Benchmarking regional and subregional influences on Shiraz fine wines
Abstract
Sensory properties of Shiraz wines from the Barossa Valley, McLaren Vale, Heathcote, Yarra Valley, Canberra District and Hunter Valley have been detailed and compared. Selection of wines utilised a new sensory method, Pivot©Profile. A comprehensive chemical investigation was performed using a range of targeted and untargeted chemical analyses on selected wines and correlated with sensory data to identify compound attributes associated with typicality from the geographical indications (GI) selected. In addition, the project has overlayed key climatic data from the specific region/subregion and vineyard to help explain the chemical and sensory differences reported in the representative wines.
Summary
This project was instigated to investigate multiple questions pertaining to the ability to differentiate fine Australian Shiraz/Syrah wines from regions with high esteem for this grape variety. The project sought to:
- Demonstrate the suitability of a rapid sensory method for characterisation of wine
- Characterise the chemical differences and similarity of fine Shiraz wines from selected regions through targeted and untargeted metabolomic approaches of analysis
- Correlate the chemical and sensory composition of selected wines to the climatic indices associated with the vineyards from which the wines were sourced
The performance of the recently developed rapid sensory descriptive method Pivot© Profile (PP) was assessed with a set of 17 Shiraz/Syrah red wines using two groups; firstly 49 international sommeliers and then 11 Australian winemakers. Premium examples of Australian regional wines were first selected by the authors and other experts familiar with regional styles. The PP results were also compared to results from descriptive analysis (DA) performed by a trained panel. This investigation demonstrated for the first time that PP and DA provide similar insights into the sensory properties of products, and confirmed that PP with expert panelists allows a rapid understanding of the main sensory differences amongst samples.
The sommelier event was well-received and was able to the showcase the quality and diversity of a range of premium Australian Shiraz/Syrah wines. Utilising the sensory method PP and summarising the current investigation exposed an influential group of wine trade professionals to Australian wine research activities and provided an insight into their preferences.
Sets of wines (22 to 28 wines) from six prominent Australian Shiraz producing regions (Barossa Valley, McLaren Vale, Yarra Valley, Heathcote, Canberra District, and the Hunter Valley) were evaluated by groups of local winemakers using the PP method to obtain maps of their sensory characteristics. Three or four wines from each region were then selected using cluster analysis and twenty-two wines evaluated using sensory descriptive analysis. The regional PP assessments provided a sensory fingerprint of the variability of each of the regions studied and identified sensory characteristics that typified the largest groups of wines of each region. The descriptive analysis highlighted sensory characteristics that distinguished the wines from the different regions.
The comparison of PP with DA demonstrated that a rapid assessment of wine using the PP method is both robust and valuable. The PP method has already been used for other Wine Australia projects beyond the current investigation due to the effectiveness and efficacy demonstrated.
The same 22 wines were analysed by 70 chemical measures and 17 site- and season-specific climate indices were determined. From cluster analysis of compositional data wines were well grouped by region of origin. Distinctive chemical fingerprints exist for the regions studied, and the climatic profiles were strongly associated with key compounds influencing sensory differences. Multivariate analyses showed that wines with stalky/cooked vegetal sensory properties had higher cinnamate esters and dimethylsulfide, relating to later budbreak and harvest day. Wines with higher monoterpenes were associated with floral aroma. High radiation measures were linked to higher tannin, colour density, norisoprenoid compounds and phenylethyl acetate and stronger dark fruit/dried fruit and tannin/colour attributes. High rainfall indices were related to generally low intensity of most sensory attributes and most compositional measures.
The volatile compounds of the chosen wines were also analysed using an untargeted metabolomics approach for gas chromatography time-of-flight mass spectroscopy to provide an overview of the wine volatilome. Several esters identified were found to be important enabling discrimination of wines based upon geographic origin, including ethyl hexanoate, ethyl heptanoate, ethyl dodecanoate, ethyl 2-phenylacetate, ethyl 2-methylbutanoate and ethyl 3- methylbutanoate. These compounds were found to be generally present in higher concentrations in warmer regions, i.e. Hunter Valley, McLaren Vale and Barossa. Ethyl cinnamate and ethyl dihydrocinnamate both showed highest abundance in Yarra Valley wines and least abundance in Hunter Valley wines. The presence of these cinnamates may contribute to differences in sensory attributes based on GI climatic differences, but winemaking practices (e.g. use of whole bunch) was also considered as being important. Overall, the results from the untargeted chemical analyses suggest that the regional compositional differences in varietal wines may be influenced by all processes in the entire wine production chain. However, the chemical basis underlying the regional typicality of Australian Shiraz wines was highlighted and specific volatile compounds that may be associated with a region are identified.
The separation of regional wines using non-volatile compounds was not achieved to same degree as using volatile compounds. However, when the apparent number of compounds in the model was reduced to 26, reasonable regional separation was able to be achieved. A hierarchical cluster analysis confirmed that most wines from the same GIs grouped together. The identification of the non–volatile compounds was not able to be performed to any great extent, however dihydroxybenzoic acid-β-D-glucopiranoside was measured at higher concentration in Hunter Valley wines. An integration of the sensory characteristics, volatile markers and non-volatile chemical markers found through targeted analyses and climatic indices resulted in both regional similarities and distinctions amongst the Shiraz wines. The six regions could be divided into 4 distinct clusters, namely, Canberra District/Yarra Valley, Heathcote, Barossa/McLaren Vale and Hunter Valley. Hunter Valley wines were the most distinct in terms of sensory and compositional characters compared to other regions. A series of heat and water related climate indices modelled for the past 20 years showed distinctive patterns associated with each region, especially for Barossa, Hunter Valley, McLaren Vale and Yarra Valley. The correlations found between climate indices and the overall sensory and chemical composition of wines produced in a region have also been described.
The project was an excellent example of the collaboration between two of Australia’s leading wine research organisations, the AWRI and NWGIC. It also enabled Wes Pearson, AWRI Senior Scientist, to successfully complete his PhD candidature as a student of Charles Sturt University. The wonderful support of members of the Wine Australia marketing team, including Andreas Clark, Stuart Barclay, Aaron Ridgway and Mark Davidson is also acknowledged. Their support and enthusiasm enabled findings of this investigation to be presented in both the United States, Hong Kong as well as in Australia to a diverse audience.