Assessment of grape smoke exposure

Abstract

The level of glycosidically bound phenolic compounds present in grapes can be used as an indication of the overall level of smoke exposure.  However, correlations between levels of grape markers at low to moderate levels and the sensory outcomes of resulting wines have hitherto been unknown.  This project establishes sensory outcomes of wine made from a range of smoke exposure levels at low to moderate smoke exposure for Shiraz and Pinot Noir dry table and Rosé wine styles.  Varying levels of extraction of grapes for Chardonnay wines were also investigated.  Excellent correlations between grape markers and sensory ratings for smoke aroma and flavour are presented. 

Summary

 Bushfire smoke contains volatile phenol compounds from the thermal decomposition of plant lignin. Grape berries exposed to smoke can absorb these compounds and store them as glycoconjugate forms that are subsequently released during fermentation and wine storage.  High levels of phenolic compounds in wines can result in potent aromas such as ‘smoky’, ‘dirty’, ‘burnt’, ‘medicinal’ and ‘ashy’.

The scale and extent of the catastrophic bushfires in vintage 2020 resulted in an extraordinary number of samples submitted at the same time to the few analytical services equipped to provide commercial analysis of grape smoke exposure marker compounds. Consequently, significant delays were encountered in the delivery of analytical results to grape growers in New South Wales. 

This project was undertaken to:

1. Establish the analytical capability to analyse grape smoke exposure markers at Charles Sturt University using methods similarly available through the Australian Wine Research Institute (AWRI).
2. Correlate the levels of grape smoke markers to markers in finished wines made from grapes with low to moderate levels of smoke marker compounds (i.e. in the grey zone).
3. Correlate the levels of grape smoke exposure markers to sensory ratings of wines made at varying levels of smoke exposure.
4. Identify the potential for 2D correlative spectroscopy as a tool to measure grape and wine smoke maker compounds.

The work reported in this project could not have been undertaken without the close collaboration and generous support from the Australian Wine Research Institute.  Identification of smoke-exposed grapes in vineyards prior to harvest was made known to the research team via Dr Mango Parker and the Commercial Services team at AWRI, along with the transfer of intellectual property for the analysis of grape glycoside compounds.  Additional support is acknowledged from the NSW growers who generously allowed harvest from their vineyards and the NSW Wine Industry Association for assistance in the coordination and communication of outcomes arising from this work.  The kind provision of 2Dshige software by Shigeaki Morita, Kwanse-Gakuin University is also gratefully acknowledged.

Successful transfer, modification and integration of methods for the analysis of six grape glycosides by LCMS using an expanded set of deuterated internal standards was undertaken at Charles Sturt University using the AWRI method. Inter-laboratory proficiency testing has confirmed the validity of the methods at Charles Sturt University.   Approximately 3 mg of each deuterated phenol glycoside standard has been secured, which is sufficient for approximately 3000 analyses.

In all wines produced in this project, a reduction in phenolic glycosides and a concurrent increase in volatile phenols from grape to wine were observed, reflecting the transformation from the former to the latter through the vinification process. Concentrations of glycosides and volatile phenols in wine followed the trends observed in grapes i.e. concentrations increased from low to high levels concomitant with the level of grape smoke exposure.

Good correlations were observed between grape and wine smoke marker compounds, indicating that as the levels of smoke marker compounds in grapes increase, the corresponding counterparts will increase in wine.  

Rosé style winemaking substantially reduced the concentration of volatile phenols, decreasing from 55% to 70% in Pinot Noir and 44% to 74% in Shiraz wines, compared to their dry wine counterparts.  However, this does not directly translate into wines with improved sensory outcomes. Rosé style wines for both cultivars were also perceived to have higher levels of smoke aroma and flavour, with diminished fruitfulness indicating that this wine style is less suited for smoke exposed grapes.  There was a greater array of more apparent fruit characters in the red table wines compared to the rosé wines. This appeared to be the reason why the sensory panel reported that Shiraz wines made from grapes with low to moderate smoke exposure were not significantly different to a control wine. The fruit attributes in the Pinot Noir wines were rated as having a lower intensity, and only one Pinot Noir wine, that with the lowest glycoside level, was rated as not being significantly different to the control for smoke aroma or taste. The higher fruitfulness in finished dry tables wines (in particular Shiraz) suggests that with careful winemaking, production of wines with commercial potential from grapes with modest levels of smoke exposure is possible. 

Reducing pressing yields from 500 to 400 L/tonne resulted in small but consistent reductions in total volatile phenol and phenol glycosides concentration in Chardonnay wines.  Overall, the sensory differences between Chardonnay wines with different extraction levels were inconsistent, likely reflecting varying fruit qualities evident from parcels of fruit with similar total phenol glycoside analysis. Encouragingly, the wines produced from grapes with low smoke exposure and the majority of the high exposure grapes were not rated as significantly higher in smoke aroma and flavour than a control wine. While the economic viability of decreased pressing yields would need to be assessed, the investigation suggests it might be possible to make commercially suitable Chardonnay wines from low smoke exposure.

Increased smoke exposure of grapes was clearly associated with increasing levels of smoke flavour and aroma when wines were compared to clean control wines.  Sensory descriptive analysis undertaken in the absence of control wines indicated practical winemaking outcomes could be achieved with Shiraz and Chardonnay at modest levels of smoke exposure, provided that sufficient fruit characters were also present.  This observation was not apparent for wines from grapes with extreme levels of smoke exposure, regardless of wine style and cultivar.

Targeting predefined bands of total phenol glycosides for grape harvest based on analytical values provided up to two weeks in advance proved to be difficult when final grape analysis was undertaken.  Consequently, the goal to select fruit within a range of total glycoside values from low to high could not be achieved for all cultivars.  Meaningful comparisons are still evident from the results, and it is of particular interest to compare wines from grapes with similar levels of total glycosides but from differing vineyards and growing conditions. This comparison indicated that grape potential fruitfulness is an important consideration in winemaking outcomes.  

These observations suggest that grapes exposed to low, and low-moderate levels of bushfire smoke have the potential to be vinified into acceptable wines, provided that judicious winemaking decisions are made that exemplify the fruitfulness and structure of the wines and limit overall skin contact, macerating and extraction.  A balance between fruitfulness and skin contact is thus required for winemaking from smoke-exposed grapes.  At high levels of smoke exposure, the resulting smoke-derived compounds may attribute a masking effect to the fruitfulness of wines, thereby making such wines unacceptable from a consumer perspective.  Wines with high potential for complex fruitful styles may therefore be more readily amenable to commercial production if they have only been exposed to low and low-moderate levels of bushfire smoke.

Finally, smoke taint markers targeted in grapes using contemporaneous analytical methods have good correlations with smoke aroma and flavour in Pinot Noir and Shiraz. Total grape phenolic glycoside appeared to be a good indicator for levels of smoke taint compounds as well as smoky aroma and flavour in finished wines.