Abstract
The influence of season and 50% cluster thinning (CT) and 50% leaf removal (LR) on skin anthocyanin profile was studied in cv. Nebbiolo in the Barbaresco wine production area (northwestern Italy). Climatic conditions over two years strongly interacted with treatments: 2000 was hot and rainy; 2001 was cooler and very dry. The leaf area/yield ratio value was higher in 2000 in response to good water availability, and it was optimal only in 2000 LR vines and in 2001 CT vines, which displayed the highest total soluble solid content at harvest. In comparison with control vines, anthocyanin concentration increased only in 2000 CT vines, even if final concentration was not significantly higher, because of its reduction before harvest. The 2001 drought reduced vine vegetative development, sugar production (−8.3%), and total anthocyanin concentration (−20.7%) in respect to 2000. Individual anthocyanin concentration varied as a consequence of chemical structure, canopy management practices, and seasonal climatic conditions. Thus, the peonidin-3-glucoside/malvidin-3-glucoside ratio varied on average from 1.9 in 2000 to 1.2 in 2001, and it was higher in treated vines than in control vines. Anthocyanin accumulation can be divided into two distinct phases: an initial phase of rapid increase influenced primarily by vine vegetative conditions and secondarily by cultural practices and a second “finishing” phase during which climatic conditions played a fundamental role. 3′-Substituted anthocyanin biosynthesis is likely more strongly influenced by climatic conditions and cultural practices than is 3′,5′-substituted anthocyanin biosynthesis.
- Vitis vinifera L.
- leaf removal
- cluster thinning
- total anthocyanins
- peonidin-3-glucoside/malvidin-3-glucoside ratio
In Vitis vinifera anthocyanins are present as monoglucoside forms of methoxylated and/or hydroxylated anthocyanidins. Number and type of substituents in the B ring of the anthocyanin molecule identify different anthocyanins: cyanidin-3-glucoside (Cy) and peonidin-3-glucoside (Pn) with two substituents (3′-substituted), delphinidin-3-glucoside (Dp), petunidin-3-glucoside (Pt), and malvidin-3-glucoside (Mv) with three substituents (3′,5′-substituted). Two different enzymes are involved in 3′- and 3′,5′-substituted form biosynthesis, the flavonoid 3′-hydroxylase and the flavonoid 3′,5′-hydroxylase, respectively (Boss et al. 1996, Bogs et al. 2006).
In some varieties, 3′-substituted forms prevail, while in others the 3′,5′-substituted forms are prevalent. Furthermore, some varieties are rich in esterified forms, which are completely absent in others (Macheix et al. 1990). In Vitis vinifera anthocyanin monoglucosides can be acylated by acetic acid, caffeic acid, or p-coumaric acid.
These different profiles are linked to the presence of metabolic enzymes that vary among varieties. The total amount of anthocyanins in berry skins is a genetic characteristic (i.e., pink, red, or black skin variety), but many other factors may influence skin color intensity at harvest.
As reported by several authors, light, temperature, and the interactive effects of light and temperature highly influence anthocyanin accumulation in berry skins (Downey et al. 2006, Yamane et al. 2006). In warm climates, high light exposure can increase the concentration of phenolics and anthocyanins because of the higher activity of phenylalanine ammonia-lyase enzyme (Roubelakis-Angelakis and Kliewer 1986). In hot-climate regions, a high degree of bunch exposure is not conducive to optimal anthocyanin accumulation in berries because the level of related temperature can be too high (Kliewer 1977, Haselgrove et al. 2000, Bergqvist et al. 2001, Spayd et al. 2002). Other factors, particularly those linked to soil composition (Yokotsuka et al. 1999), water availability (Esteban et al. 2001, de la Hera Orts et al. 2005), agronomical practices (Jackson and Lombard 1993), and virus presence in the vine (Guidoni et al. 1997), can modify berry anthocyanin amount at harvest.
Some authors have analyzed individual anthocyanin accumulation patterns during ripening (Keller and Hrazdina 1998, Yokotsuka et al. 1999), as influenced by climatic variables (Cacho et al. 1992) or clone (Roggero et al. 1986), but little is known about the effect of agronomical treatments on these patterns. In Cabernet Sauvignon, some individual anthocyanin modifications have been indicated as a function of nitrogen availability (Keller and Hrazdina 1998); anthocyanin profile was influenced by soil composition (Yokotsuka et al. 1999) and cluster thinning increased the concentration of Pn and the Pn/Mv ratio in Nebbiolo (Guidoni et al. 2002). In a hot climate, bunch shading increased the percentage of 3′,5′-substituted anthocyanins and of Mv coumaroyl derivative in Shiraz (Haselgrove et al. 2000, Downey et al. 2006), and irrigation increased the concentration of the five principal anthocyanins at harvest (Esteban et al. 2001). These observations indicate that anthocyanin profile can partially be modified by environmental and climatic factors, which likely influence the level of expression of metabolic enzymes. However, to our knowledge, there have been no studies on the effects of these variables on the activity of the flavonoid 3′-hydroxylase and flavonoid 3′,5′-hydroxylase enzymes.
Cultural practice effects on berry color have long been studied; among them, leaf removal and cluster thinning, two different methods of modifying leaf area/yield ratio and fruit-zone microclimate, could potentially improve grape quality (Hunter et al. 1991, Jackson and Lombard 1993, Dokoozlian and Hirschfelt 1995). Cluster thinning, in particular, can advance fruit ripening (Guidoni et al. 2002, Petrie and Clingeleffer 2006), but it does not exert any effect on maturity in warm and dry conditions and in vigorous vineyards (Keller et al. 2005).
The Nebbiolo cultivar is very sensitive to terroir and is characterized by high vigor and reduced berry skin color. Nebbiolo is not widespread, but in its original area (Piedmont, northwestern Italy) it is used to produce high-quality aged wines such as Barolo and Barbaresco, which have problems in color stability during aging. The anthocyanin profile of Nebbiolo is typical, with peonidin-3-glucoside the most prevalent (Guidoni et al. 1997, 2002). Peonidin-3-glucoside is a very reactive anthocyanin, particularly in wine (Gerbi et al. 2002); conversely, malvidin-3-glucoside is less reactive. In the Nebbiolo cultivation area, leaf removal and cluster thinning are two cultural practices used to improve cluster health and to reduce yield (for production of many Italian wines, crop load is controlled by law), respectively.
Furthermore, almost all the most important international cultivars studied in depth (e.g., Cabernet Sauvignon, Merlot, Shiraz, Tempranillo) are characterized by an anthocyanin profile where Mv and its derivatives are predominant and 3′-substituted anthocyanins are present in small amounts. No studies have been conducted on the anthocyanin profile evolution of the grapevine varieties where 3′-substituted anthocyanins prevail. The aim of our work was to investigate the effects exerted by leaf removal and cluster thinning on total anthocyanin amount and on the individual anthocyanin accumulation pattern of cv. Nebbiolo berry skins over the course of two experimental seasons.
Materials and Methods
Plant material and experimental conditions.
The experiment was performed in 2000 and 2001 on Vitis vinifera cv. Nebbiolo vines grafted onto 420A rootstock and planted in a hillside vineyard in Neive (production area of Barbaresco wine, Piedmont, Italy). The vineyard, located 281 m asl, was planted in 1977 on a loamy soil with a pH of 7.9. Vine spacing was 1.2 x 2.8 m; vines were pruned to a single 12-bud cane and shoots were vertically trained. The vineyard, planted on a 30% slope, was south-exposed, and vine rows were oriented east/northeast–west/southwest. The vineyard was managed according to standard viticultural practices for the cultivar and the region. Climatic conditions of the two seasons, 2000 and 2001, were very different, particularly rainfall amount and July–August maximum temperatures (Table 1⇓).
The effects of three agronomical treatments were evaluated: cluster thinning (CT), leaf removal (LR), and a control (no treatment). For CT, 50% of the clusters of each plant were removed 5 weeks after full bloom in 2000 and 7 weeks after full bloom in 2001. The distal cluster was removed, leaving only one bunch per shoot at most. Intervention time was not the same in the two years because of problems with work organization, but other experiments demonstrated that cluster-thinning timing had little or no influence on berry weight (Petrie et al. 2000, Keller et al. 2005) and total anthocyanin (Guidoni and Argamante 2003).
For LR, 50% of leaves were removed 5 weeks after full bloom and again on new growth 3 weeks later. Leaves at every other node on primary and lateral shoots were removed. Each treatment was applied on three replicate groups of four contiguous vines. A completely randomized block design was used in the experiment. Two weeks before harvest, leaf area (LA) was estimated (Smart 1985).
From the time some berries (regardless of treatment) began to develop color (veraison = time 0) to commercial harvest, 100 berries were randomly collected every 10 to 15 days from vines of each treatment and weighed. Thirty berries were randomly chosen from this group and further divided into three groups of 10 berries, which were used as triplicates for skin anthocyanin analysis. After skin and seed removal, the mesocarp of berries was homogenized and centrifuged 20 min at 3000 rpm at 20°C. The concentration of total soluble solids (TSS) was assessed on the supernatant using a PR-10 electronic refractometer (Atago, Tokyo, Japan).
Analytical procedure.
For anthocyanin extraction and analysis, a published method was used (Guidoni et al. 2002). A hydroalcoholic buffer at pH 3.2 containing 2 g/L Na2S2O5 was used as skin extractant solution. For HPLC analysis formic acid:water (10:90, v/v, as solvent A) and formic acid:methanol:water (10:50:40, v/v/v as solvent B) were used. The method allowed the extraction of the total amount of skin anthocyanin. Individual anthocyanins were identified comparing the retention time of each chromatographic peak with available data in literature (Di Stefano et al. 1995). The concentration of individual anthocyanins was expressed as mg kg−1 of berry fresh weight using malvidin-3-O-glucoside chloride (Extrasynthèse, Genay, France) as external standard. Total anthocyanin amount was calculated as the sum of the concentrations of the free and derivative forms of anthocyanins.
At harvest the number and the weight of the clusters and yield per vine were measured. Data were submitted to analysis of variance and means were separated by the Duncan test by SAS statistical software (SAS Institute, Cary, NC).
Results
Treatment effects on vine vegetative parameters.
In 2000, leaf area (LA) was similar in control and cluster thinned (CT) vines (Table 2⇓); in leaf removal (LR) vines, LA was 47% lower than control vines. CT vine yield was 16% lower than control and LR yield. In 2001, when drought and July–August high temperature reduced the vegetative development of all vines, LR caused a 26% LA decrease per vine. CR reduced yield at harvest by 13% in 2000 and by 36% in 2001. LA/yield ratio value was lower in 2001 than in 2000 in all treatments because of decreased LA. In both years, as expected, LA/yield ratio was significantly lower in LR vines as compared with CT vines; in 2001 it was significantly lower also in control vines. In 2000 the LR vine LA/yield ratio was ~1.5 m2 kg−1. In 2001 in CT and control vines, this ratio was slightly different, whereas in the other treatments it was higher (control and CT vines in 2000) or lower (LR vines in 2001) (Table 2⇓). At harvest, berry weight was not significantly different among treatments and years (Table 3⇓).
Treatment effects on total soluble solids.
Total soluble solids (TSS) were higher in 2000 than in 2001 (+2 Brix, at harvest). In 2000 the TSS increase was regular and similar in all treatments until 45 days postveraison (dpv) (Table 3⇑), after which TSS in CT and control vine berries showed no further increase, whereas TSS increased in LR vine berries (+2 Brix during the final 10 days of observation), leading to significantly higher TSS at harvest (~+1.4 Brix). On the contrary, in 2001, TSS of CT berries was higher than that of control and LR berries at harvest. During the drier year, the reduced cropload of CT vines favored a better LA/yield ratio, allowing a higher solute accumulation. Furthermore, at 30 dpv, average TSS was higher than in 2000, but as the subsequent increase was limited, it was lower than in 2000 at harvest (22 Brix in 2001, 24.3 Brix in 2000).
Treatment effects on total anthocyanin accumulation.
In 2000, anthocyanin accumulation started rapidly and the trend was similar for all treatments until 45 dpv, after which no further accumulation was detected (Table 3⇑). Anthocyanin accumulation started earlier in CT berries; indeed, anthocyanin concentration was significantly higher already at the first sampling date. At harvest, anthocyanin concentration was similar among treatments (714 mg kg−1 in CT, 624 mg kg−1 in the control, and 654 mg kg−1 in LR).
From 30 to 56 dpv in 2001, total anthocyanin concentration did not increase, except in LR vines (Table 3⇑), probably because of the prolonged drought (Table 1⇑). In CT and in control vines, particularly during the last week of ripening, a significant increase in anthocyanin accumulation was detected, leading to the maximum concentration at harvest (557 and 509 mg kg−1, respectively). In 2001, anthocyanin amount at harvest was similar in all treatments, but it was significantly lower compared with 2000 (−22 and −23% for CT and LR and −19% for the control) and with the average values (740 mg kg−1) observed at the same site (data not published).
Treatment effects on individual anthocyanin accumulation.
In 2000, with total anthocyanin concentration higher than in 2001 at harvest, only average values of Pn (+68% in 2000 versus 2001) and Cy (+29%) (3′-substituted anthocyanins) were higher than in 2001, whereas those of 3′,5′-substituted anthocyanins (Dp, Pt, and Mv) and acylated forms were not (Table 4⇓). As observed for total anthocyanins, individual anthocyanin concentrations increased during the first period of accumulation, except for Cy concentration, which showed little variation during the ripening period. Note that Cy and Pn, average concentrations were already high at the first sampling date (56 mg kg−1 corresponding, for Pn, to about 30% of total amount at harvest), whereas 3′,5′-substituted anthocyanin concentrations were very low (about 8 mg kg−1, corresponding, for Mv, to 6% of total amount). The Nebbiolo anthocyanin profile at harvest (40% Pn, 25% Mv, 5 to 6% Dp and Pt, 11% Cy, and 10 to 13% total derivatives) was very different from that observed at veraison, when ~80% of color was represented by 3′-substituted anthocyanins.
At 30 dpv a similar concentration of 3′,5′-anthocyanin had accumulated, independently from treatment (except for Mv in LR vines) and year (Table 4⇑). On the contrary, the average concentration of 3′-substituted anthocyanin free forms was lower in 2001 (87 mg kg−1) compared with 2000 (124 mg kg−1). In 2001, LR and CT treatments increased Pn and Cy concentrations. These concentrations were higher than in the control not only at 30 dpv but also at 56 (Pn) or 66 (Cy) dpv.
In 2000, independently from treatment, anthocyanin accumulation was stable or decreased for all treatments after 45 dpv except for Cy, Pn, and total derivative forms in LR vines (Table 4⇑). In LR vines in particular, Mv was significantly reduced in the week before harvest. In 2001, after a plateau phase (from 30 to 56 dpv), 3′,5′-substituted anthocyanin concentration significantly increased in CT and the control vines; 3′,5′-substituted anthocyanins in LR vines and 3′-substituted anthocyanins in all treatments, except for Pn in the control, were stable until harvest. At harvest 3′,5′-substituted anthocyanin concentration was similar in the two years, while 3′-substituted anthocyanin concentration, particularly that of Pn, was lower in 2001 than in 2000, under the same treatment. Thus, the lower 2001 anthocyanin total concentration at harvest was due to the lower Pn and Cy (Table 4⇑). Consequently, the ratio between Pn and Mv, the two most abundant anthocyanins in Nebbiolo skins, varied as influenced by season and treatments. In particular, at harvest, the average Pn/Mv ratio was 1.9 in 2000, the wetter year (1.8 in control, 2.1 in LR, and 1.9 in CT) and 1.2 in 2001, the drier year (1.0 in control, 1.1 in LR, and 1.4 in CT).
Discussion
An important interaction between cultural practices and the seasonal conditions was detected (Table 5⇓). TSS and anthocyanin reduced amounts and anthocyanin accumulation delay detected in 2001 are a possible consequence of drought which, limiting leaf development, reduced vine photosynthetic capacity and dry matter production (Escalona et al. 2003, Gomez-del-Campo et al. 2003, de la Hera Orts et al. 2005). The limited LA may have caused a higher sunlight cluster exposition and, consequently, an excessive berry temperature increase. As can be seen in the literature, high temperatures can hinder regular anthocyanin biosynthesis (Kliewer 1977, Haselgrove et al. 2000, Berqgvist et al. 2001, Spayd et al. 2002).
In our study, cluster thinning in 2000 advanced veraison with respect to the control as seen in anthocyanin concentration at 0 and 17 dpv (Table 3⇑), but at harvest no differences were detected. In a year when meteorological conditions induced high vigor (2000), berries of CT vines did not accumulate higher sugar contents with respect to the berries of control vines probably because cluster thinning contributed to the LA/yield ratio beyond 1.5 (Table 2⇑), generally considered the optimum for a fully ripe winegrape (Kliewer and Dokoozlian 2005). In the drier year (2001), cluster thinning, improving the LA/yield ratio, resulted in higher sugar accumulation and earlier ripening with respect to the control, also evident from the significant differences detected 30 dpv in total anthocyanin concentration (Table 3⇑). The detected value of LA/yield ratio in 2001 CT vines (2.07 m2 kg−1) indicates that for a vigorous variety such as Nebbiolo, the optimum of this ratio could be closer to 2.00 than to 1.5. As vine performance is also linked to the equilibrium between leaf area and crop load, so cluster-thinning results can vary, depending on whether or not crop limitation is useful to improve this equilibrium.
The 2000 LR vines, with a LA/crop weight ratio nearer to the optimum, showed a slower but more prolonged increase in TSS than did the control vines; consequently, TSS in LR vines at harvest was significantly higher than in the other treatments. No significant differences were detected in anthocyanin concentration with respect to control vines. In 2001, LR vines did not accumulate less sugar and anthocyanin than the other treatments notwith-standing their reduced LA, probably because they suffered less from drought given their fewer leaves (Gomezdel-Campo et al. 2003).
From veraison on, there were two phases of sugar and anthocyanin accumulation. During the first phase, which occurred for approximately 6 weeks, ~90 to 100% anthocyanins and ~75 to 80% sugars were accumulated; the accumulation trend of these two classes of compounds is almost parallel and significantly correlated (R2 = 0.81), as observed by others (Fournand et al. 2006). Climatic conditions and cultural practices exert a huge influence on this first phase, affecting photosynthetic activity, vine vegetative development, LA/crop weight ratio, and fruit-zone microclimate.
The second phase, which could be defined as a “finishing touch,” occurred for approximately 3 weeks and ended at harvest. The accumulation trends of anthocyanins and TSS were no longer parallel, and under different climatic conditions or different management practices they can be very different. In 2000 TSS accumulation occurred until 45 dpv in all treatments, after which TSS was enhanced only in LR vines; in 2001 only CT vines showed an increase in TSS during the last phase of ripening. Changes in anthocyanin were less significantly correlated to those in sugar in both years (R2 = 0.68), suggesting that the factors influencing the accumulation of these two classes are different. Previous studies have shown that climatic factors, such as light and temperature, exert relevant influence on polyphenol synthesis but minor effects on skin sugar amount (Wicks and Kliewer 1983). In the current study, sugar was analyzed only in pulp, but it seems possible to draw the same conclusion. Sugar accumulation was certainly influenced by vine photosynthetic efficiency (we stress the importance of a foliar apparatus still equilibrated with the yield and efficient during the last phases of ripening), but final anthocyanin concentration seemed, especially during the second phase, more strongly influenced by light and temperature, which can directly influence their biosynthesis (Keller and Hrazdina 1998, Berqgvist et al. 2001, Spayd et al. 2002).
A different behavior of individual anthocyanin accumulation —the higher differences of 3′-substituted anthocyanin concentration with respect to 3′-5′-substituted anthocyanin concentration—has been observed as a function of year and treatments (Table 4⇑, Table 5⇑). The observed abundance of 3′-, but not of 3′,5′-, substituted anthocyanins at the beginning of veraison could be analytical evidence that the biosynthesis of former occurred before that of the latter, consistent with previous references (Bogs et al. 2006, Jeong et al. 2006, Castellarin et al. 2006). Furthermore, in 2001, after a plateau phase from 30 to 56 dpv, 3′,5′-substituted anthocyanins accumulated again from 56 to 66 dpv when temperatures fell, partially limiting the negative effect of the drought, whereas 3′-substituted anthocyanin concentration, even in cooler climatic conditions, showed no further variation except for peonidin in CT vines (Table 4⇑). The detected increase of Pn concentration in CT vines is due to the berry weight reduction (Table 2⇑) and not to a real accumulation. These aspects could indicate a different sensitivity to environmental conditions and/or a different onset of the flavonoid 3′-hydroxylase and flavonoid 3′,5′-hydroxylase enzymes. Our hypothesis seems to be shared by other authors who have supposed that climatic variables could exert their influence by repressing the Cy biosynthesis without influencing the Mv biosynthesis (Keller and Hrazdina 1998), but to our knowledge, no studies have examined the impact of temperature and light on the activity of these enzymes.
As the berry skin anthocyanin profile is mainly important in determining wine color intensity and hue, and Pn is more reactive and less stable than Mv in Nebbiolo wine (Gerbi et al. 2002), it seems possible that some cultural practices, in modifying the fruit-zone microclimate, may also modify the Pn/Mv ratio, playing an important role in both color intensity and stability of wines. Thus, the Pn/Mv ratio could be used as an indicator of wine color stability. When the Pn/Mv ratio is lower in berry skins, wines could show higher color stability, but this circumstance seems to occur when anthocyanin accumulation is limited. The lower values of the ratio detected in 2001 than in 2000, particularly in control vines, could indicate that climatic conditions such as water deficit and high temperature may influence individual anthocyanin accumulation. These findings agree with the results reported by other authors who have observed that weather conditions can modulate the anthocyanin profile of a variety (Cacho et al. 1992, Esteban et al. 2001).
Conclusions
Cluster thinning and leaf removal effects on berry ripening (TSS, anthocyanin concentration, and anthocyanin profile) were strongly influenced by yearly weather conditions that can also influence vine vegetative behavior. Furthermore, the effects varied depending on how cultural practices influenced the equilibrium between leaf area and yield of the vines. In particular, the adopted practices influenced the berry skin anthocyanin profile, indicating that anthocyanin profile is not strictly dependent on genotype but can be modulate by environmental conditions. This aspect should be taken into account when estimating the enological quality of grapes. Finally, it would be interesting to further study of the effects of climatic variables on the activity of enzymes catalyzing the anthocyanin biosynthesis.
Footnotes
Acknowledgments: The authors thank Ufficio Agrometereologico Regione Piemonte for meteorological data and the Antichi Poderi dei Gallina winery of Neive for hosting the trial.
- Received January 2007.
- Revision received June 2007.
- Revision received July 2007.
- Copyright © 2008 by the American Society for Enology and Viticulture