Maceration Variables Affecting Phenolic Composition in Commercial-scale Cabernet Sauvignon Winemaking Trials

Transcription

1 Maceration Variables Affecting Phenolic Composition 93 Maceration Variables Affecting Phenolic Composition in Commercial-scale Cabernet Sauvignon Winemaking Trials Alejandro Zimman, 1 William S. Joslin, 2 Mark L. Lyon, 3 Jeffrey Meier, 4 and Andrew L. Waterhouse 5 * The effects of different maceration variables on phenolic composition in the production of commercial-scale wines were studied after aging and bottling. These variables included temperature, pomace contact time, addition of oak chips, and addition of color enzyme. Increasing temperature and pomace contact time increased the total amount of proanthocyanidins, while temperature also increased the contribution of copigmentation to the color. Color enzyme addition increased the amount of proanthocyanidins but did not contribute to an increase in color. Four vineyards located in different Californian viticultural areas were used for the experiments, and fruit composition at one site had an overwhelming effect on phenolic composition compared to the maceration treatments. Key words: Proanthocyanidins, colored proanthocyanidins, maceration, temperature, oak chips, rotary fermentor In red wine, sensory properties such as color and taste depend on phenolic composition. Color in young red wines is mainly a function of anthocyanin concentration, ph, sulfur dioxide (SO 2 ) concentration, and copigmentation [19]. Once red wine ages, color stability depends on the formation of polymeric pigments, a product of proanthocyanidins (also known as condensed tannins or tannins) and anthocyanins [24]. Additionally, other anthocyanin reaction products have been found [2]. The flavan-3-ols in wine contribute to its bitter taste and astringent perception. Monomers are more bitter than astringent, but an increase in molecular weight by condensation of monomers enhances astringency versus bitterness [15]. Phenolic compounds are extracted from skins and seeds, and their extraction is influenced by winemaking procedures. Several studies regarding phenolic extraction have investigated events that take place during fermentation and aging. Hilgard described a color maximum before the completion of fermentation in an 1887 article [6]. The spectrophotometric calculations summarized by Somers and Evans [25] have been used by many authors to describe the evolution of color, anthocyanins, and polymeric pigment during fermentation and aging [11,23,26,27]. The introduction of chromatography to quantify flavonoids has also provided useful information [11,12,13,14]. Several studies have focused, and in some cases compared, different enological 1 Graduate Student and 5 Professor, Department of Viticulture and Enology, University of California, One Shields Avenue, Davis, CA 95616; Winemakers, 2 Wente Vineyards, 5565 Tesla Road, Livermore, CA 94550, 3 Sebastiani Vineyards, 389 Fourth Street East, Sonoma, CA 95476, and 4 J. Lohr Winery, 1000 Lenzen Avenue, San Jose, CA *Corresponding author [ alwaterhouse@ucdavis.edu] Acknowledgments: This research was supported by grants from the American Vineyard Foundation. The authors thank the collaborating wineries, Sebastiani Vineyards, Wente Vineyards, and J.Lohr Winery. Manuscript submitted May 2001; revised August 2001 Copyright 2002 by the American Society for Enology and Viticulture. All rights reserved. practices such as pomace contact [1,7,9,13,21,23,28], fermentation temperature [4,16,17,21], effects of different grape components [20], enzyme treatments [29], mode of agitation [3], and carbonic maceration [28]. In general, studies comparing treatments are small-scale fermentations with two or three treatments and use only one source of grapes. This study compared the effect of common winemaking procedures on phenolic composition in commercial-scale production. The scale of the experiment was determined by available equipment to be 20 tons per treatment. This precluded fermentation replications, as only five similar tanks were available at cooperating sites. The specific procedures included addition of oak chips, extended pomace contact, use of color enzymes, use of a rotary fermentor, heating at the end of the fermentation but before press [5], and conducting fermentations at different temperatures. A new procedure that quantifies proanthocyanidins by normal-phase HPLC [8] and a modified spectrophotometric analysis [10] were used to measure phenolic composition. Materials and Methods Grapes. Cabernet Sauvignon from four California regions (Lodi, Paso Robles, Sonoma, and Monterey) were harvested during October and November 1998 for commercial-scale wine production (one hundred tons at each vineyard). Degree days were calculated using the University of California Statewide Integrated Pest Management Project database ( axp.ipm.ucdavis.edu) based on the nearest weather station to each vineyard. The calculation started on 1 January 1998 to the harvest date, using 10 C as the lower limit. Except for Monterey, each vineyard was sampled right before harvest, with 54 clusters picked randomly and divided into three groups. One hundred berries were randomly selected from each group. Weight 93

2 94 Zimman et al. and volume were measured, and number of seeds was counted in each group. Monterey grapes were sampled prior to crushing and not in the vineyard. Lodi and Paso Robles fruit were transported to Wente Vineyards in Livermore, California, and Sonoma and Monterey fruit were transported to Sebastiani Vineyards in Sonoma, California. Winemaking. Five different treatments were carried out with each fruit lot. During crushing the must from each vineyard was sequentially distributed in tanks to avoid fruit variability among the treatments. The treatments follow. Control: Twenty tons of fruit were destemmed, crushed, and placed in 22,700-L tanks. No SO 2 was added and yeast (Fermivin, DSM Food Specialities USA, Charlotte, NC) was inoculated at a concentration of 0.24 kg/1000 L. Fermentation temperature was maintained below 29.4 C and two pump overs (378 L/min for 15 min) were performed daily using open hoses to wet and break up the cap. After seven days all wines were dry (0 Brix). Wines were pressed up to 2 bar in a bladder press. Malolactic bacteria (Viniflora-Oenos, CHR Hanse, Denmark) was inoculated if no spontaneous fermentation occurred. Free SO 2 was then adjusted to 30 mg/l and ph to 3.6 with tartaric acid. Wines were put in four barrels for 14 months. The barrels were clean, uniform (same lot), and had four years of previous use to reduce the influence of variable oak compounds. After seven months the wine was racked and SO 2 was checked and adjusted if necessary. After 14 months of barrel aging, the barrel lots were blended and bottled. Oak chips: Loose French oak was added right after must was pumped into the tank in a concentration of 3.6 g/l. After fermentation was completed, the wine with the pomace and oak chips was pressed off and matured as in the control. Extended maceration: This treatment was the same as control except that after dryness the pomace was in contact with the wine for 20 days before pressing. During the extended maceration the cap was wetted every four days and then dry ice was added to exclude oxygen and prevent spoilage. Color enzyme addition: This treatment was only for Sonoma and Monterey grapes. Preparation was the same as the control, but Rapidase (DSM Food Specialities USA, Charlotte, NC) was added as prescribed during the first day of fermentation in order to increase color extraction. Rotary fermentor: The rotary fermentor was used only for the Lodi and Paso Robles grapes. Treatment was the same as in the control but fermentation was performed in a Mueller rotary tank (Mueller, Springfield, MO), which rotated twice in each direction every three hours, during the first four days. The fermenting must (11 to 14 Brix) was then pressed and the fermentation was completed in a 22,700-L tank. Rotary tanks contain a spiral blade inside to break up the cap. Heat at the end: After fermentation was conducted as in the control, two pump overs were performed each day for two days. Each additional pump over was carried out with a heat exchanger to bring pumped wine to 41 to 49 C, continuing the pumping over until the tank reached 32 C. Wine was pressed and matured as the control. Measurement of temperature during fermentation. Grapes from Paso Robles harvested on October 1999 were used for this experiment, repeating the conditions of the 1998 control. Three temperature loggers (StowAway Tidbi T, Onset, Bourne, MA) were introduced into the tank by attaching the loggers to a steel stick at 0.6, 0.9, and 1.2 m from the top of the cap. Immediately before the first pump over of the day, the temperature from the tank probe located at 0.5 m above the bottom of the tank was also recorded. Simultaneously, a thermometer was inserted at 30 cm from the top surface of the cap at five different points and an average temperature was calculated. Fermentation maximum temperature. Grapes harvested in October 1998 from a different section of the Paso Robles vineyard and fermented at J. Lohr Winery, Paso Robles, California, were used for this experiment, repeating the condition of the 1998 control. However, fermentation temperature was kept below 24 and 32 C in two separate tanks. Analyses. All wines were analyzed after 14 months of barrel aging plus four months of bottle aging. Free and total SO 2 were determined with the Ripper procedure. Titratable acidity (TA) expressed in tartaric acid was calculated by titration to ph 8.2. Volatile acidity (VA) expressed as acetic acid was determined with a cash-steam distillation. Alcohol concentration was calculated with an ebulliometer. All methods are described by Ough and Amerine [18]. Proanthocyanidin and colored proanthocyanidin quantification was done in duplicate according to Kennedy and Waterhouse [8], except that no caffeine was added to improve the recovery of skin proanthocyanidins. For wines no recovery improvement by adding caffeine was observed. Chromatography was performed on a Hewlett-Packard 1090 model (Palo Alto, CA) coupled with a photodiode array detector recording chromatograms at 280 and 520 nm. The column was a 250 x 4 mm LiChrospher Si-60, 5 µm particle size (Merck, Darmstadt, Germany). Concentration was calculated with epicatechin (Sigma, St. Louis, MO) as external standard. In addition, proanthocyanidins were divided in two groups: low molecular weight proanthocyanidins (LMWP) and high molecular weight proanthocyanidins (HMWP), based on the retention time of the dimer to tetramer from cacao beans [22]. Therefore, LMWP contained compounds with longer retention time than an epicatechin dimer and smaller than an epicatechin pentamer. HMWP contained compounds with the same and larger retention time than the pentamer. Colored proanthocyanidins were calculated based on peak area at 520 with a retention time equal or larger than the cacao bean dimer and expressed in malvidin- 3,5-diglucoside (Pfaltz & Bauer, Waterbury, CT) equivalents. These structures were also divided, as above, into low molecular weight colored proanthocyanidins (LMWCP) and high molecular weight colored proanthocyanidins (HMWCP). Analysis of variance and t-test were done by the Minitab 10.5Xtra Power statistical program (Minitab, State College, PA), calculating the least significant difference (LSD) for mean values at 0.05 level. Spectrophotometric analysis was done in duplicate following a modified protocol of Somers and Evans [25] by Boulton [10]. The absorbance of the wine at 520 nm can be attributed to

3 Maceration Variables Affecting Phenolic Composition 95 three different fractions: polymeric pigment, copigmentation, and free anthocyanins. Standard deviation for all the values was below 0.03 absorbance units. Results and Discussion There were inevitable differences among vineyards (Table 1). The main parameter in the harvest decision for this experiment was an optimum grape ripeness by Brix. Vineyards had different spacing, rootstocks, and clones. From the degree-days measurement, the Lodi vineyard was in the warmest region whereas the Monterey vineyard was in the coolest, as expected. Volume and weight data obtained from a random sampling of the vineyard indicated that Lodi contained the largest berries and Monterey the smallest. It is possible that the berry values for Monterey were affected by the breakdown of the larger berries and the inability to obtain random samples from the vineyard, resulting in a smaller berry value. There was no significant difference in the number of the seeds among vineyards. The decision point for pressing (time or density) in such experiments is always problematic, but in each lot, fruit dryness was attained within the same time. Comparison of the control wine composition from each vineyard (Table 2) indicates similar values for free SO 2 and ph. Discrepancy between Brix and alcohol among the vineyards and the control wines can only be attributed to the difficulty to obtain an accurate field Brix value in commercial-scale experiments. Measurement of cap temperature during fermentation. Cap and liquid temperatures had different values throughout fermentation (Figure 1). Maximum cap temperature registered with the loggers was 37.8 C and the minimum 22.4 C, with a pattern of steady increases and sharp decreases resulting from pump over. This common enological operation reduced the cap temperature as much as 11.9 C in 12 min. The average temperature during fermentation for each of the three loggers was 27.3, 29.4, and 29.6 C. Because of the reduction in the cap height, the logger placed deeper in the cap (number 3) came in contact with the liquid on 24 October and did not have high variations afterward. Temperatures measured by inserting Table 1 Viticultural parameters of the vineyards used in the experiments. Degree Harvest Brix at TA ph Volume/ Weight/ Seeds/ Location Spacing Rootstock Clone days ( C) (1998) harvest berry (ml) berry (g) berry Lodi Bilateral cordon own + Dog Ridge Oct c * 1.12 c x 12 + Salt Creek Paso Bilateral cordon 5C Australian Oct bc 1.02 bc 1.31 Robles 6 x 10 Sonoma GBC AxR Oct ab 0.95 ab 1.3 Monterey Bilateral cordon own unknown Nov a 0.87 a x 7 *Means followed by the same letter are not significantly different (LSD, 0.05). Centigrade Table 2 Chemical composition of control wines after 14 months of barrel aging and four months in bottle. Free SO 2 Total SO 2 Volatile acidity TA Alcohol Wine (mg/l) (mg/l) (g/l) (g/l) ph (%) Lodi Paso Robles Sonoma Monterey the thermometer into the outer surface of the cap prior to pump over indicated values similar to the loggers. Daily values obtained from the tank probe, in contact only with the juice, were much lower, with a narrow range between 20 and 22.8 C. Measured temperatures using the tank probe did not represent the temperature in the cap. This measurement agrees with a temperature difference in cap versus liquid reported by Ough and Amerine [17], although the temperature profile over time differs considerably. The effective temperature of the maceration, or extractive process, may be mainly controlled by the temperature in the cap, where the seeds and skins are present. The cap temperature during fermentation is a difficult variable to control and can easily rise far above the bulk tank temperature at this scale. It is also possible that differences found when Logger #1 Logger #2 Logger #3 Tank Probe Cap Thermometer 20 Oct 21 Oct 22 Oct 23 Oct 24 Oct 25-Oct 26-Oct 27-Oct 28-Oct Day Figure 1 Temperature measured during fermentation with loggers, tank probe, and thermometer inserted in the cap.

4 96 Zimman et al. comparing treatments at different scales are due to different cap temperatures in large- versus small-scale fermentors due to heat exchange capacity of the mass of the cap. In addition, for the purpose of planning maceration experiments, other parameters such as regime of pump overs per day, position of the temperature controller, fermentation kinetics, and cooling capabilities could affect temperature, thus influencing phenolic extraction. Fermentation maximum temperature. The results in bottled wine for two tanks where the tank fermentation temperature was maintained below 24 or 32 C according to the tank probe are shown in Table 3. Setting the fermentation maximum temperature 8 C higher significantly increased the amount of HMWP but did not affect the amount of LMWP or any of the colored proanthocyanidins fractions analyzed by HPLC. Although wines from both tanks had a similar absorbance value at 520 nm, the contribution to the color was different. The wine fermented at the lower temperature had a higher level of anthocyanins but copigmentation contributed less to the color. There was no difference in polymeric pigment values. These results do not agree with previous work by Gao et al. [4], who found that higher fermentation temperatures produced higher pigmented polymers, measured by HPLC, but did not produce any difference in free anthocyanins. Differences such as cultivar type, size of experiment, time of analysis, and analytical method make it difficult to compare results. A difference of 8 C controlled by the tank probe appears to be enough to cause differences in wine composition. For the experiment fermented at higher temperature, it is expected that cap temperature would also be higher, given that the other conditions (tank size, pump over, temperature control, cooling system, and so on) were controlled. If any of these variables change, a direct comparison between two fermentation temperatures may not be appropriate. Heat at the end. Results indicate that there was a significant increase in proanthocyanidins when compared with the same vineyard control for three of the four vineyards studied (see Table 4). Increases ranged from insignificant (51 mg/l) in Monterey to 31% (566 mg/l) in the Lodi fruit. The changes in proanthocyanidins were caused by an increase in the HMWP. In colored proanthocyanidins, only the Lodi HMWCP fraction increased and the Sonoma LMWCP fraction decreased. Heating at the end of fermentation affected the contribution of copigmentation, free anthocyanins, and polymeric pigment to wine color, but the effects were not consistent. Each of these factors increased in Lodi and Paso Robles wines, producing higher color. For Sonoma and Monterey wines, there was an increase in copigmentation and a decrease in free anthocyanins. These discrepancies among vineyards could be attributed to the fruit composition. Heating at the end of fermentation had a similar effect as the higher temperature of fermentation for two parameters: it increased the amount of HMWP and the contribution of copigmentation to the color. Extended maceration. The extended maceration significantly increased the amount of proanthocyanidins for all vineyards (Table 4). Increases ranged from 17% (277 mg/l) in Paso Robles to 41% (1188 mg/l) in Sonoma. In this case, increments for LMWP were significant for all vineyards except Paso Robles. Increments in HMWP were significant for all the vineyards. Compared to the control, the percentage increase of LMWP was always lower than HMWP. For colored proanthocyanidins by HPLC, Sonoma was the only vineyard with a significant increase caused by the HMWCP fraction. Conversely, spectrophotometric results for polymeric pigment indicated an increase in all the vineyards. The difference between colored proanthocyanidins and polymeric pigment results may indicate that there are differences in colored polymers analyzed by different techniques, which need to be addressed. The color at 520 nm was higher than the control in Lodi and Paso Robles but lower in Sonoma and Monterey. These discrepancies cannot be explained other than different fruit composition not defined here have modified the result for extended maceration. Increase in proanthocyanidins by extended contact time has been well documented in previous papers [7,9]. Unlike treatments involving increases in temperature, the extended maceration also increased the LMWP in most of the vineyards. Rotary fermentor. In comparison to the control, the rotary fermentor significantly decreased the total amount of proanthocyanidins and the colored proanthocyanidins measured by HPLC and the polymeric pigment measured with spectrophotometer (Table 4). The percentage of decrease was higher for HMWP than for LMWP. The rotary fermentor also reduced the amount of free anthocyanins, but it highly increased copigmentation in Paso Robles. The color in the rotary fermentor was lower than the control in Lodi and almost the same in Paso Robles. These results agree with Fischer et al. [3], who reported that rotary tanks reduce proanthocyanidins and pigmented proanthocyanidins compared to mechanical pump over. The authors proposed that rotary tanks produce a strong mechanical maceration that yields components with low degree of poly- Table 3 Proanthocyanidins (LMWP and HMWP) in mg/l epicatechin equivalents, colored proanthocyanidins (LMWCP and HMWCP) in mg/l malvidin-3,5-diglucoside, and color in absorbance units at 520 nm of two wines fermented at two different maximum temperatures. Proanthocyanidin Colored proanthofractions: cyanidin fractions: Contribution to the color by: Maximum Colored Abs. Polymeric Free temp. Proanthocyanidins LMWP HMWP a proanthocyanidins LMWCP HMWCP 520 nm pigment Copigmentation anthocyanins 24 C C a Indicates a significant difference at p < 0.05.

5 Maceration Variables Affecting Phenolic Composition 97 Table 4 Proanthocyanidins (LMWP and HMWP) in mg/l epicatechin equivalents, colored proanthocyanidins (LMWCP and HMWCP) in mg/l malvidin-3,5-diglucoside, and color in absorbance units at 520 nm using different winemaking practices and different source of grapes. Proanthocyanidin Colored proanthocyanidin fractions: fractions: Contribution to the color by: Vineyard Colored Abs. Polymeric Free Treatment Proanthocyanidins LMWP HMWP proanthocyanidins LMWCP HMWCP 520 nm pigment Copigmentation anthocyanins Lodi Control 1812 j * 496 de 1316 kl 491 defgh 171 cde 320 fg Oak chips 1961 i 474 e 1487 i 511 defg 156 efgh 355 def Extended maceration 2287 fg 570 c 1717 g 519 def 169 def 350 ef Rotary fermentor 1401 l 406 f 995 n 406 ij 144 fgh 262 i Heat at end 2378 f 514 d 1864 f 569 cd 162 efg 407 cd Paso Robles Control 1605 k 501 de 1104 m 432 ghi 169 def 263 hi Oak chips 1496 kl 403 f 1093 m 413 hij 144 fgh 269 ghi Extended maceration 1882 ij 505 de 1377 jk 417 ih 148 efgh 269 ghi Rotary fermentor 1153 m 376 f 777 o 334 j 131 h 203 j Heat at end 1965 i 516 d 1449 ij 451 fghi 162 efg 289 ghi Sonoma Control 2915 d 656 b 2259 d 741 b 222 a 519 b Oak chips 2616 e 632 b 1984 e 636 c 196 abc 440 c Extended maceration 4103 a 792 a 3311 a 835 a 219 a 616 a Color enzyme 3596 b 666 b 2930 b 775 ab 201 ab 574 ab Heat at end 3381 c 646 b 2735 c 728 b 189 bcd 539 b Monterey Control 1832 j 561 c 1271 l 456 efghi 138 gh 318 fgh Oak chips 1827 j 560 c 1267 l 550 d 162 efg 388 cde Extended maceration 2153 h 634 b 1519 i 421 hi 143 gh 278 ghi Color enzyme 2173 gh 563 c 1610 h 534 de 148 efgh 386 cde Heat at end 1883 ij 573 c 1310 kl 511 defg 146 efgh 365 def *Means followed by the same letter are not significantly different (LSD, 0.05). merization. Our results suggest that the rotary fermentor simply did not extract as many proanthocyanidins as the control due to the shorter contact time. It is possible that this process reduced the amount of colored proanthocyanidins because there was not enough time for the extraction from skins or because the amount of anthocyanins extracted was lower, reducing the amount of colored proanthocyanidins formed during aging. Oak chips. The presence of oak chips during fermentation had no clear effect on any of the parameters studied. HMWP was increased in Lodi but decreased in Sonoma. LMWP was only decreased in Paso Robles. Oak chips also reduced the HMWCP fraction for Sonoma but increased it for Monterey. From the spectrophotometric results, only Lodi showed an increase in color resulting from an increase in polymeric pigment and copigmentation. Sonoma and Monterey, however, had a decrease in free anthocyanins that reduced color. After fermentation, oak chips turn red, which could account for the decrease in free anthocyanins in all the vineyards except Lodi. In addition, oak chips introduce new components (such as gallic acid, hydrolyzable tannins, and polysaccharides) that could alter the equilibria of the phenolic compounds already present in the must. Color enzyme. The color enzyme treatment for Sonoma and Monterey produced an increase in total proanthocyanidins due only to an increase in the HMWP fraction. Spectrophotometric results indicate a decrease in absorbance at 520 nm for Sonoma and a similar value for Monterey. In both vineyards free anthocyanins were lower than the control, which agrees with results reported by Wightman et al. [29]. These authors also observed that enzymes caused an increase in colored proanthocyanidins. That result was found only in the HMWCP and the polymeric pigment result in Monterey. It seems that enzymes do not increase the color of the wine in the long term but do increase the amount of proanthocyanidins that could favor colored proanthocyanidin formation. Grape composition. Wines produced using a vineyard located in Sonoma had the highest concentration of proanthocyanidins and color regardless of the winemaking operation. That also happened for colored proanthocyanidins with the exception of Lodi-heat at the end and Sonoma-oak tannins. In terms of the fraction contributing to color, polymeric pigment and free anthocyanins were also higher in Sonoma than in the other vineyards. The highest copigmentation values were found in the Monterey and Sonoma wines. Polymeric pigment is the fraction that contributes more to the color of wine after wines have been aging for 18 months. The percentage ranges from 38% in Monterey-control to 65% in Paso Robles-oak chips. Free anthocyanins or any other pigment that are bleached by SO 2 are still responsible for 22 to 34% of the wine color. Copigmentation is especially high in all the Monterey wines (30 to 38%), compared to the other wines with percentages that ranged from 10 to 29%.

6 98 Zimman et al. Grape composition is another variable that needs to be taken into account when studying phenolic extraction and aging. In some cases the winemaking operation did not have the same effect using grapes from different sources. Conclusions This study shows how several parameters affected the phenolic composition in commercial-scale wines analyzed after 18 months of aging. Temperature was an important parameter that affected proanthocyanidin content and color, but it was not fully controlled during fermentation. Extended pomace contact time, like additional heat, increased the amount of HMWP but also increased the LMWP fraction. Shorter pomace contact time occurring in rotary fermentors had the opposite effect, reducing proanthocyanidin concentration. The color enzyme treatment did not increase the color but increased the amount of proanthocyanidins. Regardless of the treatment used, wines produced using grapes from a vineyard located in Sonoma yielded higher proanthocyanidin amounts, colored proanthocyanidins, and absorbance at 520 nm, indicating that grape composition can overwhelm any treatment applied. In some cases treatments had inconsistent effects among vineyards, and it is believed that fruit composition or variables not accounted for in the experiments are responsible for the different responses. Literature Cited 1. Auw, J.M., V. Blanco, S.F. O Keefe, and C.A. Sims. Effect of processing on the phenolics and color of Cabernet Sauvignon, Chambourcin, and Noble wines and juices. Am. J. Enol. Vitic. 47: (1996). 2. Bakker, J., and C.F. Timberlake. Isolation, identification, and characterization of new color-stable anthocyanins occurring in some red wines. J. Agric. Food Chem. 45:35-43 (1997). 3. Fischer, U., M. Strasser, and K. Gutzler. Impact of fermentation technology on the phenolic and volatile composition of German red wines. Int. J. Food Sci. Technol. 35:81-94 (2000). 4. Gao, L., B. Girard, G. Mazza, and A.G. Reynolds. Changes in anthocyanins and color characteristics of Pinot noir wines during different vinification processes. J. Agric. Food Chem. 45: (1997). 5. Gerbaux, V. Etude de quelques conditions de cuvaison susceptibles d augmenter la composition polyphenolique des vins de Pinot Noir. Rev. Oenologues 69:15-18 (1993). 6. Hilgard, E.W. The extraction of color and tannin during red-wine fermentation. Calif. Agric. Exp. Stat. Bull. 77:1-3 (1887). 7. Kantz, K., and V.L. Singleton. Isolation and determination of polymeric polyphenols in wines using Sephadex LH-20. Am. J. Enol. Vitic. 42 (4): (1991). 8. Kennedy, J.A., and A.L. Waterhouse. Analysis of pigmented highmolecular mass grape phenolics using ion-pair, normal-phase highperformance liquid chromatography. J. Chromatogr., A 866:25-34 (2000). 9. Kovac, V., E. Alonso, M. Bourzeix, and E. Revilla. Effect of several enological practices on the content of catechins and proanthocyanidins of red wines. J. Agric. Food Chem. 40: (1992). 10. Marco, J.E. Copigmentation: Evolution of pigments and cofactors during maturation, and their relation to color in Cabernet Sauvignon and Pinot noir wines. Thesis, University of California (1999). 11. Mayen, M., J. Merida, and M. Medina. Free anthocyanins and polymeric pigments during the fermentation and post-fermentation standing of musts from Cabernet Sauvignon and Tempranillo grapes. Am. J. Enol. Vitic. 45: (1994). 12. Mayen, M., J. Merida, and M. Medina. Flavonoid and nonflavonoid compounds during fermentation and postfermentation standing of musts from Cabernet Sauvignon and Tempranillo grapes. Am. J. Enol. Vitic. 46: (1995). 13. Merida, J., L. Moyano, C. Millan, and M. Medina. Extraction of phenolic compounds in controlled macerations of Pedro Ximenez grapes. Vitis 30: (1991). 14. Nagel, C.W., and L.W. Wulf. Changes in the anthocyanins, flavonoids and hydroxycinnamic acid esters during fermentation and aging of Merlot and Cabernet Sauvignon. Am. J. Enol. Vitic. 30: (1979). 15. Noble, A.C. Bitterness in wine. Physiol. Behav. 56 (6): (1994). 16. Ough, C.S., and M.A. Amerine. Studies on controlled fermentation. V. Effects on color, composition, and quality of red wines. Am. J. Enol. Vitic. 12:9-19 (1961). 17. Ough, C.S., and M.A. Amerine. Studies with controlled fermentation. VI. Effects of temperature and handling on rates, composition and quality of wines. Am. J. Enol. Vitic. 12: (1961). 18. Ough, C.S., and M.A. Amerine. Methods for Analysis of Musts and Wines, 2nd ed., p Wiley-Interscience, New York (1988). 19. Ribereau-Gayon, P. The anthocyanins of grapes and wines. In Anthocyanins as Food Colors. P. Markakis (Ed.), pp Academic Press, New York (1982). 20. Ribereau-Gayon, P., and J.C. Milhe. Recherches technologiques sur les composes phenoliques des vins rouges. Conn. Vigne Vin 4:63-74 (1970). 21. Ribereau-Gayon, P., P. Sudraud, J.C. Milhe, and A. Canbas. Recherches technologiques sur les composes phenoliques des vins rouges. Conn. Vigne Vin 4: (1970). 22. Rigaud, J., M.T. Escribano-Bailon, C. Prieur, J.M. Souquet, and V. Cheynier. Normal-phase high-performance liquid chromatographic separation of procyanidins from cacao beans and grape seeds. J. Chromatogr., A 654 (2): (1993). 23. Scudamoresmith, P.D., R.L. Hooper, and E.D. McLaran. Color and phenolic changes of Cabernet Sauvignon wine made by simultaneous yeast/bacterial fermentation and extended pomace contact. Am. J. Enol. Vitic. 41 (1):57-67 (1990). 24. Somers, T.C. The polymeric nature of wine pigments. Phytochemistry 10: (1971). 25. Somers, T.C., and M.E. Evans. Spectral evaluation of young red wines: anthocyanin equilibria, total phenolics, free and molecular SO 2, chemical age. J. Sci. Food Agric. 28: (1977). 26. Somers, T.C., and M.E. Evans. Evolution of red wines. I. Ambient influences on colour composition during early maturation. Vitis 25:31-39 (1986). 27. Somers, T.C., and K.F. Pocock. Evolution of red wines. III. Promotion of the maturation phase. Vitis 29 (2): (1990). 28. Timberlake, C.F., and P. Bridle. The effect of processing and other factors on the colour characteristics of some red wines. Vitis 15 (1):37-49 (1976). 29. Wightman, J.D., S.F. Price, B.T. Watson, and R.E. Wrolstad. Some effects of processing enzymes on anthocyanins and phenolics in Pinot noir and Cabernet Sauvignon wines. Am. J. Enol. Vitic. 48:39-48 (1997).