Increasing thiol aroma molecules such as 3-SH, 3-SHA and 4-MSP can significantly impact the flavor-aroma profile of specific wines especially Rose’ and Sauvignon Blanc. Stabulation is a process of holding partially clarified juice at very low temperature for coordinated times/temperatures prior to fermentation initiation. A revolutionary new enzyme application has also been developed by Laffort which can also enhance the thiol development in select wine/yeast combination vinifications either alone or in combination with Stabulation.
- Does stabulation impact thiol aroma development in Sauvignon Blanc or Rose’ wines?
- Does a specific enzyme addition impact thiol aroma development in Sauvignon Blanc or Rose’ wines?
Strategy: In parallel tanks use two different process regimens varying the temperature and holding time of Sauvignon Blanc or Rose’ juice lots from the same vineyard. Keep fermentation process conditions/additions as similar as possible.
See Thiol Enzyme Protocol below for enzyme option.
The purpose of this trial is to compare a process method, Stabulation, to traditional processing OR to compare an enzyme treatment, Thiol Enzyme, either with or without Stabulation processing with thiol aroma-flavor development as the measurable by tasting panel and E-Nose.
Control treatment vs Experimental for this trial is as follows:
Traditional juice processing vs Stabulation
Traditional juice processing vs Thiol Enzyme application
Traditional juice processing vs Stabulation AND Thiol Enzyme application
Varietal Application – Rose’ or Sauvignon Blanc
Juice should be identically sourced and processed in parallel for trial and experimental treatments
Finished wines will be evaluated for sensory impact by ARC member tasting panels and by E-Nose by Bruce Zoecklein at Virginia Tech
The general procedural guidelines listed below may not be ideal for all participating wineries and some variation may occur. The winery variability was discussed in project harmonization meetings before harvest and all changes were determined to be acceptable. The slight processing variations builds robustness into the trial and in no way eliminates the individual trial iteration from proper comparison with other trials.
- In parallel tanks use two different process regimens varying the temperature and holding time of Rose’ or Sauvignon Blanc juice.
- Keep fermentation process conditions/additions as similar as possible.
- Source control and experimental fruit from the same vineyard
- See attached Thiol Enzyme Protocol for enzyme option.
- Rose’ (many red grape varietal options including blends)
- Sauvignon Blanc
Dosage(s) for Thiol Enzyme Trial Option
- Thiol Enzyme at 4-6 g/hL added right before yeast addition, 150-300 NTU in juice for fermentation
Press condition guidelines – (from Christophe Rossi, Provence style Rose’ production)
- use Lafase Press enzyme to maximize free run juice and limit maceration
- use Champagne cycle press program
- destem into press-add 25ppm SO2
- add 25ppm SO2 before pumping over in juice tank
- add 5 kg dry ice for each ton fruit in the press – top fill add layers, axial fill add in pump
- Do not settle any solids before allocating juice to stabulation tank (if press protocol is different than above some solids management may be needed)
- Run must through heat exchanger after destemming to facilitate cooling (less oxidation than juice)
Stabulation : holding juice at low temperature
- Juice Tank inerting, purge tank with CO2 before adding juice from press
- Agitation: Stir by Dry Ice addition or CO2 cane every 12 hours or other non-oxidizing mechanism
- Fining (optional: bentonite, Vegecoll, Polylact, Gelarom…)
After stabulation time allow juice to increase in temperature to 10°C (around 3-4 days), rack off stabulation lees, add Laffort Thiol Enzyme if doing the enzyme trial option, yeast juice after proper preparation of selected yeast, ferment as usual.
Sample – Time points – Documentation
- Juice YAN and juice chemistry pre-fermentation (pH, TA, SO2 other routine analyses)(Target 250-300 ppm YAN for fermentation)
- Document yeast selection, preparation and yeast nutritional plan, all addition dose rates
- Document all cellar manipulations (topping, QC/QA checks, movements, etc) and dates of all measurements and additions.
- Thiol Aroma samples pre-blending/bottling (minimum 125 mL with 100ppm SO2 added)(SARCO Lab)
- Sensory evaluation pre-bottling/pre-blending. E-Nose testing at Virginia Tech University (Set aside 2 x 750 bottles of control and treatment)(time TBD) Sensory evaluation of wines at blending/bottling – as late as possible with trial lots kept separate. Difference and Preference testing by ARC member panels
- Thiol Volatiles – SARCO Lab
- YAN and Juice Chemistry – to be done by participating winery
- Sensory Evaluation – by ARC member sensory panel
- E-Nose – by Bruce Zoecklein, Virginia Tech University
Thiol Enzyme Protocol
The purpose of this trial is to determine the impact of a new enzyme formulation referred to as the “thiol enzyme” on the amount of thiol precursors and thiol aroma compared with Stabulation or no post settling treatment. In both cases Sauvignon Blanc juice will have been clarified with a settling enzyme. After acoholic fermentation samples will be collected in 750 ml glass wine bottles to be collected by Laffort’s staff.
Stabulation Trial commitments/completed trials – 16 / 11
Thiol Enzyme Trial commitments/completed trials – 8 / 4
ARC member tasting panel opportunities – 4 (Edna Valley, Sonoma Co. Paso Robles, Santa Rosa)
Presentation of Stabulation Trial Wines – 3 Events, 6 wines presented (control and treatment for each)
NVGG Rootstock – Bogle, Ravenswood
WBM Innovation & Quality – Chamisal, Wild Horse
ARC Strategic Planning Meeting – Wild Horse, Van Duzer
There were both qualitative (sensory evaluation) and quantitative (ENose analysis, Thiol Aroma analysis-SARCO Laboratory) measurables for the stabulation and thiol enzyme trials. A sensory evaluation was performed at a series of regional tasting events. The sensory evaluations were structured and data analyzed by Larry Brooks, Instructor for Sensory Analysis at Fresno State University. All of these tests were presented to the groups as a triangle/preference test. This is a test in which three glasses are presented blind to the tasters. Two of the three are identical, and one is different. The tasters have two tasks. First to identify correctly which two are similar and which one different. Secondly to make a preference between them. This type of test is recommended when there is a small group of technical tasters as was the case here. The probability of guessing correctly with this test is low – only one in six.
The number of correct answers to achieve a statistical variance of less than 5% depends on the number of tasters. For ten tasters it is seven. For twenty it would be eleven. Results were held to the 5% standard, which some hold to be overly rigorous for sensory work. It remains the scientific standard, but it should be kept in mind that this is a very high bar for the sensory realm.
ARC Member Winemaker Tasting Panel Sensory Evaluation
Over the course of four months four different ARC member tasting panels were convened where stabulation and/or thiol enzyme trial wines were evaluated. A total of 16 wines were evaluated in 21 separate tastings. In those 21 tasting opportunities there were 12 examples of statistically significant differences detected. In a cumulative data analysis, of the 12 wines determined to be different the preference was 7 for the standard control process and 5 for the stabulation/thiol enzyme treatment. The 9 stabulation wines determined to be different showed 5 standard control wines preferred and 4 stabulation treatment preferred. The small data set of thiol enzyme trials had 3 wines determined to be different with 2 standard control preferred and 1 thiol enzyme treatment preferred.
Figure 1. Sensory evaluation data for Stabulation and Thiol Enzyme Trials.
Applied Research Cooperative 2016 Stabulation Trial Report – Electronic Nose
Dr. Bruce Zoecklein, ARC Coordinator
Introduction to Electronic Nose technology
Electronic nose (ENose) analysis can quickly evaluate and classify volatiles. We have used this technology to evaluate grape maturity and vineyard management practices on grape and wine volatiles (Devarajan et al. 2011, Gardner et al. 2015, Zoecklein et al. 2011). Although ENose analysis of wine may be difficult due to ethanol interferences (Ragazzo-Sanchez et al. 2006), studies by our group and others have shown its usefulness in classifying wines from grapes grown in different locations or growing conditions (Di Natale et al. 1996, Devarajan et al. 2011), aroma composition after bottle opening (Di Natale et al. 1996), use of aged barrels at various oak toasting levels (Chatonnet and Dubourdieu 1999), identification or discrimination of treatments or additions (Lozano et al. 2005) and identification of specific wine volatiles Gardner et. al 2009, Santos et al. 2004). Additionally, proper methodology can help minimize ethanol or water interferences and improve discrimination of specific volatile compounds (Santos et al. 2004, Gardner 2009, Devarajan et al. 2011).
Electronic nose analysis.
A conducting polymer 32-sensor electronic nose (ENose; Cyranose 320, Smiths Detection, London, England) was used to evaluate the wines used in the stabulation studies conducted by ARC (Applied Research Cooperative). Aliquots (10-mL) were pipetted into 40-mL clear glass vials sealed with Teflon/silicone 3-mm septa (MicroLiter Analytical Supplies).
Vials were placed in a 30°C water bath for 20 minutes, and then pierced with a 21-gauge syringe needle to allow the electronic nose needle into the vial headspace. A pre-determined method (Gardner 2009) allowed for a 30-second purge time and 20-second sampling time. An ethanol sample, specific to the ethanol concentration of each wine, was used to create a baseline for the electronic nose, minimizing ethanol concentration differences between treatment and control wines.
Presentation of ARC stabulation trial results
Analysis data are presented as a canonical plot that depicts the maximum amount of variation among samples based on ENose evaluation. Significant differences at the 95% level are indicated by non-intersecting ellipses, which represent the 95% confidence intervals and group means. The wines are found to be significantly different from one another when plots show non-intersecting ellipses of each analysis set. Results are based on volatile-sensor association. Each data point represents 32 sensor readings. This study used a baseline alcohol standard to minimize the impact of alcohol as previously described. The ENose used for this evaluation is composed of polymers in varying oxidative and reductive states, which allows for selectivity of specific groups of compounds. Each numeric value represents an individual sensor from the 32-sensor array detection system.
Figure 2. ENose Data Plot, Red is treatment, Green is control
Each of the 12 paired control and treated samples submitted for ENose analysis exhibited similar profiles to the plot shown in figure 2. The plot (Fig. 2) was generated by ENose for the control and treatment wine samples submitted by Trinchero Family Estates. This trial was a Thiol Enzyme trial on Sauvignon Blanc juice utilizing the Laffort commercial yeast Zymaflore X5. The ENose analysis clearly shows tight grouping of the data points within the specific group ellipses. The non-intersection between the treatment and control ellipses indicates that there are significant differences in the subpopulations of volatiles detected with the ENose instrument which supports the differentiation seen in the sensory evaluations by ARC member winemaker tasting panels.
Thiol Volatile Aroma Analysis – SARCO Laboratory, Bordeaux France
The varietal aromas of Sauvignon Blanc are the most well-known of the non-Muscat grape varieties. The main descriptors used to describe these aromas are known to be derived from specific molecules:
- 4MMP (4-mercaptopentan-2-one) was the first identified. It has a strong odour of box tree and broom.
- 3MHA (3-mercaptohexyl acetate) has box tree and passionfruit nuances.
- 3MH (3-mercaptohexan-1-ol) is responsible for grapefruit and passionfruit odors
Wines made from premium Alsatian grape varieties such as Gewürtzaminer, Muscat, Riesling, Pinot Gris, as well as wines from Colombard, Chenin Blanc, Sauvignon Blanc and Sémillon infected by noble rot, can present a sauvignon-like character. In all those wines, the volatile thiols can reach and exceed their perception thresholds. Also, in some wines, levels can be higher than those found in Sauvignon Blanc.
Extensive analyses carried out at SARCO laboratory confirmed the presence of 3MH and its acetate (at concentrations sometimes well above the perception threshold) in many rosé wines made from the most common red grape varieties: Merlot, Cabernet Sauvignon, Cabernet Franc, Grenache, Syrah, Cinsault, Mourvèdre, Pinot Noir, and Malbec, irrespective of their geographical origin.
Samples of both control and treated wines were properly preserved with SO2 and sent to the SARCO Laboratory in Bordeaux France for volatile thiol analysis. The samples were analyzed by GC/MS and the 3 main volatile thiol marker molecules were quantified.
Figure 3. Stabulation Trial Volatile Thiols Measurement
Wines from the Stabulation trials had a mixed response to the treatment which is not unexpected. Variables such as grape varietal, vineyard farming practices, temperature and duration of stabulation period, CO2 mixing method and timing, yeast strain selection rehydration and nutritional supplementation can all play a role in the final measurable outcome.
The data from the trials presented in figures 3 and 4 is the cumulative amount of the three main volatile thiols measured by GC/MS for control and treatment samples. The control for each winery was set as their standard practice for rose’ winemaking while the treatment consisted of either a Stabulation procedure or a Thiol Enzyme application with slight process variation as selected by the individual winery to allow for equipment and logistic differences among the participating ARC wineries.
As illustrated in figure 3, of the eight Stabulation trials five exhibited a significant increase in the measured volatile thiols while one exhibited a decrease, two trials had no significant difference. Interestingly the sensory preference for the wines did not always follow the higher levels of volatile thiol components.
Figure 4. Thiol Enzyme Trial Volatile Thiols Measurement
Wines from the Thiol Enzyme treatment had a more consistent response with four of five exhibiting a significant increase in the measured volatile thiol compounds. The absolute response in terms of total volatile thiols expressed was much greater in the Thiol Enzyme trials than in the Stabulation trials suggesting that the mechanisms of the two applications may be different or that the thiol potential in the different vineyard/varietal materials was significantly different. At four ARC winery locations a Stabulation procedure and Thiol Enzyme treatment were performed independently with three of the four exhibiting a much greater response to the Thiol Enzyme treatment than the Stabulation procedure (figure 5). Once again, the sensory data did not follow the increased levels of thiol aroma molecules with two of the three trials showing significantly increased thiols having the control preferred in the blind tastings.
Figure 5. Differential Response to Thiol Enzyme Treatment or Stabulation
Case Study A – Wild Horse Winery and Vineyards
ARC member Wild Horse Winery and Vineyards chose a Pinot Noir Rose’ as their stabulation trial wine. A 22 ton load of Pinot Noir clone 459 from Rancho Sisquoc in Santa Maria was received directly to press and the resulting volume split into two equal sized fermenters. No enzyme was used in the fruit processing.
The control fermentor was settled for 24 hours then racked and inoculated with Zymaflore X5 yeast under temperature control at 55°F. The stabulation fermentor was chilled to 34 °F for 12 days before racking and inoculation. The stabulation fermenter was mixed by dry ice addition every 12 hours throughout the stabulation period. The fermenter was warmed to 50 °F, inoculated with Zymaflore X5 yeast and temperature controlled at 55°F during fermentation. All additions were the same for both control and stabulation fermenters.
Figure 6. Wild Horse Winery Juice and Wine Chemistry
Figure 7. Brix Depletion and Temperature Plots for Control and Stabulation Fermentations
The juice and wine chemistry for the control and stabulation trials were not significantly different, and the fermentation curves were fairly typical for temperature controlled fermentations (Figure 7) so sensory evaluation should be uninfluenced. The resulting Pinot Noir wines were tasted blind by an ARC winemaker panel on three separate occasions. The first tasting resulted in no significant detection of difference while at the subsequent two tastings there was a statistically significant difference detected. In both of the tastings where a significant difference was detected, the stabulation wine was preferred both times. The ENose data for the Wild Horse wines was very similar to the profile in Figure 2 with both ellipses well separated indicating measurable differences. The GC/MS analysis at SARCO Laboratory in Bordeaux France showed a decrease in volatile thiols in the stabulation wine, which may have been due to compromised sample integrity based upon the other observances throughout the trials and the sensory differentiation clearly indicated with the fresh samples used for the tasting panels.
Case Study B – Bogle Vineyards
ARC member Bogle Vineyards chose a Sauvignon Blanc as their stabulation trial wine. A 171 ton lot of Sauvignon Blanc was harvested on August 9 and 10 and pressed yielding 31,676 gallons which was allocated into a 9126 gallon control settling tank and a 22,550 settling tank with 60ppm SO2 added to the respective volumes.
The control settling tank was cooled to 50°F for 2 days, then racked and inoculated with VIN 13 yeast under temperature control at 55°F. The stabulation tank was chilled to 32°F for 18 days before racking and inoculation. The stabulation tank was mixed by dry ice addition every 12 hours throughout the stabulation period. The fermenter was warmed to 50°F, inoculated with vin13 yeast and temperature controlled at 55°F during fermentation. All additions were the same for both control and stabulation fermenters.
Figure 8. Bogle Vineyards Juice and Wine Chemistry
Figure 9. Brix Depletion and Temperature Plots for Control and Stabulation Fermentations
The juice and wine chemistry for the control and stabulation wines were not significantly different, and the fermentation curves were fairly typical for temperature controlled fermentations (Figure 9) so sensory evaluation should be uninfluenced. The resulting Sauvignon Blanc wines were tasted blind by an ARC winemaker panel on one occasion with a statistically significant difference detected and the control wine preferred. The ENose data for the Wild Horse wines was very similar to the profile in Figure 2 with both ellipses well separated indicating measurable differences. The GC/MS analysis at SARCO Laboratory in Bordeaux France showed a significant increase in total aromatic thiol compounds with a 250 ppm increase in 3-MH and a doubling of 3-MHA from 17 to 34 ppm. The tasting results across the range of wines and wineries did not follow the concentration of aromatic thiols measured, so the preference for the control wine is not inconsistent with the other trial observations.
The purpose of the Stabulation and Thiol Enzyme trials was to investigate the impact of increasing aromatic thiols in thiol driven varietals and winestyles. The quantitative data indicates that the treatments were able to increase the amount of the target thiol molecules in most instances. Variables which can impact the increase of thiols include winegrape varietal, vineyard quality in terms of precursor thiol molecules, yeast selection, proper yeast rehydration, yeast nutrition for strong fermentation and process options such as Stabulation and Thiol Enzyme addition. ENose evaluation of the resulting wines shows that there are clear differences between control and treated wines.
The sensory evaluations did not always follow intuition that increases in the quality parameter of aromatic thiol molecules would result in a better perception of the resulting wines. The confounding factors such as yeast and fermentation treatments, while bringing variation to this study, do allow for optimization in concert with a Stabulation or Thiol Enzyme treatment. Biolaffort recommends the use of specific thiol revealing yeast strains with the Thiol Enzyme application. The small size of the tasting panels, usually less than 15, can also cause some unexplainable variation in the perception aspect of the evaluation.
The Stabulation and Thiol Enzyme treatments can be unequivocally viewed as beneficial tools in a winemaker’s arsenal in dealing with either fruit limitations or driving specific wine styles including differentiation of finished wines from similar incoming fruit especially when taking the aforementioned variability factors into account.