Understanding and mitigating the impact of smoke exposure on grape and wine composition and wine sensory properties

Abstract

This project addressed important questions about the impact of smoke from the 2020 bushfires on wine. Vineyards with a range of smoke and fire damage were selected for an evaluation of the impact of early-season smoke. Grape composition was affected by the smoke, and wines had strong unpleasant smoky flavours that regular wine consumers strongly disliked. Concentrations of smoke marker compounds in grapes and wine related to unacceptably smoke-affected wine were determined for three major varieties. Strategies to reduce smoke taint were evaluated, and treatment of white juice with activated carbon gave the best results.

Summary

Smoke from the 2020 bushfires affected wine grapes across the ACT, NSW, SA and Victoria, with economic impacts estimated at $665 million. It is well understood that volatile phenols in smoke can be taken up by grapes and metabolised to form phenolic glycosides, which can accumulate in high concentrations during the growing season and contribute an unpleasant smoky aroma, flavour and aftertaste to wine. Smoke exposure can be detected in grapes and wine by measuring a suite of marker compounds (volatile phenols and phenolic glycosides) and comparing results with the concentrations found in non-smoke-exposed samples. However, the challenges of the 2020 vintage raised several questions that highlighted gaps in our knowledge about smoke taint. Firstly, does early-season smoke impact on wine quality in the same way as later in the season? What are the links between grape and wine composition and smoky flavours in wine, and what concentrations in grapes produce unacceptable wine? What are the most promising strategies to reduce smoke taint in wine?

The impact of early-season smoke was investigated using a comprehensive set of real smoke-exposed samples from the Adelaide Hills wine region. The Cudlee Creek fire in the Adelaide Hills occurred well before véraison, when winegrapes were small and green, approximately three months before harvest, at a phenological stage previously considered to be at low risk of smoke taint. Twenty-four vineyards with a range of smoke and fire damage were selected to evaluate the impact of early-season smoke. Grape composition was shown to be affected by the smoke. Volatile phenols and phenolic glycosides were elevated in the grapes at harvest, with concentrations matching anecdotal observations of smoke in the vineyards. Phenolic glycosides were also detected in the grapes sampled in January, four weeks after the fire, showing that early detection of smoke exposure is possible. Red wines from vineyards heavily affected by smoke had strong, unpleasant smoky flavours and were strongly disliked by regular wine consumers. White wines had less intense smoke flavour due to different wine production methods. 

This study clearly showed that smoke early in the season can have a devastating effect on wine quality, particularly for red wine styles. This information continues to be communicated to growers and winemakers so they can be alert to the risk of smoke characters from early-season smoke and make informed decisions about vineyard management and wine production. Ultimately, producers should avoid producing unacceptable smoky wine that could damage their own reputation and Australia’s reputation for quality wine. On a practical basis, the study showed that the time period during which wine grapes are at risk from smoke exposure is longer than previously thought. Ideally, controlled burns planned for fuel hazard reduction or for other reasons such as stubble management should not be undertaken near vineyards during the risk period. The wine sector should negotiate with those planning burns to ensure that bushfire risk is managed while minimising the impact of early-season smoke on wine grapes and avoiding preventable crop losses.

In addition to the early-season smoke event in the Adelaide Hills, grapes were exposed to smoke at various times during the 2019/20 vintage in many wine regions across Australia. Analysis of smoke marker compounds enabled non-smoke-exposed samples to be identified, which could be managed in the usual manner. However, for the samples with confirmed smoke exposure, it was not known what concentrations of smoke marker compounds in grapes would produce unacceptably smoke-affected wine. For the first time, these links were established with a rigorous set of 65 wines made from Chardonnay, Pinot Noir and Shiraz grapes with a range of smoke exposure. The focus was on chemical composition of grapes and wines and objective assessment of smoke flavour. Smoke flavour in wine was predicted well using smoke markers in grapes. Marker compounds key to smoke flavour were identified. The varieties differed in grape chemistry, wine chemistry and smoke flavour. Interestingly, some wines, but not all, made from smoke-exposed fruit had strong smoke flavour. The concentrations are summarised in a decision tool to guide assessment of smoke-exposed grapes, which is attached to this report in Appendix 8.

Harvesting grapes, purchasing grapes or bulk wine, and choosing processing options for grapes suspected of smoke exposure are critical decisions, with significant financial and legal implications for growers and winemakers alike. Smoke taint was the top topic of queries through the AWRI helpdesk in 2019, 2020 and 2021 and in total over the last five years, covered approximately 25% of all winemaking queries since 2017. In 2019/20 alone, the economic impact of fires was estimated at $664 million across NSW, VIC and SA. The decision tool will enable better decisions to be made about how to manage smoke-exposed grapes, alleviating uncertainty and stress. It will help protect brand reputation by avoiding the production of unacceptably smoky wine. It will minimize costly wine quality downgrades and unnecessary crop losses.

The establishment of unacceptable levels of smoke markers in Chardonnay, Pinot Noir and Shiraz grapes is a major development in the assessment of smoke-exposed grapes. These varieties represented 48% of Australia’s wine grapes in 2022. It is important to note that the critical concentrations were different in the three varieties, so the concentrations may not be applicable to other cultivars or to rosé, fortified or sparkling wine styles. Further research is required to establish concentrations that lead to unacceptably smoky wines in other varieties and wine styles. 

Strategies for reducing smoke taint in wine continue to be sought, and genuine solutions are elusive. Most treatments have negative impacts on wine flavour and colour, style and quality, and add costs and/or processes to manage smoke taint. This project focused on strategies to apply in the winery and evaluated yeast strains, glycolytic enzymes and activated carbons. The most successful strategy to date has been the treatment of white juices with activated carbon to remove phenolic glycosides and/or volatile phenols prior to fermentation. This practice was widely applied to white winemaking in 2020, reportedly with good outcomes. It was known that carbon products varied in their effectiveness, but the underlying reasons were unknown. The surface characteristics of activated carbon products were explored to allow insight into the reasons for their efficacy. Surface characteristics varied widely, and particular combinations of characteristics were required for optimal removal of volatile phenols or glycosides from a given matrix. There are few options for remediating red wine styles, and smoke taint remains a very significant risk for producers of premium full-bodied red wines. In contrast, some producers were more comfortable with a certain level of risk, and produced commercial grade white wine or non-wine products to recoup value from an otherwise lost crop. Innovative solutions are still required for red wine remediation.

Climate modelling predicts that fire events in southern Australia, where most Australian winegrapes are produced, will become more frequent and intense. Hazard reduction through prescribed burns is an important part of bushfire prevention, and the window of suitable weather and conditions for these activities is limited. It is therefore likely that prescribed burns will occur during the grapegrowing season, posing a risk to the crop. The knowledge from this project will support negotiations about prescribed burns and prepare the sector for bushfire events, increasing preparedness for climate change and reducing the risk from extreme environmental events. Increasing our knowledge about smoke taint represents an important tool to adapt to and mitigate against the effects of climate change.