How does grape composition affect wine chemistry and ultimately the sensory properties of the wine made from those grapes?
Not only isn’t there a clear answer to that question, it probably hasn’t been asked as often or as emphatically as might be expected.
We are aware of some specific and often quite pronounced links – pepper due to rotundone, for example, or floral characters from monoterpenes – but the body of general knowledge has many gaps. Dr Paul Boss from the CSIRO is not sure why, but he has a suspicion the power of yeast may have something to do with it.
‘If you look at all the volatile compounds in a grape and compare it with what’s in a wine there’s hardly any overlap’, he said. ‘The grape has a template for the yeast to work on to produce the wine, so although people know that the yeast plays an important role, the grape’s composition is the foundation that the wine is built up from.’
Image: Ian Routledge / Wine Australia
Boss and fellow investigators Jun Niimi, Susan Bastian and David Jeffery from the University of Adelaide have just completed a four-year, Wine Australia-funded project to examine the grape quality parameters that influence wine flavour and aroma.
‘We came at the project knowing we’ve done studies in the past where we’ve had trials with grapes coming from different regions or treatments, we ferment them in the same manner and they all taste quite different, so what is it about the composition of the grape which is changing the sensory attributes of those wines.
‘The whole concept was to try to find ways of measuring compounds in the fruit that will give some idea of wine style and the quality of the wine you can make from those grapes. We took the view that although the grading system is not perfect let’s measure everything we can in the grapes and see if we can relate that to the sensory outcomes in the wine.’
One of the Sensory panels assessing the project’s wines.
It was a comprehensive and multi-disciplinary project, involving analytical and synthetic chemists, biochemists, sensory scientists and international experts on multivariate statistics.
They created and analysed 12 grape metabolite data blocks for Cabernet Sauvignon and 9 for Chardonnay, with fruit gathered from regions throughout South Australia. Trained sensory and winemaker panels then looked at the resulting wines and scored them for sensory attributes or graded them in terms of quality.
The results were not definitive, but certainly encouraging.
‘We went in with a really open mind – happy to find and follow up any correlation’, Dr Boss said. ‘In a perfect world we would have come up with a list of five or six compounds you can measure to know whether it should be A grade or E grade, but unsurprisingly it’s not quite as simple as that.
‘But the correlation between some of these data blocks clearly shows we can get insights into wine style prediction by measuring compounds in the grapes.’
Individual results narrowed the field of enquiry both by confirming a number of specific correlations – such as using colour as a basis for assessing red grapes – and by highlighting classes of compounds that are of little use to measure. The research also revealed that links between berry tasting and wine sensory analysis are not entirely consistent from year to year, suggesting that it may not be best way to predict wine outcome or that a different approach to berry sensory assessment is needed.
The biggest conundrum thrown up was the different results from the two varieties. Cabernet Sauvignon revealed quite clear differences that could be linked back to the origin of the fruit, but not Chardonnay, supporting the concept that the latter is made more in the winery than the vineyard. The question is whether one or other of these is the norm.
Part of that answer – and a guide to the next steps – may be revealed by the current major Wine Australia project to understand and refine the expression of Australian Shiraz terroir.