Research Note Treatment of Barrel Wood Infected with Acetic Acid Bacteria

Research Note Treatment of Barrel Wood Infected with Acetic Acid Bacteria KARL L. WILKER ~* and MURLI R. DHARMADHIKARF Four barrel sanitizing treatments were compared for their effectiveness on wood infected with acetic acid bacteria. These treatments included hot water and solutions of chlorine, sulfur dioxide, and potassium carbonate. Pieces of stave wood ("mini-staves") in sterile flasks of wine plugged with cotton were used instead of barrels of wine. An initial experiment looked at all four treatments and a control (water rinse) using ministaves that had two sides sealed with a wet surface liner (the sides that would normally be the stave ends). The mini-staves were contaminated with a strain of Acetobacter aceti. Only the hot water treatment (85 C to 88 C water for 20 minutes) was successful in eliminating the bacteria. A second experiment studied the hot water and control treatments using mini-staves without sealed sides. These were contaminated with either a strain of A. aceti or A. pasteurianus. The hot water treatment was successful in eliminating both strains of bacteria. KEY WORDS: oak, barrel sanitizing, acetic acid bacteria Acetic acid bacteria (Acetobacter, Gluconobacter) can negatively affect wine quality because of their ability to produce acetic acid from ethanol. Acetic acid is the main volatile acid of wine and is considered to be objectionable at levels above 1.2 to 1.4 g/l (12,17). Acetic acid bacteria can also metabolize substances in wine other than ethanol and produce compounds which alter wine quality (4,12,13). A. aceti and A. pasteurianus are the species most often associated with the spoilage of wine (11,15). Gluconobacter oxydans is sometimes isolated from wine during the alcoholic fermentation but is more commonly found on spoiled grapes (15). Acetic acid is formed in low levels during the alcoholic fermentation by yeast and during the malolactic fermentation by lactic acid bacteria (1,6,8,9,20). The production of excessive quantities of acetic acid can occur through the action of spoilage yeast (21). Wines can also acquire acetic acid during storage in new barrels as a result of the chemical hydrolysis of wood hemicellulose (7). Acetic acid bacteria have not been considered a major winemaking concern because of their requirement of oxygen for growth (15). The exclusion of oxygen from wine and the presence of free sulfur dioxide are usually considered to be an adequate means of avoiding acetic acid bacteria spoilage. Recent research suggests that this may not be a entirely successful way of controlling these bacteria, especially when aging wine in wooden barrels (15). Several studies have found strains of A. aceti and A. pasteurianus capable of surviving under the semi-anaerobic conditions of wine stored in barrels (11,15). It also appears that the amount of free sulfur dioxide typically used for red wine storage in 1Associate Research Professor and 2Research Professor, Department of Fruit Science, Southwest Missouri State University, Research Campus, Mountain Grove, MO 65711. *Corresponding author. This research was conducted at the Research Campus, Southwest Missouri State University, Mountain Grove, MO 65711. Manuscript submitted for publication 30 September 1996. Copyright 1997 by the American Society for Enology and Viticulture. All rights reserved. 516 barrels is not enough to protect a wine from the metabolism of acetic acid bacteria (15). Controlling acetic acid bacteria is of particular significance when aging wine in used barrels, as they are often contaminated with these bacteria. The storage of used barrels without the occurrence of this problem is difficult because of the porous nature of wood and the fact that a 225-L barrel will absorb 5 to 6 L of wine (7). A common question concerning oak cooperage is what to do with these infected barrels. Many experts consider it impractical to sanitize barrels due to the presence of bacteria in the pores of the wood (7,19). However, barrel rehabilitation procedures can be found in the popular literature. Treatments routinely mentioned involve filling and soaking a barrel with either a chlorine or soda ash solution or hot water (5,16,19). Treating the inside of a barrel with steam or sulfur dioxide are also suggested treatments (5,7). Research results concerning the effectiveness of these treatments are limited. Most studies evaluating barrel preparation and cleaning treatments have focused on differences in the extraction of oak compounds by wine (5,14). A study was undertaken to evaluate four sanitizing treatments for eliminating acetic acid bacteria (A. aceti, A. pasteurianus) from barrel wood. These included soaking barrel wood with hot water or solutions of either chlorine, sulfur dioxide, or potassium carbonate. Pieces of oak staves ("mini-staves") contained in Erlenmeyer flasks of autoclaved wine, plugged with cotton, were used instead of barrels of wine. This was done to reduce the level of outside contamination as much as possible. Materials and Methods Bacterial strains: The A. aceti strain used in this study was isolated from a research wine produced in the Fruit Processing Laboratory of the Department of Fruit Science, at the Research Campus of Southwest Missouri State University, Mountain Grove, Missouri.

BARREL WOOD-- 517 This wine was made from Cayuga White grapes without the addition of sulfur dioxide. The A. pasteurianus strain (ATC2879) used was obtained from the American Type Culture Collection. The two strains were identified on a species level using methods described by Drysdale and Fleet (12). The cultures were maintained on mannitol agar slants, stored at room temperature (18 C to 27 C), and transferred monthly. Barrel wood: The mini-staves were obtained by sawing up barrel staves width-wise from an unused l l.4-l white oak barrel. The mini-staves cut from a single stave were kept separate from those cut from other staves. To a portion of the mini-staves, the two sides that had been cut were sealed with a wet surface liner (made for use with barrels). These sides contained "pores" (cross sections of the conductive vessels of the wood) and are normally the end portions of the stave that do not come in direct contact with wine. These tubes are plugged by tyloses which form during the conversion of sapwood to hardwood (18). Incomplete development of tyloses can result in wine leaking out the ends of staves through these pores. This condition is not totally unheard of in barrels used for wine storage. Lack of tyloses in these conductive vessels could also allow for the greater penetration of acetic acid bacteria or a sanitizing solution into a mini-stave. It was unclear if the over all effect would be to make the unsealed mini-staves more difficult or easier to sanitize then an actual barrel. With this in mind it was decided to use staves both with and without these sides sealed. The average dimensions of the mini-staves were 4.1 x 1.3 x 1.8 cm.. The mini-staves with two sealed sides were used in the first experiment. This experiment looked at the effectiveness of all four sanitizing treatments on A. aceti bacteria. The mini-staves without sealed sides were used in a second experiment, which looked at only the effectiveness of the hot water treatment on A. aceti and A. pasteurianus bacteria. Wines: The wines used in this experiment were prepared from a 1989 Seyval blanc vinified in the Fruit Processing Laboratory previously mentioned. Potassium bicarbonate was added to the wine to raise the ph to approximately 3.8. This was to facilitate the growth of acetic acid bacteria. The wine was sterilized by placing 170-mL lots into 500-mL Erlenmeyer flasks plugged with cotton, and then autoclaving at 121 C for 15 minutes. The autoclaved wines had low levels of ethanol (4.6% to 7.55%), high ph values (3.68 to 3.88 ), high oxygen concentrations (6.7 to 7.7 mg/l), low levels of free sulfur dioxide (1 to 3 mg/l) and a volume of approximately 140 to 150 ml. The wines were kept at room temperature (18 C to 27 C) throughout their use in this experiment. All of the wines were considered capable of supporting the growth of the acetic acid bacteria being studied. Inoculation of mini-staves: In the first experiment, five mini-staves (with sealed sides) where aseptically placed into each of three autoclaved flasks of wine. Each flask contained mini-staves cut from the same barrel stave, but from a different barrel stave than the other two flasks. These flasks were inoculated with a loop of wine containing an active culture of A. aceti. This culture was obtained by previously inoculating an autoclaved Erlenmeyer flask of wine with a loop of culture from a mannitol slant of A. aceti. One of the flasks containing mini-staves developed a mold problem 14 days after it was inoculated. These staves were autoclaved and then inoculated again in a different flask of wine. After the flasks showed obvious signs of acetic spoilage, they were plated to determine the number of acetic acid bacteria present. The following day, each of the staves were subjected to one of the different sanitizing treatments (control, chlorine, sulfur dioxide, potassium carbonate, hot water). Each sanitizing treatment was replicated on one stave per flask. This resulted in each treatment being replicated three times. After receiving one of the treatments, each stave was aseptically placed into a separate flask of wine. The ratio of wood surface (unsealed) to wine in each flask approximated that of a 200-L barrel (18). The wines in these flasks (containing treated staves) were plated onto bromocresol green-ethanol plates at weekly intervals to determine if acetic acid bacteria were present. In the second experiment, three autoclaved ministaves (without sealed sides) were aseptically placed into each of six autoclaved Erlenmeyer flasks of wine. Three barrel staves were used to make the "mini" staves for the six flasks. The mini-staves obtained from each individual barrel stave were put into two different flasks. Three of the flasks of mini-staves (each flask representing a different stave) were inoculated with A. aceti and three were inoculated with A. pasteurianus. The inoculations were made with loops of culture from mannitol slants. Once the flasks showed obvious signs of acetic spoilage, they were plated to determine the number of acetic acid bacteria present. Two days later, one "mini" stave per flask was subjected to a control treatment and one was subjected to a hot water treatment. After receiving one of the treatments, each stave was placed into a separate flask of wine. The ratio of wood surface to wine in each flask approximated that of a 20 liter barrel (18). The wines in the flasks containing treated staves were plated as previously described. Treatments: To one flask at a time, the infected mini-staves were aseptically removed, placed into a sterile beaker, and rinsed four times with 200 ml of sterile distilled water. The mini-staves were then aseptically removed from the beaker to receive one of the treatments. The control treatment consisted of aseptically placing a infected mini-stave in a Erlenmeyer flask of wine. The hot water treatment consisted of placing an infected mini-stave into either a sterile 400-mL beaker (first experiment) or a sterile 500-mL Erlenmeyer flask (second experiment). Hot distilled water (85 C to 88 C) was placed into either the beaker (250 ml of water) or Erlenmeyer flask (300 ml of water). An additional beaker or Erlenmeyer flask of hot water without a infected "mini" stave was monitored with a thermometer to determine changes in temperature. The infected

518 ~ WILKER and DHARMADHIKARI "mini" stave was left in the hot water for 20 minutes. The temperature of the water after 20 minutes varied between 58 C and 60 C. The mini-stave was then aseptically placed into a Erlenmeyer flask of wine. The chlorine treatment consisted of aseptically placing an infected mini-stave into a capped 500 ml Erlenmeyer flask containing 300 ml of chlorine solution. The chlorine solution had approximately 250 mg/l available chlorine and was produced by adding recently purchased household bleach (5.25% sodium hypochlorite) to distilled water. Citric acid was added to lower the ph of the solution to approximately 7.00. The ministave was left in this solution for 24 hours. The ministave was aseptically removed from the Erlenmeyer flask, rinsed with 200 ml sterile distilled water and placed into a sterile 400-mL beaker. The mini-stave was soaked for five minutes in a solution of sulfur dioxide (300 mg/l) and citric acid (500 mg/l) to allow the sulfur dioxide to react with any residual chlorine. The mini-stave was rinsed with 200 ml of sterile distilled water and soaked for five minutes in an additional 200 ml of the same type of water. The mini-stave was then aseptically placed into a Erlenmeyer flask of wine. The alkaline treatment consisted of aseptically placing a infected mini-stave into a capped Erlenmeyer flask containing 300 ml of a hot alkaline solution (62 C). The alkaline solution was made by adding potassium carbonate (1.25 g/l) to hot distilled water (65 C). The mini-stave was left in this solution for 24 hours. The mini-stave was then aseptically removed from the Erlenmeyer flask, rinsed with 200 ml sterile distilled water and placed into a sterile 400 ml beaker. The mini-stave was soaked for five minutes in a solution of sulfur dioxide (300 mg/l) and citric acid (500 mg/ L) to allow the citric acid to react with any residual potassium carbonate. The mini-stave was rinsed with 200 ml of sterile distilled water and soaked for five minutes in the same type of water. The mini-stave was aseptically placed into a Erlenmeyer flask of wine. The sulfur dioxide treatment consisted of aseptically placing a infected mini-stave into a capped Erlenmeyer flask containing 300 ml of a sulfur dioxide solution. The sulfur dioxide solution had a concentration of free sulfur dioxide of approximately 250 mg/l (determined by aeration oxidation method) and a ph of 3.04. It was prepared by adding 300 mg/l of potassium metabisulfite and 500 mg/l of citric acid to distilled water. The mini-stave was left in the solution for 24 hours. The mini-stave was then aseptically removed from the Erlenmeyer flask, rinsed with 200 ml of sterile distilled water and placed into a sterile 400 ml beaker. The mini-stave was soaked for five minutes in 200 ml of sterile distilled water and then aseptically placed into a flask of wine. The flasks of treated mini-staves were kept at room temperature (18 C to 27 C) with weekly plating of the wines to monitor for the presence of Acetobacter bacteria. The first plating of each flask occurred one week after it was treated. The flasks were plated for 10 weeks. Plating of a individual flask was discontinued after that flask produced a plate completely covered with acetic acid bacteria (greater than 300 cfu plate). If a flask produced a plate that was less than completely covered with acetic acid bacteria it was plated again the following week. The treatment of a mini-stave was considered unsuccessful if acetic acid bacteria were detected in the flask of wine containing the treated mini-stave, during the 10-week plating period. Detection of acetic acid bacteria: The presence of acetic acid bacteria was determined by spread-inoculating 0.1 ml of wine over the surface of plates of bromocresol green-ethanol agar (11). All flasks were gently swirled prior to plating. The flasks of wine containing treated mini-staves were plated without duplication or dilution. Preliminary work found that if growth of acetic acid was to occur in these wines it was relatively rapid. This made monitoring for the presence of acetic acid bacteria in these wines relatively easy. The flasks of wine containing spoiled mini-staves were plated in duplicate after diluting in 0.1% peptone water. Plates were incubated at 30 C for two to four days. The growth of colonies and the changing of the color of the plate from bluish-green to yellow and then back to bluish-green indicated the presence of acetic acid bacteria (11). Chemical analyses: Total and free sulfur dioxide (aeration oxidation) and ph were analyzed as described by Amerine and Ough (2). Oxygen and ethanol (HPLC), and acetic, lactic, and malic acids (HPLC) were determined as described by Zoecklein et al. (21). Results and Discussion The flasks of wine inoculated with A. aceti in the first experiment (containing five mini-staves with sealed sides) each had acetic acid bacteria populations of approximately 104 cfu/ml of wine the day before the sanitizing treatments. The wines in the flasks were turbid, exhibited strong aromas of acetic acid and ethyl acetate, and had film growth on their surfaces. The condition of the staves was thought to be beyond what one would normally consider recoverable for wine use. Ten weeks after the sanitizing treatments in the first experiment, the only flasks of wine that were not contaminated with acetic acid bacteria (less than 10 cfu/ml) were the three hot water treatment flasks (Table 1). Considering how susceptible these flasks of wine were to the growth of acetic acid bacteria, it appeared that the hot water treatment successfully sanitized these staves. The control, alkaline, and chlorine treatment flasks all had acetic acid bacteria present (10 or more cfu/ml) the week after the treatments. Only one of the sulfur dioxide treatment flasks had acetic acid bacteria present a week after the treatments. The other two flasks were found to have acetic acid bacteria present when they were plated a week later. All of the flasks that tested positive for acetic acid bacteria eventually took on the appearance and smell of the original flasks inoculated with A. aceti. Heat treatment flasks remained clear and did not smell of acetic spoilage.

BARREL WOOD m 519 Table 1. Detection of acetic acid bacteria on "mini-staves", infected with Acetobacter aceti, after undergoing a sanitizing treatment. Treatments Weeks prior to detection of acetic acid bacteria Control A 1 12 Hot Water A ~3 B C Alkaline A 1 Chlorine A 1 Sulfur dioxide A 2 B 2 1Treatment replication. 2Greater than 10 acetic acid bacteria cfu/ml of wine after one week. 3Less than 10 acetic acid bacteria cfu/ml of wine after 10 weeks. Once it was determined that the hot water treatment was successful in treating the sealed mini-staves, a second experiment was conducted to determine if the hot water treatment was capable of sanitizing unsealed mini-staves. In addition to mini-staves contaminated with the same strain ofa. aceti used in the first experiment, mini-staves contaminated with a strain of A. pasteurianus were also used. The day before the hot water treatment, each of the six flasks of wine and mini-staves had acetic acid bacteria populations of approximately 10 4 cfu/ml of wine. They were similar in aroma and appearance to the flasks of spoiled staves in the first experiment. One of the A. pasteurianus flasks was contaminated with a yeast population of approximately 10 4 cfu/ml of wine. Ten weeks after the sanitizing treatments in the second experiment, the heat treatment flasks remained free of acetic acid bacteria. The control flasks tested positive for acetic acid bacteria after just one week. All of the control flasks eventually became cloudy and resembled the original flasks that had been inoculated. Two of the A. aceti heat treatment flasks and two of the A. pasteurianus heat treatment flasks became cloudy by the end of the 10-week evaluation period. These flasks contained a fine sediment, but no film growth was observed. When these wines were examined under a microscope they were found to contain bacteria in the form of long chains made up of short rods. The heat treatment wines were analyzed for malic and lactic m m acid. Three of the cloudy wines had undergone a malolactic fermentation while one of the cloudy wines (A. aceti treatment group) and the two clear wines had not. The wines that underwent a malolactic fermentation had elevated acetic acid levels (1.1-1.5 g/l). The wines that did not undergo a malolactic fermentation had only trace levels of acetic acid (0.1 g/l). None of the heat treatment flasks of wine had aromas commonly associated with acetic spoilage. It appears that the majority of the acetic acid, in the wines that had undergone a malolactic fermentation, was produced by lactic acid bacteria. Other researchers have recorded the production of similar levels of acetic acid by lactic acid bacteria in high ph wines (10). The control flask, for the mini-staves inoculated with A. pasteurianus and containing a yeast contaminant, was found to have yeast present when it was plated a week after the sanitizing treatments. The heat treatment flask for these staves did not have any yeast present (less than 10 cfu/ml) during the 10-week plating period. The effectiveness of the hot water treatment, in contrast to the chemical based treatments, appears to be due to the ability of the heat to penetrate the surface of the staves and kill the acetic acid bacteria. Actual contact with the acetic acid bacteria by the hot water is not necessary as long as the heat is able to reach the bacteria. The chemical based treatments require that the solution actually come in contact with the bacteria. These solutions may have difficulty in reaching the bacteria in the pores of the wood. The film produced by the bacteria may also act as a barrier to the solutions but may be less effective in protecting them from heat. The three chemical based solutions examined are likely able to kill the acetic acid bacteria they come in contact with; however, their inability to reach all of the acetic acid bacteria present reduces their usefulness. Also the use of chlorine is less commonly suggested today than it once was, due to concerns regarding chlorine residue and the formation of compounds that contribute offflavors to wine (3). The hot water treatment has the advantage of no chemical residues, but will tend to leach out desirable oak compounds intended for extraction by wine. This could be minimized by limiting the period of time the barrel is full of hot water. The effectiveness of the twenty minute contact period in this study suggests that the length of time the hot water needs to be in a barrel may be relatively short. However, it would be important that the hot water come in contact with all of the contaminated surfaces of a barrel such as the bung hole and the area surrounding it. No attempt was made to determine if the staves treated with hot water contributed a spoiled character to the wines, due to the extraction of compounds absorbed on their surface during the spoilage process. This is likely to occur with barrel wood that has had extended contact with wine spoiled to the extent that occurred in this experiment. The use of wood so extensively spoiled was done in an effort to provide a extreme test for the sanitizing treatments in regards to the

520 m WILKER and DHARMADHIKARI presence of acetic acid bacteria. It was not chosen to determine if wood spoiled to this extent should be reused. Conclusions While the hot water treatment was effective on all nine of the mini-staves treated in this study, it is not known if it is as equally effective on actual barrels. Some wine barrels may have acetic acid bacteria in places protected from the heat of a hot water treatment. However, it does appear to be the most effective barrel sanitizing treatment currently being used. The hot water treatment would seem most appropriate as a means of reducing the build up of acetic acid bacteria in used barrels, rather than as a means of restoring barrels that have undergone extensive acetic acid bacteria spoilage. Literature Cited 1. Amerine, M. A., H. W. Berg, R. E. Kunkee, C. S. Ough, V. L. Singleton, and A. D. Webb. The Technology of Winemaking (4 th ed.). AVI Publishing Co., Inc., Westport, CT (1980). 2. Amerine, M. A., and C. S. Ough. Wine and Must Analysis. John Wiley and Sons, Canada (1974). 3. Amon, J. M., R. F. Simpson, and J. M. Vandepeer. A taint in woodmatured wine attributable to microbiological contamination of the oak barrel. Austral. NZ Wine Ind. J. 2(2):35-37 (1987). 4. Benda, I. Wine and brandy. In: Prescott and Dunn's Industrial Microbiology (4 th ed.). G. Reed (Ed.). pp 2291-4022. AVI Publishing Co. Inc., Westport, CT (1982). 5. Boidron, J. Preparation and maintenance of barrels. In: The Barrel and the Wine. pp 71-82. Seguin Moreau Tonnellerie, USA (1994). 6. Bousbouras, G. E., and R. E. Kunkee. Effect of ph on malo-lactic fermentation in wine. Am. J. Enol. Vitic. 22:121-126 (1971). 7. Chatonnet, P. Barrel aging of red wines. In: The Barrel and the Wine. pp 91-109. Seguin Moreau Tonnellerie, USA (1994). 8. Corison, C. A., C. S. Ough, ~. W. Berg, and K. E. Nelson. Must acetic acid and ethyl acetate as mold and rot indicators in grapes. Am. J. Enol. Vitic. 30:130-134 (1979). 9. Davis, C. R., and M. J. Reeves. Acid formation during fermentation and conversation of wine. In: Proceedings Second International Cool Climate Viticulture and Oenology Symposium, Auckland, New Zealand. R. Smart, R. Thorton, S. Rodriquez, and J. Young (Eds.). pp 308-312. New Zealand Society for Viticulture and Oenology, Auckland West, New Zealand (1988). 10. Davis, C. R., D. J. Wibowo, T. H. Lee, and G. H. Fleet. Growth and metabolism of lactic acid bacteria during and after malolactic fermentation of wines at different ph. Appl. Environ. Microbiol. 51:539-545 (1986). 11. Drysdale, G. S., and G. H. Fleet. Acetic acid bacteria in some Australian wines. Food Technol. Austral. 37:217-220 (1985). 12. Drysdale, G. S., and G. H. Fleet. Acetic acid bacteria in winemaking: a review. Am. J. Enol. Vitic. 39:143-154 (1988). 13. Drysdale, G. S., and G. H. Fleet. The growth and survival of acetic acid bacteria in wines at different concentrations of oxygen. Am. J. Enol. Vitic. 40:99-105 (1989). 14. Hoey A. W., and I. D. Codrington. Oak maturation of table wine. In: Proceedings of Sixth Australian Wine Industry Technical Conference. T. H. Lee (Ed.) pp 261-266 Australian Industrial Publishers Pty. Ltd., Adelaide, Australia (1987). 15. Joyeux, A., S. Lafon-Lafourcade, and P. Rib~reau-Gayon. Evolution of acetic acid bacteria during fermentation and storage of wine. Appl. Environ. Microbiol. 48:153-156 (1984). 16. Knox, M. Barrel preparation and maintenance. Practical Winery 8(2):62-64. 17. Margalith, P, Z. Flavour Microbiology. pp 167-8. Charles Thomas Publishers, Springfield, IL (1981). 18. Singleton, V. L. Some aspects of the wooden container as a factor in wine maturation. In: Chemistry of Winemaking. Advances in Chemistry, Vol. 34 A. D. Webb (Ed.). pp 254-277. American Chemical Society, Washington, DC (1974). 19. Singleton, V. L. Using wooden cooperage in the winery today. In: Proceedings of the 13 t" Pennsylvania Wine Conference. pp 9-18. The Pennsylvania State University, University Park (1981). 20. Wibowo, D., R Eschenbruch, C. R. Davis, G. H. Fleet, and T. H. Lee. Occurrence and growth of lactic acid bacteria in wine: a review. Am. J. Enol. Vitic. 36:302-313 (1985). 21. Zoecklein, B. W., K. C. Fugelsang, B. H. Gump, and F. S. Nury. Production Wine Analysis. Chapman Hall, NY (1990).