|
A Technical Guide for Wine Producers
|
|
RECENT ARTICLES | WYNBOER HOME
Effect Of Yeast Strain And Lees Contact On Chenin Blanc Wine Quality
Johann Marais and Neil Jolly, ARC Infruitec-Nietvoorbij, Stellenbosch
Key words: Yeast strain, lees contact, wine quality, Chenin blanc
This study forms part of a more general investigation into Chenin blanc. Aspects such as berry size and ageing potential of Chenin blanc wine are discussed in other articles.
INTRODUCTION
Chenin blanc plantings and tonnage pressed amount to approximately 18% of the total wine grapes in South Africa. In view of this prominent role played by Chenin blanc in the South African wine industry and as a result of competition with other white cultivars and other wine producing countries in overseas markets, there has been renewed interest in Chenin blanc as a white wine cultivar. Recent local and overseas competitions have shown that this cultivar, when cultivated and treated correctly, is able to produce high quality wines. Because cultivars of a neutral type, such as Chenin blanc, are mainly dependent on the presence of fermentation flavourants, viticultural and oenological factors that promote the development of positive fermentation components play an important role. In this study the effects of yeast strain and lees contact on Chenin blanc wine quality were investigated. Lees contact may be beneficial to certain styles of wine to impart more complexity and body (Zoecklein et al., 1997; Fornairon-Bonnefond et al., 2002).
MATERIAL AND METHODS
The investigation was conducted over four seasons (2001 - 2004). During the 2001 season pilot trials were conducted on Nietvoorbij grapes, which included yeast strains NT 45, NT 116, NT 7, VIN 13, VIN 7 and N 96 (from Anchor Bio-Technologies), RV 1, D 47, Lalvin QA 23 and Lalvin GRE (from Lallemand South Africa). These yeast strains were selected as a result of recommendations by winemakers and yeast manufacturers. Based on the results obtained, the remainder of the investigations concentrated on four yeast strains, namely VIN 13, N 96, QA 23 and NT 116. This selection was based on the successful completion of fermentation, ester production, the formation of lees character and overall wine quality. During the 2002 to 2004 seasons grapes from the Villiera estate were used. The lees concentration with and without the addition of enzymes also varied during the investigation.
During the final season (2004) three treatments, in combination with yeast strains VIN 13, N 96, QA 23 and NT 116, were conducted as follows. Immediately after fermentation the control wines of each yeast strain were racked, stabilised and bottled according to standard Nietvoorbij small scale white wine vinification techniques. The second group of wines (standard lees treatment) were left on the full lees component in 20 litre stainless steel containers. The third group of wines (enzyme treatment) were also left on the full lees component in 20 litre stainless steel containers. To these wines were added 10 g/hl of a commercial enzyme ("Rapidase Filtration", DSM, France) in order to accelerate yeast cell autolysis.
In the last two instances the full lees component was removed after fermentation and aerated at 15oC for 48 hours by means of a shaking machine to prevent the formation of undesirable sulphury flavours (Lavigne-Cruège & Dubourdieu, 2001). After re-addition the containers of these two treatments were rolled weekly to achieve the blending of wine and lees. The wines were regularly monitored sensorially for the formation of lees character up to the time of bottling.
The following analyses were done during lees contact, i.e. total extract and total nitrogen (FAN). After five months on the lees the wines were cold stabilised, filtered, bottled and stored at 15oC.
The wines were analysed gas chromatographically 11 months after harvesting and sensorially evaluated by a panel of six experienced judges. The entire prodecure was conducted in triplicate.
RESULTS AND DISCUSSION
During the pilot trials in the 2001 season it was established that yeast strains RV 1 and D 47 were definitely conducive to the formation of sulphur, and these yeast strains could therefore be eliminated without further ado. It was further established that the formation of undesirable sulphury compounds could generally be prevented by aerating the lees directly after fermentation. With regard to body, yeast strain GRE fared exceptionally well, but could not be used because the wines did not ferment entirely dry. Yeast strains QA 23, N 96, VIN 13 and NT 116 produced the highest levels of acetate esters and corresponding fruitiness and were therefore automatic choices.
The pilot trials during the 2001 season and the trails during the two subsequent seasons formed the basis for the successful completion of the investigation during the 2004 season. The majority of the results from the previous seasons were confirmed by the latter investigation. For the purpose of this article, only the final results of the 2004 season will therefore be presented.
Directly after fermentation yeast strains VIN 13 and NT 116 showed the most fruitiness and highest overall wine quality, which corresponded with the highest acetate ester and lowest higher alcohol levels (Table 1). The trends of the sensorial data for individual yeast strains and lees treatments after lees contact (Table 2) differ slightly with regard to fruitiness and overall wine quality from the data directly after fermentation (Table 1), but correspond with those of the combined data (Tables 3 and 4). Although no statistical differences between yeast strains could be observed after five months’ lees contact, the trends clearly indicated that yeast strain QA 23 produced the most fruitiness, lees character and highest overall wine quality (Table 3). It seems like yeast strain N 96 produced wines with slightly more body than the other yeast strains. With regard to the comparison between the control and the two lees treatments, the enzyme treatment produced the most lees character, the most body and statistically the highest overall wine quality (Table 4). As could be expected the control wines showed the most fruitiness, seeing that no lees character was involved that could possibly have masked it. This may be further ascribed to the fact that the control wines were left at 0oC for the most part of the lees contact period, while the lees treatments were kept at 15oC throughout. The above trends were also reflected in the individual results (Table 2).
In all instances the levels of sulphury flavours were insignificantly low, indicating that the aeration of the lees directly after fermentation was successful (Tables 2, 3 and 4).
With regard to the total extract and total nitrogen (FAN) levels that were monitored on an ongoing basis during lees contact, it is obvious that there were only minor differences between yeast strains (Figures 1 and 2). The trends of the total extract and FAN levels of the two lees treatments corresponded, therefore only the combined lees/enzyme treatment is shown. The FAN levels increased gradually throughout, indicating that a certain degree of yeast cell autolysis did take place. The total extract reached a plateau after two months, but decreased with cold stabilisation, filtration and bottling at five months. This indicates that certain components, deriving from the lees, may be lost which will then go hand in hand with a decrease in mouthfeel or complexity. Despite this occurrence the treated wines still showed higher total extract levels than the control wines at the end of the trial period. This confirms the sensorial observation of more body and lees character in the treated wines (Table 4).
CONCLUSIONS AND RECOMMENDATIONS
In general the observations over the entire period of investigation may be summarised as follows: The development of sufficient lees character was a problem throughout. This may be ascribed to the fact that the lees contact took place in steel containers (instead of wooden barrels where subtle taking up of oxygen takes place), or that Chenin blanc (perhaps only in this investigation) simply does not have the ability to form sufficient lees character, or a combination of the two aspects. Nuances of bread, sparkling wine, citrus and buttery flavours were nevertheless observed. The wines of the two lees treatments were consistently more full-bodied than the control treatments. It seems as though lees character may be lost to a certain extent during fining, filtration and/or bottling. The wines also appear to show fairly little ageing character after 11 months, which might possibly signify that such wines could have a longer shelf life.
According to the results obtained under the conditions of this investigation, yeast strain QA 23 with enzyme treatment during lees contact appeared to offer the best parameters for the preparation of Chenin blanc wines where lees character and more body are required. It is very important to aerate the lees directly after fermentation to prevent the formation of undesirable sulphury compounds. Further good results will probably be obtained if lees contact takes place in wooden barrels. Ongoing investigations into this matter are recommended.
Literature references
Fornairon-Bonnefond, C., Camarasa, C., Moutounet, M. & Salmon, J.-M., 2002. New trends on yeast autolysis and wine ageing on lees: A bibliographic review. J. Int. Sci. Vigne Vin 36, 49-69.
Lavigne-Cruège, V. & Dubourdieu, D., 2001. The aptitude of wine lees for eliminating foul-smelling thiols. Aust. Grapegrower & Winemaker 450, 37-44.
Zoecklein, B. W., Marcy, J. E. & Jasinski, Y., 1997. Effect of fermentation, storage sur lee or post-fermentation thermal processing on White Riesling (Vitis vinifera L.) glycoconjugates. Am. J. Enol. Vitic. 48, 397-402.
Further enquiries may be addressed to: Dr Johann Marais, tel. (021) 809-3096, email MaraisJ@arc.agric.za or Dr Neil Jolly, tel. (021) 809-3060, email JollyN@arc.agric.za
|
|
Wynboer is incorporated in WineLand, magazine of the SA wine producers.
Visit our sister sites:
|
*C = Control
*C = Control, S = Standard lees treatment, E = Enzyme treatment.
* All values are the averages of three repetitions and the tasting data of six judges.
* All values are the averages of three repetitions and the tasting data of six judges.
| UP |
COPYRIGHT (C) 2000 WineLand |