemicellulose, the composition of which will vary perhaps usefully from species to species. The 12 woods used in this research, including American oak, were chosen on several criteria: botanical similarly to oak, exclusivity to New Zealand, and historical association with New Zealand. The woods were cut to chips measuring about 10 x 20 x 2.5 mm. The moisture content was measured after dry heating to 110°C. Fresh samples of chips were heated (toasting in the context of wine) to 200°C for 2 hours, 210°C for 3 hours, called light and heavy toasting respectively. Weight loss was determined. The colour of the untreated and toasted wood chips was measured in Hunter colour space, yielding data on lightness (L*), hue angle (the basic colour) and saturation (the intensity of colour). The moisture content of oak was the lowest of all the woods. The weight loss of oak chips at 200°C was much greater than that of other woods, but the colour change did not indicate losses due to severe charring. Overall, each wood behaved in a distinctive way to the toasting treatments, with some charring much more than others. Hue was the least affected, indicating that the basic colour of the woods was little changed by toasting. Light and saturation generally decreased strongly, particularly on heavy toasting. Colour was thus being lost and less light reflected. An unoaked chardonnay was infused with toasted chips at the rate of 5 g.L-1 for two weeks at room temperature, and later decanted. At all stages exposure to air was minimised. The 25 treatments (2 x 12 plus the unwooded control) were first assessed by a panel comprising eight experienced wine tasters and 29 AUT staff members who claimed some knowledge of wine flavour. This qualitative/semi-quantitative analysis required tasters to assess the wines in terms of 12 descriptors commonly associated with oaked wines (boxes were ticked for ‘sweet oak’, ‘smokey’, ‘vanilla’ etc.), and to choose the three most liked and the three least liked. 6 Confidential A principal component analysis of a correlation matrix of descriptors was used to summarise panelist’s opinion. The first two principal components explained 53 % of the variation and served to group descriptors into four quadrants, which were each associated with different woods and toasting levels. Most liked were totara light (toast), kahikatea heavy, manuka heavy and American oak light. Macrocarpa light toast was almost universally disliked. On the basis of liking and association with New Zealand, five woods and chosen toasting levels and the control were selected for hedonic trials (1 to 9 liking scale) with 180 consumers (age range and gender were identified) in six retail wine shops. The decreasing numerical of liking by treatment was totara (6.49), control, manuka, American oak, kahikatea, radiata pine (5.47), with an overall significant effect (P < 0.001) for treatment. Tukey’s test revealed that only totara and the control treatments were outstanding (P < 0.05). Retail wine shop as a factor was marginally significant. Older consumers liked the wines more (P < 0.05), as did females (P < 0.001). There were no significant interactions between any of the factors. Because of the difficulties in sourcing totara, manuka appears to be the most viable alternative to oak as a wine flavouring in the New Zealand context.
| Identifer | oai:union.ndltd.org:ADTP/284461 |
| Date | January 2008 |
| Creators | Mahajan, Ishita |
| Publisher | AUT University |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
Page generated in 0.0015 seconds