Development of quality assessment Tools for chardonnay in relation to Grape, juice and wine composition

Abstract

Chardonnay is an important wine varietal and continues to be the main white wine variety grown in Australia, second only to Shiraz in total annual crush. The main gauges of white grape quality in the field, and the parameters used when making harvest decisions, are sugar and acid contents in the berry, despite the fact that one of the key components of the quality of a wine is its aroma. In the case of Chardonnay wines, aroma tends to be due to a complex mixture of esters, C13-norisoprenoids, monoterpenoids, malolactic fermentation characters and oak volatiles; so this variety is impacted by more inputs than other white wine varieties. There is a need for objective tools to determine quality, both in grape and wine, and development of such tools implies understanding the underlying factors that modulate Chardonnay wine aroma and flavour. This includes aspects such as terroir and winemaking techniques, but also recognising compositional variables that can be measured in the berries to guide production of desired wine styles. This thesis comprises a number of studies that examine different factors that affect Chardonnay wine quality based on grape composition. As a preliminary step to examine current trends in production that could affect final wine quality, a survey was distributed amongst Chardonnay wine producers in Australia. Over 150 respondents, or around 10% of all Chardonnay producers in Australia, participated in the survey. From the responses it was observed that production of Chardonnay wines was concentrated in the ultra- ($25-50) and super-premium ($15-24) categories rather than the lower priced segments. More importantly, the survey showed that although producers still relied mainly on total soluble solids (i.e. °Brix) and titrable acidity to determine grape maturity, berry tasting and grape flavour intensity were weighing as heavily in their decisions, especially when assigning a quality grade to the fruit. This emphasises the importance of the aroma-related compositional traits of the berries in the decisionmaking process, as well as that of understanding the chemical differences between berries from different quality grades. Other findings from the survey included trends in yeast selection, use of oak during fermentation and ageing according to price category, and changes in current Chardonnay wine styles. Expert ratings and medals are currently one of the main ways by which the quality of wines is judged. A study was therefore undertaken to investigate the volatile molecules responsible for the differences between the quality ratings of commercial Chardonnay wines of different quality levels to pinpoint compounds that could be used to discriminate between higher and lower quality Chardonnay wines. The wine selection comprised oaked and non-oaked wines, ranging in price point, sales volume and wine writer quality score, sourced from Australia’s main wine producing regions, which were assessed by a panel of eight industry experts. Correlation of sensory and chemical data indicated that nine volatile compounds were significantly and positively correlated to quality score and price, including oak and age-related volatiles (Z)- and (E)-oak lactones, furfural and diethyl succinate. Eleven volatiles were found to be negatively and significantly correlated to quality score, amongst them hexyl acetate, β-damascenone, 3-methyl butyl acetate, ethyl butanoate, ethyl hexanoate, hexanoic acid. Results also showed significant correlations between quality score and production method and vintage, where samples fermented in oak barrels and older than one vintage were preferred by the experts and thereby scored higher. In contrast, younger, more fruit-driven and simpler wines were all scored lower. Quality of any wine depends on the quality and constitution of the grapes used in its production. Origin (or terroir) can have a large influence on the composition of the grapes through the effects of weather, topography and soil composition. As the main focus of the thesis, grapes sourced from different geographical regions in South Australia were studied for their effects on the composition of the ensuing wines. Berries collected during the 2014-2016 harvests from the Barossa Valley, Eden Valley, Clare Valley, the Adelaide Hills, Langhorne Creek, McLaren Vale and the Riverland were analysed and vinified to produce research-scale wines. Results from 2014 showed that it was possible for a trained panel to discriminate between wines from three distinct regions in South Australia (Adelaide Hills, Eden Valley and the Riverland) through sensory descriptive analysis, when the main attributes associated with each of these wines were generated. Both positive and negative sensory attributes were related to volatile compounds, which were used as markers of quality in small-scale fermentation wines. Discrimination between the different juices was possible based solely on their chemical composition (using compounds such as vitispirane, C6-compounds, 5-methylfurfural, C6-C10 acids and guaiacol, along with Zn, pH, TA and °Brix). Warmer weather in the Clare and Barossa Valleys related to higher quantities of hydrolytically-released β-ionone, vitispirane, 5-methylfurfural, guaiacol and 2,6- dimethoxyphenol and certain fatty acids but lower quantities of 1-hexanol, (Z)-linalool oxide and linalool. Chemical data from the other two vintages was assessed to validate and further refine the understanding of the regional effects, allowing models to be proposed that link Chardonnay grape composition and quality. Industry requires rapid, low cost tools that can withstand the high throughput of vintage. To this effect, in addition to research based methodologies, work was started on the development of prediction models using partial least squares regression and mid-infrared (MIR) spectroscopy that could differentiate between grape juice from different origins and quality levels. Good separation was obtained in 2014 and 2016 between the tested regions (Adelaide Hills, Barossa Valley, Clare Valley, Eden Valley, Langhorne Creek and Riverland) using the fingerprint region (1500-800 cm-1). Classification according to grade based solely on the fingerprint region proved more challenging although models were especially successful at classifying A-grade grapes both years. Future improvement and further work is required on these methodologies before they can be used routinely by the industry, but this initial work offers guidance and the promise of realising a rapid tool for discriminating grapes of different origins or quality levels. The results of this thesis give insight into the different parameters affecting Chardonnay grape and wine quality, and contribute to the determination of objective indices that can be used by producers to grade their fruit, thereby potentially guiding their practices in the vineyard and winery. This knowledge can be used to further enhance understanding of the link between grape and wine composition and quality, and aid in the development of rapid assessment tools for the industry.