Abstract
The influence of timing and method of basal defoliation on the profile and content of anthocyanins and flavonols in Tempranillo grapes was investigated. Basal leaf removal was manually and mechanically performed at two phenological stages, prebloom and fruit set. Phenolic composition was determined in grape extracts by HPLC-UV-Vis and 13 anthocyanin and flavonol compounds were identified and quantified. Regardless of the timing and method of defoliation, basal leaf removal led to more ripened fruit in terms of higher soluble solids and reduced acidity and favored the accumulation of flavonols and anthocyanins, which was related to the increase in total leaf area per yield observed in defoliated vines. For anthocyanins, there was a significant relationship between their concentrations and the larger relative skin mass observed in berries of defoliated vines. In general, the enhancement in flavonols and anthocyanins observed in berries from basal-defoliated vines tended to be greater when defoliation was conducted mechanically; yet overall, no evident differences between prebloom and fruit-set defoliation were found.
Effective yield regulation has been of interest to many in the wine industry and has focused on reducing the worldwide wine surplus (28 million hL; OIV 2009) and improving grape quality. Winter pruning and cluster thinning have been the most widespread practices in viticulture for crop control. Cluster thinning is expensive, due to high labor requirements, and may not always ensure an increase in grape quality (Chapman et al. 2004).
Prebloom basal leaf removal has been investigated in recent years as a viticultural practice to regulate yield components (Poni et al. 2006) and improve canopy microclimate (Tardaguila et al. 2010). Conducted around bloom (that is, earlier than traditional basal leaf removal usually performed from fruit set to veraison), prebloom defoliation is based on the functional relationship between yield and availability of carbohydrates at this stage (Caspari et al. 1998). The carbohydrates required for flowering and berry set may derive from the overwintering reserves in the trunk and other perennial organs of the vine and from leaf photosynthesis. However, there is consensus among researchers that leaves are the main carbohydrate source for inflorescence development during flowering (Lebon et al. 2008). Removal of basal leaves depletes the total carbohydrate availability. As a consequence, crop regulation is achieved through reduced fruit set, leading to smaller and looser clusters (Poni et al. 2006, Intrieri et al. 2008, Tardaguila et al. 2010). In these studies, grape composition improved in defoliated vines, as soluble solids and total anthocyanin concentrations increased as well. The enhancement of sunlight exposure as a result of basal leaf removal has also been related to improvement of grape quality (Smart and Robinson 1991), leading to fruit richer in soluble solids, anthocyanins, and phenols. Research has shown a positive effect of sunlight favoring the accumulation of polyphenols in the berries, mainly anthocyanins and flavonols (Spayd et al. 2002, Tarara et al. 2008). However, excessive sunlight exposure might also lead to berry color decrease (Haselgrove et al. 2000, Bergqvist et al. 2001) or even to berry sunburn (Bergqvist et al. 2001, Spayd et al. 2002), especially in warm climate regions.
Several studies have focused on the effects of traditional leaf removal, conducted between fruit set and veraison with the aim of improving fruit light exposure and ventilation, on the anthocyanin content of berries. Anthocyanin concentration increased in berries of defoliated vines in Pinot noir and remained unaltered or increased in other cultivars such as Merlot and Cabernet franc (Mazza et al. 1999) and Nebbiolo (Chorti et al. 2010). In Nebbiolo, 3′-hydroxylated anthocyanins (cyanidin-3-glucoside and peonidin-3-glucoside) were more sensitive to changes in solar radiation and temperature than 3′,5′-substituted anthocyanins (malvidin-3-glucoside, petunidin-3-glucoside, and delphinidin-3-glucoside) (Chorti et al. 2010).
Anthocyanins and other phenolic compounds such as flavonoids and nonflavonoids are important indicators of grape quality. Anthocyanins are typically located in the first cellular layers of the hypodermis (Moskowitz and Hrazdina 1981) and are responsible for grape and wine color. The flavonols are synthesized in the vacuoles of the outer layers of epidermis (Downey et al. 2003) and are considered natural sunscreen protectors for the berries.
The lack of information on the effects of prebloom basal leaf removal on the phenolic content of berries opens an important field of research. This work aims to investigate the influence of timing (prebloom or fruit set) and method (manual or mechanical) of basal leaf removal on the phenolic content of Tempranillo grape berries, focusing on anthocyanins and flavonols.
Materials and Methods
Treatments and experimental layout.
The study was conducted in a commercial Vitis vinifera L. cv. Tempranillo vineyard (clone 43 grafted onto 110 rootstock) located in Ollauri (lat. 42°31′N; long. 2°49′W, 527 m asl), La Rioja (Spain), over three consecutive seasons: 2007, 2008, and 2009. Vines were planted in 1996 in a clay-loam soil at a spacing of 2.70 m × 1.15 m, with 3220 vines/ha and trained to a vertically shoot-positioned (VSP) system. Each vine was winter pruned to retain six 2-node spurs. Vines were not irrigated during the growing season and shoots were trimmed once at the end of July, before veraison.
The experimental design involved a control and four treatments: (1) control or nondefoliated; (2) manual removal of the first eight basal leaves at prebloom (stage 19; Coombe 1995) (Man-PB); (3) manual removal of the first eight basal leaves at fruit set (stage 27; Coombe 1995) (Man-FS); (4) mechanical defoliation at prebloom (Mec-PB); and (5) mechanical defoliation at fruit set (Mec-FS). Laterals, if present, were not removed during manual defoliation. Prebloom treatments were conducted on 29 May, 13 Jun, and 4 Jun, and fruit-set treatments were on 15 Jun, 4 Jul, and 19 Jun in 2007, 2008, and 2009, respectively. Mechanical leaf removal was performed with a tractor-mounted pulsed-air leaf remover (Collard, Bouzy, France), which operated by blowing compressed air strongly enough to tear off a whole leaf or fragments of leaf blades. The leaf remover was driven at ~0.5 km/hr and blew off the leaves around the basal 60 cm of foliage, in the fruiting zone. The machine operated in two passes, once on each side of the canopy.
Treatments were arranged in a completely randomized design that consisted of five replicates of 20-vine plots for each treatment. Within each replicate plot, five vines were randomly chosen and marked one month before bloom. Additionally, for each vine, a representative shoot was also randomly tagged and the basal cluster of each labeled shoot was marked. These five tagged basal clusters per replicate plot were used for the analysis of phenolic composition. Weather conditions as daily trends of total rainfall and mean air temperatures were recorded over the three seasons by an automated meteorological station located near the vineyard.
Cluster exposure and canopy porosity.
One week before harvest, appraisal of canopy porosity and cluster exposure was carried out in 2008 and 2009 using digital image analysis, based on a proposed methodology (Tardaguila et al. 2010). For each treatment, the 25 labeled vines were photographed between 07:00 and 09:00 hr one week before harvest. Two different hue classes were established: clusters and canopy porosity given as foliage gaps. To avoid the influence of yield on the percentage of cluster pixels in the image, the ratio of pixels of cluster/yield per vine was also calculated.
Leaf area.
At harvest, for each labeled shoot, main, lateral, and total leaf area were determined using a method based on the relationship between weight-to-surface leaf disc (Smart and Robinson 1991). All main and lateral leaves per tagged shoot were separately removed and weighed. One hundred discs (2 cm diam) were cut from these leaves and weighed. The weight of leaf discs was compared to the weight of main and lateral leaves, allowing for calculation of main and lateral leaf area per shoot. In 2009, the amount of leaf area removed by manual and mechanical defoliation treatments at prebloom and fruit set was also estimated using a power regression between the length of the main vein of each main leaf and its leaf area (y = 1.703 ×1.883, R2 = 0.94***). This regression was established from a set of 100 Tempranillo main leaves using a leaf area meter (LI-3100C; LI-COR, Lincoln, NE).
Berry tissues and fruit composition.
Grapes from all treatments were hand-picked on 14 Oct 2007, 16 Oct 2008, and 23 Sept 2009. Harvest date was chosen on the basis of the control treatment grapes reaching 22 Brix for total soluble solids. The five tagged basal clusters per replicate plot were manually destemmed and berries were mixed. Of this pool of berries per replicate, two sets of five representative subsamples of 50 berries each were randomly selected for the measurement of soluble solids and acidity and the analysis of total anthocyanins and phenols. The total soluble solids concentration (Brix) was determined using a temperature-compensating digital refractometer (Atago, Tokyo, Japan) and titratable acidity and pH were determined according to OIV methods (OIV 1990). Additionally, in 2007, another set of 300 berries per treatment was separated for berry tissue determinations. Berries for tissue and phenolic analysis were weighed, frozen, and stored at −18°C until berry dissection or phenolic analysis. Berry tissues were assessed by a previous methodology (Poni et al. 2009). Each berry was individually weighed, sliced in half with a scalpel, which was also used to carefully remove the skin from the flesh and seeds, without rupturing the hypodermal cells. Seeds were then manually removed from the flesh, counted, and both skin and seeds were rinsed in deionized water, blotted dry, and weighed using a high-precision (0.01 mg) analytical scale (CP225D; Sartorius, Goettingen, Germany). Anthocyanins and total phenols were analyzed for each berry subsample using an established method (Iland et al. 2004) that comprises three stages: homogenization, maceration-extraction, and absorbance measurement. Berries were allowed to partially thaw prior to homogenization, and temperature was kept <5°C at all times. Each subsample of 50 berries was homogenized using an Ultra Turrax grind mixer (IKA, Staufen, Germany) at high speed (14,000 rpm for 1 min). Anthocyanin concentration was expressed as mg per berry and per gram fresh berry mass, whereas total phenols were expressed as absorbance units (AU) at 280 nm per berry and per gram fresh berry mass.
Grape phenolics.
In 2007 and 2008, the hydroalcoholic (ethanol:water, 50/50 v/v at pH 2.0) grape extracts obtained from the extraction method of Iland et al. (2004) were analyzed by HPLC. For each treatment, 25 grape extracts were obtained. Since the anthocyanin and total phenol values obtained by this method did not show statistical differences (p = 0.236 and 0.173, respectively) among the five subsamples corresponding to a given replicate plot, these five extracts were blended together into a new single extract. Thus, for each defoliation treatment, five “blended” grape extracts were analyzed by HPLC. These five final extracts per treatment were filtered through PTFE discs (0.45 μm) and subjected to HPLC-DAD on an Agilent modular 1100 liquid chromatograph (Waldbronn, Germany) equipped with a G1313A injector, G1311A HPLC quaternary pump, on-line G1379A degasser, G1316A oven, G1315B photodiode array detector, and Agilent Chemstation software. A LiChrosphere 100 RP-18 reverse-phase column (5 μm packing, 250 × 4 mm i.d.) was used, protected with a guard column of the same material (Scharlab, Barcelona, Spain) thermostated at 30°C. The HPLC-DAD conditions were as described by Donovan et al. (1998). The solvents were (A) 50 mM NH4H2PO4 at pH 2.6 (adjusted with o-phosphoric acid), (B) acetonitrile/solvent A (80:20, v/v), and (C) 200 mM o-phosphoric acid at pH 1.5 (adjusted with sodium hydroxide), establishing the following gradient: isocratic 100% A in 5 min, from 100 to 92% A and from 0 to 8% B in 3 min, from 92 to 0% A, from 8 to 14% B and from 0 to 86% C in 12 min, from 14 to 16.5% B and from 86 to 83.5 % C in 5 min, from 16.5 to 21.5% B and from 83.5 to 78.5% C in 10 min, from 21.5 to 50% B and from 78.5% to 50% C in 35 min, from 50 to 100% B in 5 min, from 0 to 100% A in 4 min, at a flow rate of 0.5 mL/min. Deionized water was purified with a Milli-Q water system (Millipore, Bedford, MA) before use. Acetonitrile of HPLC-gradient grade, o-phosphoric acid of analytical reagent grade, and ammonium phosphate of analytical reagent grade were purchased from Scharlab. All other chemicals (analytical-reagent grade) were obtained from Panreac (Mollet del Vallès, Spain). Spectra were recorded from 250 to 600 nm.
Quantification was carried out by peak area measurements at 365 nm for flavonols and at 520 nm for anthocyanins. External standard calibration curves were drawn. Likewise, flavonol content was expressed as quercetin-3-rutinoside (y = 0.0252 × + 0.4957; r2 = 0.9997) and anthocyanin content as malvidin-3-glucoside (y = 0.0084× + 0.7767; r2 = 0.9997)equivalents. The commercial standards used were quercetin-3-rutinoside (quercetin-3-rutinoside trihydrate 95% HPLC; Sigma Aldrich, St. Louis, MO) and malvidin-3-glucoside (malvidin-3-glucoside chloride > 95% HPLC; Extrasynthèse, Genay, France). Individual phenolic compounds were tentatively identified according to their order of elution, retention times of pure compound, and the information and characteristics of the UV-vis spectra published in the literature (Monagas et al. 2005, Gómez-Alonso et al. 2007). Those compounds with undetermined identity were consecutively named as polyphenol followed by a Roman numeral (I, II, etc.). Each measurement was run in triplicate.
Total flavonols included the concentrations of polyphenols I and II, myricetin-O-glucoside, quercetin-O-galactoside, quercetin-O-glucuronide, and quercetin-O-glucoside. Total anthocyanins were calculated as the sum of delphinidin-3-glucoside (DpGl), cyanidin-3-glucoside (CyGl), petunidin-3-glucoside (PtGl), peonidin-3-glucoside (PnGl), malvidin-3-glucoside (MvGl), malvidin-3-(6-acetyl)-glucoside (MvGlAc), cyanidin-3-(6-p-coumaryl)-glucoside (CyGlCm), petunidin-3-(6-p-coumaryl)-glucoside (PtGlCm), and malvidin-3-(6-p-coumaryl)-glucoside (MvGlCm).
Statistical analysis.
Analysis of variance was performed using the InfoStat statistical package (Professional 2007 edition, Córdoba, Argentina). Year was considered as a random variable and the error term for the defoliation treatments was the year × treatment interaction mean square. Mean separation between treatments was accomplished with the Student-Newman-Keuls test. Principal component analysis was also used to examine the effects of yield and vegetative growth on fruit composition using InfoStat.
Results
Weather.
Heat accumulation, expressed as growing degree days (GDD) from 1 Apr to 31 Oct, indicated that 2009 was the warmest year (Table 1), with higher T mean values during the summer. Over the same period, total rainfall was higher in 2008 (448 mm) than in 2007 (250 mm) and 2009 (183 mm), respectively. In spring, total rainfall in May 2008 (159 mm) was 2.5 times and 6.5 times that in May 2007 (61.2 mm) and May 2009 (29.8 mm), respectively.
Leaf area.
Amount of leaf area (LA) removed in 2009 was similar for mechanical prebloom (Mec-PB) and fruit set (Mec-FS) treatments, yet greater at fruit set in manual defoliation (Man-FS) (Table 2). It accounted for 20.4% (Mec-FS) to 32% (Man-FS) of the sum of removed LA and total LA at harvest. Total LA did not vary between control and defoliated vines in any season and the only differences among treatments were observed in 2008, where total LA per shoot in Man-PB vines was 1508 cm2 larger than in Mec-PB vines (Table 2). Main LA was reduced in defoliated vines compared to the control, but no differences due to method or timing of defoliation were consistently observed. By contrast, lateral LA increased in manual treatments compared to the control in the three seasons and compared to mechanical treatments in 2007 and 2008. Yet, no differences between prebloom and fruit-set timings were detected.
Canopy characteristics.
Overall, basal defoliation induced a significant increase in cluster exposure and canopy porosity (Table 3). The effectiveness of defoliation in providing a better exposure of the clusters together with increased canopy porosity was very similar between prebloom and fruit set, but appeared to be emphasized with mechanical, as compared to manual, defoliation.
Berry tissues and grape composition.
Flesh, skin, and seed to berry ratios of grape berries in season 2007 are shown (Table 4). The main differences were for the relative skin mass, which increased 0.7% in manual treatments and from 0.4 to 1.8% in mechanical treatments, with the highest value corresponding to Mec-FS berries. Overall, no differences between the two timings were observed, although there was a trend toward lower seed number per berry when vines were defoliated at fruit set.
Basal leaf removal led to increased soluble solids concentrations at harvest regardless of season (Table 5). The effect on the other parameters was less evident, as increased pH and lower titratable acidity values were found mostly for mechanical treatments. Differences in soluble solid concentrations due to timing and method of defoliation were minor and fluctuated among years, although mechanical and prebloom treatments tended to produce higher values.
Anthocyanins and total phenols.
Anthocyanins were markedly enhanced in berries from defoliated shoots at both timings in all seasons. There was a tendency toward greater effects due to mechanical defoliation in 2008 (Table 6), although no differences between prebloom and fruit-set treatments were detected. For total phenols, the differences among treatments were minimal across seasons.
Phenolic monomers.
Four flavonols were identified in the grape extracts of Tempranillo in both seasons: myricetin-O-glucoside, quercetin-O-galactoside, and quercetin-O-glucuronide + quercetin-O-glucoside, which coeluted (Table 7). The polyphenol I and II compounds were also quantified at 365 nm, like flavonols. Apart from these two compounds, whose nature remains undetermined, only flavonol glycosides were found in the Tempranillo grape extracts at 365 nm, whereas no flavonol aglycones were recorded. Individual and total flavonol concentrations were higher in 2007 than in 2008, except for the coeluted quercetin-O-glucuronide + quercetin-O-glucoside. The effect of early defoliation on the flavonol concentration in berries slightly differed between seasons, although similar trends were observed. Basal leaf removal induced a significant increase of total flavonol concentrations in 2008, mainly due to the enhancement of quercetin glycosides, whereas no changes were detected in the previous season (Table 7). The major flavonol enrichment was found in Mec-PB berries, which were double in quercetin-glycosides and had 75% more flavonols than control berries. Some between-season discrepancies were found for the glucuronide and glucoside quercetin derivative concentrations in the grape extracts of control and defoliated vines. In 2007, defoliation induced a decrease in these flavonols, whereas in 2008, the opposite was observed. In general, neither the time nor the method of defoliation had an impact on the total flavonol content of Tempranillo berries except for quercetin-O-galactoside, which had higher concentrations in berries from mechanically defoliated vines (Table 6). Overall, the major flavonol in the Tempranillo grape extracts was myricetin-O-glucoside, followed by quercetin-O-glucoside and quercetin-O-glucuronide.
Nine anthocyanins were identified in the Tempranillo grape extracts (Table 8), the most abundant being the 3-glucosides of malvidin (MvGl), delphinidin (DpGl), and petunidin (PtGl). Like the flavonols, total anthocyanin concentrations in the berries were higher in 2007 than in 2008 for all treatments, including the control. Basal leaf removal led to significant enhancement of total anthocyanins (15 to 47%) in both seasons, although the increase was more pronounced in 2007. In 2007, grape extracts of defoliated vines were 340 to 600 mg malvidin-3-glucoside/kg grapes richer in anthocyanins than grape extracts of control vines, representing a 27 to 47% increase. Such an increase was also observed for eight individual anthocyanins in 2007 and four compounds in 2008, and the major effects corresponded to the 3-glucoside anthocyanins (Table 8).
The timing of leaf removal did not influence pigment concentrations except for petunidin-3-glucoside-p-coumarate (PtGlCm) in both years, and cyanidin-3-glucoside (CyGl), peonidin-3-glucoside (PnGl), and the p-coumarate of cyanidin-3-glucoside (CyGlCm) in 2008. For the accumulation of these specific anthocyanin compounds in berries, in general, prebloom defoliation was more effective than fruit-set leaf removal. In 2008, mechanical treatments led to higher DpGl, PtGl, and total anthocyanin (17%) concentrations in the berries than manual treatments. This outcome confirms the higher total anthocyanins in the berries (mg/g berry) of mechanical treatments compared to manual treatments in 2008 as determined by visible spectrophotometry. The concentration of total anthocyanins and total flavonols in berries determined by HPLC was plotted against the relative skin mass of berries in 2007 (Figure 1). Whereas a significant polynomial regression (R2 = 0.94, p < 0.05) was found for total anthocyanins, with a maximum at 11.0% of relative skin mass, no relationship was observed between this berry parameter and the concentration of total flavonols.
Principal component analysis.
To better distinguish factors that had a greater impact on the anthocyanin and phenol concentrations in berries among the different treatments, multivariate analysis by principal component analysis (PCA) was conducted on data from the three seasons (Figure 2). Yield, total leaf area per yield (TLA/yield), and berry mass data from seasons 2007 and 2008 included in the PCA have been reported in a companion study (Diago et al. 2010). PC1 and PC2 explained 76.3% and 15.5% of the variation in the data respectively. PC1 ranged from yield at the positive end of the axis to TLA/yield, at the negative end, whereas PC2 was mainly related to berry mass. Anthocyanins and total phenols were closely related to total soluble solids (Pearson correlation coefficients (r) were 0.91 and 0.75, respectively, at p < 0.05) in the upper left quadrant. Furthermore, all three fruit composition parameters were positively correlated to TLA/Y (r values from 0.73 to 0.84, at p < 0.05) and negatively to yield (r values from −0.72 to −0.87, at p < 0.05). Control and basal defoliation treatments were distributed along PC1, from the positive end, where control was placed, toward the negative end, where both Mec-PB and Mec-FS were located. Manual defoliation treatments were situated close to the origin of PC1, between the control and mechanical defoliation treatments.
Discussion
Leaf area and canopy characteristics.
Despite the substantial leaf area restriction imposed by manual and mechanical defoliation (65% of LA removed on average at the timing of defoliation), the vines successfully recovered due to intrinsic vine vigor and achieved similar total LA at harvest to those of control, non-defoliated vines. This canopy regrowth was mainly due to increased lateral formation and development, especially in manual treatments, as found elsewhere (Poni et al. 2006, Intrieri et al. 2008, Palliotti et al. 2011). The lowered lateral regrowth in mechanically defoliated vines compared to manually defoliated vines in 2007 and 2008 may be explained by the leaf-remover blowing effect on growing or incipient lateral tips at the time of defoliation, preventing further development during the season.
The observed increase in cluster exposure and canopy porosity seems to be a feasible and expected outcome, as the viticultural practice of defoliation causes a direct and generally long-lasting effect on canopy leaf density in the fruiting zone. However, the compensating growth of retained primary leaves and improved formation of laterals in response to prebloom and fruit-set defoliation might lead to partial cluster reshading. Similar findings were reported in Graciano and Carignan cultivars over two seasons (Tardaguila et al. 2010).
Berry tissues and fruit composition.
The increase in the relative skin mass in Tempranillo berries of vines following prebloom and fruit-set defoliation are in good agreement with previous results in Barbera (Poni et al. 2009) and Sangiovese (Palliotti et al. 2011), although the percentage of increase in the present work (0.7% on average) was lower than the 3.6% reported by Palliotti et al. (2011) over three consecutive seasons. The long-term adaptation mechanism of berry skin thickening as a response to precocious and prolonged cluster exposure caused by basal leaf removal, which was postulated by Poni et al. (2009), seems to be confirmed in Tempranillo berries.
As observed in the PCA biplot (Figure 2), the enhancement in must soluble solids reflected changes in the final total leaf area/yield ratio (TLA/Y) (values for the same vines monitored in the present work were reported in Diago et al. 2010). Another factor that might have influenced this maturity boost in defoliated vines is the “quality” of their foliage compared to control plants. Compensatory leaf recovery of the vines as a response to early leaf pulling by promoting both primary and lateral growth could have led to a “younger,” photosynthetically more active canopy (Palliotti et al. 2000) than the control. Increased leaf assimilation rates in leaves from lateral shoots in prebloom defoliated vines comparted to non-defoliated vines have been reported (Palliotti et al. 2011).
Anthocyanin and total phenol concentrations essentially increased in Tempranillo berries as a result of basal leaf removal; similar findings were reported in leaf-pulling trials performed with cultivars such as Sangiovese (Poni et al. 2006, Intrieri et al. 2008, Palliotti et al. 2011) and Barbera (Poni et al. 2009). The increase in TLA/yield (Diago et al. 2010) and the more advanced ripening stage in terms of total soluble solids in the defoliated vines of Tempranillo seems to have favored the phenolic increase in the berries, regardless of berry mass, as this parameter showed no correlation with anthocyanins and total phenols in the PCA. Additionally, the improvement of cluster exposure and canopy porosity might have also helped in promoting the accumulation of anthocyanins and phenols in the berries of defoliated vines, as there is general consensus on the positive effect of sunlight on phenol accumulation in berries, mainly flavonols (Haselgrove et al. 2000, Spayd et al. 2002, Tarara et al. 2008). Intrinsically related to cluster exposure is the temperature of the clusters across the season, since prolonged cluster temperatures over 30°C may result in anthocyanin synthesis inhibition (Haselgrove et al. 2000, Spayd et al. 2002). Although the cluster temperature was not measured in the present work, the removal of secondary shoots on count nodes and some shoots from noncount nodes at prebloom in Barbera vines did not induce differences in cluster temperature but led to increased sugars and anthocyanins in the berries (Bernizzoni et al. 2011).
Phenolic monomers.
There were some between-season discrepancies in glucuronide and glucoside quercetin derivative concentrations in the grape extracts of control and defoliated vines. In 2007, leaf pulling induced a decrease in these flavonols, whereas in 2008 the opposite was observed. Environmental factors may affect the production, transport, and accumulation of flavonoids in grapes in a typical bell-shaped manner, and they may improve the final amount of flavonoids only when present at optimal levels (Braidot et al. 2008). A decrease of flavonoid biosynthesis may occur when endogenous or exogenous factors, such as water and thermal stress, are limiting or excessive. The higher temperatures and lower rainfall in season 2007 compared to 2008 might have impacted the differential biosynthesis of these quercetin glycosides and the total concentrations of flavonols and anthocyanins in grape extracts between the two years. In this regard, the climatic conditions before veraison determine the total phenolics found in skin cells (Cadot et al. 2011), especially for flavonols, since their synthesis occurs from berry set until harvest and their concentration in berries is higher at the initial stages of berry development (Downey et al. 2003).
The increase in total and individual anthocyanins, except for malvidin-3-glucoside, determined by HPLC in berries of basal defoliated vines of Tempranillo is in good agreement with results for Sangiovese (Palliotti et al. 2011) and Tempranillo (Diago et al. 2012) wines, in which higher concentrations of individual anthocyanin glucosides, including malvidin-3-glucoside, and other glycosylated pigments were found in wines from prebloom and fruit-set defoliated vines than in control wines. Another factor potentially involved in higher anthocyanin concentration in berries of early defoliated vines is increased relative skin mass (Poni et al. 2009, Palliotti et al. 2011). Whereas the anthocyanin concentration in berries increased with relative skin mass, reaching a peak around 11%, this factor did not show any relationship with total flavonol concentration. Similarly, Downey et al. (2003) did not find increased flavonol in Shiraz berries of increased relative skin mass. In the skin, flavonols accumulate in the vacuoles of the epidermis and outer hypodermis (Flint et al. 1985), whereas anthocyanins accumulate in outer and inner layers of the hypodermis (Moskowitz and Hrazdina 1981). Whether the postulated adaptation mechanism of berry skin thickening (Poni et al. 2009) in prebloom and fruit-set defoliated berries occurs from increased cell division and/or cell enlargement, the more widespread localization of anthocyanins in cells of several layers of the hypodermis compared to the more localized flavonols would provide anthocyanins with more space to accumulate. Fontes et al. (2011) reported that the anthocyanin content in the outer hypodermal layer soon approaches saturation. Subcellular localization of flavonoids in the skins of grape berries is a complex phenomenon (Hardie et al. 1996) and additional cellular processes may also underlie the adaptation mechanism to precocious and prolonged light exposure (Poni et al. 2009). Whereas the thickness of the skin and the total number of skin cells containing phenolics do not change from veraison onward (Cadot et al. 2011), the type of cell changes, as does the morphology and distribution of the anthocyanin-containing vacuolar structures, mainly due to light exposure (Irani and Grotewold 2005).
Both light and temperature are capable of altering the anthocyanin acylation pattern and proportion (Tarara et al. 2008) as well as differentially affecting individual anthocyanin synthesis (Haselgrove et al. 2000). Cortell and Kennedy (2006) also suggested that ultraviolet radiation favored the synthesis of trihydroxylated anthocyanins, malvidin-3-glucoside, delphinidin-3-glucoside, and petunidin-3-glucoside versus the dihydroxylated pigments. In our work, these three anthocyanins were found in higher concentrations in 2007 (warm and sunny) than in 2008 (cool and rainy), whereas peonidin-3-glucoside and cyanidin-3-glucoside exhibited only minor differences.
The flavonol and anthocyanin profiles of the Tempranillo extracts recorded in this work coincide with descriptions of Tempranillo berries in previous studies (Gómez-Alonso et al. 2007). The absence of flavonol aglycones in berries has also been reported in Shiraz and Chardonnay (Downey et al. 2003) and Tempranillo (Gómez-Alonso et al. 2007). In this regard, Cheynier et al. (1998) suggested that flavonols exist in grape berries only as glycoside derivatives.
The enhancement of flavonoid concentrations in the berries in response to prebloom and fruit-set defoliation is a positive outcome. Anthocyanin concentration in the fruit is closely related to anthocyanin concentration in the wine and final wine quality. In this regard, increased concentrations of anthocyanins and flavonols were found in wines of Tempranillo prebloom and fruit-set defoliated vines over two consecutive seasons (Diago et al. 2012). Moreover, flavonols may play a major role as cofactors in copigmentation processes in wine (Boulton 2001) resulting in improved color density (Diago et al. 2012). In this regard, some of the most stable associations between anthocyanins and flavonols in the wines occur between the main anthocyanins such as malvidin-3-glucoside and the flavonol quercetin-O-glucoside (Downey et al. 2003).
Conclusions
In Tempranillo vines, the restriction in the carbohydrate supply caused by basal defoliation, either manual or mechanical at prebloom or fruit set, was effectively compensated by the plant, primarily through increased lateral regrowth. Nevertheless, cluster exposure and canopy porosity improved in defoliated vines, which contributed to obtaining more ripened fruit of increased anthocyanin and total phenol concentrations in grapes. Basal leaf removal favored the accumulation of flavonols and anthocyanins, and these phenolic compounds were related to increased leaf-to-fruit ratios and higher relative skin mass only for anthocyanins, but overall no differences between prebloom and fruit-set defoliation were found. This outcome is interesting, as a 15 to 20 day window from prebloom to fruit set is open for viticulturists who want to defoliate their vines for crop regulation and grape quality improvement.
Acknowledgments
The authors thank the Agencia de Desarrollo Económico de La Rioja ( ADER-2006-I-ID-00157) and the Ministerio de Ciencia e Innovación ( AGL2007-60378) for financial support.
The authors also thank the Agrupación de Bodegas Centenarias y Tradicionales de Rioja (ABC) and New Holland for their assistance and help; and Nicola Libelli and Alvaro Garrido for their help with berry dissection and the HPLC work, respectively.
- Received November 2011.
- Revision received March 2012.
- Accepted March 2012.
- Published online September 2012
- ©2012 by the American Society for Enology and Viticulture