European Food Research and Technology

, Volume 220, Issue 5, pp 597–606

Effect of ripening stage of grapes on the low molecular weight phenolic compounds of red wines

Authors

    • Dpto de Biotecnología y Ciencia de los AlimentosUniversidad de Burgos
  • M Luisa González-San José
    • Dpto de Biotecnología y Ciencia de los AlimentosUniversidad de Burgos
Original Paper

DOI: 10.1007/s00217-004-1106-z

Cite this article as:
Pérez-Magariño, S. & González-San José, M.L. Eur Food Res Technol (2005) 220: 597. doi:10.1007/s00217-004-1106-z

Abstract

Different red wines were elaborated to study the effect of the date of the grape harvest on the levels of individual low molecular weight phenolic compounds, which are chiefly responsible for the wine color. Two red grape varieties and two consecutive years were studied at three different harvesting stages of grapes, and the changes during the 18 months of wine aging (12 months in oak barrels and 6 months in the bottle) were also followed. The results showed that the wines made from grapes harvested 1 week later than the usual date generally had higher contents of some simple phenols, which can act as cofactors that can maintain the color intensity and violet tonalities in aged wines. Besides, these wines had lower levels of caftaric and coutaric acids, which are two of the main substrates for oxidation and browning processes.

Keywords

AgingCofactorsColorRed grapesLow molecular weight phenolsRed winesRipening

Introduction

Many physical and biochemical changes that take place during the grape ripening process are important to obtain grapes with certain special characteristics. In general, these changes do not have the same effect as the different compounds of the grapes. Therefore, the date the grapes are harvested is critical to the production of quality wines.

The traditional indices that have been commonly used to establish the best time to harvest grapes are berry weight, sugar and acid content, and the relation between sugars and acidity [14]. These indices are related to the highest fruit yields, but these harvest data do not coincide with the optimum properties of the grapes for producing quality wine [2, 5], especially as they do not consider important parameters such as volatile compounds and the phenolic composition of the grapes [6, 7].

The importance of taking phenolic compounds into account when determining the harvest date, especially with a view to producing quality wines, is because they are well known to contribute to several sensorial attributes of wines, such as color, body, bitterness, and astringency [8, 9].

The phenolic components in grapes change during the ripening process, but they do not follow the same trends as the sugars and acids, and so their peak concentration is not usually reached at the same time as peak sugar concentrations. Earlier studies [4, 10] showed that phenolic compounds accumulated during berry ripening, reaching peak concentration levels about 1 week before the peak sugar concentrations (usual harvest date). In this time, a slight fall is observed, after which there is a new increase or accumulation, more gradual and less intense as well as variable depending on each of the individual phenolic compounds concerned [1114]. Then, the maximum concentrations are usually found 1 week or later than for the sugars [10].

Low molecular weight phenols play an important role in determining the sensorial characteristics of wines, although they are present in low amounts. Some of these compounds can act as cofactors, also called copigments, stabilizing the color of anthocyanins [15, 16], and besides they also have an influence on certain characteristics, such as astringency and bitterness [17, 18]. Different studies showed that the increase in the concentration of cofactors leads to color intensification (hyperchromic shift), and also to a bathochromic shift in the wavelength of maximum absorbance, providing a blueshift owing to the formation of colored pigments [1820]. Copigments include a large variety of structurally unrelated compounds, such as flavonoid and non-flavonoid phenols, and organic acids [15].

A recent review [18] shows that cinnamic acids and quercetin glycosides are the compounds mainly involved in copigmentation processes with anthocyanins, which affect the color of red wines. The cited paper summarizes different studies which indicated that a higher concentration in these compounds provided greater color enhancements and a shift in the wavelength of the maximum absorbance.

To our knowledge, there have been no studies on the influence of the grape harvest date on the low molecular weight phenol composition of wines, nor on the relationship to the wine quality or the wine evolution during wood aging. Therefore, the aim of this work was to study the influence of the date of the grape harvest on the levels of low molecular weight phenols of red wines, and also on the evolution of phenols and on wine quality during the aging period for 1 year in wood barrels and 6 months in the bottle. Two different varieties, three stages of ripening, and two consecutive vintages were studied.

Materials and methods

Chemicals

Gallic acid, protocatechuic acid, vanillic acid, caffeic acid, syringic acid, coumaric acid, and vanillic aldehyde were purchased from Sigma–Aldrich (St. Louis, MO, USA), and the flavanol monomers catechin and epicatechin were from Extrasynthèse (Lyon, France).

Milli-Q water, formic acid (Merck, Darmstadt, Germany), and methanol and acetonitrile (Lab-Scan, Dublin, Ireland) were used in high-performance liquid chromatography (HPLC) analyses.

Winemaking process

This study was carried out with wines elaborated from two important red grape varieties, Tinto Fino (TF) and Cabernet Sauvignon (CS). These grapes were cultivated in vineyards from the “Ribera del Duero” Spanish Designation of Origin. Grapes from two consecutive vintages were used for this study.

The grapes were harvested at three different stages of ripening. The first harvest date was decided by the criteria of the enologist, based mainly on the sugar content, but also taking into account acidity and the grape sanitary condition. The other two harvest dates were 1 and 2 weeks after the conventional date. During the further 2-week period there were no significant climatic changes, so the ripening process continued adequately.

The wines for each variety at each ripening stage were elaborated in duplicate batches in stainless steel vats with a capacity of 1,200 kg using traditional red-wine-making methods. After being harvested by hand and selected to remove damaged grapes, the grapes were destemmed and transferred to the vats and a small amount of SO2 (0.04 g/l) was added. The total polyphenol index (TPI as measured by absorbance at 280 nm), density, and temperature (between 25 and 28 °C) were monitored and controlled daily during fermentation. The wines were drawn off when a maximum in the TPI values was recorded during two consecutive days. This moment coincided with nearly complete consumption of the reducing sugars ( below 3 g/l), with the maceration time being between 12 and 14 days. After alcoholic fermentation, the wines were racked and transferred into barrels for malolactic fermentation and wood aging. The wines were aged in barrels made from new medium–high char American oak for 12 months. After 1 year in the wood, the wines were bottled and allowed to continue aging in the bottle for 6 months in dark cellars at controlled temperature and relative humidity levels.

Extraction of low molecular weight phenols

A preliminary separation step and a partial concentration step are required in order to quantify the individual low molecular weight phenols. Column separation was performed on Amberlite XAD-2 resin (Sigma–Aldrich, St. Louis, MO, USA) with a particle size of 150–250 μm according to the method of DiStefano and Cravero [21] with certain slight modifications [22], to obtain two fractions of low molecular weight phenols. Fraction A consisted of water-soluble components, gallic acid being the only one of interest in this study. Fraction B, extracted using ether, comprised the low molecular weight phenols and flavanol monomers. Following column separation, the fractions were evaporated to dryness in a rotary vacuum evaporator (T<35 °C) and immediately redissolved in a known volume of methanol (2 ml). These extracts were diluted 20:80 (methanol/water) and filtered through a Millex-HV 0.45-μm membrane filter (Millipore Co., Bedford, MA), prior to their HPLC analyses.

HPLC diode-array detection mass spectrometry analyses

The liquid chromatograph equipment employed was an Agilent Technologies LC–MS Series 1100, with a diode-array detection system and a mass detector. The method used was a slightly modified version of the method described by Pérez-Magariño et al. [23]. A Spherisorb ODS2 column (250 mm × 4.6-mm inner diameter, particle size 3 μm) and a guard column of the same material were used. The mobile-phase solvents were 4.5% formic acid in water (solvent A) and solvent A/acetonitrile (9:1 v/v) (solvent B). The conditions employed were a linear gradient at a flow rate of 0.7 ml/min from 0 to 47% solvent B in 40 min; from 47 to 100% in 35 min; and isocratic conditions for a further 20 min.

Figure 1 shows the HPLC chromatograms of both fractions recorded at 280 nm of the low molecular weight phenols identified in aged red wines.
Fig. 1

High-performance liquid chromatography chromatograms of fractions A and B recorded at 280 nm. Gallic acid (1); protocatechuic acid (2); caftaric acid (3); catechin (4); coutaric acid (5); vanillic acid (6); caffeic acid (7); fertaric acid (8); syringic acid (9); vanillin (10); epicatechin (11); coumaric acid (12)

The low molecular weight phenols that were clearly identified in the two fractions were gallic acid, protocatechuic acid, vanillic acid, caffeic acid, syringic acid, coumaric acid, caftaric acid, coutaric acid, fertaric acid, vanillic aldehyde (vanillin), and the flavanol monomers catechin and epicatechin. All of them were identified by comparing their retention time, UV–vis spectra, and mass spectra with their respective standard or with published data [23]. They were quantified and expressed as milligrams per liter of the corresponding compound. Since no commercial standards were available for the tartaric esters (caftaric acid, coutaric acid, and fertaric acid), they were quantified and expressed as milligrams per liter of the esterified caffeic acid, coumaric acid, or ferulic acid, respectively.

Chromatic parameters

The CIELab parameters [24] were determined and calculated using the illuminant D65 and a 10º standard observer, following the OIV suggestions [25]. These spectrophotometric measurements were carried out using a Beckman model DU-650 diode-array spectrophotometer (Las Rozas, Spain) with quartz cuvettes of 1-mm path length, with the CIELab parameters being calculated automatically by a suitable program installed in the spectrophotometer (Analytical Development Center, Beckman Instruments, 1995).

All the analyses described previously were carried out in duplicate and periodically during wine aging. Different samples were taken at 0, 2, 8, and 12 months after aging in wood, and 18 months after aging (12 months in wood and 6 months in the bottle).

Sensorial analyses

Sensorial analyses were carried out by ten expert tasters in order to determine the differences among wines made from grapes harvested at the three different dates studied. These analyses were realized in wines that had not been aged and in 18-month aged wines, specially related to color, astringency, and structure.

Statistical analyses

Statistical analyses of the data were carried out using analysis of variance (ANOVA) and the least significant difference (LSD) test to determine statistically different values at a significance level of α=0.05. All statistical analyses were performed using the Statgraphics Plus 4.0 statistical package for Windows (1999).

Results and discussion

The wines made in each of the 2 years displayed very similar levels of the individual quantified phenols and color parameters. The α values shown in Tables 1 and 2 indicate that no statistically significant differences were found between the data from the two vintages in each compound, color parameter, and each wine elaborated from grapes harvested at different stages of ripening. This fact could be explained because the climatic conditions were very similar in the two vintages studied, reaching comparable maturity indices in both years. Thus, the grape composition in terms of sugars, average berry weight and total phenols was very similar [7]. Besides, the ripening process was exhaustively controlled and the culture practices were adapted to the ripening process in order to obtain similar grape quality and composition every year .
Table 1

Mean values of low molecular weight phenols ± standard deviation of red wines that had not been aged from the two vintages. Cabernet Sauvignon (CS); Tinto Fino (TF)

Compounds

1st CS harvest

2nd CS harvest

3rd CS harvest

1st TF harvest

2nd TF harvest

3rd TF harvest

Protocatechuic acid (mg/l)

b 7.93±0.14a

a 6.42±0.10

a 6.57±0.36

B 8.25±0.16b

A 5.83±0.07

B 8.22±0.27

0.4197c

0.8448

0.3215

0.4244

0.7880

0.3171

Vanillic acid (mg/l)

b 9.10±0.16

c 13.09±0.08

a 6.55±0.26

B 12.17±0.11

A 11.42±0.62

C 16.41±0.16

0.2664

0.6451

0.3315

0.7429

0.3035

0.8450

Caffeic acid (mg/l)

b 2.70±0.15

c 3.57±0.07

a 2.36±0.15

A 4.84±0.24

C 9.95± 0.16

B 7.20±0.15

0.3242

0.5934

0.3758

0.2530

0.3823

0.9921

Syringic acid (mg/l)

a 7.75±0.10

b 8.34±0.07

b 8.39±0.08

A 7.08±0.21

A 7.12±0.08

B 10.61±0.17

0.7297

0.9913

0.9925

0.3228

0.9643

0.4479

Gallic acid (mg/l)

a 29.74±0.08

b 35.22±0.12

c 36.04±0.13

B 22.26±0.16

C 32.45±0.35

A 20.27±0.35

0.8056

0.4882

0.1999

0.3930

0.1412

0.1722

Coumaric acid (mg/l)

a 2.87±0.08

b 3.82±0.11

a 3.00±0.09

A 5.00±0.15

C 11.21±0.13

B 10.84±0.08

0.9780

0.5493

0.5731

0.3459

0.4987

0.6465

Caftaric acid (mg/l)

a 24.43±0.08a

a 24.10±0.34

b 26.06±0.11

C 16.18±0.18

B 10.31±0.11b

A 8.89±0.10

0.7313c

0.2652

0.3078

0.4057

0.3135

0.7760

Coutaric acid (mg/l)

b 5.96±0.08

a 5.54±0.08

b 5.86±0.11

C 10.04±0.09

B 5.35±0.11

A 7.83±0.10

0.7313

0.9924

0.2812

0.9689

0.3027

0.5824

Fertaric acid (mg/l)

a 0.40±0.08

b 0.88±0.09

c 2.14±0.10

B 1.12±0.08

C 1.92±0.21

A 0.82±0.10

0.9560

0.5093

0.4081

0.9098

0.1609

0.7364

Catechin (mg/l)

c 101.13±0.52

b 93.95±0.48

a 89.48±1.91

C 22.23±0.17

B 17.32±0.25

A 11.85±0.13

0.2831

0.9820

0.2960

0.9475

0.3178

0.2886

Epicatechin (mg/l)

a 50.01±1.04

b 56.65±1.26

b 58.20±0.49

A 18.24±0.74

B 23.44±1.32

B 22.95±0.35

0.2980

0.2725

0.3160

0.2887

0.2701

0.5860

a,bValues with the same letter for each grape variety were not significantly different.

cα values of analysis of variance (ANOVA) results of vintage effect in each parameter and harvest date

Table 2

Mean values of CIELab parameters ± standard deviation of red wines that had not been aged from the two vintages.

Compounds

1st CS harvest

2nd CS harvest

3rd CS harvest

1st TF harvest

2nd TF harvest

3rd TF harvest

L*

c 49.59±0.14a

a 46.17±0.09

b 47.09±0.29

B 60.99±0.01

A 59.42±0.10b

C 62.63±0.20

0.6616c

0.2816

0.3591

0.1469

0.2502

0.2446

a*

b 63.62±0.02

a 63.10±0.11

c 64.77±0.20

C 50.41±0.01

B 49.91±0.08

A 48.12±0.12

0.1961

0.2740

0.2802

0.0925

0.3450

0.3934

b*

a 1.49±0.06

b 4.41±0.07

a 1.22±0.11

B -6.32±0.03

C -4.67± 0.04

A -6.53±0.04

0.3903

0.3187

0.6735

0.2784

0.2877

0.2468

C*

a 63.64±0.01

a 63.26±0.12

b 64.78±0.21

C 50.80±0.01

B 50.13±0.08

A 48.56±0.13

0.1728

0.2623

0.2866

0.1075

0.3478

0.3873

h*

b 1.33±0.06

c 4.00± 0.06

a 1.08±0.10

B -7.15±0.04

C -5.35± 0.04

A -7.73±0.03

0.3611

0.3041

0.6680

0.2489

0.2793

0.2742

a,bValues with the same letter for each grape variety were not significantly different.

cα values of ANOVA results of vintage effect in each parameter and harvest date

Since no vintage effect was detected, this factor was not taken into account and the wines from both vintages were considered as replications.

The ANOVA and the LSD test results for the wines that were not aged showed that the “harvest date” effect was detectable (Table 1), but that it was different for each individual phenolic compound.

Similar results were obtained in the wines from both varieties, and, in general, higher levels of the simple phenolic acids were found in the wines made from the maturer grapes (second and third harvests), with the exception of protocatechuic acid. This fact could be interesting in the aging process, since some of those phenolic compounds act as copigments, and then they can improve the color characteristics of wines [18, 19, 26].

The most significant difference between both kinds of single-variety wines studied was detected in tartaric ester levels (caftaric, coutaric, and fertaric acids). So, the TF wines made from the grapes with the lowest degree of ripening were richer in tartaric esters. However, the wines made from the CS grape variety yielded contrary results, with the wines made from the grapes collected on the last harvest date being richer in tartaric esters than the wines made from the grapes harvested on the first and second dates. These results can be due to variety differences, since the synthesis capacity varies among grape variety [12, 27, 28].

The changes in flavanol monomers, catechin and epicatechin, during the ripening stages selected were different for each compound. The wines made from the least ripe grapes (first harvest) had greater concentrations of catechin than those made from the riper grapes (second and third harvests). These results are in agreement with those found by Fernández-de-Simón et al. [29] and Jordao et al. [30]. However, these authors also found a decrease of epicatechin during the ripening process, which was not detected in the wines of this study, which showed a slight increase of the levels of epicatechin in both cases.

The wines also had differences in color characteristics (Table 2). The wines of both varieties made from the grapes collected on the second harvest date had the lowest lightness (L*), and presented the highest color intensity, as was expected, owing to a strong correlation between both variables [31]. In general, these wines also had higher tonality (h*) and lower red tones (a* values). In addition, no important differences among wines were observed in the chroma values (C*). Although these color differences were not great, the results were corroborated by the sensorial analyses. The expert tasters found that wines made from the grapes collected on the second harvest date showed higher color intensity values and violet tones, but they did not find differences in tonality (yellow–red tones).

These wines were aged for 18 months (12 months in wood and 6 months in the bottle) and the evolution of the compounds and color parameters studied was followed during the aging process.

It is well known that during wood aging and bottle storage, the phenolic composition of wines shows significant changes, which are correlated with their chroma characteristic, flavor, and mouthfeel sensation. These changes are correlated with the kind of barrel wood, the time of aging, and the own composition of the wines, among other parameters. Then, it is important to have in mind that all the wines studied were aged in the same conditions (kind of barrels, degree of char, time of aging, cellar, etc).

The evolution of the levels of each compound studied is showed in Figs. 2, 3, and 4, which point out that the evolution was not always similar in all the wines studied. Varietal differences, such as harvest date effects, were detected, except for the evolution of catechin and epicatechin, which showed a continuous decrease during aging in all the wines studied. This fact is directly correlated with their participation in the formation of polymerized pigments, such as previously described [7].
Fig. 2

Changes of protocatechuic, syringic, and coumaric acids of Cabernet Sauvignon (CS) and Tinto Fino (TF) wines during the aging period. 1st, 2nd, and 3rd refer to the first, second, and third harvest dates.

The evolution of the levels of syringic acid, coumaric acid, vanillic acid, caftaric acid, coutaric acid, and even gallic acid and caffeic acid were similar among the wines. The levels of syringic acid increased more or less continuously with similar tendency in all the wines studied (Fig. 2b). Coumaric acid levels showed similar evolution by variety (Fig. 2c), in CS wines they remained more or less constant, while in TF wines they increased strongly depending on the sort of wine, with the wines made from the ripest grapes showing the highest increase. The evolution of vanillic acid did not show a variety nor a ripening stage factor. After a significant decrease during the first few months, the levels remained more or less constant (Fig. 3b).
Fig. 3

Changes of caffeic, vanillic, and gallic acids of CS and TF wines during the aging period.

Fig. 4

Changes of caftaric, coutaric, and fertaric acids of CS and TF wines during the aging period.

Gallic acid levels dropped over the course of aging in all the TF wines and in two wines of CS (Fig. 3c). This drop was very large during the first months of aging. The results related to vanillic and gallic acid were surprising, since higher levels of both acids were expected, inasmuch as the wines were aged in wood. One possible explanation could be that gallic acid is a very reactive compound and was turned into the galloylated derivatives [7], leading to a decrease in free gallic acid. The decline in vanillic acid levels may also be caused by condensation with other phenolics.

The levels of caftaric acid showed a greater decrease in the CS wines than in TF ones, and no significant differences were observed among wines made from grapes at different stages of ripening (Fig. 4a). The levels of coutaric acid remained constant in TF wines, as well as in the CS wines from the third harvest date, while the levels in the CS wines from the first and second harvest date showed a slight decrease (Fig. 4b). These results could be related to the fact that these compounds are very reactive and usually take part in oxidation processes [32]. The evolution of fertaric acid was different by wine variety and ripening stage (Fig. 4c). This level of this compound showed a high increase in the CS wines, as in the TF wines from the third harvest date, while the levels of fertaric acid in TF wines from the first and second harvest dates remained more or less constant.

A similar evolution was detected for protocatechuic acid, with the levels in the wines from the third harvest date of TF and CS showing a similar trend to increase during aging (Fig. 2a). However, the CS wines from the first and second harvest dates showed a significant decrease in this compound during the first months of aging.

The cited differences should be correlated with the particular composition of each wine studied, which could be influenced by the varietal and ripening stage factors under consideration.

Relating to chromatic parameters, all the wines showed similar evolutions during aging; however, the varietal effect was clearly detected, and only a slight effect of harvest date was observed. Figure 5 shows the evolution of L*, a*, and h* parameters, which represent the changes in the chromatic characteristics of the wines. It is necessary to point out that h* is highly correlated with b* [31] and with the tonality, the classical chromatic parameter [33], which is well known to enologists.
Fig. 5

Changes of CIELab parameters, L*, a*, and h* of CS and TF wines during the aging period.

The L* values increased during aging, which indicates that the aged wines were lighter, showing a decrease in their color intensity. As expected, the a* values (red tones) decreased during the aging process. This fact is due to the well-known decrease of free anthocyanin levels during aging of wine [7, 34]. There was also an increase in tonality (h*), associated with the oxidation processes taking place during aging in wood. The changes of the h* values were more important in the first months of aging, occurring at the same time that the levels of caftaric acid decreased. Then, it is possible that the increase of the h* value was due to the formation of brown pigments by oxidation reactions in which are involved, between other compounds, some simple phenols such as caftaric and coutaric acids, which are compounds very susceptible to oxidation in wines [32].

The sensorial analysis of these aged wines showed that tasters found differences among wines, pointed out that wines from the second harvested grapes had more color intensity and blue tones, and they were more structured and equilibrated, characteristics also found in the wines from the third harvested grapes, but with slightly lower values. In addition, the wines made from the third harvest date for CS and TF were evaluated with slightly higher values of yellow tones.

According to the uniform trends exhibited by all the wines, the observations concerning the effect of the degree of grape ripening on the low molecular weight phenols and the color of the wines that had not been aged were also generally applicable in the case of the 18-month aged wines. So, the initial differences found among the wines were maintained more or less with aging. Therefore, the aged wines made from maturer grapes (second and third harvest date) were, in general, also richer in low molecular weight phenols, with the exception of protocatechuic acid. Then, since the aging conditions were similar (oak barrels, cellars, temperature, and humidity levels), it is possible to affirm that the initial composition of the grapes really determines the final characteristics of aged wines.

Taking into account the results obtained, it seems clear that the control of the harvest date is quite beneficial to the quality and stability of wines, having a great influence on the final characteristics of wine, which are extremely important for product acceptance and evaluation. In summary, the best characteristics are related to the fact that wines made with riper grapes had lower levels of caftaric acid and coutaric acid, the principal substrates for oxidation and browning [32], and higher contents of some of the simple phenols which can act as anthocyanin copigments [18, 26], contributing to maintain the color intensity and violet tonalities in wood-aged wines.

Acknowledgements

The authors wish to thank the Comisión Interministerial de Ciencia y Tecnología for funding provided for this study under project FEDER 1FD-1319. The authors are also grateful to Tomás Postigo from the “Pago de Carraovejas” winery for providing the wines for this study.

Copyright information

© Springer-Verlag 2005