• gas chromatography-mass spectrometry;
  • grapevines;
  • guaiacol;
  • smoke taint;
  • Vitis vinifera;
  • volatile phenols


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background and Aims:  Grapevine smoke exposure has been reported to produce smoke aromas in wine, resulting in ‘smoke taint’. This study describes the application of smoke to field-grown grapevines between veraison and harvest to investigate the effect of timing and duration of smoke exposure on wine composition and sensory attributes.

Methods and Results:  Smoke was applied to grapevines as either a single smoke exposure to different vines at veraison or at 3, 7, 10, 15, 18 or 21 days post-veraison or repeated smoke exposures to the same vines at veraison and then at 3, 7, 10, 15, 18 and 21 days post-veraison. Gas chromatography-mass spectrometry analysis of guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol showed elevated levels in all wines produced from fruit from smoked grapevines. Repeated smoke exposures had a cumulative effect on the concentration of these compounds. A trained sensory panel identified the aromas of ‘burnt rubber’, ‘smoked meat’, ‘leather’ and ‘disinfectant’ in all wines derived from smoke-exposed grapevines but not in control wines.

Conclusions:  Smoke application to field-grown grapevines between veraison and harvest can influence the accumulation of volatile phenols and intensity of smoke aromas in resultant wines. A peak period of vine sensitivity to smoke at 7 days post-veraison is identified. Repeated smoke exposures have a cumulative effect.

Significance of the Study:  This is the first study to report the deliberate and controlled smoke application to field-grown grapevines demonstrating the timing and duration of smoke exposure to significantly affect wine chemical and sensory characters.


best estimate threshold


free amino nitrogen


gas chromatography-mass spectrometry


principal component analysis


total soluble solids


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Postharvest smoke exposure of grapes has been shown to influence the chemical composition and sensory characteristics of wine with the potential to cause an apparent ‘smoke taint’ (Kennison et al. 2007). However, to date, smoke has not been deliberately applied to grapevines in a field situation to determine the impact on grape and wine composition under controlled conditions. Furthermore, the effect of timing and duration of grapevine smoke exposure on the development of smoke taint in wine has not been previously investigated. As such, scientific literature relating to the in-field exposure of grapevines to smoke in the development of smoke taint is scant despite the issue's high relevance to viticulture in Australia and overseas.

Smoke is a highly complex substance, comprising particulate matter, carbon monoxide, carbon dioxide, polycyclic aromatic hydrocarbons, ozone (O3), various oxides of nitrogen and sulfur as well as a multitude of volatile and semi-volatile organic compounds (McKenzie et al. 1994, Nolte et al. 2001, Radojevic 2003, Reisen and Brown 2006). The composition of smoke can vary greatly depending on both the fuel source and pyrolytic conditions, in particular combustion temperature, oxygen availability, and moisture content (Baltes et al. 1981, Maga 1988, Simoneit et al. 1993).

Smoke can impart desirable organoleptic properties to foods. These are largely attributed to the presence of smoke-derived volatile compounds including phenols, carbonyls, acids, esters, lactones, pyrazines, furan and pyran derivatives (Maga 1988, Wittkowski et al. 1990, McKenzie et al. 1994, Guillén et al. 1995, Guillén and Ibargoitia 1998, Fine et al. 2001). Of these volatiles, guaiacol and 4-methylguaiacol are considered to be key smoke components (Baltes et al. 1981, Wittkowski et al. 1992). They are derived from the thermal degradation of wood lignin during combustion and exhibit ‘smoky’, ‘musty’, ‘caramel’, ‘burning’, ‘sweet’, ‘phenolic’, ‘sharp’, and ‘smoked sausage’ aromas and flavours (Baltes et al. 1981, Boidron et al. 1988, Wittkowski et al. 1992, Rocha et al. 2004).

In wine, guaiacol and 4-methylguaiacol typically originate from oak barrel fermentation and/or maturation (Boidron et al. 1988, Maga 1989, Swan 2004) at concentrations of up to 100 and 20 µg/L for guaiacol and 4-methylguaiacol, respectively (Pollnitz et al. 2004). However, a range of volatile phenols, including guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol, have recently been identified in juice, unwooded wine, acid and enzyme hydrolysates prepared from smoke-affected grapes (Kennison et al. 2007, 2008). Because these compounds were absent from the corresponding control samples (i.e., unsmoked grapevines), their origin was attributed directly to smoke exposure.

The aroma descriptors, aroma detection thresholds and wine concentrations reported for guaiacol, 4-methylguaiacol, 4-ethylguaiacol, and 4-ethylphenol are shown in Table 1. Because guaiacol exhibits the lowest aroma detection threshold (Boidron et al. 1988) and was the most abundant volatile phenol detected in smoke-tainted wines (Kennison et al. 2007, 2008), it is considered to be of greatest importance. Boidron et al. (1988) reported aroma thresholds for guaiacol in various media – 5.5 µg/L in water, 20 µg/L in model wine, 95 µg/L in white wine and 75 µg/L in red wine – whereas Simpson et al. (1986) reported a lower detection threshold, just 20 µg/L, for guaiacol in a dry white table wine. However, the detection thresholds for guaiacol and 4-methylguaiacol may in fact be even lower than these earlier data. Indeed, Eisele and Semon (2005) suggest that guaiacol is by far more potent. They determined the best estimate threshold for guaiacol to be 0.48 µg/L in water and 0.91 µg/L in apple juice, with even lower taste detection thresholds reported at 0.17 µg/L and 0.24 µg/L in water and apple juice, respectively. In a previous study involving the postharvest application of smoke to grape bunches, a perceptible ‘smoke taint’ was still evident after considerable blending to achieve sub-threshold concentrations of guaiacol and 4-methylguaiacol (Kennison et al. 2007). This suggests that guaiacol and 4-methylguaiacol might not be solely responsible for smoke taint in wine (Kennison et al. 2007).

Table 1.  Aroma descriptors, aroma detection thresholds and wine concentrations reported for guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol.
CompoundAroma descriptorsAroma detection threshold (µg/L)Wine concentration (µg/L)
WaterModel wineWhite wineRed wine
GuaiacolSmoky, phenolic§, chemical§0.48§§ 5.52095 20750–100
4-methylguaiacolToasted, ash103065650–20††
4-ethylguaiacolSmoky, spicy, toasted, bread§254770150 110‡‡2–437
4-ethylphenolHorsy, stable, phenolic§13044011001200 605‡‡2–2200

The effects of grapevine smoke exposure on the composition and sensory properties of wine are currently not well understood. This study was therefore undertaken to address this knowledge gap, and investigates the effect of both timing and duration of grapevine smoke exposure on wine quality.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Trial establishment and smoke application

The trial was sited in the locality of Capel in the Geographe region of Western Australia. The site was selected based on an infrequent history of smoke exposure and location away from forested areas. No incidences of externally derived smoke were observed at this site throughout the duration of the experimental period.

Purpose-built greenhouse-type tents measuring 6 m long × 2.5 m high × 2 m wide were constructed from galvanised steel framing to enclose grapevines for the application of smoke. The tents were covered with a greenhouse-grade Solarweave plastic (Gale Pacific, Australia) designed to enable plant photosynthesis and productivity. Smoke was generated by the combustion of dry barley straw in a (50 L) lidded metal drum for 30 min. A remotely controlled variable speed air pump was used to force air through an inlet pipe into the lidded drum, which subsequently forced smoke through an outlet tubing into the tent. Particulate matter (PM10, i.e., ≤10 µm in diameter) within the tent was monitored using a DustTrack laser photometer (TSI Model 8520; TSI Inc., St. Paul, Minnesota, USA) to maintain a maximum PM10 level of 200 µg/m3 for the duration of each smoke treatment, a level considered comparative to a high-pollution incident (Reisen and Brown 2006).


Two independent smoke experiments were conducted using Vitis vinifera cv. Merlot grown in a commercial vineyard. These were a single smoke exposure (for 30 min) of field-grown grapevines that was applied at either veraison or 3, 7, 10, 15, 18, 21 or 24 days post-veraison. Alternatively, field-grown grapevines received eight consecutive smoke exposures (for 30 min each) applied to the same vines at the beginning of veraison then at 3, 7, 10, 15, 18, 21 and 24 days post-veraison. For each smoke experiment, a control (i.e., unsmoked) treatment was also established, where grapevines were enclosed in tents (as earlier) but without the addition of smoke. Each treatment was replicated three times.


At harvest, the three replicates per treatment (approximately 15 kg each) were harvested on the same day for fruit analysis and wine production. Samples of smoked and control grape juice were analysed for total soluble solids (TSS) by refractometry (Iland et al. 2004) and for free amino nitrogen (FAN) by methods described by Dukes and Butzke (1998). Each fruit replicate was harvested at an average TSS of 21.6 ± 1.8 °Brix, crushed, destemmed inoculated with Saccharomyces cerevisiae EC1118 yeast (200 mg/L) (Lallemand Inc., Montreal, Canada) and fermented in 15-L fermentation vessels. Fermenting musts were plunged twice daily and the wine was pressed from the skins when the TSS approached 0 °Baumé. All wines were stored in 4.6 L enclosed glass fermenters at 15°C until the residual sugar was below 2 g/L. After the wines were racked from gross lees, they were inoculated with Leuconostoc oenos (10 mg/L) (Vinaflora Oenos, Chr. Hansen, Denmark) for malolactic fermentation. On the completion of malolactic fermentation, as determined by quantitative malic acid analysis, wines were racked from lees, free SO2 was adjusted to 30 ppm and the wines were cold stabilised (28 d at 2°C). Wines were then filtered (5 µm) and bottled. The alcohol content of final wines was measured by ebulliometry (Iland et al. 2004).

Quantitative determination of guaiacol, 4-methylguaiacol, 4-ethylguaiacol, 4-ethylphenol, furfural, 5-methylfurfural, eugenol and vanillin

Guaiacol, 4-methylguaiacol, 4-ethylguaiacol, 4-ethylphenol, furfural, 5-methylfurfural, eugenol and vanillin were quantified by gas chromatography-mass spectrometry (GC-MS) using methods reported previously (Spillman et al. 1997, Pollnitz 2000, Pollnitz et al. 2000, 2004, Kennison et al. 2008).

Sensory analysis

Sensory analysis of experimental wines (smoked and control) was conducted by a panel of eight trained judges comprising four males and four females aged between 21 and 30 years. Panellists were selected on the basis of interest and availability, having experienced at least 100 h of tertiary wine sensory education and being regular wine consumers, non-smokers, of good health, and able to detect smoke aroma of red and white smoked wines at predetermined thresholds ascertained by Kennison et al. (2007). Wines were assessed for aroma only (not tasted on the palate) so as to reduce any potential negative health impacts associated with the tasting of smoke-tainted wine and in accordance with ethical requirements for conducting sensory experiments (Meilgaard et al. 2007).

Panellists underwent eight quantitative descriptive analysis training sessions (two per week) prior to formal evaluation (Meilgaard et al. 2007). Descriptive aroma terms, based on wines used in the study, were generated by panellists with the panel consensus on six descriptive terms. Utilising experimental wines as references, panellists were trained to measure the smoke aroma presence and intensity on an unstructured 100-point line scale.

Formal evaluation of two wine replicates from both the single and repeated smoke experiments (i.e., 22 wines in total) was conducted over four sessions, each held at the same time on different days. Wine samples (20 mL) were presented to panellists at room temperature in three digit coded ISO standard tasting glasses in a randomised order. All glasses were covered with glass covers to avoid contamination of the testing area and other samples. To avoid sensory fatigue, panellists were required to leave the testing area to an external environment (for 10 min) after evaluating each sample.

Statistical methods

All data were analysed using SPSS version 14.0 for Windows (SPSS Inc., Chicago, Illinois, USA). Analysis of variance (ANOVA) was used to analyse chemical data at the 5% level of significance (P < 0.05). Wine sensory data were analysed by ANOVA and principal component analysis (PCA).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Effect of smoke exposure on grapes and wine

Smoke was applied to field-grown grapevines as either a single smoke exposure applied at either veraison or at 3, 7, 10, 15, 18, 21 or 24 days post-veraison; or as eight smoke exposures applied to the same vines at veraison and then at 3, 7, 10, 15, 18, 21 and 24 days post-veraison. Fruit from unsmoked (control) vines obtained a higher average fruit TSS level (22.3 °Brix) than fruit from vines subjected to smoke application (Table 2). The TSS was lowest in grapes from vines that had received repeated smoke exposures (19.3 °Brix). These vines also produced the lowest fruit yields (11 kg/vine) as compared with the mean fresh fruit weight of 15.3 kg/vine for all other treatments (Table 2). Repeated smoke exposure to vines resulted in the development of necrotic lesions on leaves, effects that were not seen on control grapevines or grapevines exposed to single smoke applications (not shown).

Table 2.  Yield, total soluble solids (TSS) and free amino nitrogen (FAN) of grapes and alcohol content of wine derived from smoked and control (unsmoked) grapevines.
Yield (kg/vine)TSS (°Brix)FAN (mg/L)Alcohol (% v/v)
  1. †Smoked grapevines were subjected to either a single smoke application or eight repeated smoke applications between veraison and harvest (i.e., at 0, 3, 7, 10, 15, 18, 21 and 24 days post-veraison). Means followed by the same letter within columns are not significantly different at P ≤ 0.05, n = 3.

Control 17.0a22.33a87.2f12.80a
Single smoke exposure at:0 day post veraison12.1b20.93b96.0ef12.47ab
3 days post veraison16.1a19.73cd112.8bcd11.23c
7 days post veraison15.9a21.07b117.3b12.13b
10 days post veraison15.7a19.40d114.0bc10.93cd
15 days post veraison15.9a21.03b102.8cde12.07b
18 days post veraison14.8a21.07b100.1def12.00b
21 days post veraison16.1a20.67bc108.6bcde11.93b
24 days post veraison15.2a21.23b101.7cde12.53ab
Repeated smoke exposure 11.0b19.33d134.4a10.57d
LSD (5%) 2.361.01813.30.6478

FAN in grapes at harvest was significantly higher in fruit from vines subjected to repeated smoke exposures (134.4 mg/L, P < 0.05) compared with all other treatments, with the unsmoked treatments lowest in FAN (87.2 mg/L) (Table 2). Fruit from vines subject to repeated smoke exposures also had higher total SO2 (18 mg/L) and pH (3.8) values in grape juice at harvest compared with all other single smoke and control (unsmoked) treatments.

The fermentation rate of must was faster for grapes from vines exposed to repeated smoke exposures. In comparison with wines from the unsmoked (control) treatment that completed fermentation after 12 days, wines from the repeated smoke exposure treatment took only 8 days to complete fermentation (Figure 1). Fermentation rates of musts from all other smoke treatments were also faster (10–12 days). Ethanol content was up to 17% lower in wines made from fruit from grapevines subject to repeated smoke exposures (10.6% v/v) than in the other wines, and the highest ethanol content was in the wines from non-smoke-exposed control vines (12.8% v/v) (Table 2). Similarly, all wines vinified from grapes from vines exposed to a single smoke exposure contained intermediate ethanol concentrations between 10.9 and 12.5% v/v.


Figure 1. Fermentation curves for fruit harvested from grapevines exposed to repeated smoke applications (○) from veraison to harvest (i.e., at 0, 3, 7, 10, 15, 18, 21 and 24 days post-veraison) and control (unsmoked) grapevines (●). Mean values from three replicates; standard errors are obscured by symbols and so are not shown but are <0.5 in all cases.

Download figure to PowerPoint

Quantitative determination of smoke-derived volatiles in grapes and wine

The volatile phenols guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol were selected as analytes of interest based on their reported contribution to the aroma and flavour of smoked food products (Baltes et al. 1981) and provenance in smoke-tainted wines (Kennison et al. 2007, 2008). Vanillin, eugenol, furfural and 5-methylfurfural were included in the quantitative GC-MS analysis but because of being detected at levels that have a negligible effect on aroma properties (i.e., ≤5 µg/L for vanillin, eugenol, and 5-methylfurfural and ≤40 µg/L for furfural) are not discussed further.

GC-MS analysis detected large and significant differences in volatile phenol composition between control wines and wines derived from smoked grapevines (Table 3 and Figure 2). The highest levels of smoke-derived volatile phenols occurred in wines made from vines repeatedly exposed to smoke, i.e., 388 µg/L of guaiacol, 93 µg/L of 4-methylguaiacol, 16 µg/L of 4-ethylguaiacol and 58 µg/L of 4-ethylphenol. In contrast, control wine contained between 4 µg/L of guaiacol and non-detectable (<1 µg/L) levels of the other phenols. Wine derived from repeatedly smoked grapevines contained at least four-fold higher volatile phenol concentrations than any of the wines derived from grapevines that received a single smoke treatment.

Table 3.  Mean concentrations and standard errors of guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol in wines made from fruit harvested from smoked and control (unsmoked) grapevines.
TreatmentConcentration (µg/L)
  • Smoked grapevines were subjected to eight repeated smoke exposures applied between veraison and harvest (i.e., at 0, 3, 7, 10, 15, 18, 21 and 24 days post-veraison).

  • For each analyte, means followed by the same letter are not significantly different at P ≤ 0.05, n = 3.

  • n.d., not detected; tr., trace (i.e., positive identification but <1 µg/L); SE, standard error.

Figure 2. Guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol concentrations in wine made from grapes harvested from grapevines exposed to single smoke applications between veraison and harvest (i.e., at 0, 3, 7, 10, 15, 18, 21 or 24 days post-veraison); C, control. Data are means (n = 3). Error bars show two standard errors of the mean.

Download figure to PowerPoint

Importantly, the timing of grapevine smoke exposure was found to influence the concentration of smoke-derived volatile phenols in wine. For the experiments involving single smoke applications, the highest concentrations of guaiacol, 4-methylguaiacol, 4-ethyphenol and 4-ethylguaiacol corresponded to wine derived from grapevines exposed to smoke 7 days post-veraison (Figure 2). Guaiacol concentration increased from 7 µg/L for smoke exposure at the initial onset of veraison, peaked at 92 µg/L for smoke exposure at 7 days post-veraison, then decreased to between 34 and 61 µg/L for each subsequent smoke exposure until harvest. Similar compositional trends were observed for 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol.

Sensory analysis of experimental wines

Quantitative descriptive analysis of wines made from fruit of grapevines exposed to smoke identified aromas of ‘burnt rubber’, ‘smoked meat’, ‘leather’, and ‘disinfectant and hospital’ as being associated with smoke taint (Boidron et al. 1988, López et al. 1999). Aromas of ‘red berry fruits’ and ‘confection’ were identified as wine characters common in distinguishing control wines from unsmoked vines. ANOVA showed both the wines and the panellists to be sources of wine sensory variation for all aroma attributes (P < 0.001) (Table 4). However, there was no significant variation among the replicates (i.e., wine by replication and panellist by replication), indicating consistency in panellist rating between sessions and wines. Other sources of variation resulted from the wine by panellist interactions for the aroma descriptors of burnt rubber (P < 0.05), red berry fruits (P < 0.001) and confection (P < 0.001), indicating that individual panellists have different thresholds or levels of sensitivity to the aromas of these compounds.

Table 4.  Analysis of variance for wine sensory attribute ratings of ‘burnt rubber’, ‘smoked meat’, ‘leather’, ‘disinfectant’/‘hospital’, ‘red berry fruits’ and ‘confection’ for wine (W), panellist (P), replicate (R), wine by panellist (W × P), panellist by replicate (P × R) and wine by replicate (W × R).
Aroma descriptorWine (W)Panellist (P)Replicate (R)W × PP × RW × R
  1. F ratios are shown as sources of variation. Significance indicated by *P < 0.05, **P < 0.01 and ***P < 0.001.

Burnt rubber14.34***3.71***0.131.69*0.440.65
Smoked meat21.56***3.13***0.020.970.650.52
Red berry fruits9.65***6.19***1.372.18***0.350.41

Wines vinified from grapes of vines receiving repeated smoke exposures attained higher scores for off-aromas (‘burnt rubber’, ‘smoked meat’, ‘leather’, and ‘disinfectant and hospital’) compared with all other wines (Figure 3). Compared with wines made from fruit of vines exposed to eight smoke applications, wines made from fruit produced from unsmoked (control) vines exhibited significantly higher scores for ‘confection’ and ‘red berry fruits’ aromas (P < 0.001). Wines from single smoke exposure experiments revealed the full range of aroma characters from ‘red berry fruits’ and ‘confection’ through to ‘smoked meat’, ‘burnt rubber’, ‘leather’, and ‘disinfectant and hospital’ as displayed in the PCA biplot (Figure 4). The PCA of mean aroma results from single-smoked wines showed that principal component 1 (PC1) accounted for 93% of the overall variation and principal component 2 (PC2) accounted for 4% of the variation (Figure 4). PC1 is largely characterised by the contrast of positive loadings on smoke-like aromas (‘leather’, ‘burnt rubber’, ‘smoked meat’, and ‘disinfectant and hospital’ aromas) and negative loadings on fruit and wine aromas (‘red berry fruit’ and ‘confection’ aromas) (Table 5). PC2 is further defined with a positive loading for the ‘disinfectant and hospital’ aroma and with negative loadings for smoke and fruit aroma descriptors. The smoke-like aroma descriptors of ‘leather’, ‘burnt rubber’ and ‘smoked meat’ were highly correlated with each other (r = 0.74 to 0.79). Likewise, there was a high correlation between Merlot wine character aroma descriptors of ‘red berry fruits’ and ‘confection’ (r = 0.87). The aroma descriptor of ‘disinfectant and hospital’ was negatively correlated with wine character aromas (r = −0.55 to −0.59) and weakly correlated with smoke-like aromas (r = 0.48 to 0.52).


Figure 3. Polar coordinate (cobweb) graph of mean aroma intensity ratings of ‘burnt rubber’, ‘smoked meat’, ‘leather’, ‘disinfectant and hospital’, ‘red berry fruits’ and ‘confection’ for wines made from grapes from grapevines exposed to eight smoke applications (—) (applied at veraison and then at 3, 7, 10, 15, 18, 21 and 24 days post-veraison) and control (unsmoked) treatments (----) (least significant difference (LSD)), P < 0.001 are indicated in parenthesis for each term). Aroma descriptor values for unsmoked (control) wines range from 0 (‘smoked meat’) to 0.3 (‘burnt rubber’).

Download figure to PowerPoint


Figure 4. Principal component analysis (PCA) biplot of mean wine sensory scores from experiments that applied a single smoke exposure to field-grown grapevines between veraison and harvest (i.e., at 0, 3, 7, 10, 15, 18, 21 or 24 days post-veraison) and wine derived from control (unsmoked) vines (C). Aroma descriptors are indicated by arrows labelled DH (‘disinfectant and hospital’), L (‘leather’), BR (‘burnt rubber’), SM (‘smoked meat’), CF (‘confection’) and RBF (‘red berry fruits’).

Download figure to PowerPoint

Table 5.  Factor loadings on principal component 1 (PC1) and principal component 2 (PC2) for aroma descriptors of ‘burnt rubber’, ‘smoked meat’, ‘leather’, ‘disinfectant’/‘hospital’, ‘red berry fruits’ and ‘confection’ for wines derived from single smoke exposure applied to grapevines at veraison or at 3, 7, 10, 15, 18, 21 or 24 days post-veraison.
Aroma descriptorPC1PC2
Burnt rubber0.41−0.38
Smoked meat0.41−0.48
Red berry fruits−0.41−0.42

The sensory properties of each experimental wine varied depending on the timing of smoke application. Single smoke exposure to grapevines at either 7 or 10 days post-veraison led to more intense ‘leather’, ‘smoked meat’, ‘burnt rubber’, and ‘disinfectant and hospital’ aromas in resultant wines at higher levels than other timings of smoke exposure (Figure 4). Single smoke exposure to grapevines at veraison and at 18, 21, and 24 days post-veraison subsequently produced wines with high aromas of ‘red berry fruits’ and ‘confection’ and low smoke aroma characteristics.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Previous research has shown that the postharvest smoke exposure of grapes affects the chemical composition and aroma of wine, leading to the development of perceivable smoke taint aromas (Kennison et al. 2007). The current study demonstrates that field exposure of grapevines to smoke can lead to the development of smoke taint in wine, the timing of smoke exposure to grapevines can influence the chemical and sensory properties of resultant wine, and repeated smoke exposure has a cumulative effect on the concentration of smoke-derived volatile compounds in resultant wines.

The concentration of smoke taint indicator compounds (guaiacol, 4-methylguaiacol, 4-ethyphenol and 4-ethylguaiacol) measured in wine varied depending on the timing and number of smoke exposures that the vines received. Levels of these compounds in wines from grapevines subjected to a single smoke exposure at the onset of veraison were low, with guaiacol levels comparable to those found in control treatments (6.7 µg/L). However, smoke exposure at 7 days post-veraison resulted in higher levels of guaiacol (92 µg/L). Later smoke exposures resulted in significant levels of taint, but the concentrations of the indicator compounds were always less (63–93%) than wine corresponding to smoke exposure at 7 days post-veraison. The reasons for variation in sensitivity to smoke exposure during the post-veraison period are currently unclear. During veraison, changes occur in assimilate partitioning of sugar uptake and metabolism (Conde et al. 2007) and in phloem unloading from symplastic to apoplastic pathways (Zhang et al. 2006). The chemical and structural characteristics of grape cell walls also change during this period (Nunan et al. 1998, Mullins et al. 2000). The peak uptake of volatile smoke components by fruit following smoke exposure at 7 days post-veraison may be related to these changes in berry physiology. An alternative hypothesis is that the peak period of uptake identified in this experiment may be independent of ontogeny and could also relate to changes in sensitivity to uptake of the compounds at the leaf level associated with short-term environmental effects on vine physiology such as vine water status. These hypotheses are the subject of an ongoing study.

Repeated smoke exposures to field-grown grapevines led to accumulation of smoke compounds to high levels. Irrespective of when smoke exposures were applied to vines, from the period of veraison to harvest, the effects on the levels of guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol in wines were additive. Indeed, the sum of smoke taint-related compounds detected in wines generated from single smoke exposure treatments closely approximates the compound levels detected in wines generated from repeated smoke exposures (Figure 5). These results imply that repeated or prolonged vineyard smoke exposure, which can occur from frequent fire events, over the post-veraison period will potentially have a cumulative negative effect on resultant wine quality and value.


Figure 5. Sum of guaiacol, 4-methylguaiacol (4-MeG), 4-ethylguaiacol (4-EG) and 4-ethylphenol (4-EP) concentration in eight wines made from Merlot grapes from vines that each received a single field-based smoke exposure (applied at veraison or at 3, 7, 10, 15, 18, 21 or 24 days post-veraison as indicated by bands) versus compound levels detected in wine from Merlot grapes from vines that received repeated smoke exposures (applied at veraison then again at 3, 7, 10, 15, 18, 21 and 24 days post-veraison).

Download figure to PowerPoint

Quantitative descriptive wine aroma analysis demonstrated an increase in smoke-related aromas described as ‘burnt rubber’, ‘smoked meat’, ‘leather’, and ‘disinfectant and hospital’ in wines from the repeated smoke exposure experiment. These aromas clearly dominated the wine's sensory profile, overpowering any ‘confection’ and ‘red berry fruits’ aromas (Figure 3). Smoked wine from the repeated smoke exposure experiment contained guaiacol and 4-methylguaiacol at levels well in excess of the highest published aroma detection thresholds for red wine (of 75 and 65 µg/L, respectively), although 4-ethylguaiacol and 4-ethylphenol were present at sub-threshold concentrations (Boidron et al. 1988). In contrast, the volatile phenols were either not detected or detected at trace levels only in control wines. Therefore, as in previous studies, the source of volatile phenols can be attributed directly to smoke (Kennison et al. 2007, 2008). Furthermore, the accentuation of smoke taint in wine derived from the repeatedly smoked grapevines is correlated with the increased levels of guaiacol and 4-methylguaiacol, with little or no contribution from 4-ethylguaiacol or 4-ethylphenol.

Smoke-like aromas were also present to various degrees in all wines vinified from the single smoke exposure experiment regardless of smoke application timing and resultant compound level in wine. The levels of guaiacol, 4-methylguaiacol, 4-ethylguaiacol and 4-ethylphenol were detected by the panellists in all these wines below reported aroma thresholds for red wine (Boidron et al. 1988, Chatonnet et al. 1992) except for smoke application at 7 days post-veraison that produced wine with guaiacol above the reported aforementioned aroma detection threshold in red wine of 75 µg/L. Panellists detected elevated smoke aromas in the 7 days post-veraison wines, although smoke aromas were also detected in all other wines made from fruit of vines exposed to single smoke applications – even though their compound levels were below published aroma thresholds. It should be noted that there is some conjecture regarding the detection threshold for guaiacol; certainly there is disagreement between published thresholds, i.e., 75 µg/L reported by Boidron et al. (1988) and 20 µg/L reported by Simpson et al. (1986). A recent workshop demonstrated that all 60 delegates presented with 70 unidentified wines including a control and a wine spiked with guaiacol (20 µg/L) were able to discern the wines, giving descriptors for the latter such as ‘smoky’ and ‘burnt bacon’ (Mark Sefton, pers. comm., 2007). Given the intricate and complex nature of smoke, indicator compounds play an important role in assessing the extent and impact of smoke taint on wine quality. In the present study, the volatile phenols, together with sensory analysis, have been proven as effective smoke taint markers. However, it is acknowledged that with time, additional volatile compounds will likely be identified as components of smoke, which are also responsible for the discernable aroma attributes of smoke-tainted wine.

Grapevine smoke exposure leads to increased levels of FAN in grapes, an effect most evident following repeated smoke applications. Interestingly, grapes harvested from repeatedly smoked grapevines also fermented the most rapidly, in agreement with previous studies (Kennison et al. 2007). While the increase in ferment rate may be associated with the increased FAN in must (Henschke and Jiranek 1993, Bell and Henschke 2005), the basis for this increase is unclear. Some direct contributions from nitrogenous smoke compounds is possible as research has demonstrated the uptake and assimilation, by nitrite reductase, of nitrogenous compounds (NO and NO2) by plants (Hosker and Lindberg 1982, Nussbaum et al. 1993, Stulen et al. 1998, Takahashi et al. 2001); however, it is likely that any such contribution would be very minor based on simple mass balance. Increased FAN may also be linked to the injury response of grapes following high levels of smoke exposure, for example, a biochemical response to necrotic lesion that developed on laminae following repeated smoke treatment (Heath 1980).

Field-based smoke exposure to grapevines showed an adverse effect on grape ripening (e.g., sugar accumulation) irrespective of the timing and duration of smoke application. In a recent study, the stomatal conductance, CO2 assimilation rate and intercellular CO2 levels of Chrysanthemoides monilifera were reduced for 5 h following smoke exposure for 1 min, with 24 h required to achieve physiological recovery to control levels (Gilbert and Ripley 2002). Additionally, the presence of SO2 and O3 in smoke has been shown to induce stomatal closure in grapevines (Rosen et al. 1978). It is therefore conceivable that the photosynthetic capacity of grapevines decreases following smoke exposure, which in turn inhibits grape maturation and ripening. Furthermore, these physiological effects would be further exacerbated by the loss of photosynthetically active leaf area because of the formation of necrotic lesions on laminae (Heath 1980), as occurred in the current study with grapevines subjected to repeated smoke treatments.

In summary, the deliberate application of smoke to field-grown grapevines between veraison and harvest affected yield, grape composition (sugar accumulation and FAN), wine composition, wine sensory properties, and most importantly, wine quality. The volatile phenol levels and intensity of smoke taint in wine was influenced by both the timing and the duration of grapevineexposure to smoke. For single smoke treatments, the highest levels of volatile phenols were observed in wines corresponding to smoke exposure 7 days post-veraison. For repeated smoke treatments, a cumulative effect on smoke-derived volatile phenol concentrations was observed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors wish to acknowledge the invaluable technical assistance of Bob Frayne, Glynn Ward, Sue Wills and Eric Wootton from the Department of Agriculture and Food, Western Australia and Capel Vale vineyard. This research was supported by Australia's grape growers and winemakers through their investment body the Grape and Wine Research and Development Corporation, with matching funds from the Australian Federal Government (Project RD 05/02-3).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Baltes, W., Wittkowski, R., Söchtig, I., Block, H. and Tóth, L. (1981) Ingredients of smoke and smoke flavor preparations. In: The quality of food and beverages. Eds. G.Charalambous and G.Inglett (Academic Press: New York) pp. 119.
  • Bell, S-J. and Henschke, P.A. (2005) Implications of nitrogen nutrition for grapes, fermentation and wine. Australian Journal of Grape and Wine Research 11, 242295.
  • Boidron, J.N., Chatonnet, P. and Pons, M. (1988) Influence du bois sur certaines substances odorantes des vins. Connaissance de la Vigne et du Vin 22, 275294.
  • Chatonnet, P., Dubourdieu, D., Boidron, J.N. and Pons, M. (1992) The origin of ethylphenols in wines. Journal of the Science of Food and Agriculture 60, 165178.
  • Conde, C., Silva, P., Fontes, N., Dias, A.C.P., Tavares, R.M., Sousa, M.J., Agasse, A., Delrot, S. and Gerós, H. (2007) Biochemical changes throughout grape berry development and fruit and wine quality. Food 1, 122.
  • Dukes, B.C. and Butzke, C.E. (1998) Rapid determination of primary amino acids in grape juice using an o-phthaldialdehyde/N-acetyl-L-cysteine spectrophotometric assay. American Journal of Enology and Viticulture 49, 125134.
  • Eisele, T.A. and Semon, M.J. (2005) Best estimate aroma and taste detection threshold for guaiacol in water and apple juice. Journal of Food Science 70, 267269.
  • Fine, P.M., Cass, G.R. and Simoneit, B.R.T. (2001) Chemical characterization of fine particle emissions from fireplace combustion of woods grown in the northeastern United States. Environmental Science and Technology 35, 26652675.
  • Gilbert, M.E. and Ripley, B.S. (2002) The effect of smoke on the photosynthetic gas exchange of Chrysanthemoides monilifera. South African Journal of Botany 68, 525531.
  • Guillén, M.D. and Ibargoitia, M.L. (1998) New components with potential antioxidant and organoleptic properties, detected for the first time in liquid smoke flavoring preparations. Journal of Agricultural and Food Chemistry 46, 12761285.
  • Guillén, M.D., Manzanos, M.J. and Zabala, L. (1995) Study of a commercial liquid smoke flavoring by means of gas chromatography/mass spectrometry and Fourier transform infrared spectroscopy. Journal of Agricultural and Food Chemistry 43, 463468.
  • Heath, R.L. (1980) Initial events in injury to plants by air pollutants. Annual Reviews 31, 395431.
  • Henschke, P.A. and Jiranek, V. (1993) Yeasts: Metabolism of nitrogen compounds. In: Wine microbiology and biotechnology. Ed. G.H.Fleet (Harwood Academic Publishers: Sydney) pp. 77164.
  • Hosker, R.P. Jr and Lindberg, S.E. (1982) Review: atmospheric deposition and plant assimilation of gases and particles. Atmospheric Environment 16, 889910.
  • Iland, P., Bruer, N., Edwards, G., Weeks, S. and Wilkes, E. (2004) Chemical analysis of grapes and wine: techniques and concepts (Patrick Iland Wine Promotions Pty. Ltd.: Campbelltown, South Australia) pp. 3043.
  • Kennison, K.R., Gibberd, M.R., Pollnitz, A.P. and Wilkinson, K.L. (2008) Smoke-derived taint in wine: the release of smoke-derived volatile phenols during fermentation of Merlot juice following grapevine exposure to smoke. Journal of Agricultural and Food Chemistry 56, 73797383.
  • Kennison, K.R., Wilkinson, K.L., Williams, H.G., Smith, J.H. and Gibberd, M.R. (2007) Smoke-derived taint in wine: effect of postharvest smoke exposure of grapes on the chemical composition and sensory characteristics of wine. Journal of Agricultural and Food Chemistry 55, 1089710901.
  • López, R., Ferreira, V., Hernández, P. and Cacho, J.F. (1999) Identification of impact odorants of young red wines made with Merlot, Cabernet Sauvignon and Grenache grape varieties: a comparative study. Journal of the Science of Food and Agriculture 79, 14611467.
  • Maga, J.A. (1988) Flavour chemistry of wood smoke. In: Smoke in food processing. Eds. J.A.Maga (CRC Press: Boca Raton, FL) pp. 4988.
  • Maga, J.A. (1989) The contribution of wood to the flavor of alcoholic beverages. Food Reviews International 5, 3999.
  • McKenzie, L.M., Hao, W.M., Richards, G.N. and Ward, D.E. (1994) Quantification of major components emitted from smouldering combustion of wood. Atmospheric Environment 28, 32853292.
  • Meilgaard, M.C., Civille, G.V. and Carr, B.T. (2007) Sensory evaluation techniques, 4th edn (CRC Press: Boca Raton, FL).
  • Mullins, M.G., Bouquet, A. and Williams, L.E. (2000) Developmental physiology: lowering and fruiting. In: Biology of the grapevine. Eds. M.G.Mullins, A.Bouquet and L.E.Williams (Cambridge University Press: UK) pp. 112146.
  • Nolte, C.G., Schauer, J.J., Cass, G.R. and Simoneit, B.R.T. (2001) Highly polar organic compounds present in wood smoke and in the ambient atmosphere. Environmental Science and Technology 35, 19121919.
  • Nunan, K.J., Sims, I.M., Bacic, A., Robinson, S.P. and Fincher, G.B. (1998) Changes in cell wall composition during ripening of grape berries. Plant Physiology 118, 783792.
  • Nussbaum, S., Von Ballmoos, P., Gfeller, H., Schlunegger, U.P., Fehrer, J., Rhodes, D. and Brunold, C. (1993) Incorporation of atmospheric 15NO2-nitrogen into free amino acids by Norway spruce Picea abies (L.) Karst. Oecologia 94, 14321939.
  • Pollnitz, A.P. (2000) The analysis of volatile wine components derived from oak products during winemaking and storage. PhD Thesis, The University of Adelaide, Australia.
  • Pollnitz, A.P., Pardon, K.H. and Sefton, M.A. (2000) Quantitative analysis of 4-ethylphenol and 4-ethylguaiacol in red wine. Journal of Chromatography A 874, 101109.
  • Pollnitz, A.P., Pardon, K.H., Sykes, M. and Sefton, M.A. (2004) The effects of sample preparation and gas chromatograph injection techniques on the accuracy of measuring guaiacol, 4-methylguaiacol and other volatile oak compounds in oak extracts by stable isotope dilution analyses. Journal of Agricultural and Food Chemistry 52, 32443252.
  • Radojevic, M. (2003) Chemistry of forest fires and regional haze with emphasis on Southeast Asia. Pure and Applied Geophysics 160, 157131.
  • Reisen, F. and Brown, S.K. (2006) Implications for community health from exposure to bushfire air toxics. Environmental Chemistry 3, 235243.
  • Rocha, S.M., Rodrigues, F., Coutinho, P., Delgadillo, I. and Coimbra, M.A. (2004) Volatile composition of Baga red wine: assessment of the identification of the would-be impact odourants. Analytica Chimica Acta 513, 257262.
  • Rosen, P.M., Musselman, R.C. and Kender, W.J. (1978) Relationship of stomatal resistance to sulfur dioxide and ozone injury in grapevines. Scientia Horticulturae 8, 137142.
  • Simoneit, B.R.T., Rogge, W.F., Mazurek, M.A., Standley, L.J., Hildemann, L.M. and Cass, G.R. (1993) Lignin pyrolysis products, lignans, and resin acids as specific tracers of plant classes in emissions from biomass combustion. Environmental Science and Technology 27, 25332541.
  • Simpson, R.F., Amon, J.M. and Daw, A.J. (1986) Off-flavour in wine caused by guaiacol. Food Technology in Australia 38, 3133.
  • Spillman, P.J., Pollnitz, A.P., Liacopoulos, D., Skouroumounis, G.K. and Sefton, M.A. (1997) Accumulation of vanillin during barrel-aging of white, red, and model wines. Journal of Agricultural and Food Chemistry 45, 25842589.
  • Stulen, I., Perez-Soba, M., De Kok, L.J. and Van der Eerden, L. (1998) Impact of gaseous nitrogen deposition on plant functioning. New Phytologist 139, 6170.
  • Swan, J. (2004) Barrel profiling: helping winemakers attain quality and consistency. Practical Winery and Vineyard. 26(4), 1622.
  • Takahashi, M., Sasaki, Y., Ida, S. and Morikawa, H. (2001) Nitrite reductase gene enrichment improves assimilation of NO2 in Arabidopsis. Plant Physiology 126, 731741.
  • Wittkowski, R., Baltes, W. and Jennings, W.G. (1990) Analysis of liquid smoke and smoked meat volatiles by headspace gas chromatography. Food Chemistry 37, 135144.
  • Wittkowski, R., Ruther, J., Drinda, H. and Rafiei-Taghanaki, F. (1992) Formation of smoke flavor compounds by thermal lignin degradation. In: Flavor precursors: thermal and enzymatic conversions. Eds. R.Teranishi, G.R.Takeoka and M.Güntert (American Chemical Society: Washington, DC) pp. 232243.
  • Zhang, X.Y., Wang, X.L., Wang, X.F., Xia, G.H., Pan, Q.H., Fan, R.C., Wu, F.Q., Yu, X.C. and Zhang, D.P. (2006) A shift of phloem unloading from symplasmic to apoplasmic pathway is involved in developmental onset of ripening in grape berry. Plant Physiology 142, 220232.