EFFECT OF TIMING OF LEAF REMOVAL ON YIELD PARAMETERS, GRAPE AND WINE QUALITY OF VITIS VINIFERA L. CV. 'SAUVIGNON BLANC'

Transcription

1 UNIVERSITY OF NOVA GORICA SCHOOL OF VITICULTURE AND ENOLOGY EFFECT OF TIMING OF LEAF REMOVAL ON YIELD PARAMETERS, GRAPE AND WINE QUALITY OF VITIS VINIFERA L. CV. 'SAUVIGNON BLANC' DIPLOMA THESIS Daniela MARKOVIC Mentors: doc. dr. Branka Mozetič Vodopivec prof. dr. Paolo Sivilotti Nova Gorica, 2016

2 II

3 ZAHVALA Zahvaljujem se mentorjema doc. dr. Branki Mozetič Vodopivec in prof. dr. Paolu Sivilottiju za mentorstvo in pomoč pri izvedbi praktičnega dela in pri sestavi diplomskega dela. Diplomsko delo je nastalo v okviru aktivnosti, ki so se izvajale v sklopu projekta VISO. Projekt VISO je sofinanciran v okviru Programa čezmejnega sodelovanja Slovenija Italija iz sredstev Evropskega sklada za regionalni razvoj in nacionalnih sredstev. III

4 POVZETEK Z odstranjevanjem listov v predelu grozdov se lahko spremeni mikroklima grozdov, ki lahko izzove spremembe v sekundarnem metabolizmu vinske trte, ki so odvisne tudi od časa izvedbe ukrepa. Uravnavanje sinteze aromatskih komponent je zanimivo pri belih sortah kot je sorta 'Sauvignon blanc', kjer so metoksipirazini in tioli tisti, ki oblikujejo značilno aromo vina te sorte. Eksperiment diplomske naloge smo zastavili na Oslavju (Gorica, Italija) v vinogradu sorte grozdja 'Sauvignon blanc' (Vitis vinifera L.). Želeli smo preveriti vpliva odstranjevanja listov v predelu grozdja pred in po cvetenju na količino in kakovost pridelka ter aromatske lastnosti grozdja in vina. Količinski parametri pridelka so pokazali majhne vplive izvedenih ukrepov razlistanja; pozno razlistanje na količino pridelka ni vplivalo, v nasprotju z razlistanjem pred cvetenjem, kjer smo opazili rahlo zmanjšanje pridelka po trsu in mase grozdov. Tudi pri osnovnih zrelostnih parametrih smo opazili manjše spremembe, predvsem večje vsebnosti topne suhe snovi pri razlistanju pred cvetenjem. Pri aromatskih komponentah smo zaznali vpliv razlistanja pred cvetenjem na količino prekurzorjev tiolov tako v grozdju kot tudi vinu; večjo vsebnost 4-merkapto-4- metilpentan-2-on (4MMP) in nižje koncentracije 3-merkaptoheksan-1-ol (3MH). S poznejšim odstranjevanjem listov smo zmanjšali koncentracije 4MMP, na vsebnost 3MH ta ukrep ni vplival. Tudi s senzorično oceno vina smo potrdili pomembne vplive razlistanja pred in po cvetenju na določene note v aromi vina 'Sauvignon blanc'. Ključne besede: Vitis vinifera L. cv. 'Sauvignon blanc', razlistanje pred cvetenjem, razlistanje po cvetenju, aromatske komponente, tioli, LC-MS IV

5 SUMMARY Leaf removal changes cluster microclimate, and thus modifications in the secondary metabolism are triggered, but differently as regard to the timing of application. In case of white grape varieties, the interest on the modification of aroma characteristics is searched, and in case of 'Sauvignon blanc', the shift in concentration of methoxypyrazines and thiols is responsible for the bouquet of the wines. An experimental trial was set up in Oslavia (Gorizia, Italy) with the aim to evaluate how pre-flowering and post-flowering leaf removal applied on 'Sauvignon blanc' vines could change the aromatic occurrence in grapes and wines. The yield parameters were slightly affected by the treatments of leaf removal; while late leaf removal did not impact on production, pre-flowering leaf removal slightly promoted a reduction on yield and cluster weight. Also basic maturation parameters were slightly changed, mainly higher soluble solids in case of pre-flowering leaf removal. As regard aromatics, both thiol precursors in grapes and thiols in wines revealed to be shifted in case of pre-flowering leaf removal, with higher values of 4-mercapto-4- methyl-pentan-2-one and lower values of 3-mercaptohexan-1-ol. In case of late leaf removal lower concentration of 4MMP were revealed and similar of 3MH. Also the degustation of the wines revealed some peculiarities of pre-flowering and late leaf removal treatments for some important sensorial notes of 'Sauvignon blanc' wines. Key words: Vitis vinifera L. cv. 'Sauvignon blanc', pre-flowering leaf removal, postflowering leaf removal, aromatic compounds, thiols, LC-MS V

6 VI

7 TABLE OF CONTENTS POVZETEK... IV SUMMARY... V TABLE OF CONTENTS... VII LIST OF TABLES... IX LIST OF FIGURES... X ABBREVIATIONS... XI 1 INTRODUCTION Hypothesis THEORETICAL BASES The timing of leaf removal Effect of leaf removal on aromatic compounds Thiols and light Vitis vinifera L. 'Sauvignon blanc' MATERIALS AND METHODS Materials Methods DETERMINATION OF LEAF AREA DETERMINATION OF YIELD PARAMETERS DETERMINATION OF BASIC MATURATION PARAMETERS GRAPE BERRY COLLECTION AND DETERMINATION OF THIOL PRECURSORS IN GRAPES MICROVINIFICATIONS, MUST ANALYSIS AND QUANTIFICATION OF THIOLS IN WINES VII

8 3.2.6 SENSORY ANALYSIS OF WINES STATISTIC PROCESSING OF THE DATA RESULTS AND DISCUSSION Leaf area components Yield parameters Basic maturation parameters Thiol precursors in grapes Thiols in wines Sensory analysis of the wines CONCLUSIONS LITERATURE VIII

9 LIST OF TABLES Table 1: Leaf area components before and after pre-flowering leaf removal performed on Sauvignon blanc vines in Oslavia (Gorizia, Italy) on 21 may UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) Table 2: Leaf area components before and after the second intervention of leaf removal on Sauvignon blanc grown in Oslavia (Italy) on 20 June UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) Table 3: Leaf area components at harvest on Sauvignon blanc grown in Oslavia (Italy). UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) IX

10 LIST OF FIGURES Figure 1: Clusters and leaves of Sauvignon blanc (photo Miha Godec, archive project Interreg VISO)... 6 Figure 2: Location of the experimental site (Oslavia, Gorizia, Italia) (quantumgis map processing)... 8 Figure 3: Completely randomized experimental design set in the vineyard of Sauvignon blanc in Oslavia (Gorizia, Italy). UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal... 9 Figure 4: Scheme of the treatments (UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal)... 9 Figure 5: Collection of samples of thiol precursor analyses using liquid nitrogen to immediately froze the berries (photo Miha Godec, archive project Interreg VISO ) Figure 6: Effect of leaf removal on cluster number (A), average cluster weight (B), yield (C) and leaf area-to-yield ratio (D) on Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) Figure 7: Effect of leaf removal on total soluble solids, titratable acidity and ph during maturation (A, C, E) and at harvest time (B, D, F) of Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season The arrows identify harvest time. Bars represent 95% confidence interval (n=3) Figure 8: Effect of leaf removal on the concentration of cys-4mmp (A), cys-3mh (B), glut-4mmp (C) and glut-3mh (D) at harvest time in the grapes of Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) Figure 9: Effect of leaf removal on the concentration of 4MMP (A) and 3MH (B) in the wines of Sauvignon blanc produced in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) Figure 10: Effect of leaf removal on the sensory characteristics of c the wines of Sauvignon blanc produced in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) X

11 ABBREVIATIONS UNT = untreated vines LR = leaf removal ELR = pre-flowering leaf removal LLR = post-flowering leaf removal IBMP = 3-isobutyl-2-methoxypyrazine 4MMP = 4-mercapto-4-methyl-pentan-2-one 3 MH = 3-mercaptohexan-1-ol Cys-4MMP = 4-S-cysteinyl-4-methylpentan-2-one Glut-4MMP = 4-S-glutathionyl-4-methylpentan-2-one Cys-3MH = 3-S-cysteinyl hexan-1-ol Glut-3MH = 3-S-glutathionyl hexan-1-ol XI

12

13 1 INTRODUCTION The implementation of leaf removal on white grape varieties was supposed to be detrimental for the quality, mainly for aroma compounds. In any case, in the North- Eastern part of Italy in several seasons the abundant rain also during harvest can damage the clusters since Botrytis cinerea or sour rot develop over the clusters resulting in negative modifications of the grape quality. For this reason, leaf removal proved to be crucial since the improvement of canopy microclimate could hinder or restrain the development of rots even more than selective pesticides. Nowadays winegrowers are aware that the timing of leaf removal is also important. The best results are obtain when the technique is applied at berry-set, since the grape berries have still time to adapt to sunlight, reducing the occurrence of sunburns. A lot of emphasis has been given in the last decades to pre-flowering leaf removal, but there is a lack of knowledge on the effect of this technique on aroma concentration. Both pre- and post-flowering leaf removal techniques are known to impair modifications in the primary and secondary metabolism. Among aroma compounds, thiols are particularly important for Vitis vinifera L. cv. 'Sauvignon blanc' and they account for notes of box-tree, grapefruit, blackcurrant, catpee, passionfruit and many others. To our knowledge, there are no contributions on the effects of leaf removal over the biosynthesis of thiols in grapes, thus the present thesis wanted to investigate how the timing of leaf removal could affect the occurrence of these secondary metabolites in grapes and thereafter how they could impair on the quality of the wines. 1.1 Hypothesis To improve the quality of the grapes a better cluster microclimate is needed, thus the aim of the trial was: 1. To evaluate two timings of leaf removal - pre-flowering and berry-set leaf removal and their effects on yield parameters (berry weight, cluster weight, cluster number, yield 1

14 per vine and leaf area/yield ratio) in a vineyard of Vitis vinifera cv. 'Sauvignon blanc' cultivated in the Italian D.O.C. Collio region; 2. To ascertain how the timing of leaf removal affect the accumulation of soluble solids or the degradation of acids during maturation and at harvest; 3. As regard thiols and thiol precursors, to understand how pre-flowering or berry-set leaf removal could differently change their biosynthesis and the relative amount of 4- mercapto-4-methyl-pentan-2-one and 3-mercaptohexan-1-ol in grapes and wine; 4. To evaluate the effect of the timing of leaf removal on the sensory properties of the wines. 2

15 2 THEORETICAL BASES 2.1 The timing of leaf removal During the growing season many agricultural practices can be undertaken in the vineyard with different purposes, so starting at budburst bud removal, shoot elimination, leaf removal and cluster thinning can be applied with the aim to reach the wanted goal. Because they are applied during the active growing of the canopy, they can affect the physiology of the plant with important modifications on yield and grape quality. Among the techniques that can be applied during the summer season, leaf removal is a widespread practice used with the aim of primarily improving cluster microclimate but also influencing yield quality (Mescalchin et al., 2008). Practically, as related with the timing of application, 2-8 leaves are pulled from the basal part of the shoots. When leaf removal is performed at berry-set or at veraison, only 2-3 leaves surrounding the clusters are eliminated, and normally the leaves above the clusters are retained in order to protect them from the damaging effects of sunlight (Mescalchin et al., 2008). As late leaf removal is preformed, as much problems of sunburns will happen (Genovese et al., 2010). Different story in case of earlier applications; when applied in a very early stage - pre-flowering time leaf removal causes a reduction of berry set, since there is the lack of nutrients going into the cluster: the vine sacrifices some flowers because of the stressful situation (Poni et al., 2009). Thus, the first visible result that could be observed is a reduction of the yield, as shown by Poni et al. (2006) in 'Sangiovese' and 'Trebbiano' or by Sternad Lemut et al. (2013) in 'Pinot Noir'. Several experiments investigated the effect of leaf removal over basic primary metabolism with contrasting results ascertained. In experiments carried out in Slovenia, Krajniger (2014), Soban (2015) and Turk (2014) showed that the application of leaf removal prior to flowering resulted in slight modifications in primary metabolism in grapes. But the results are not always consistent also in other experiments; preflowering leaf removal accounted for a higher content of sugars in 'Sangiovese' and 'Trebbiano' (Poni et al., 2006) and in 'Tempranillo' in Rioja (Diago et al., 2012), while no differences were showed on 'Pinot Noir' in Oregon (Lee and Skinkis 2013) and 'Tempranillo' in Valencia (Risco et al., 2014). Similar behavior was reported for 3

16 titratable acidity; pre-flowering leaf removal was not affecting the parameter in some experiments carried out on 'Pinot Noir' in Oregon (Lee and Skinkis 2013), 'Trebbiano' (Poni et al., 2006) and 'Tempranillo' in Valencia (Risco et al., 2014), while an increase was observed on 'Sangiovese' (Poni et al., 2006) and a reduction in the case of a 'Tempranillo' in the area of the Rioja (Diago et al., 2012). More interesting outcomes were found as regard the secondary metabolism. Diago et al. (2012), Gatti et al. (2012), Matsuyama et al. (2014), Poni et al. (2006), Soban (2015) and Turk (2014) reported an increase in the concentration of anthocyanins with the preflowering defoliation, while Risco et al. (2014) did not find any change for the variety 'Tempranillo' in the area of Valencia. Diago et al. (2012) and Sternad Lemut et al. (2013) have shown that early leaf removal, especially when carried out pre-flowering time, increases the synthesis of flavonols in berries, and moreover Feng et al. (2015) have shown that the increase of flavonols was also linked to the intensity of defoliation. 2.2 Effect of leaf removal on aromatic compounds As regards the aromatic component, there are not many experiments that have considered the effect of defoliation. Belancic et al. (2007) and Zoecklein et al. (1992) showed a positive effect of leaf removal on the occurrence of certain aromatic compounds in grapes, while Arnold and Bledsoe (1990) highlighted a reduction of vegetal notes while keeping the fruit character of 'Sauvignon blanc' grapes. Feng et al. (2015) working on 'Pinot Noir', could see that berry-set defoliation had a positive effect on the concentration of some aromas including the -damascenone. On the contrary, Scafidi et al. (2013) found that by shielding the grapes of 'Grillo' with the dark box a higher concentration of aromatic compounds could be obtained as compared with grapes normally exposed to the sun. A large number of experiments investigated the role of defoliation, but more generally of light, on the accumulation/degradation of methoxypyrazines in grapes. The methoxypyrazines, accumulate from the first stages after flowering, reaching a maximum prior to veraison and then decreased during ripening. Ryona et al. (2008) were able to verify that the shading of the clusters led to a greater accumulation of methoxypyrazines before veraison, while Kock et al. (2012) ascertained that late 4

17 treatment of shading of the bunch did not lead to any change in the concentration of these aromatic substances. By comparison with a control in which the plants were not defoliated, Šuklje et al. (2014) demonstrated that the defoliation carried out in postflowering time led to a reduction of the concentration of 3-isobutyl-2-methoxypyrazine in 'Merlot' grapes at harvest, while in the case of 'Sauvignon blanc' the difference was negligible. Krajniger (2014) reported that pre-flowering leaf removal maintained a higher level of IBMP in grapes in both 'Cabernet Sauvignon' and 'Merlot', while when performed at berry-set the concentration was reduced. 2.3 Thiols and light 3-Mercaptohexan-1-ol (3MH) is a volatile thiol and a key contributor to the distinct odor of grapefruit or passion fruit in 'Sauvignon blanc' wines as well as 'Colombard', 'Muscat', 'Sylvaner', 'Pinot blanc' (Tominaga et al., 1998). The sensory notes presented by 'Sauvignon blanc' wines, spread from asparagus notes, cat-pee, flinty, through to the more tropical and fruity typical aromas of grapefruit and passion fruit (Ribereau-Gayon, 2005), and also become off-flavors at too high concentrations. The most representative compounds of 'Sauvignon blanc' wine aroma are 4-mercapto-4- methyl-pentan-2-one (4MMP), 3-mercaptohexan-1-ol acetate (3MHA), 3- mercaptohexan-1-ol (3MH). Many others were discovered recently but the most important are the three above mentioned. There is still a lot of confusion around the actual notes of thiols, since the increase of their concentration accounts for a change in the perceived aroma. Tominaga et al. (1998) reported that 3MH is a key contributor of grapefruit or passion fruit in 'Sauvignon blanc', but also the tropical aroma has been associated with the same compound and also with 3MHA. Differently, elderberry was found to be related with high concentrations of 4MMP. In respect of the thiol molecules responsible for the aroma of 'Sauvignon blanc' - but also other varieties - unfortunately there are still no results published, even if the exposure of the grapes to UV radiation seems positive in increasing the concentration of the precursors of these molecules in grapes (Kobayashi et al., 2011). 5

18 2.4 Vitis vinifera L. 'Sauvignon blanc' 'Sauvignon blanc' is a white winegrape variety (figure 1) probably originated in the western France, now successfully cultivated in many emerging and established winegrowing areas all over the world (Brancadoro et al., 2012). Figure 1: Clusters and leaves of Sauvignon blanc (photo Miha Godec, archive project Interreg VISO) France and Italy apart, Sauvignon is grown in different East-Europe Countries: Croatia, Serbia, Hungary, Czech Republic, Slovakia, Rumania, Moldova and Ukraine, while outside Europe in California, Chile, Argentina, Brazil, Uruguay, South Africa, Australia and New Zealand. In each terroir, the particular climatic and pedological condition impair characteristic notes to 'Sauvignon blanc' wines. Thus in New Zealand, a cool climate area, asparagus, gooseberry and green pepper notes account for the bouquet of most of the wines, while the Poully-Fume and Sancerre are recognized because of flint (pierre à fusil) aroma and the evolution in oak barrels. In the other viticultural areas, as related with terroir the particular bouquet of 'Sauvignon blanc' wines is related with elderberry, grapefruit, blackcurrant, passionfruit and/or the combination with other more 6

19 fruity aromas. In Northern Italy 'Sauvignon blanc' is cultivated in Friuli Venezia Giulia, Trentino Alto Adige and Lombardy, while in Slovenia we can find the variety both in the Western and in the Eastern part of the country. The wine-style obtained in Western Slovenia is more similar to Friuli, while in the Eastern cooler winegrowing areas the wines are more similar to the nearer Styria. Typically 'Sauvignon blanc' is produced as varietal, but in the area of Bordeaux is blended with Semillon also in the area of Sauternes where the grapes are harvested only when B. cinerea infects the berries. 7

20 3 MATERIALS AND METHODS 3.1 Materials The experiment trial was carried out in the seasons 2013 in Oslavia (North-Eastern part of Italy) in the Italian "Collio" winegrowing area, on Vitis vinifera L. cv. 'Sauvignon blanc' (Figure 2). A vineyard of Vitis vinifera L. cv. 'Sauvignon blanc' clone R3, with an age of 33 years was selected for the experiment. The vineyard was planted with 3787 plants/he (1,20 m between the vines and 2,20 m between the rows), adopting the traditional Cappuccina training system. Figure 2: Location of the experimental site (Oslavia, Gorizia, Italia) (quantumgis map processing) Within the surface of the vineyard, three rows were selected, and three treatments with three replicates each were identified adopting a completely randomized experimental design (figure 3). 10 following representative and healthy vines were then selected for each plot and marked with a visible label. The treatments under comparison were set as follows (figure 4): 8

21 UNT (untreated), no leaves were removed in any time. In the graphs are represented with a pale grey colour; ELR (early leaf removal), leaf removal was performed approximately 15 days before flowering removing 4-to-6 leaves for each shoot (21 May 2013). In the graphs represented using a red colour; LLR (late leaf removal), leaf removal performed approximately 15 days after flowering just after berry-set removing 2-to-3 leaves for each shoot (20 June 2013). In the graphs presented with a pale green colour. Row 1 UNT ELR LLR Row 2 ELR LLR UNT Row 3 UNT ELR LLR Figure 3: Completely randomized experimental design set in the vineyard of Sauvignon blanc in Oslavia (Gorizia, Italy). UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal Figure 4: Scheme of the treatments (UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal) 9

22 3.2 Methods DETERMINATION OF LEAF AREA Similarly to Soban (2015), Tronkar (2014) and Turk (2014) leaf area was determined using an indirect method. A first set of leaves, collected randomly from the base and the top of the shoots, both from main and lateral shoots, was collected. The main vein of each leaf was measured and the area was assessed scanning the leaves and processing each image throughout binarisation and determination of the relative percentage of black and white by means of ImageJ (Schneider et al., 2012). A regression between main vein (x) and leaf area (LA) was then calculated (LA=0.753x x; R 2 =0,9152). At the time of pre-flowering, post-flowering leaf removal and at harvest, the main veins of all the leaves of one plant were measured, separating them by shoot and, within the shoot, dividing main and lateral leaves. A second regression was then assessed between the number of leaves/shoot (n) and leaf area/shoot (SA) both for main shoots (SA main = n; R 2 =0,7853) and lateral shoots (SA laterals = 54.15n; R 2 =0,8988) DETERMINATION OF YIELD PARAMETERS On 04 th September 2013, harvest date, 10 vines/plot were harvested, counting the number of clusters/vine and weighting the production using a portable electronic fishing hanging scale Senior OCS-20A (Shenzhen Kingroot Industry Co. Ltd., Guangdong, China). The average cluster weight was thereafter computed rating the yield by the number of clusters. The equilibrium between canopy and crop was evaluated assessing leaf area-to-yield ratio (m 2 /kg) on 3 vines/plot using the data of leaf area collected at harvest DETERMINATION OF BASIC MATURATION PARAMETERS From each plot, representative samples of 50 berries, randomly selected from different parts of the clusters within the plots, were collected during five data-points from veraison to harvest and also one week after harvest, in plastic bags immediately cooled 10

23 in a portable fridge. The samples were thereafter immediately hand-squeezed and analyzed for total soluble solids ( Brix), titratable acidity (as tartaric acid g/l) and ph, malic and tartaric acid (g/l) using a WineScanTM FT120 Basic (FOSS, Hillerød, Danimarca) GRAPE BERRY COLLECTION AND DETERMINATION OF THIOL PRECURSORS IN GRAPES A second set of berries were collected in urine flasks and immediately frozen with liquid nitrogen in the field in order to block any oxidation and so further increase of thiol precursors (figure 5). The samples were thereafter stored in dry ice until they were stored in -80 C freezer. Figure 5: Collection of samples of thiol precursor analyses using liquid nitrogen to immediately froze the berries (photo Miha Godec, archive project Interreg VISO ). The frozen grape samples were grinded to a powder using a IKA M20 Universal analytical mill (Königswinter, Germany), using again liquid nitrogen to avoid oxidation. The grape berry powder was thereafter collected in 50 ml-falcon tubes and stored back at -80 C until the analysis. For the analysis of thiol precursors a slightly modified 11

24 method as described by Lisjak et al. (2013) was used. The powder was transferred into ice-cold, deoxygenated methanol in ratio 1:4 (w/v) for the extraction. After cleaning and concentration using ion exchange and styrene-divinylbenzene columns, the analyses of thiol precursors were performed by means of a liquid chromatography coupled to tandem mass spectrometric detector (Agilent Technologies, Palo Alto, USA) in the laboratory of Agricultural Institute of Slovenia MICROVINIFICATIONS, MUST ANALYSIS AND QUANTIFICATION OF THIOLS IN WINES The grapes collected from the different plots were pooled, destemmed and added with 30 ppm of SO 2 and dry ice to prevent the oxidation. The grapes were thereafter filled in an experimental UltraPress (Skrlj, Batuje, Slovenia) and the extraction of the must obtained by pressing constantly at 3 bars for minutes. The liquid spiked from the bottom of the UltraPress in 5 L tanks, was constantly added with dry ice in order to prevent oxidation. The must was thereafter stored overnight at 5 C to facilitate the settling of the solids. The following day, the liquid was poured off and divided in three aliquots of ml filled in 1L LMR flasks for the fermentation. The fermentation was performed using Alchemy II yeast (Anchor wine yeast, South Africa) at 15 C. The LMR flasks were covered with air-lockers filled with water in order to allow the excess of CO 2 to be removed during fermentation. From the beginning to the end of fermentation, must samples were collected every 3 days to check the advancement of the process, analyzing the content of sugar residues and alcohol by means of the WineScanTM FT120 Basic. At the end of fermentation, the wines were racked, and samples of 100 ml supplemented with 50 ppm of SO 2 and frozen at -20 C until thiol analyses. The remaining wine was mixed, according with the treatment and bottled in 0,375 L bottles. The determination of volatile thiols 4-mercapto-4-methylpentan-2-one (4MMP), 3- mercaptohexan-1-ol (3MH) and 3-mercaptohexylacetate (3MHA) was performed by gas chromatography coupled with mass spectrometry, using the method of Tominaga et al. (1998) as modified by Šuklje et al. (2013). 12

25 3.2.6 SENSORY ANALYSIS OF WINES For wine tasting we used the scorecard advanced M29/BST for white wines of the Centro Studi Assaggiatori of Brescia (Italy). Born in 1990, nowadays it is the most important center of sensory analysis in Italy, as it connects effectively the research, firms and tasters. Advanced Bst : is a descriptive test with high useful information that generates quantitative and hedonic profiles. The scorecard has been structured with 10 levels from 0 to 9, where 0 corresponded to no perception of the descriptor. The advanced Bst scorecard was the best tool in order to examine the wine aroma, but also for the evaluation of brandy, coffee, balsamic vinegar and beer. Sensory analysis was held on 18 December 2013 in the tasting room of the School of Viticulture and Enology in Lanthieri Mansion, Vipava. The panel was formed by 10 persons among wine experts and professionals STATISTIC PROCESSING OF THE DATA The average data of each parameter are represented in the graphs, and the vertical bars represent the 95% confidence interval. Camussi et al. (1995) reports that this statistical index can be used to claim differences among treatments similarly as using the Fisher LDS method. 13

26 4 RESULTS AND DISCUSSION 4.1 Leaf area components At the beginning of the season, when the time for pre-flowering leaf removal came, approximately 1000 cm 2 /vine of leaf area was evaluated in all three treatments, with the predominance of the ones on main shoots around cm 2 /plant and just the first leaves originated from lateral shoots accounting for cm 2 /plant (table 1). Before the start of the experiment there were no significant differences between treatments for both main and lateral leaves. After the first interventions of leaf removal, performed on ELR treatment, a 46% reduction of main leaves was ascertained. As compared with LLR and UNT, a significant reduction of main and of course of total leaf area was obtained. Table 1: Leaf area components before and after pre-flowering leaf removal performed on Sauvignon blanc vines in Oslavia (Gorizia, Italy) on 21 may UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) treatment main shoots (cm 2 /shoot) laterals (cm 2 /shoot) total leaf area /shoot (cm 2 /shoot) total leaf area after ELR (cm 2 /shoot) UNT 892±35 260± ± ±59 % reduction ELR 801±58 261± ±74 567±65 46,6 LLR 811±39 277± ± ±44 After one month from the first operation, leaf area was more than doubled, mainly due to the increase in leaves grown on lateral shoots (+ 600% on the average of the three treatments), and we could still observe a trend of a lower leaf area on the main shoots of the ELR treatment (table 2). As follows the leaf pulling in the LLR treatment, a 6,72% reduction of leaf area was ascertained after the elimination of 2-3 leaver/shoot. 14

27 Table 2: Leaf area components before and after the second intervention of leaf removal on Sauvignon blanc grown in Oslavia (Italy) on 20 June UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) treatment main shoots (cm 2 /shoot) laterals (cm 2 /shoot) total leaf area /shoot (cm 2 /shoot) total leaf area after LLR (cm 2 /shoot) UNT 1571 ± ± ± ±736 ELR 1301 ± ± ± ±654 % reduction LLR 1799 ± ± ± ± 733 6,72 In the summer season, when the shoots overcame the last wire, two operations of shoot trimming were performed, thus reducing the leaf area at the end of the season as compared with what ascertained on 20 June (table 3). Table 3: Leaf area components at harvest on Sauvignon blanc grown in Oslavia (Italy). UNT, untreated; ELR, pre-flowering leaf removal; LLR, post-flowering leaf removal. Data ± 95% confidence interval (n=3) treatment main shoots (cm 2 /shoot) laterals (cm 2 /shoot) total leaf area (cm 2 /shoot) UNT 981 ± ± ± 213 ELR 1167 ± ± ± 233 LLR 1119 ± ± ± 226 No substantial differences in the leaf area of the main shoots was ascertained, while in case of the lateral leaves a significant higher value was observed in case of the treatment ELR. This increase eventually resulted in a greater total leaf area of ELR treatment in comparison with the control UNT, while LLR remained intermediate. Regrowth of leaves on lateral shoots was recorded by Diago et al. (2012) in 'Tempranillo' and by Pastore et al. (2013) on 'Sangiovese' as follows as pre-flowering leaf removal, while Tardaguila et al. (2010) in 'Graciano' and 'Carignan', and Soban (2015) and Turk (2014) in 'Refošk' did not find any difference in such component of leaf area. 15

28 4.2 Yield parameters The treatments applied in the vineyard did not affect the number of clusters, since no operations of cluster thinning were carried out (figure 6A). The average cluster weight was slightly reduced in case of ELR treatment, while no effects of after-flowering leaf removal (LLR) were ascertained (figure 6B). As related with the variation of cluster weight, also yield was slightly reduced in case of ELR while no differences were found between UNT and LLR. cluster number A average cluster weight (g) B 0 UNT ELR LLR 0 UNT ELR LLR yield (kg) 3,5 3,0 2,5 2,0 1,5 1,0 0,5 C Leaf area / yield (m 2 /kg) 2,5 2,0 1,5 1,0 0,5 D 0,0 UNT ELR LLR 0,0 UNT ELR LLR Figure 6: Effect of leaf removal on cluster number (A), average cluster weight (B), yield (C) and leaf area-to-yield ratio (D) on Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) 16

29 The ratio leaf area-to-yield (figure 6D) was found within the range of optimality (as proposed by Kliewer and Dokoozlian, 2005 within 1.0 and 1.4 m 2 /kg) for the treatments UNT and LLR, while significantly higher in the case of ELR. As compared with UNT, the slightly higher values of the ratio in case of LLR were mainly related with the leaf area component since the yield was nearly the same. As opposite, in case of ELR, the significantly higher ratio was due both to slightly higher leaf area and slightly lower yield. 4.3 Basic maturation parameters The grape sampling started around the end of véraison time. At the first date of sampling the average value of total soluble solids was 11,1 Brix (figure 7A), the titratable acidity 27,7 g/l (figure 7C) and the ph 2,42 (figure 7E). During maturation the amount of total soluble solids increased and reached the maximum level at harvest time, around 20,1 Brix, and remained the same one week later (figure 7A). As regard titratable acidity, a steep decrease was observed till the 22 nd August followed by a slight reduction till harvest, reaching a value of 7,3 g/l (figure 7C). A further decrease happened in the following week after the harvest. The behaviour of ph was the opposite as compared with titratable acidity; from the end of véraison till 22 nd August, a steep increase in ph was observed, but thereafter the increase was less pronounced reaching a value of 3,16 at harvest time (figure 7E). A slightly increase was also shown in the following week sampling. ELR reported significantly higher content of total soluble solids as compared with UNT, while LLR showed no differences between both ELR and UNT (figure 7B). Titratable acidity at harvest was similar among treatments, even if a slightly higher value was measured in case of ELR (figure 7D). As related with the values just reported, ELR showed significantly lower values of ph, while UNT and LLR reached similar ph values at harvest (figure 7F). 17

30 Total soluble solids ( Brix) Titratable acidity (g/l) 24 A B ,6 C E D F ,0 8,0 7,0 6,0 5,0 4,0 3,0 2,0 3,4 Total soluble solids ( Brix) Titratable acidity (g/l) 3,3 3,3 ph 3,0 2,7 3,2 3,1 ph 2,4 2,1 3,0 2,9 7/8 14/8 21/8 28/8 4/9 11/9 UNT ELR LLR Figure 7: Effect of leaf removal on total soluble solids, titratable acidity and ph during maturation (A, C, E) and at harvest time (B, D, F) of Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season The arrows identify harvest time. Bars represent 95% confidence interval (n=3) Also other experiments carried out on different varieties showed limited influence of pre-flowering leaf removal on basic technological parameters (Krajniger, 2014; Soban, 18

31 2015; Turk, 2014). As opposite, a positive effect of pre-flowering leaf removal was shown in case of highly productive varieties ('Sangiovese', 'Tempranillo') by Pastore et al. (2013) and Diago et al. (2012), with an increase of total soluble solids. While no effects of pre-flowering leaf removal were verified by Krajniger (2014) for 'Merlot' and 'Cabernet Sauvignon', Soban (2015) and Turk for 'Refošk' (2014) on titratable acidity, in case of highly producing varieties a reduction of same parameter was shown (Diago et al., 2012; Pastore et al., 2013). As opposite as titratable acidity, preflowering leaf removal was shown to increase the values of ph in 'Sangiovese' grapes. 4.4 Thiol precursors in grapes Both timings of leaf removal were profitable for a modification of the thiol profile in the grapes at harvest, with a particular behavior in case of ELR and LLR as compared with UNT (figure 8). The concentration of 4-S-cysteinyl-4-methylpentan-2-one (Cys-4MMP) was significantly higher in case of ELR and lower for LLR, while intermediate concentration was measured in UNT grapes (figure 8A). As opposite, no differences were shown for 3-S-cysteinyl hexan-1-ol (Cys-3MH), with slightly lower values in case of ELR (figure 8B). Similarly to Cys-4MMP, similar trend among treatments were shown in case of 4-S-glutathionyl-4-methylpentan-2-one (Glut-4MMP), with significant higher contents in case of ELR and lower for LLR (figure 8C), but not in comparison to UNT. As regard 3-S-glutathionyl hexan-1-ol (Glut-3MH) a different behavior was shown, with the significantly lowest content for UNT and significantly higher for both ELR and mainly LLR grapes (figure 8D). Also the difference between ELR and LLR was significant. There is a lack of knowledge about the effects of leaf removal on thiol precursors. A study of Kobayashi et al. (2011) pointed out the positive effects of UV radiation in order to obtain an increase of thiol precursors. Since leaf removal increase the exposure of the grapes to the light, we can speculate that an increase of UV exposure was obtained in case of ELR and LLR, positively affecting the concentration of some thiol precursors. 19

32 cys-4mmp (ng/kg) 5,0 4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0 UNT ELR LLR A cys-3mh (ng/kg) 6,8 5,9 5,0 4,1 3,2 2,3 1,4 0,5 UNT ELR LLR B 0,65 C 39 D glut-4mmp (ng/kg) 0,55 0,45 0,35 0,25 0,15 glut-3mh (ng/kg) ,05 3 UNT ELR LLR UNT ELR LLR Figure 8: Effect of leaf removal on the concentration of cys-4mmp (A), cys-3mh (B), glut- 4MMP (C) and glut-3mh (D) at harvest time in the grapes of Sauvignon blanc grown in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) 4.5 Thiols in wines The concentration of 4MMP in wines was in average 5,4 ng/l while in case of 3MH was much higher with mean values of 445 ng/l. As already observed for thiol precursors, the concentration of thiols in wines was significantly affected by the timing of leaf removal. Looking at 4MMP, significantly higher values were observed for ELR, in comparison with both UNT and LLR (figure 9A). As opposite, the concentration of the same thiol was the lowest for LLR. Different behaviour was observed in case of 20

33 3MH, with slightly similar values in case of UNT and LLR, while significantly lower values were shown for ELR (figure 9B). Since there are no studies regarding the effect of leaf removal on thiols, we can just speculate that the effect of light exposure is positive in order to obtain a higher concentration in the grapes (Kobayashi et al., 2011). In case of ELR the lower concentration of 3MH and the higher concentration of 4MMP can be related with a transient delay in the maturation of the grapes, possibly created by the stressfull effect of pre-flowering leaf removal. 4MMP (ng/l) 10,0 9,0 8,0 7,0 6,0 5,0 4,0 3,0 2,0 1,0 0,0 UNT ELR LLR A 3MH (ng/l) UNT ELR LLR B Figure 9: Effect of leaf removal on the concentration of 4MMP (A) and 3MH (B) in the wines of Sauvignon blanc produced in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) 4.6 Sensory analysis of the wines The aroma quality of the wine represent a preview of what was confirmed by the taste. The different samples, served at 10 C, were examined by the panel as blind samples. The descriptors used in the scorecard were analysed and the most typical for 'Sauvignon blanc' and the most significant are presented in the figure 10. As regard alcoholic perception (figure 10), no difference among treatments were found, but a slightly lower value was revealed in case of ELR. 21

34 The acidic perception is a must of 'Sauvignon Blanc' as it is combined with its aromas and flavors. All samples were tested with a good level of acidity, being higher in the treatment UNT and significantly lower in LLR. Higher acidity is responsible for more freshness as it was observed in UNT. The bitterness descriptor was judged at low levels for all tested wines, with not significant differences among treatments. As regard wine body, similar evaluation was given to UNT and ELR wines, while slightly nearly significant higher values were given by tasters to LLR wines. Moving on aroma descriptors, eldelberry flower is a typical aromatic characteristic of the variety 'Sauvignon blanc'. By comparing the wines examined, a significant higher value was observed in case of the treatments UNT while similar evaluation was given to ELR and LLR. The aromatics related with other flower notes were nearly similar among treatments, but slightly lower in ELR and mainly in LLR treatments. No difference was ascertained by panelists as regard citrus fruits, even if a slightly higher value was obtained by LLR. More similar valuation was given by tasters for the tropical fruits descriptor, related with the reminiscence of lychee, pineapple, passion fruit, grapefruit. A regard the box tree note, another typical aromatic of 'Sauvignon blanc', again no differences were revealed in the comparison of wines, but slightly lower values were received by LLR and mainly ELR wines. 22

35 10 8 alcoholic perception 10 8 acidic perception 10 8 bitterness UNT ELR LLR body UNT ELR LLR eldelberry flowers UNT ELR LLR other flowers UNT ELR LLR citrus fruits UNT ELR LLR tropical fruits UNT ELR LLR box tree UNT ELR LLR 0 UNT ELR LLR 0 UNT ELR LLR Figure 10: Effect of leaf removal on the sensory characteristics of the wines of Sauvignon blanc produced in Oslavia (Gorizia, Italia) in the season Bars represent 95% confidence interval (n=3) 23

36 5 CONCLUSIONS To summarise the results presented, I will answer the hypothesis as described in the dedicated chapter: 1. To evaluate two timings of leaf removal pre-flowering and berry-set leaf removal and their effects on yield parameters (berry weight, cluster weight, cluster number, yield per vine and leaf area/yield ratio) in a vineyard of Vitis vinifera cv. 'Sauvignon blanc' cultivated in the Italian D.O.C. Collio region. With the early intervention of pre-flowering leaf removal, a 46% reduction of leaf area on main shoots was ascertained, while in case of late leaf removal the reduction was negligible. No substantial differences in the leaf area of the main shoots was ascertained at harvest, while in case of the lateral leaves a significant higher value were observed in case of the treatment ELR. This increase eventually resulted in a greater total leaf area of ELR treatment in comparison with the control UNT, while LLR remained intermediate. The average cluster weight was slightly reduced in case of ELR treatment, while no effects of LLR were ascertained. As related with the variation of cluster weight, also yield was slightly reduced in case of ELR while no differences were observed between UNT and LLR. The ratio leaf area-to-yield was found within the range of optimality for the treatments UNT and LLR, while significantly higher in the case of ELR. 2. To ascertain how the timing of leaf removal affect the accumulation of soluble solids or the degradation of acids during maturation and at harvest. ELR reported significantly higher content of total soluble solids as compared with UNT, while LLR showed no differences between both ELR and UNT. As opposite, titratable acidity at harvest was similar among treatments, even if slightly higher values were measured in case of ELR. As related with the values just reported, ELR showed significantly lower values of ph, while UNT and LLR reached similar ph values at harvest. 24

37 3. As regards thiols and thiol precursors, to understand how pre-flowering or berryset leaf removal could differently change their biosynthesis and the relative amount of 4MMP and 3MH in grapes and wine. The concentration of Cys-4MMP and Glut-4MMP was significantly higher in case of ELR and lower for LLR, while average concentration was measured in UNT grapes. As regard Cys-3MH no differences were shown among treatments, while for Glut-3MH, both leaf removal treatments were profitable to increase the amount as compared with UNT, particularly in case of LLR. Similarly to Cys-precursors, pre-flowering and post-flowering leaf removal accounted for a shift in the thiol profile, with higher levels of 4MMP and lower of 3MH in case of the former treatment, and lower values of 4MMP in case of the latter treatment. 4. To evaluate the effect of the timing of leaf removal on the sensory properties of the wines. The wines of the untreated sample were evaluated with a higher acidity in comparison with both leaf removal treatments. As regard the aromatic descriptors, in the untreated wines significantly higher scores were shown for elderberry flowers and slightly higher for box tree, while no differences were found between leaf removals. As a trend, a better perception of the body roundness was revealed in case of the pre-flowering leaf removal wines, while similar scores were obtained by UNT and ELR. Good ratings and no significant differences among treatments were obtained for the perceptions of the citrus and tropical fruits. 25

38 6 LITERATURE Arnold R. A., Bledsoe A. M. (1990) The Effect of Various Leaf Removal Treatments on the Aroma and Flavor of Sauvignon blanc Wine. American Journal of Enology and Viticulture, vol. 41, pages Belancic A., Agosin E. (2007) Methoxypyrazines in Grapes and Wines of Vitis vinifera. American Journal of Enology and Viticulture, vol. 58, no 4, pages Brancadoro L., Tamai G., Sartori E., Anaclerio F., Forti R., Colautti M., Petrussi C., Ceccon D. (2012). Quaderni tecnici VCR 08. I cloni originali VCR di Sauvignon. Rauscedo, Italy: Vivai Cooperativi Rauscedo. Camussi A., Möller F., Ottaviano E., Sari Gorla M. (1995) Metodi statistici per la sperimentazione biologica. Bologna: Zanichelli. Diago M. P., Ayestaran B., Guadalupe Z., Poni S., Tardaguila J. (2012) Impact of prebloom and fruit set basal leaf removal on the flavonol and anthocyanin composition of Tempranillo grapes. American Journal of Enology and Viticulture, vol. 63, no. 3, pages Feng H. F., Yuan F., Skinkis P.A., Qian M.C. (2015) Influence of cluster zone leaf removal on Pinot noir grape chemical and volatile composition. Food Chemistry, vol. 173, pages Gatti M., Bernizzoni F., Civardi S., Poni S. (2012) Effects of cluster thinning and preflowering leaf removal on growth and grape composition in cv. Sangiovese. American Journal of Enology and Viticulture, vol. 63, no. 3, pages Genovese M., Nuzzo V., Shackel K. A., Matthews M. A. (2010) Scottature solari su vite, come e quando insorgono e come evitarle. Informatore Agrario, vol. 66, no 23, page 45. Kliewer W. M., Dokoozlian N. K. (2005) Leaf area/crop weight ratios of grapevines: influence on fruit composition and wine quality. American Journal of Enology and Viticulture, vol. 56, no.2, pages

39 Kobayashi H., Takase H., Suzuki Y., Tanzawa F., Takata R., Fujita K., Kohno M., Mochizuki M., Suzuki S., Konno, T. (2011) Environmental stress enhances biosynthesis of flavor precursors, S-3-hexan-1-ol-glutathione and S-3-hexan-1-ol-L-cysteine, in grapevine through glutathione S-transferase activation. Journal of Experimental Botany, vol. 62, pages Koch A., Ebeler S.E., Williams L.E., Matthews M. (2012) Fruit ripening in Vitis vinifera: light intensity before and not during ripening determines the concentration of 2-methoxy-3-isobutylpyrazine in Cabernet Sauvignon berries. Physiologia Plantarum, vol. 145, pages Krajniger I. (2014) Vpliv zgodnjega odstranjevanja listov pri sortah Vitis vinifera 'Cabernet Sauvignon' in 'Merlot' na osnovne kakovostne parametre in vsebnost metoksipirazinov v grozdju.»diplomsko delo«, Nova Gorica: Univerza v Novi Gorici. Lee J., Skinkis P. A. (2013) Oregon Pinot noir grape anthocyanin enhancement by early leaf removal. Food chemistry, vol. 139, pages Lisjak K., Janeš L., Velikonja Bolta Š., Bavčar D., Čuš F, Požgan F., Vanzo A. (2013) Determination of thiol precursors in Sauvignon Blanc grapes. V: Biasioli, F. (ed.). Book of abstracts. San Michele a/adige - Italy: Research and Innovation Centre Fondazione Edmund Mach, pages Matsuyama S., Tanzawa F., Kobayashi H., Suzuki S., Takata R., Saito H. (2014) Leaf removal accelerated accumulation of delphinidin-based anthocyanins in Muscat Bailey A [Vitis labruscana (Bailey) and Vitis vinifera (Muscat Hamburg)] grape skin. Journal of the Japanese Society for Horticultural Science, vol. 83, no. 1, pages Mescalchin E., Bottura M., Cainelli R., Fellin F., Gobber M., Lucin R., Margoni M., Mattedi F., Michelotti F., Patton A., Penner F., Ribolli F. (2008) Sfogliare precocemente la vite per evitare scottature e botrite. Informatore Agrario, vol. 64, no 17, pages Pastore C., Zenoni S., Fasoli M., Pezzotti M., Tornielli G. B., Filippetti I. (2013) Selective defoliation affects plant growth, fruit transcriptional ripening program and flavonoid metabolism in grapevine. BMC Plant Biology, vol. 13, no. 1, page 1. 27

40 Poni S., Casalini L., Bernizzoni F., Civardi S., Intrieri C. (2006) Effects of early defoliation on shoot photosynthesis, yield components, and grape composition. American Journal of Enology and Viticulture, vol. 57, no. 4, pages Poni S., Bernizzoni F., Civardi S., Libelli N. (2009) Effects of pre-bloom leaf removal on growth of berry tissues and must composition in two red Vitis vinifera L. cultivars. Australian Journal of Grape and Wine Research, vol. 15, pages Ribéreau-Gayon P., Dubourdieu D., Donèche B., Lonvaud, A. (2005) Handbook of Enology. The Microbiology of Wine and Vinifications. Vol. 1, 2 nd Edition. Chichester, England: John Wiley & Sons Ltd. Risco D., Pérez D., Yeves A., Castel J.R., Intrigliolo D.S. (2014) Early defoliation in a temperate warm and semi-arid Tempranillo vineyard: vine performance and grape composition. Australian Journal of Grape and Wine Research, vol. 20, no. 1, pages Ryona I., Pan B.S., Intrigliolo D.S., Lakso A.N., Sacks G.L. (2008) Effects of cluster light exposure on 3-isobutyl-2-methoxypyrazine accumulation and degradation patterns in red wine grapes (Vitis vinifera L. Cv. Cabernet Franc). Journal of Agricultural and Food Chemistry, vol. 56, pages Scafidi P., Pisciotta A., Patti D., Tamborra P., Di Lorenzo R., Barbagallo M. G. (2013). Effect of artificial shading on the tannin accumulation and aromatic composition of the Grillo cultivar (Vitis vinifera L.). BMC Plant Biology, vol. 13, page 175. Schneider C. A., Rasband W. S., Eliceiri K. W. (2012) NIH Image to ImageJ: 25 years of image analysis. Nature Methods, vol. 9, no. 7, pages Soban M. (2015) Change in yield parameters and grape anthocyanin profile of Vitis vinifera l. 'Refošk' as affected by cluster thinning and pre-flowering leaf removal.»diplomsko delo«, Nova Gorica: Univerza v Novi Gorici. Sternad Lemut M., Sivilotti P., Franceschi P., Wehrens R., Vrhovsek U. (2013). Use of metabolic profiling to study grape skin polyphenol behavior as a result of canopy 28