Effect of Severity of Leaf and Crop Removal on Grape and Wine Composition of Merlot Vines in Hawke s Bay Vineyards

Effect of Severity of Leaf and Crop Removal on Grape and Wine Composition of Merlot Vines in Hawke s Bay Vineyards Petra D. King, 1 * Daniel J. McClellan, 1 and Richard E. Smart 2 Abstract: Manipulation of vine balance is widely practiced to enhance grape and wine quality. Reported benefits from crop thinning are ambiguous. Studies were undertaken over two growing seasons (2006 2008) in Hawke s Bay, a cool-climate region of New Zealand, to investigate the influence of basal leaf and crop removal on fruit and wine composition and the indices of vine balance associated with optimum quality. Three levels of basal leaf removal in the fruiting zone were used. Crop removal achieved average reductions of 15% with moderate treatment and 35% with severe treatment below the average nonthinned yield of 17.5 t/ha. Treatments were undertaken at the preveraison growth stage on Merlot vines. Leaf removal had no effect on fruit composition, but levels of total anthocyanins and the flavonol quercetin-3-glucoside were enhanced in the wines. In contrast, crop removal increased sugar concentration and decreased titratable acidity of the must. While crop removal had no effect on wine anthocyanins, the proportions of anthocyanins as malvidin-3-glucosides and total phenolics in the wines were significantly increased. Relative to accepted indices of vine balance, the study vines were unbalanced with excessive shoot growth and dense canopies. A crop reduction of ~6.0 t/ha brought the capacity of the exposed leaf area into balance with that required to ripen fruit to optimum maturity and produce high-quality wines. Key words: leaf and crop removal, Merlot, grape and wine composition 1 Faculty of Applied Sciences, Business and Computing, Eastern Institute of Technology, Hawke s Bay, 4142 New Zealand; and 2 Smart Viticulture, 31 North Corner, Newlyn near Penzance TR185JG, Cornwall, UK. *Corresponding author (email: pking@eit.ac.nz; fax:+64 6974 8910) Acknowledgments: The authors gratefully acknowledge financial support for this research from the Business Links Fund of the New Zealand Tertiary Education Commission. The authors thank Karen Ball for technical assistance. Manuscript submitted Feb 2012, revised May 2012, accepted Jun 2012. Publication costs of this article defrayed in part by page fees. Copyright 2012 by the American Society for Enology and Viticulture. All rights reserved. doi: 10.5344/ajev.2012.12020 Merlot is considered a problem winegrape variety in cool-climate regions like Hawke s Bay, New Zealand (mean January temperature 19.5 C). The variety has large clusters leading to a potential of excess crop and lack of ripening. Further, Merlot has large leaves and can be vigorous in many situations, and so canopy shading is common. Management practices commonly used to overcome these problems involve manipulation of canopy architecture through trellising and training systems, leaf removal to enhance leaf and fruit exposure to light, and crop removal to balance the ability of the vine leaf area to ripen fruit. In Hawke s Bay vineyards, considerable leaf and crop removal is undertaken each season as standard management procedure between set and harvest. Leaf removal in the fruit zone will increase cluster exposure to light but does little to improve the leaf shading occurring within the canopy overall. Conversion to a divided canopy is another alternative, yet is rarely used in Hawke s Bay. There is a considerable body of research on leaf and crop removal strategies for many cultivars and growing conditions and their influence on grape composition and wine quality. Results report both positive effects and neutral or insignificant outcomes. The benefits accruing from leaf removal through reducing shading and improving fruit exposure are relatively unequivocal. Levels of anthocyanins and phenolics in Cabernet Sauvignon berries and wine were enhanced by leaf removal over the period postbloom to veraison (Hunter et al. 1995, Zoecklein et al. 1996, Joscelyne et al. 2007). Positive effects with Cabernet Sauvignon were found on levels of grape glycosides (Zoecklein et al. 1996) and wine norisoprenoid (Lee et al. 2007). With Merlot, leaf removal at veraison enhanced levels of phenolics and color density of grapes (Mazza et al. 1999), and total Merlot skin monomeric anthocyanins and the flavonol quercetin-3-glucoside increased with cluster exposure (Spayd et al. 2002). Studies on Shiraz showed that leaf removal increased levels of quercetin-3 glucoside in berries (Haselgrove et al. 2000), while cluster shading decreased wine color and anthocyanin and tannin levels and altered sensory attributes, particularly β-damascenone glycosides (Ristic et al. 2007). In Pinot noir, cluster shading decreased levels of skin tannins (Cortell and Kennedy 2006), but in ripe Shiraz there was no significant effect of shading on either skin or seed tannin levels (Downey et al. 2004), although both studies noted decreases in skin flavonols from shading. Reductions in berry anthocyanins were reported in some seasons from fruit shading and anthocyanin composition was also significantly altered by shading (Cortell and Kennedy 2006, Downey et al. 2004). Overall, increased fruit exposure to sunlight benefits color and levels of total skin monomeric anthocyanins of red-fruited grape cultivars, particularly in cooler temperature growing regions where the negative effect of excessive berry temperature on grape composition is not an issue. There are numerous examples of positive effects from cropload reduction on grape and wine quality for a wide range of cultivars and include advanced maturity, decreased acidity, 500

Influence of Leaf and Crop Removal on Merlot 501 and increased anthocyanins and phenolics in Cabernet Sauvignon (Petrie and Clingeleffer 2006). Cluster thinning increased sugar and enhanced wine color, aroma, and flavor in Pinot noir (Reynolds et al. 1996, 2007), while concentrations of soluble solids, skin anthocyanins, and flavonoids were increased by cluster thinning of Nebbiolo grapes (Guidoni et al. 2002). Decreasing yield per vine significantly increased Shiraz wine color and phenolics and improved levels of metabolites which contribute to wine aroma and flavor (Wolf et al. 2003). However, cluster removal has produced neutral or negative effects on wine composition. Few differences were found in Cabernet Sauvignon wines from cluster thinning at veraison (Chapman et al. 2004), although a yield reduction by varying winter bud numbers influenced astringency and vegetative flavors of wine, indicating that the time of crop-load manipulation was important. Cluster thinning of Cabernet Sauvignon in Napa over three seasons also had minimal effects on must sugar, ph, titratable acidity, free amino N, total N, and wine color, phenolics, and aroma (Ough and Nagaoka 1984), with seasonal effects greater than crop-load effects. Thinning clusters on Cabernet Sauvignon at veraison reduced yield and induced early ripening of grapes but not higher grape quality based on TA, ph, and levels of skin anthocyanins. (Nuzzo and Matthews 2006). Studies of cluster thinning on Cabernet Sauvignon in Washington State had no effect on vine growth, berry weight, and fruit composition (Keller et al. 2005), and the authors concluded that cluster thinning should be considered as a band aid management option when exceptional yield potential coincides with a cool growing season. In terms of a rationale for these conflicting results, it is accepted that vine balance between vegetative and fruit growth is critical to the ability of the vine to properly ripen fruit and so produce high-quality wine. There is a considerable body of research investigating the effects of crop load and vine balance on fruit composition, and a range of indices of vine balance are reported. Critical values of such ratios are proposed to adequately ripen fruit and promote high-quality wine (Smart and Robinson 1991, Jackson and Lombard 1993, Howell 2001, Kliewer and Dokoozlian 2005, Nuzzo and Matthews 2006). These indices of vine balance are not always presented in the literature on crop-removal responses, making it difficult to draw conclusions on relationships among vine balance, grape ripening and composition, and wine quality. The objectives of the present study were to evaluate the effects of a range of leaf and crop removal levels on Merlot vine growth, fruit ripening, and grape and wine composition and to interpret the outcomes with reference to assessments of vine balance. Materials and Methods The study investigated the influence of canopy manipulation procedures on grape and wine quality over the 2006 2007 and 2007 2008 growing seasons in Hawke s Bay in the North Island of New Zealand. (Growing seasons subsequently identified by the year of harvest, 2007 or 2008.) Study plots were selected in a commercial Merlot vineyard (clone 181, 3309 rootstock) within the Bridge Pa Triangle subregion. The vineyard was planted in 1999 at 1.6 m x 2.4 m spacing on Takapau sandy loam on gravels and was drip irrigated. Growth and overall canopy vigor of the vines was even within the block. A replicate was a panel of five vines, and two adjacent measurement vines within the panel were selected on the basis of uniformity of trunk circumference (12.6 ± 1.2 cm). The vines surrounding the measurement vines in the panel acted as buffers. There were four replicates of each treatment arranged in a randomized block design. Vines were winter pruned to two canes and ~25 buds. Shoots were trained to a vertical shoot-positioned (VSP) system. In both seasons, nil, moderate, and severe treatments were established to give a range of levels of fruit exposure to light through leaf removal (LR) in the fruiting zone (main and lateral leaves) and through crop removal (CR) of clusters. The severe LR treatment produced high levels of exposure to sunlight. Cluster counts were made on 24 measurement vines selected randomly from the trial area to establish mean cluster number per vine. For the nil CR treatment, mean numbers were 39.7 (SD 5.7) and 45.5 (SD 4.6) clusters per unthinned vines in 2007 and 2008, respectively. For the moderate and severe CR treatments in both years, crop removal of apical clusters only was undertaken in late January at E L stage 33 (berries still hard and green). In 2007, 10 and 20 clusters per vine were removed in the moderate and severe CR treatments, respectively. In 2008, 14 and 27 clusters per vine were removed in the moderate and severe CR treatments, respectively. In both seasons LR was performed within two days of CR, with the two basal leaves removed from shoots on moderate LR vines and all leaves removed to the second cluster plus the leaves from a further two nodes from shoots on severe LR vines. All other canopy manipulations and viticulture inputs were undertaken according to typical vineyard commercial practices and performed uniformly across all treatments. Canopy density was measured using point quadrant analysis (PQA) (Smart and Robinson 1991) in early February 2007 and late January 2008, immediately after leaf removal. PQA was undertaken across two measurement vines in each of three reps, selected randomly from the four reps of each treatment, so that the canopy areas of six vines were assessed in each LR treatment. The number of leaves intercepted by 150 insertions over a 3 m length by a slender rod pushed through the canopy was estimated at each of five heights above the cordon wire within the canopy: at 75 mm above the cordon wire in the fruit zone and at 275, 410, 510, and 690 mm. This provided the estimate of average leaf layer number (LLN). Vine or canopy leaf area A T (m 2 ) was calculated from the equation A T = LLN[Wh/k], where W and h represent the width and height of the canopy (m), respectively, and the parameter k (0.6) is the canopy extinction coefficient, which represents the ratio of the horizontally projected shadow area to the total leaf area (one-sided) of the vine canopy (Green 2008). Canopy surface area was calculated as per the methodology detailed by Smart and Robinson (1991) based on 2.4 m vine spacing, a 1.3 m high canopy at trimming, and 0.4 m canopy width. Photosynthetically active radiation (PAR) within the fruiting zone at each leaf removal treatment was measured within

502 King et al. one hour of solar noon under cloudless conditions on 25 Feb 2008 using an Accupar LP-80 ceptometer (Decagon Devices, Pullman, WA). Rows were east-west oriented so that the northern side of the row received more direct sunshine than the southern side. Measurements were made along the five vines within the plot, on both sides of the canopy, in the fruiting zone so that 8 m length of canopy was measured per plot. The ceptometer probe was held in the horizontal plane against the clusters with the light diodes in a vertical plane facing into the interrow area. The probe was shaded by leaves in the nil LR treatment. At the same time as the PAR flux density measurements were made within the canopy, ambient PAR was recorded from the midrow at the same height as the top of the vine canopy. Canopy temperatures were continuously measured within the block in the fruit zone using e Temp Thermocron temperature loggers (Rollex Group, Victoria, Australia), which recorded data every 30 min. The loggers were shielded within well-ventilated polystyrene coffee cups suspended upside down and attached to the cordon wire. Harvest was undertaken on 16 Apr 2007 and 4 Apr 2008 at the commercial harvest times. Grapes were cool-stored at harvest until detailed yield, cluster, and berry parameters were measured. Samples of 100 berries were taken from 20 clusters (5 berries/cluster) randomly selected from each replicate, and berry anthocyanins were determined based on the AWRI industry standard methodology using spectrophotometric methods (Iland et al. 1996, 2000). Wines were made from each of the crop removal and leaf removal treatments and these were replicated four times. Grapes from the two measurement vines in each replicate were destemmed and crushed using an Enoveneta crusher destemmer (Enoveneta S.p.A., Padova, Italy). Samples of must were taken immediately after crushing and total soluble solids (TSS) determined using a digital refractometer (Index Instruments, Cambridge, UK). Must titratable acidity (TA) and ph were measured using an autoanalyzer (785 DMP Titrino; Metrohm, Riverview, FL). Additions of SO 2 at a rate of 30 mg/kg were made using post-crush weight. The addition of sucrose was made to raise Brix to the maximum attained within the trial. Fermentations were in food-grade 5-L plastic containers. Must was warmed to 15 C and inoculated with Lalvin EC1118 yeast (Lallemand Australia, Edwardstown, South Australia) at a rate of 250 mg/ kg must. Di-ammonium phosphate (DAP) was added at 300 mg/l. Superfood (Pacific Rim Oenology, Blenheim, NZ) a yeast nutrient, was added at 250 mg/l of must. Fermentations were undertaken in a room maintained at 26 C. Wine was fermented in contact with skins for 18 days, with the cap plunged daily. Wine was pressed in a Pillan 20- L, 4 Bar, hydraulic press (Enotecnica Pillan, Vicenza, Italy). Marc was pressed to give 70% yield of post-crush weight in liters. After pressing, 60 mg/l SO 2 was added. Wines were allowed to settle for two days, racked, and then returned to 0 C storage for four weeks before a final racking. Wine was sparged with NO 2 gas for 5 min before bottling. Bottles were sparged with N 2 before filling by siphon. At bottling, samples were taken from each wine and analyzed using high-performance liquid chromatography (HPLC) following the methods of Harbertson et al. (2003). All data were analyzed for statistical significance by twoway ANOVA using Minitab 15 statistical software (Minitab Inc., State College, PA). Results Canopy measurements. Point quadrant analysis (PQA) was determined in the fruiting zone and for the total canopy for each treatment in 2007 and 2008, after the leaf removal treatments had been imposed. Data are presented (Table 1) from the nil CR treatment only for each level of leaf removal. LLN for the total canopy area was averaged over five heights and presented as average LLN. Both the nil and moderate LR treatments were shaded in the fruit zone, with the severe LR treatment considerably enhancing the fruit exposure to light. Severe LR removed ~2 m 2 leaves /vine from the fruiting zone. The ratios of canopy leaf area to surface area (LA:SA) confirm the undesirably high canopy densities present in all these Merlot vines, regardless of leaf removal treatments and high levels of leaf shading present. Canopy PAR levels. Light intensity as PAR flux density was measured in a horizontal plane with sensors aligned vertically in the fruit zone on 25 Feb 2008 (Table 2). The PAR levels reaching the fruit zone on the sunny northern side of the row were significantly lower where leaves were not removed. It is generally accepted that the light compensation point for grapevine leaves for photosynthesis is between 15 and 30 μmol m 2 /s (Smart 1986), indicating that PAR levels Table 1 Effect of leaf removal (LR) treatment on Merlot vine canopy characteristics in February 2007 and 2008, Hawke s Bay, NZ. LR treatment Fruit zone Total canopy Gaps Interior leaves Interior clusters Leaf area Surface (%) LLN a (%) (%) LLN a (m 2 /vine) area 2007 2008 2007 2008 2007 2008 2007 2008 2007 2008 2007 2008 2007 2008 Nil 0a b 0a 2.5c 3.2c 31.5b 41.8c 69.4b 71.5c 3.1 3.2 12.89 13.2 4.6 4.7 Moderate 11.1a 8.3a 1.6b 1.6b 21.9b 17.3b 37.7ab 29.9b 2.9 2.8 12.0 11.7 4.3 4.2 Severe 62.2b 45.0b 0.3a 0.3a 0a 2.3a 3.7a 0a 2.6 2.6 10.9 10.7 3.9 3.8 Optimum values c 20 40% <1 1.5 <10% <40% <1 1.5 <1.5 a Leaf layer number. b Values with the same letter do not differ significantly. c Optimum values according to Smart and Robinson (1991).

Influence of Leaf and Crop Removal on Merlot 503 were very low for all treatments apart from severe LR on the sunny side. Light intensity was also influenced by canopy side effects. PAR levels in the fruit zone at the severe LR were ~13% lower on the southern side than the northern side, and only 2% of PAR in the fruit zone. Harvest data. Yield data for the 2007 and 2008 harvest seasons confirm that significant crop-load differences were achieved by cluster removal and that the crop loads per treatment were similar in the two seasons (Table 3). Overall, yield per vine was significantly reduced by severe CR at each level of leaf removal, although cluster numbers on the moderate vines did not always differ significantly from the nil vines. There were also some differences in cluster numbers per treatment in the two seasons. These results demonstrate the difficulty in consistently achieving similar crop levels when a standard percentage of clusters is removed from each vine. Cluster weights were significantly increased overall as expected for most severe CR treatments, regardless of LR treatment. Table 2 Effect of leaf removal (LR) treatment and canopy side on photosynthetically active radiation (PAR) values. Canopy light reading position/ treatment Ambient PAR (μmol m 2 /s) PAR fruiting zone (μmol m 2 /s) % Ambient PAR measured % PAR reduction relative to severe Sunny side north Nil LR 1903 16.7 a a 0.8 a 94.7 Moderate LR 1927 56.9 ab 3.0 ab 81.9 Severe LR 1859 315.0 b 16.9 b 0 Shaded side south Nil LR 1926 18.3 a 1.0 a 54.5 Moderate LR 1972 31.0 a 1.6 a 29.6 Severe LR 1949 40.2 a 2.1 a 0 a Values with the same letter down the column do not differ significantly (p > 0.05). Table 3 Effect leaf removal (LR) and crop removal (CR) treatments on Merlot yields at harvest in the 2007 and 2008 seasons. Treatment Grape yield (kg/vine) Clusters/ vine Cluster wt (g) 2007 2008 2007 2008 2007 2008 Nil LR Nil CR 6.5 b a 6.7 b a 48.0 b 48.8 b 135.0 a b 137.3 ab b Moderate CR 5.6 b 5.5 b 37.0 ab 39.8 a 151.0 ab 138.8 ab Severe CR 4.7 a 3.6 a 31.6 a 21.9 a 148.7 ab 164.4 b Moderate LR Nil CR 6.6 b 6.9 b 43.4 b 49.6 c 152.1 ab 139.1 ab Moderate CR 6.1 b 6.0 b 40.4 b 40.8 b 151.0 ab 147.1 ab Severe CR 4.0 a 4.2 a 25.5 a 25.3 a 156.9 ab 166.0 b Severe LR Nil CR 7.0 b 6.1 b 48.9 c 43.5 b 143.1 a 140.2 ab Moderate CR 4.9 b 5.6 ab 34.3 b 35.7 a 142.9 a 156.9 b Severe CR 4.4 a 4.7 a 24.6 a 30.3 a 178.9 b 155.1 b a Values with the same letter down a column do not differ significantly (p > 0.05). b Values with the same letter down a column regardless of level of leaf removal do not differ significantly (p > 0.05). Pruning data and vine balance. The pruning data at winter 2007 and 2008 are shown together with some indices of vine balance (Table 4). Mean cane weight and pruning weight per meter row length did not differ significantly among the treatments in both seasons (data not shown), likely a reflection of constant shoot trimming to a canopy height of ~1200 mm and hedged sides. Since there was no effect of treatment on pruning weight, yield:pruning weight ratios reflected yield differences. With respect to vegetative growth, pruning weight and mean cane weight values were up to three-fold those cited as optimal (Smart and Robinson 1991), indicating that the vines were overvigorous (Table 5). The vines were also unbalanced in favor of vegetative growth considering leaf area:fruit weight values and also canopy surface area:fruit weight, both with less than optimal values. The indices also indicate substantial canopy shading, with higher than optimal values of leaf area per meter canopy length and leaf area:canopy surface area. Fruit ripeness. Leaf removal treatments had no significant effect on fruit composition in terms of TSS, TA, and ph at harvest in either season (Table 6). Comparison of temperatures within the trial vine canopies between the two seasons (data not presented) show 2008 monthly maxima higher and minima lower than 2007. The cumulative growing degree Table 4 Effect of leaf removal (LR) and crop removal (CR) treatments on vine balance of Merlot vines in 2007 and 2008. Yield: pruning wt (kg) Leaf area: yield (m 2 /kg) Fruit yield: canopy surface area (kg/m 2 ) Treatment 2007 2008 2007 2008 2007 2008 Nil LR Nil CR 4.0 b a 4.5 c 2.0 a 1.6 a 1.6 b 1.6 b Moderate CR 2.9 a 3.9 b 2.3 a 1.9 a 1.4 a 1.3 b Severe CR 2.4 a 2.3 a 2.7 b 2.7 b 1.1 a 0.9 a Moderate LR Nil CR 4.2 b 5.2 c 1.8 a 1.4 a 1.6 b 1.7 b Moderate CR 3.0 a 3.6 b 2.0 a 1.6 a 1.5 b 1.4 b Severe CR 2.3 a 2.6 a 3.2 b 2.3 b 1.0 a 1.0 a Severe LR Nil CR 4.0 b 4.5 b 1.5 a 1.5 a 1.7 b 1.5 b Moderate CR 2.9 a 3.6 a 2.2 a 1.6 a 1.2 a 1.3 b Severe CR 2.6 a 3.6 a 2.5 b 2.0 b 1.1 a 1.1 a a Values with the same letter do not differ significantly (p > 0.05). Index Table 5 Vine balance indices of Merlot vines in the present study and optimal values. Present range Optimal range a Mean cane wt (g) 60 70 20 40 Pruning wt/m canopy (kg/m) 0.86 0.3 0.6 Yield/pruning wt (kg/kg) 2.5 5.0 5 10 Leaf area/yield (m 2 /kg) 1.5 3.0 0.8 1.2 Leaf area/m canopy length (m 2 /m) 6 8 2 5 Leaf area/canopy surface area (m 2 /m 2 ) 2.2 3.1 <1.5 a Values that optimize grape yield and wine quality and reduce disease incidence (Smart and Robinson 1991).

504 King et al. days base 10 C (GDD) over the period 1 Oct to 31 Mar were identical over the two seasons and were ~50 GDD higher between 1 Oct and harvest in 2007 due to the later harvest time. Overall ripeness levels at the different treatments were similar in both seasons. Cluster removal enhanced ripening in terms of increased TSS and decreased TA levels in both years, but significantly increased ph for the 2007 harvest Table 6 Effect of leaf removal (LR) and crop removal (CR) treatments on Merlot must composition at harvest in 2007 and 2008. TSS (Brix) ph Titratable acidity (g/l) Treatment 2007 2008 2007 2008 2007 2008 Nil LR Nil CR 23.2 a a 24.1 ab 3.45 a 3.53 a 5.49 b 4.96 a Moderate CR 24.2 b 24.2 ab 3.56 b 3.54 a 4.70 b 4.93 a Severe CR 24.8 b 24.6 b 3.5 c 4.31 a 4.53 a 4.64 a Moderate LR Nil CR 22.7 a 23.4 ab 3.47 a 3.47 a 5.3 b 5.1 b Moderate CR 23.8 b 24.1 ab 3.40 a 3.52 a 4.83 a 4.80 ab Severe CR 24.4 b 24.7 b 3.62 b 3.59 a 3.60 a 4.26 ab Severe LR Nil CR 22.8 a 23.2 a 3.43 a 3.51 a 5.41 b 5.24 b Moderate CR 24.3 b 23.52 ab 3.51 b 3.53 a 4.65 a 4.79 ab Severe CR 24.7 b 23.9 ab 3.57 c 3.56 a 4.61 a 4.60 ab a Values with the same letter do not differ significantly (p > 0.05). only. There was less difference in ripeness between moderate and severe CR treatments than between nil and moderate. Chemical analysis of wines. Total and individual monomeric anthocyanins were determined in 2007 and 2008 wines (Table 7). Overall levels of anthocyanins were ~20% lower in 2008 compared to 2007, despite similar TSS levels. Cluster removal had no significant effect on anthocyanin levels in either season. In 2007, severe LR significantly increased anthocyanin compared to nil or moderate LR. In 2008, however, anthocyanin levels were significantly different between nil and moderate LR treatments, but moderate and severe treatments did not differ. Overall, there were fewer significant differences in glucoside levels in 2008 wines than in 2007 (Table 7). There were differences in the proportions of malvidin compared to the other glucosides; cluster removal significantly reduced malvidin proportions at each LR treatment in the 2007 vintage. The proportions of delphinidin, petunidin, and peonidin- 3-glucoside were also increased by CR treatment irrespective of LR treatment. The proportions of cyanidin appeared to be increased by moderate and severe CR. In contrast, LR had no significant influence on the proportions of monomer glucosides, except moderate and severe LR treatments increased the proportions of peonidin and the severe LR treatment increased cyanidin. The 2007 wines were analyzed for quercetin-3-glucoside concentration (Table 8). The severe LR treatment resulted Table 7 Effect of leaf removal (LR) and crop removal (CR) treatments on total anthocyanins and proportions of 3-monomeric glucocides in Merlot wines in 2007 and 2008. Year/ treatment Total anthocyanins (mg/l) a 3-Glucosides as % total anthocyanins b Malvidin Delphinidin Petunidin Peonidin Cyanidin C C C C L C L 2007 nil LR Nil CR 432.5 a c 69.8 b 10.0 a 12.4 a 6.5 a a 0.9 a a Moderate CR 438.8 a 66.5 a 11.7 b 13.9 b 6.7 ab a 1.2 a a Severe CR 442.3 a 62.1 a 13.3 b 15.0 b 8.0 b a 1.6 a a 2007 moderate LR Nil CR 439.5 a 70.5 b 8.5 a 13.1 a 7.4 a b 0.5 b a Moderate CR 472.9 a 64.4 a 12.1 b 13.8 b 8.2 ab b 1.5 b a Severe CR 467.7 a 59.6 a 14.1 b 15.1 b 9.0 b b 2.2 b a 2007 severe LR Nil CR 468.2 b 69.0 b 10.3 a 12.2 a 7.1 a b 1.5 b b Moderate CR 531.5 b 64.6 a 12.1 b 13.6 b 8.0 ab b 1.8 b b Severe CR 517.2 b 60.6 a 13.1 b 15.0 b 8.9 b b 2.4 b b 2008 nil LR Nil CR 362.6 a 64.1 a 12.8 a 13.9 a 7.5 a a 1.7 a a Moderate CR 341.5 a 61.7 a 13.1 a 14.4 a 8.7 a a 2.0 a a Severe CR 338.7 a 55.6 a 16.2 a 15.8 a 9.8 a a 2.5 a a 2008 moderate LR Nil CR 357.5 b 64.3 a 12.3 a 13.7 a 7.9 a a 1.8 a a Moderate CR 386.6 b 61.1 a 13.8 a 14.5 a 8.4 a a 2.1 a a Severe CR 387.4 b 56.6 a 15.2 a 14.8 a 9.0 a a 2.4 a a 2008 severe LR Nil CR 356.8 ab 61.2 a 13.4 a 14.4 a 8.9 a a 2.2 a a Moderate CR 373.1 ab 63.5 a 12.4 a 13.8 a 8.3 a a 2.0 a a Severe CR 363.6 ab 59.1 a 14.1 a 14.8 a 9.5 a a 2.4 a a a Anthocyanins as malvidin equivalents. b C: significance of CR treatments; L: significance of LR treatments; there were no significant differences among LR treatments for malvidin, delphinidin, and petunidin. c Values with the same letter down the column within years do not differ significantly (p > 0.05).

Influence of Leaf and Crop Removal on Merlot 505 in significantly (p < 0.01) higher quercetin levels than the nil or moderate LR treatment. In terms of total phenolics in the wines for both years, the severe CR treatment increased levels of phenolics relative to the nil CR treatment. Moderate CR relative to nil CR significantly increased total phenolics in 2007 only. In both years, leaf removal had no effect on phenolics. Table 8 Effect of leaf removal (LR) and crop removal (CR) treatments on levels of quercetin-3-glucoside in Merlot wines in 2007 and total phenolics in 2007 and 2008. Treatment Quercetin-3- glucoside (mg/l) Cluster effects Phenolics expressed as catechin equivalents (mg/l) Leaf effects 2007 2008 NIL LR Nil CR 13.2 a a a b 4243.6 a c 4862.1 a Moderate CR 11.9 a a 4548.5 b 5013.3 ab Severe CR 11.4 a a 4958.2 b 5376.3 b Moderate LR Nil CR 14.7 a a 4226.8 a 4782.2 a Moderate CR 15.2 a a 4547.0 b 5273.3 ab Severe CR 11.0 a a 4959.8 b 5556.5 b Severe LR Nil CR 20.4 a b 4322.2 a 4893.1 a Moderate CR 21.8 a b 5158.4 b 5109.4 ab Severe CR 20.5 a b 5213.8 b 5286.0 b a No significant differences found. b Values with the same letter down the column do not differ significantly. c Values with the same letter down the column within levels of LR do not differ significantly. Discussion The results confirm the trial site as consisting of vigorous vines that were highly vegetative and with dense, shaded canopies. LLN values were greater than two-fold the optimum value, and levels of interior leaves were up to four-fold the optimum value, which indicated substantial shading. Measured light levels in the fruit zone indicated that inner leaves were heavily shaded and received PAR at about light compensation values. Severe LR treatment in the fruiting zone reduced canopy leaf area by ~15 to 20%, or ~2 m 2 /vine, resulting in enhanced PAR flux density measured at the fruit zone. There were no effects of LR treatment on ripeness levels. As expected, the severe CR treatment decreased yield per vine and increased cluster weight by ~35% and up to 25% in 2007 and 2008, respectively. Ripeness at harvest was significantly enhanced in terms of increased TSS and ph and decreased TA, particularly in 2007, with the ripest fruit produced at the lowest crop load. These outcomes for Merlot vines are similar to findings with other red cultivars. Enhanced ripening rates in terms of TSS, TA, and ph have been reported with Pinot noir (Reynolds et al. 1996), Cabernet Sauvignon (Nuzzo and Matthews 2006, Petrie and Clingeleffer 2006), and Nebbiolo (Guidoni et al. 2008) The moderate CR treatment was just as effective as the severe CR treatment in enhancing ripeness. Moderate CR vines carried on average 1.3 kg more fruit per vine than the severe CR vines, with similar levels of ripeness. These results have economic implications for vineyard operators. In terms of specific effects on wine composition, CR treatments had no influence on anthocyanin levels. The absence of an anthocyanin response to crop removal is at variance with findings with Cabernet Sauvignon (Petrie and Clingeleffer 2006) and Nebbiolo (Guidoni et al. 2002), where reducing yield increased anthocyanin levels. Conversely, leaf removal in the fruit zone produced significant increases of ~15% in total anthocyanin levels. In Cabernet Sauvignon and Shiraz wines, fruit-zone shading reduced both total phenolics and total anthocyanins by 50 to 60% (Joscelyne et al. 2007). Overall, there were fewer significant differences in wine glucoside levels in the 2008 vintage than in 2007, which may be related to the lower total anthocyanins in 2008 vintage wines. There were differences in the proportions of malvidin- 3-glucoside compared to the other glucosides: crop removal significantly reduced malvidin at each level of leaf removal in the 2007 vintage, whereas reducing crop load resulted in an increase in the other monomerics measured. The proportions of the different glucosides were not influenced by anthocyanin levels. Concentrations of cyanidin, peonidin, and petunidin-3-glucosides were found to increase in Nebbiolo grapes that were cluster-thinned one week after bloom (Guidoni et al. 2002) relative to unthinned grapes. In contrast to the Merlot results presented here, concentrations of malvidin-3-glucoside and acylated anthocyanins in Nebbiolo were unaffected by crop removal. It is possible that the timing of crop removal has an influence on the outcomes. Leaf removal had a greater influence on the proportions of the monomeric anthocyanins in the Merlot wines in 2007 than in 2008. The earlier harvest date in 2008, in which total anthocyanins were lower than in 2007, may have contributed to this result. Cluster shading in the absence of leaf removal appeared to reduce the proportion of the delphinidin, petunidin, peonidin, and cyanidin-3-glucosides, although the shading effects were significant for only the latter two glucosides. Unlike the crop removal effects on the proportion of anthocyanins present in the malvidin form, the malvidin proportion appeared unaffected by leaf removal. These Merlot results have both similarities and variance from other reported findings. In one study, shading of Merlot clusters had inconsistent effects on the proportions of the individual glucosides between two seasons studied (Spayd et al. 2002), although there was a lower proportion of delphinidin, cyanidin, petunidin, and peonidin and a higher proportion of malvidin in shaded clusters in a single year only. Shading Pinot noir grapes reduced the proportions of delphinidin, cyanidin, petunidin, and malvidin relative to unshaded, but there was a large increase in peonidin glucosides (Cortell and Kennedy 2006). In contrast, bunch shading increased the proportion of malvidin, petunidin, and delphinidin-3-glucosides and of the coumarylated form of malvidin in Shiraz grapes (Haselgrove et al. 2000, Downey et al. 2006). Varietal differences may contribute to these effects. Increased light exposure of fruit with the severe LR treatment increased quercetin levels in the wines. Similar light

506 King et al. effects have been reported with Merlot (Spayd et al. 2002), Pinot noir (Price et al. 1995), and Shiraz (Haselgrove et al. 2000, Downey et al. 2004), while Cortell and Kennedy (2006) found model extractions had a lower concentration of flavonols in shaded treatments, similar to the nil LR treatment reported here. It has been hypothesized that quercetin-3-glucoside may have a role in copigmentation of wine as a cofactor (Price et al. 1995), and thus play a role in color stabilization. According to one report, color enhancement of wines from copigmentation was typically 4 to 6 times greater than that expected from the pigment concentration alone (Boulton 2001). Quercetin glycoside cofactors have been shown to cause significant shifts in the wavelength at which maximum absorbance occurred (Asen et al. 1972), increasing absorbance by up to 200%. As such, enhanced light exposure from leaf removal is likely to have a positive influence on wine color through the elevation of cofactor levels. In both years there was no benefit to phenolic levels from leaf removal at the E L 33 stage of development, indicating no effect on the development of phenolics from fruit exposure to light. This result is at variance with a report for Merlot (Mazza et al. 1999), where removal of basal leaves at bloom increased total phenolics by 15 to 18% over two seasons. The authors also reported similar benefits for Cabernet franc and Pinot noir, indicating perhaps that timing of leaf removal may be critical. This conclusion is validated by the findings of Cortell and Kennedy (2006) who reported that shading Pinot noir clusters from the E L 27 stage (berries ~2 mm diam) resulted in a substantial decrease in skin proanthocyanidins. Reducing Merlot crop load through crop removal at veraison in this current study significantly increased total phenolics in the wines in both seasons. Such an effect was also shown for Merlot, Cabernet franc, and Pinot noir wines (Mazza et al. (1999), although crop removal at bloom resulted in higher levels than at veraison. However, the authors found that leaf removal had a greater effect on phenolics than crop removal. Enhanced levels of phenolics were also reported for Cabernet Sauvignon following crop removal (Petrie and Clingeleffer 2006). There are a range of research-validated indices or indicators of vine balance with critical values proposed to adequately ripen fruit and promote high-quality wine. While there is general agreement as to the level of crop load at which fruit ripening and quality are optimized, factors such as climate, trellis training system, and whether irrigated or dry-land production is adopted appear to influence these balance indicators. The indices found here confirm the vines were unbalanced, with the indices such as mean cane weight, yield:pruning weight, leaf layer number, and leaf area:fruit yield all indicating excessive cane growth and vigorous and dense canopies with considerable shading. The effects of the moderate and severe CR treatments were to remove ~1 to 1.2 kg and 2.2 to 2.4 kg of fruit per vine, respectively. Fruit ripeness was enhanced by crop removal but the level of ripeness was unaffected by the level of crop removal. Severe and moderate CR treatments reduced yield by 36% and 15%, respectively, on average, and resulted in a reduction of the fruit yield:canopy surface area ratio to within the optimum range and the vines into balance. Because of the dense canopy, the amount of surface area was inadequate to ripen the total crop load without severe CR treatment. Without this crop reduction, ripening rate was slowed, maturity was delayed, and wine composition affected. In contrast, the crop load as indicated by yield:pruning weight ratio and the leaf area:yield ratio was decreased and increased, respectively, relative to optimal values (Smart and Robinson 1991, Kliewer and Dokoozlian 2005). However, the values found here for Merlot are very similar to those found for dry-farmed Cabernet Sauvignon in the Napa Valley (Nuzzo and Matthews 2006), where yield reductions of 25 to 75% were evaluated. It would appear that the concept of a universally applicable index of vine balance may be problematic. The VSP trellis system used here encourages shading for vigorous vines (see Smart and Robinson 1991). While the yield:pruning weight and leaf area:fruit yield ratios would indicate the vines were undercropped, the positive fruit composition to reducing crop load is best understood from examining the change brought about by crop removal to the yield:canopy surface area index. The reduction in crop brought the capacity of the exposed leaf area into balance with that required to ripen fruit to optimum maturity and produce high-quality wines. Single-canopy VSP training systems are the dominant method of managing vine growth in Hawke s Bay. Despite extreme canopy-trimming procedures adopted to restrict shoot and leaf growth during the season, the vine characteristics of high LLNs and dense canopies with excessive shading are the consequences of high vigor. Crop reduction is widely adopted as a means of coping with the deficiencies of the training system. The fruiting capacity of grapevines in any climatic region is largely determined by total leaf area (Smart and Robinson 1991, Kliewer and Dokoozlian 2005) and also by the percentage of the total leaf surface area that is exposed to full sunlight, other factors not being limiting. Conclusions The canopy intervention procedures universally undertaken in the vineyards of Hawke s Bay are an artifact of the VSP training system widely used in the vineyards. Without leaf removal to enhance fruit exposure and a reduction in crop load through crop removal, the dense, heavily shaded canopies are unable to ripen fruit to optimum maturity. This is a particular issue given the predominance of red-fruited cultivars (Cabernet Sauvignon and Merlot) grown in Hawke s Bay where green characters associated with below-optimum fruit ripeness are a consistent fault in some wines in some seasons. Literature Cited Asen, S., R.N. Stewart, and K.H. Norris.1972. Co-pigmentation of anthocyanins in plant tissues and its effect on color. Phytochemistry 11:1139-1145. Boulton, R. 2001. The copigmentation of anthocyanins and its role in the color of red wine: A critical review. Am. J. Enol. Vitic. 52:67-87.

Influence of Leaf and Crop Removal on Merlot 507 Chapman, D.M., M.A. Matthews, and J.X. Guinard. 2004. Sensory attributes of Cabernet Sauvignon wines made from vines with different crop yields. Am. J. Enol. Vitic. 55:325-334. Cortell, J.M., and J.A. Kennedy. 2006. Effect of shading on accumulation of flavonoid concentration in (Vitis vinifera L.) Pinot noir fruit and extraction in a model system. J. Agric. Food Chem. 54:8510-8520. Downey, M.O., J.S. Harvey, and S.P. Robinson. 2004. The effect of bunch shading on berry development and flavonoid accumulation in Shiraz grapes. Aust. J. Grape Wine Res. 10:55-73. Downey, M.O., N.K. Dokoozlian, and M.P. Krstic. 2006. Cultural practice and environmental impacts on the flavonoid composition of grapes and wine: A review of recent research. Am. J. Enol. Vitic. 57:257-268. Green, S.R. 2008. Measurement and modelling the transpiration of fruit trees and grapevines for irrigation scheduling. In Proceedings of the Fifth International Symposium on Horticultural Crops. I. Goodwin and M.G. O Connell (eds.). Acta Hortic. 792:321-327. Guidoni, S., P. Allara, and A. Schubert. 2002. Effect of cluster thinning on berry skin anthocyanin composition of Vitis vinifera cv. Nebbiolo. Am. J. Enol. Vitic. 53:224-226. Guidoni, S., A. Ferrandino, and V. Novello. 2008. Effects of seasonal and agronomical practices on skin anthocyanin profile of Nebbiolo grapes. Am. J. Enol. Vitic. 59:22-29. Harbertson, J.F., E.A. Picciotto, and D.O. Adams. 2003. Measurement of polymeric pigments in grape berry extracts and wines using a protein precipitation assay combined with bisulfite bleaching. Am. J. Enol. Vitic. 54:301-306. Haselgrove, L., D. Botting, R. van Heeswijck, P.B. Hoj, P.R. Dry, C. Ford, and P.G. Iland. 2000. Canopy microclimate and berry composition: The effect of bunch exposure on the phenolic composition of Vitis vinifera L cv. Shiraz grape berries. Aust. J. Grape Wine Res. 6:141-149. Howell, G.S. 2001. Sustainable grape productivity and the growth-yield relationship: A review. Am. J. Enol. Vitic. 52:165-174. Hunter, J.J., H.P. Ruffner, C.G. Volschenk, and D.J. Le Roux. 1995. Partial defoliation of Vitis vinifera L. cv. Cabernet Sauvignon/99 Richter: Effect on root growth, canopy efficiency, grape composition, and wine quality. Am. J. Enol. Vitic. 46:306-314. Iland, P.G., W. Cynkar, I.L. Francis, P.J. Williams, and B.G. Coombe. 1996. Optimization of methods for the determination of total and red-free glycosyl glucose in black grape berries of Vitis vinifera. Aust. J. Grape Wine Res. 2:171-178. Iland, P., A. Ewart, J. Sitters, A. Markides, and N. Bruer. 2000. Techniques for chemical and quality monitoring during winemaking. Patrick Iland Wine Promotions, Campbelltown, Adelaide. Jackson, D.I., and P.B. Lombard. 1993. Environmental and management practices affecting grape composition and wine quality A review. Am. J. Enol. Vitic. 44:409-430. Joscelyne, V.L., M.O. Downey, M. Mazza, and S.E.P. Bastian. 2007. Partial shading of Cabernet Sauvignon and Shiraz vines altered wine color and mouthfeel attributes, but increased exposure had little impact. J. Agric. Food Chem. 55:10888-10896. Keller, M., L.J. Mills, R.L. Wample, and S.E. Spayd. 2005. Cluster thinning effects on three deficit-irrigated Vitis vinifera cultivars. Am. J. Enol. Vitic. 56:91-103. Kliewer, W.M., and N.K. Dokoozlian. 2005. Leaf area/crop weight ratios of grapevines: Influence on fruit composition and wine quality. Am. J. Enol. Vitic. 56:170-181. Lee, S.H., M.J. Sao, M. Rui, J.P. Cotta, D.E. Block, N.K. Dokoozlian, and S.E. Ebeler. 2007. Vine microclimate and norisoprenoid concentration in Cabernet Sauvignon grapes and wine. Am. J. Enol. Vitic. 58:291-301. Mazza, G., L. Fukumoto, P. Delaquis, B. Girard, and B. Ewert. 1999. Anthocyanins, phenolics, and color of Cabernet Franc, Merlot, and Pinot Noir wines from British Columbia. J. Agric. Food Chem. 47:4009-4017. Nuzzo, V., and M.A. Matthews. 2006. Response of fruit growth and ripening to crop level in dry-farmed Cabernet Sauvignon on four rootstocks. Am. J. Enol. Vitic. 57:314-324. Ough, C.S., and R. Nagaoka. 1984. Effect of cluster thinning and vineyard yields on grape and wine composition and wine quality of Cabernet Sauvignon. Am. J. Enol. Vitic. 35:30-34. Petrie, P.R., and P.R. Clingeleffer. 2006. Crop thinning (hand versus mechanical), grape maturity and anthocyanin concentration: outcomes from irrigated Cabernet Sauvignon (Vitis vinifera L.) in a warm climate. Aust. J. Grape Wine Res. 12:21-29. Price, S.F., P.J. Breen, M. Valladao, and B.T. Watson. 1995. Cluster sun exposure and quercetin in Pinot noir grapes and wine. Am. J. Enol. Vitic. 46:187-194. Reynolds, A.G., S. Yerle, B. Watson, S.F. Price, and D.A. Wardle. 1996. Fruit environment and crop level effects on Pinot noir. III. Composition and descriptive analysis of Oregon and British Columbia wines. Am. J. Enol. Vitic. 47:329-339. Reynolds, A.G., J. Schlosser, D. Sorokowsky, R. Roberts, J. Willwerth, and C. de Savigny. 2007. Magnitude of viticultural and enological effects. II. Relative impacts of cluster thinning and yeast strain on composition and sensory attributes of Chardonnay Musqué. Am. J. Enol. Vitic. 58:25-41. Ristic, R., M.O. Downey, P.G. Iland, K. Bindon, I.L. Francis, M. Herderich, and S.P. Robinson. 2007. Exclusion of sunlight from Shiraz grapes alters wine color, tannin and sensory properties. Aust. J. Grape Wine Res. 13:53-65. Smart, R.E., and M. Robinson. 1991. Sunlight into Wine. A Handbook for Winegrape Crop Management. Winetitles, Adelaide. Spayd, S.E., J.M. Tarara, D.L. Mee, and J.C. Ferguson. 2002. Separation of sunlight and temperature effects on the composition of Vitis vinifera cv. Merlot berries. Am. J. Enol. Vitic. 53:171-182. Wolf, T.K., P.R. Dry, P.G. Iland, D. Botting, J. Dick, U. Kennedy, and R. Ristic. 2003. Response of Shiraz grapevines to five different training systems in the Barossa Valley, Australia. Aust. J. Grape Wine Res. 9:82-95. Zoecklein, B., T.K. Wolf, C. Yoder, and Y. Jasinski. 1996. Effect of crop level on Chardonnay and Cabernet Sauvignon (Vitis vinifera L.) glycosides. In Proceedings for the Fourth international Symposium on Cool Climate Enology and Viticulture. T. Henick-Kling et al. (eds.), pp. 58-65. New York State Agricultural Experiment Station, Geneva.