Optimising grape quality and value to improve vineyard profitability and sustainability
Abstract
The objectives of this investigation were to employ predictive models of wine style based upon grape berry sugar accumulation profiles, or berry skin colour, to create a harvest decision framework. Specific wine styles (Fresh, Intermediate and Mature) were predicted for Shiraz and Cabernet Sauvignon cultivars based upon a specified number of days following cessation of sugar accumulation in the grape berry. Compositional measures of grapes, wine and wine sensory scores could be used to consistently predict the specified wine styles from the plateau of sugar accumulation irrespective of vineyard location, altitude or vine yield. Industry partners in the Barossa Valley undertook field based trials incorporating irrigation scheduling with the sugar accumulation models for harvest decision making. For the white cultivars Chardonnay and Sauvignon Blanc, grape berry skin colour measures of hue and tint angle determined using a Dyostem™ instrument were used to optimise harvest decisions for specific wine styles.
Summary
The ability to determine an optimum harvest time for specific wine styles is one of the main challenges of vineyard and winemaking decisions. Optimising grape quality and value through alignment of harvest decisions with specific wine styles were the practical goals of this project. Traditional measures of grape maturity based upon objective measures of gross composition (Brix, pH, titratable acidity) and berry colour, along with subjective assessment of berry sensory profiles are among the commonly considered indicators for deciding when to harvest grapes. Defining potential wine styles using objective measures of grape berry maturation and having the potential to predict specific wine styles two to four weeks in advance of harvest will provide improved decision making capability for vineyard and winery managers, improve winery logistical operations and provide a means to ensure a consistent wine style profile within the limits of viticultural and regional specific terroir.
This project was undertaken using a model of experimental sites within commercially operated vineyards located in the Riverina, Orange, Tumbarumba, McLaren Vale and Barossa Valley. The main series of experiments focused upon the cultivars Shiraz and Cabernet Sauvignon and was undertaken across three consecutive years. Grape berry maturation was monitored regularly from veraison, by determining a sugar accumulation profile per berry (mg/berry), and harvest decisions for winemaking determined from the cessation of sugar accumulation. Three wine styles were targeted (Fresh Fruit, Intermediate Fruit and Mature Fruit) with the timing for grape harvests determined by the number of days following a plateau of sugar accumulation. Comprehensive metabolite profiling of grapes, wines combined with winemaking inputs and wine sensory assessments were used to model the flux of compounds of interest that may determine wine style. A consistent trend in wine sensory profiles was evident when grapes were harvested from vineyards according to the indicated maturation times, and when those vineyards were managed to accommodate excessive heat and avoid water stress, irrespective of vineyard location or climate. The tempo of change in the sensory profiles of wines was consistent between vintages and sites. There was no apparent link between sugar content (Baumé) and wine styles. The three wine styles could be made from fruit harvested from the same vineyard and Baumé with varying maturity times. These observations indicate that the proposed models for grape maturity are a robust and reliable means to align specific wine styles with a harvest decision based upon berry sugar accumulation.
Clear markers of berry maturation and wine styles in red wines were identified. Dimethyl sulfide is a marker of late maturity for both Shiraz and Cabernet Sauvignon and has an important contribution to the dark stewed fruit profile of wines. Cis-3-hexenol has been recognised as consistent and potent marker of the Fresh Fruit maturity stage in Shiraz. The contribution of higher alcohol acetate esters to Shiraz wine sensory profiles was consistently observed with mature grapes. γ-nonalactone has been identified as a reliable indicator of berry shrivelling in Shiraz. Berry shrivelling, a phenomenon commonly observed in Shiraz grapes has an important and significant impact on wine sensory profiles. Identification of reliable markers enables targeted and objective measures to assist in determining optimum harvest times.
Berry colour evolution for the white cultivars Chardonnay and Sauvignon Blanc was determined using a Dyostem™ instrument expressing berry skin colour in hue degree angle. Regular measurements and monitoring of grape berry skin colour evolution from green to yellow hues enabled correlation between the hue angle and commercial wine sensory characters to be explored. The potential to use berry skin colour as an additional measure to traditional grape parameters for harvest decisions in support of specific wine styles is worthy of further investigation. The Dyostem™ instrument provided important measures of berry colour and size but feedback from vineyard managers and staff indicated the process was time consuming and onerous. Further work in this area should focus on easier to deploy image capture platforms and real time data analytics.