AJEV Papers in Press. Published online March 20, 2012.

AJEV Papers in Press. Published online March 20, 2012. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Effects of Cluster Thinning and Pre-Flowering Leaf Removal on Growth and Grape Composition in cv. Sangiovese Matteo Gatti, 1 Fabio Bernizzoni, 1 Silvia Civardi, 1 and Stefano Poni 1 * 1 Istituto di Frutti-Viticoltura, Università Cattolica del Sacro Cuore, Via Emilia Parmense 84, 29100 Piacenza, Italy. *Corresponding author (email: stefano.poni@unicatt.it; tel +390523599271; fax +390523599268) Acknowledgments: This project was partially funded by CRPV, Regione Emilia-Romagna. The authors thank David Verzoni for restyling the English text. Manuscript submitted Nov 2011, revised Jan 2012, accepted Feb 2012 Copyright 2012 by the American Society for Enology and Viticulture. All rights reserved. Abstract: Crop regulation techniques applied as pre-flowering defoliation (D) and early cluster thinning performed at the same time (ECT) or at lag-phase of berry growth (LCT) were tested over three seasons on high-yielding cv. Sangiovese (Vitis vinifera L.) and compared to non- defoliated, un-thinned control (C) vines of the same cultivar. Treatment severity consisted of removing primary leaves and any laterals developed from nodes 1 to 6 in D and by thinning 50% of clusters chosen among distal ones or those inserted on weak shoots, in the CT plots. Despite the fact that yield per vine was not as reduced in D (-32%) as in CT treatment (-45%) in comparison to C, early defoliated vines also had largely improved sugar and total anthocyanin concentrations and highest total phenolics. Yield components were also markedly affected by treatments: D vines showed smaller clusters and berries, leading in turn to improved cluster looseness, as well as higher relative skin and seed growth. While all crop-regulating treatments led to an increase in the final leaf-to-fruit ratio, parameters of technological maturity were essentially uncoupled as equally high Brix levels corresponded to the highest TA in D and, conversely, to lowest TA and highest ph in cluster thinning treatments. Overall results show that different final yield-grape composition patterns can be reached depending upon the principle used for crop regulation as a primary consequence of a diversified degree of compensation triggered on single yield components. Key words: leaf area, fruit-set, cluster weight, skin growth, leaf-to-fruit ratio 1

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 With about 67,000 hectares Sangiovese is currently the most grown Vitis vinifera L. variety in Italy and, despite a considerable decreasing trend over the last five years, it still accounts for about 11% of Italy s vineyard acreage (www.istat.it). While being primarily grown in Tuscany, Emilia-Romagna and Marche regions for making wines ranging from everyday table reds to classic, renowned DOC appellations like Chiantis and Brunellos, Sangiovese retains the notoriety of not being an easy to grow cultivar and reports of its performance overseas underscore quite variable or below-expectation results (Fish 2003). In effect, even when grown in the most suitable environments, Sangiovese requires careful handling of a natural tendency to easily over-crop due to its notably high fruitfulness of shoots regardless of their origin (primary or secondary buds, base buds). Thus, while growing Sangiovese in low-to-moderate vigour sites may contribute to reducing crop potential, yield usually needs additional regulation through traditional summer pruning operations like shoot and/or cluster thinning. However, manual cluster thinning is time- consuming, requires skilled labour and its effects are notably uneven (Ough and Nagaoka 1984). Indeed, despite significant reductions in yield, it may have no impact on grape quality, and final cropping can be higher than expected because of compensatory berry growth (Keller et al. 2005). Among the vineyard practices which can have an impact on crop level through effects conveyed by alterations of the source potential at specific growth stages while not acting directly on cluster number, pre-bloom leaf removal has gained increased popularity (Poni et al. 2006). The technique is based on the well-established relationship by which carbohydrate supply at flowering is a primary determinant of fruit set (Coombe 1962, Caspari and Lang 1996) and, as a result, a significant source limitation imposed at this stage by removing basal and fully functional leaves reduces fruit set and/or final berry size. A number of studies (Poni et al. 2006, 2

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 Sabbatini and Howell 2010, Tardaguila et al. 2010, Palliotti et al. 2011) conducted under an array of sites and genotypes, including Sangiovese (Intrieri et al. 2008), on the effects of pre- flowering leaf removal, either manual or mechanical, have shown that defoliated vines evince significant yield reduction, improved grape composition and higher wine appreciation in a wide range of cases. While the constancy of crop-controlling effects can be reasonably accounted for by the robust relationship between fruit-set and carbohydrate supply prior to flowering, grape composition enhancement embraces a series of possible intervening factors related to seasonal variations in source-sink balance (Palliotti et al. 2011), cluster microclimate (Tardaguila et al. 2010) and differential growth of berry organs (Poni et al. 2009). Indeed, seasonal monitoring of the whole-canopy, net CO 2 exchange (NCER) rate carried out with an enclosure system on both non-defoliated and pre-bloom defoliated (D) Sangiovese vines has shown that NCER/yield (shoot basis) increased by 38 % in D vines and, hence, resulted in largely enhanced carbohydrate content for ripening (Poni et al. 2008). More recently, Palliotti et al. (2010) have shown in cvs. Sangiovese and Ciliegiolo that a pre- bloom source limitation strong enough to cause the desired effects of reduced fruit set and berry growth can be achieved through timely, non-invasive application of a film-forming anti- transpirant compound. Two sprays applied at pre-bloom made it possible to limit leaf assimilation and transpiration rate by 30-70% for several weeks after treatment as compared to an unsprayed control, while allowing full leaf function recovery upon natural degradation. Overall, such techniques considerably broaden the range of practices suitable to control yield in the vineyard. Thus, based on their objective differences in terms of yield-regulating mechanisms, a realistic hypothesis is that different techniques might ultimately result in very similar cropping reduction while final grape composition would still considerably differ as a 3

79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 function of the response of specific yield components like berry number, size, relative skin growth and so on. The goal of this study was to provide a comparative three-year evaluation of effects exerted by manual cluster thinning and pre-flowering leaf removal on vegetative growth, yield components and grape composition of field grown Sangiovese grapevines. Plant material and experimental layout Materials and Methods The trial was run over three years (2007-2009) in a commercial, non-irrigated vineyard of Vitis vinifera L. cv. Sangiovese clone 12T grafted to K5BB stock established in 1999 at Tebano, Azienda Terre Naldi, (44.2 N, 11.5 E, 72 m a.s.l.), Italy. Vine spacing was 3.0 m x 1.1 m (inter- and intra-row) for a resulting density of 3,030 plants/ha and vines were trained to a vertically shoot-positioned (VSP), spur-pruned cordon trellis with a bud-load of about 15 count nodes per vine distributed on 7-8 two-node spurs per vine. The cordon was raised 1 m from the ground with three pairs of surmounted catch wires for a canopy wall extending 1.3 m above the cordon. The soil is a silty-clay-loam and its hydrological constants (field capacity and wilting point) were 38.2% and 21.5% (by volume), respectively, as calculated after Saxton et al. (1986) based on texture (10% sand, 55% loam, 35% clay), organic matter content (1.1%) and salinity (0.18 ms/cm). The resulting maximum field available water was therefore estimated at about 140 mm/m of soil depth and apparent soil density was calculated as 1.30 g/cm 3. For each season, the monthly weather records for mean daily air temperatures ( C) and daily rainfall (mm) from April to September as measured by a nearby weather station are reported in Figure 1. Three adjacent rows were selected to build a randomized complete block design (RCBD) with each row as a block. Within each row, four panels of 3-post spaces (16.5 m row length, 15 vines 4

102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 per panel) were ribbon-tagged and randomly assigned to the following treatments: defoliated (D), early cluster thinning (ECT), late cluster thinning (LCT) and non-thinned, non-defoliated control (C). Treatments D and ECT were concurrently imposed at stage H ( flowers separated, Baggiolini 1952) by manually removing primary leaves and any lateral developed from nodes 1 to 6 on each shoot and by thinning 50% of clusters. Distal clusters and those inserted on weak shoots were preferentially removed until the targeted thinning threshold was reached. LCT was applied following the same criterion during the lag-phase of berry growth (on average, 2 nd week in July). Dates for D and ECT treatments were 14 May in 2007, and 22 May in both 2008 and 2009. Ten shoots per row and treatment combination (two shoots on each of five vines) growing at the apical position on a spur were then randomly chosen and tagged for subsequent measurements. Fruit-set estimate, berry growth, yield and fruit composition Each basal cluster per tagged shoot was photographed against a dark background with a digital camera held perpendicular to the inflorescence on the day of defoliation. The regression equation between actual flower (y) number and the number of flowers counted on photo prints (x) previously determined for cv. Sangiovese by Poni et al (2006) was used to estimate initial flower number on tagged inflorescences. Defoliated leaf area and final total leaf area per shoot were determined upon measuring the surface of each lamina via a leaf area meter (LI-3000A, LI-COR Biosciences, Lincoln, NE), with main and lateral contributions being kept separate. At harvest dates on 6, 16 and 8 September in 2007, 2008 and 2009, respectively, the tagged clusters were individually picked, immediately weighed and the number of berries counted; the weight of any distal clusters was recorded 5

125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 separately. Compactness was visually estimated on each basal test cluster using code OIV 204 (Organisation Internationale de la Vigne et du Vin 1983), which ranks as 1 berries in grouped formation with many visible pedicels and as 9 misshapen berries. Must composition as soluble solids concentration (Brix), ph and titratable acidity (TA) was assessed on 50-berry samples taken from the basal clusters of tagged shoots. Total soluble solids were determined by a temperature-compensating refractometer (RX-5000 Atago-Co Ltd, Tokyo, Japan). TA was measured by a Crison Compact Titrator (Crison, Barcelona, Spain) with 0.1 N NaOH to a ph 8.2 end point and expressed as g/l of tartaric acid equivalents. A second set of 50-berry samples was taken from the same clusters and total anthocyanins and phenolics were determined after Iland (1988). The sample berries were homogenized at 20000 rpm with an Ultra-Turrax (Rose Scientific Ltd, Alberta, Edmonton, Canada) homogenizer for one min, then two grams of the homogenate were transferred to a pre-tared centrifuge tube, enriched with 10 ml aqueous ethanol (50%, ph 5.0), capped and mixed periodically for an hour before centrifugation at 959 x g for five min. A portion of the extract (0.5 ml) was added to 10 ml 1M HCL, mixed and let stand for three hours; then the absorbance values were registered at 520 nm and 280 nm on a Kontron (Tri-M Systems and Engineering Inc., Toronto, Ontario, Canada) spectrophotometer. Total anthocyanins and phenolics were expressed as mg per g of fresh berry mass. In 2008 and 2009, an additional sample of ten berries was taken from each test cluster. Each berry was weighed and then sliced in half with a razor blade, the seeds and flesh carefully removed from each berry half using a small metal spatula without rupturing any pigmented hypodermal cells and the seeds carefully separated by hand from the flesh. Both skins and seeds were rinsed in de-ionized water, blotted dry and weighed. 6

148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 Statistical analyses Vine performance data were subjected to a two-way analysis of variance using the SigmaStat software package (Systat Software, Inc. San Jose, CA, USA). Year was considered as a random variable and the error term for the treatment factor was the year treatment interaction mean square. Treatment comparison was performed each year by Student-Neuman-Keuls test at P 0.05. Year x treatment interaction was partitioned only in case of F test significance. Visual ratings of cluster compactness were subjected to square root transformation prior to analysis. Results Active heat summation calculated as base 10 C degree days (DD) from 1 April until 30 September was lowest in 2008 with 1893 DD and highest in 2009 at 2068 (Figure 1). Cumulated precipitation was highest in 2008 (252 mm); 2007 had the highest rainfall during the month preceding harvest. Final total leaf area per vine and per shoot as well as leaf area per shoot due to primary leaves were lowest in D vines (Table 1), whereas, on a shoot basis, lateral development did not differ among treatments (Table 1). The fraction of leaf area removed with D was about 33% of final total leaf on a three-year basis (not shown). Given a fairly homogeneous flower number per cluster across treatments (about 300), berry number per cluster was, in LCT and D, respectively, higher and lower as compared to C vines, whereas fruit set was reduced only by the defoliation treatment (-7.1% vs. C), which in turn led to less compact clusters (Table 1). While berry number per cluster varied considerably across seasons, D was not able to affect this parameter in the first trial year as compared to C. Conversely, defoliation led to lower berry number than C in the two following years (Figure 2). 7

170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 Cluster number per vine recorded at harvest confirmed that a thinning severity set at about 50% of the pending crop mirrored an average decline in yield per vine of 45% in both thinning treatments as compared to control due to somewhat heavier clusters (Table 2). However, cluster thinning did not affect berry size, which, conversely, was markedly reduced along with cluster weight by early defoliation. Smaller clusters in D vines, associated with unchanged cluster number per vine when compared to C, resulted in a 32% crop yield reduction. Note, however, that yield per vine and its main components (berry and cluster weight, cluster number per vine) showed a significant year x treatment interaction, whose partitioning is shown in Figures 3 A-D. Generally speaking, yield per vine differences between crop-regulating techniques and control were amplified in 2007 and 2009 and less pronounced in 2008 (Fig. 3A). Then, too, berry weight was drastically affected by D in 2007 and to a somewhat lesser extent the two following seasons (Fig.3B), whereas cluster-thinned treatments strengthened the tendency to bear heavier clusters than control in 2008 and 2009 (Fig. 3C). All treatments raised the final leaf area-to-yield ratio as compared to C, although the maximum gain was achieved by the cluster-thinned vines scoring the top value of 1.65 m 2 /kg (Table 2). Must soluble solids concentration was improved by at least two degrees Brix in all treatments, although they differed in term of their response to TA and ph. These parameters decreased and increased, respectively, in cluster thinned vines as compared to C, whereas the opposite took place for D (Table 2). Like sugar concentration, total anthocyanin concentration (mg/g) was markedly improved by all treatments, whereas total phenolics were highest in D while also scoring higher values in cluster thinned treatments as compared to control (Table 2). The two-year assessment of yield-regulation effects on berry organs confirmed that differences in total berry weight were driven by flesh weight since absolute skin weight was very 8

193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 similar among treatments (Table 3). These differences led to D vines having higher relative skin and seed growth as compared to the other treatments. Partitioning of the year x treatment interaction for fractional skin-to-berry weight ratio (Figure 4) showed that relative differences among treatments were maintained each season, although a larger effect was seen in 2008. Discussion Cluster thinning applied at two dates during the season and pre-bloom leaf removal were effective at reducing yield in Sangiovese vines while achieving remarkable improvement in sugar, total phenolic and anthocyanin concentrations. Despite this shared effect, our study underscores that physiological and yield-regulating mechanisms underpinning such responses are quite distinct and the response to specific technological parameters like TA and ph is likewise different. In the absence of major compensation from cluster yield components, yield per vine decreased about 45%, i.e. only 5% less than the applied cluster thinning severity. While this predictive power is decidedly more complicated to apply as an indirect method impacting on yield when a source limitation is used, data show that the amount of leaves removed via defoliation ( 33% of final leaf area per shoot) was very similar to the overall yield constraint exerted by this treatment (32%) and that, since cluster number per vine was unaffected, both berry number per cluster and berry size converged to reduce yield. A look at the interactive effects exhibited in Figure 2 shows that while berry number was unaffected by D in trial year 1, it became very sensitive the two following seasons. This response links to the vine s capacity to initially buffer the strong foliar stress brought about by D either through mobilising reserves from permanent organs or via significant photosynthetic compensation from the remaining leaves (Palliotti et al. 2010). However, if the treatment is reiterated over seasons, this buffer 9

216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 capacity appears to fade and the vine system becomes more susceptible to the limitation imposed by leaf removal. Quite interestingly, although final crop per vine in D was higher than in thinned treatments (12% more on average), clusters in D were nevertheless less compact and, hence, less susceptible to cluster rot, thus reinforcing the role of a single yield component in affecting both total yield and cluster morphology. Regardless of timing of cluster thinning (the unseasonably early ECL was included in this study to provide a synchronous treatment with D), there was no compensation due to increased berry growth (Fig. 3B), whereas cluster weight was heavier due to higher berry number, albeit with variability across years. The substantial increase in berry number per cluster in both thinning treatments in 2008 seems to result from a carryover effect from the 2007 thinning, which considerably and instantaneously raised the leaf-to-fruit ratio and, hence, the amount of assimilates available for reserve storage that, according to May (2004), might have quite favourable effects on the number of flowers/cluster at bud-break the following season. Indeed, the yield variability across seasons shown in Fig 3A for ECT and LCT suggests a tendency towards an alternate bearing pattern despite identical fractional reduction of pending crop each year. Lavezzi et al. (1995) applied the same thinning treatments (25 and 50% crop removed at fruit-set ) on cv. Prosecco for three consecutive seasons and ended up with no differences in yield between treatments due to a progressively strong compensation in terms of enhanced bud fertility by the thinned vines. Vine performance as the balance between yield and final must composition was indeed quite positively affected by all treatments, although different mechanisms can again be envisaged. Control vines delivered an estimated yield/ha of 22 t (three-year mean), which is definitely in excess compared to Sangiovese standards for appellation and, not surprisingly, scored just fair 10

239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 sugar concentration, colour and phenolics. This response fits with the final total leaf-to-fruit ratio of about 1 m 2 /kg, which is below the optimum threshold if relationships among total, external and exposed leaf area fractions are taken into account (Kliewer and Dokoozlian 2005, Mabrouq and Sinoquet, 1998). Note that all treatments induced a change in this ratio, which for CT treatments is quantitatively more important and primarily the outcome of the direct reduction of the pending crop, whereas in D it derives from a source limitation-triggered effect that, as shown elsewhere (Poni et al. 2006), was strong enough to limit yield more than proportionally as compared to leaf removal. Moreover, the very good grape composition scored by D supports the assumption that post-defoliation canopy function foliage was enhanced due to lower shoot age and improved photosynthetic compensation by the retained leaves (Candolfi-Vasconcelos and Koblet, 1991). On the other hand, enhanced grape composition in the CT treatments confirms Sangiovese s attitude to positively react to such a technique (Filippetti et al. 2007). A look at berry pigmentation shows that differences in relative skin weight among yield- regulating systems did not correspond to differential effects on anthocyanin concentration, which was conversely very similar. This holds with the hypothesis that having smaller berries, hence a higher skin-to-berry ratio, does not necessarily reflect an increased colour as compared to larger size berries (Roby and Mathews 2004, Walker et al. 2005). Although canopy microclimate changes were not specifically quantified in this study, a recent contribution by Di Profio et al. (2011) has reported no alteration in canopy microclimate assessed through point quadrat analysis in control and cluster thinned Cabernet Sauvignon and Cabernet Franc vines, whereas cluster exposure and canopy gaps were sharply increased in vines subjected to a pre-veraison leaf removal. Indeed, differences in cluster microclimate between C and CT vines could relate to compensating vegetative growth triggered in CT by reproductive 11

262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 sink removal; however, Keller at al. (2005) have reported in a five-year-study carried out on cvs. Cabernet Sauvignon, Riesling and Chenin blanc that cluster thinning and its timing had little or no influence on shoot growth, leaf area and pruning weight. Then, data reported in Table 1 also suggest that an even very much anticipated cluster thinning was not able to trigger a significant vegetative growth compensation as leaf area development. Therefore, hypothesis holds that treatments sharing similar cluster microclimate (i.e. C vs. ECT, LCT) did show differences in berry colour, whereas other treatment combinations (i.e. D vs. ECT, LCT) quite likely marked by more pronounced changes in cluster exposure did not evince any differences in berry pigmentation. Yet, regardless of the effects on grape composition reported in the present study, and in agreement with earlier findings related to the red cvs. Barbera and Lambrusco (Poni et al. 2009), D confirms its effectiveness in promoting skin growth at the expense of flesh growth, thereby favouring the formation of berries with a thicker skin. Besides advantages related to such a feature like improved cluster rot resistance and attenuated must loss during mechanical harvesting, constancy of its effect across an array of genotypes and environments supports the assumption that, despite the drastic yet temporary source limitation imposed at pre-flowering, developing ovaries and young berries must benefit from the improved light and thermal microclimate. Indeed, cell division at either the pericarp or exocarp level, which in seeded cultivars is reported to be usually completed within a maximum of 38-40 days after anthesis (Hardie et al. 1996), seems to be quite sensitive to temperature. In cv. Tokay, thickness and cell number of berry pericarp held at 40 C during flowering and fruit set largely decreased as compared to vines kept at 25 C (Kliewer 1977). A study by Harris et al. (1968) showed that Sultana berries grown at a mean daily minimum of 20 C in greenhouse developed more quickly than those at two field locations where the minimum daily temperatures were 18 and 16 C, 12

285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 respectively. In this study diurnal mean air temperature calculated for the days of leaf removal was 23.3, 16.3 and 23.5 C in 2007, 2008 and 2009, respectively (Figure 1), thereby suggesting that cluster temperature increase subsequent to defoliation might have promoted, especially in 2008, cell-division rates and, hence, skin formation. The data shown in Figure 4 support the finding that maximum relative skin growth and increase in comparison to C vines were larger in 2008 than in 2009. A distinctive feature of this trial was that while soluble solids and anthocyanin concentration were similarly affected by treatments, there was clearly un uncoupling with the technological maturity parameters TA and ph, which in CT treatments showed a trend expected from the standpoint of advanced maturity but D went in the opposite direction. The ability of D to enhance ripening while preserving total acidity mirrors findings previously reported for cv. Trebbiano (Poni et al. 2006) and support the viability of the technique within a scenario of global warming where maintenance of sufficient acid levels is becoming a real challenge, especially for white wines and young reds (Schultz and Stoll, 2010). Although malic and tartaric fractions of TA were not determined in this study, if we posit a link to the work by Kliewer and Schultz (1964) reporting higher amounts of 14 CO 2 incorporated into tartaric acid for berries held in full sun as compared with amounts recovered in shaded berries, it is likely that the synthesis of tartaric acid was promoted by improved light and temperature microclimate due to D, which in the long run counter-balanced higher sensitivity to temperature-driven malic acid degradation. In the end, D vines, while sharing very similar must sugar concentration and anthocyanin accumulation with CT treatments, showed higher grape phenolics at harvest. Although the mechanisms involved in this response are complex, improved light microclimate in D very likely impacted on flavonol glucoside accumulation, whose synthesis largely increases in exposed 13

308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 berries as compared to shaded ones (Price et al. 1995, Haselgrove et al. 2000). Then too, Downey et al. (2004) examined pro-anthocyanidin content and composition throughout berry development in both shaded and exposed cv. Shiraz clusters and highlighted significant differences in both content and composition throughout the intermediate stages of berry development, with shaded fruit reaching a much lower maximum in pro-anthocyanidin content than exposed fruit. A recent paper by Kemp et al. (2011) shows that fairly early timing of leaf removal (7 days after flowering) increased pro-anthocyanidin concentration in Pinot Noir wines as compared to later timing, albeit without significantly affecting the mean degree of polymerisation. Conclusions Although based on totally different principles, leaf removal conducted on cv. Sangiovese at pre- bloom confirms its being a viable alternative to traditional cluster thinning for crop regulation in the vineyard. In addition to being suitable to mechanization, its other advantages include the possibility of supporting a higher crop at very similar sugar and anthocyanin concentrations while achieving other desirable features such as lower cluster compactness and increased relative skin and seed growth. If this pattern is linked to D s evident capacity to maintain higher TA and lower ph, the overall indications point to its specific suitability for high-yielding vineyards with big compact clusters established in areas where maintenance of adequate final acidity in a scenario of ongoing global warming and more frequent peak temperatures in summer are becoming priorities. Calibration provided in our study indicates that removing the first six main leaves and any laterals developed at that time results in a crop reduction of about 32% as compared to the non-defoliated control. 14

332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 Cluster thinning, regardless of its timing of application (whether at pre-flowering or lag phase of berry growth), confirmed its effectiveness in Sangiovese at enhancing grape composition. Under our specific trial conditions, the feared effects of berry growth compensation by the retained clusters either was not seen or was limited to a year sensitive to the effect of increased berry number, whereas data of final leaf-to-fruit ratio support the assumption that cluster removal is a needed tool to adjust an otherwise source-limited vine balance. Albeit still bound to manual execution by skilled personnel and, hence, expensive, the technique seems to find its best use in high-cropping genotypes having, on average, two inflorescences per shoot and grown in vigorous yet cool environments where excessive total acidity and immature tannins at harvest can dim the prospects of making good aged red wines. Literature Cited Baggiolini, M. 1952. Les stades repérés dans le développement annuel de la vigne et leur utilisation pratique. Revue Romande d Agriculture, de Viticulture et d Arboriculture 1: 4-6. Candolfi-Vasconcelos, M.C., and W. Koblet. 1991. Influence of partial defoliation on gas exchange parameters and chlorophyll content of field-grown grapevines Mechanisms and limitation of the compensation capacity. Vitis 30:129 141. Caspari, H.W., and A. Lang. 1996. Carbohydrate supply limits fruit-set in commercial Sauvignon blanc grapevines. In Proceedings for the 4 th International Symposium on Cool Climate Enology and Viticulture, Henick-Kling T. et al. (Eds.), pp. 9-13. New York State Agric. Exp. Sta., Geneva, NY. Coombe, B.G. 1962. The effect of removing leaves, flowers and shoot tips on fruit-set in Vitis vinifera L. J. Hort. Sci. 37:1-15 Di Profio, F., A G. Reynolds and A. Kasimos. 2011 Canopy management and enzyme impacts on Merlot, Cabernet franc, and Cabernet Sauvignon. I. Yield and berry composition. Am. J. Enol. Vitic. 62:139-151. Downey, M., J. Harvey and S. Robinson. 2004. The effect of bunch shading on berry development and flavonoid accumulation in Shiraz grapes. Austr. J. Grape and Wine Res. 10:55-73. Filippetti, I., S. Ramazzotti, M. Centinari, B. Bucchetti and C. Intrieri, C. 2007. Effects of cluster thinning on grape composition: preliminary experience on Sangiovese grapevines. Acta Hort. 754:227-234. Fish, T. 2003. California Sangiovese losing ground. Wine Spect. 11. 15

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 Hardie W.L, T.P. O Brien and V.G. Jaudzems V.G. 1996. Morphology, anatomy and development of the pericarp after anthesis in grape, Vitis vinifera L. Austr. J. Grape and Wine Res. 2:97 142. Harris, J.M., P.E. Kriedemann and J.V. Possingham. 1968. Anatomical aspects of grape berry development. Vitis 7: 106 119. Haselgrove, L., D. Botting, R. van Heeswijck, P.B. Høj, 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 and Wine Res. 6:141 149. Iland, P.G. 1988. Leaf removal effects on fruit composition. In Proceedings for the Second Cool Climate Viticulture and Oenology Symp. R.E. Smart, and J.B. Robinson (eds.), pp. 137-138. Auckland, New Zealand. Intrieri, C., I. Filippetti, G. Allegro, M. Centinari and S. Poni. 2008. Early defoliation (hand vs mechanical) for improved crop control and grape composition in Sangiovese (Vitis vinifera L.). Austr. J. Grape and Wine Res. 14: 25-32. Keller, M., L.J. Mills, R.L. Wample, and S.E. Spayd. 2005. Cluster thinning effects on three deficitirrigated Vitis vinifera cultivars. Am. J. Enol. Vitic. 56: 91-103. Kemp, B.S., R. Harrison, and Creasy G.L. 2011. Effect of mechanical leaf removal and its timing on flavan-3-ol composition and concentrations in Vitis vinifera L. cv. Pinot noir wine. Austr. J. Grape and Wine Res. 17: 270-279. Kliewer, W.M., and H.B. Schultz. 1964. Influence of environment on metabolism of organic acids and carbohydrates in Vitis vinifera. II. Light. Am. J. Enol. Vitic. 15:119 129 Kliewer, W.M. 1977. Effect of high temperatures during the bloom-set period on fruit-set, ovule fertility, and berry growth of several grape cultivars. Am. J. Enol. Vitic. 4, 215 222. 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. Lavezzi A., A. Ridomi, L. Pezza, C. Intrieri C. and O. Silvestroni O. 1995. Effects of bunch thinning on yield and quality of Sylvoz-trained cv. Prosecco (Vitis vinifera L.). Riv. Vitic. Enol. 2: 35-40. Mabrouq, H., and H.Sinoquet. 1998. Indices of light microclimate and canopy structure of grapevines by 3D digitising and image analysis, and their relationship to grape quality. Austr. J. Grape and Wine Res. 4:2-13. May, P. 2004. Development after fertilisation. In Flowering and Fruitset in Grapevines. P. May (ed.), pp. 63 72. Lythrum Press, Adelaide. O.I.V.1983. Codes des caractères descriptifs des variétés et espèces de vitis. Dedon, Paris. 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. Palliotti, A., Poni S., Berrios J.G., Bernizzoni F. 2010. Vine performance and grape composition as affected by early-season source limitation induced with anti-transpirants in two red Vitis vinifera L. cultivars. Austr. J. Grape and Wine Res. 16, 426-433. 16

399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 Palliotti, A., M. Gatti and S. Poni. 2011. Early leaf removal to improve vineyard efficiency: gas exchange, source-to-sink balance and reserve storage responses. Am. J. Enol. Vitic. 62: 219-228. Poni, S, L. Casalini, F. Bernizzoni, S. Civardi, and C. Intrieri. 2006. Effects of early defoliation on shoot photosynthesis, yield components, and grape composition. Am. J. Enol. Vitic. 57: 397 407. Poni S., F., Bernizzoni, and S. Civardi. 2008. The effect of early leaf removal on whole-canopy gas exchange and vine performance of Vitis vinifera L. cv. Sangiovese. Vitis 47: 1-6. Poni, S., F. Bernizzoni, S. Civardi S. and N. Libelli. 2009. Effects of pre-bloom leaf removal on growth of berry tissues and must composition in two red Vitis vinifera L. cultivars. Austr. J. Grape Wine Res. 15: 185-193. Price, S., P. Breen, M. Valladao, and B. Watson.1995. Cluster sun exposure and quercitin in Pinot noir grapes and wine. Am. J. Enol. Vitic. 46: 187-194. Roby G., and M.A. Matthews M.A. 2004. Relative proportions of seed, skin and flesh, in ripe berries from Cabernet Sauvignon grapevines grown in a vineyard either well irrigated or under water deficit. Austr. J. Grape and Wine Res. 10: 74 82. Sabbatini, P. and G.S. Howell. 2010. Effects of early defoliation on yield, fruit composition, and harvest season cluster rot complex of grapevines. HortScience, 45: 1804-1808. Saxton, K.E., W.J. Rawls, J.S. Romberger, and R.J. Papendick. 1986. Estimating generalized soil-water characteristics from texture. Soil Sci. Soc. Amer. J. 50:1031-1036. Schultz, H.R., and M. Stoll. 2010. Some critical issues in environmental physiology of grapevines: future challenges and current limitations. Austr. J. Grape and Wine Res.16: 4 24. Tardaguila, J., F. Martinez de Toda, S. Poni, and M.P. Diago. 2010. Impact of early leaf removal on yield and fruit and wine composition of Vitis vinifera L. Graciano and Carignan. Am. J. Enol. Vitic. 61: 372-381. Walker R., Blackmore D.H., Clingeleffer P.R., Herridge G.H., Ruhl E.H., Nicholas P.R. 2005. Shiraz berry size in relation to seed number and implications for juice and wine composition. Austr. J. Grape and Wine Res. 11: 2 8. 17

Figure 1 Seasonal trends (1 April-30 September) of diurnal mean air temperature ( C) and rainfall (mm) recorded in 2007 (A), 2008 (B) and 2009 (C) near the trial vineyard. The first upward arrow indicates dates of ECT and D treatments and the second date of LCT. Bold arrows show harvest date. 18

Figure 2 Variation over years of berries per cluster for the different canopy management treatments in Sangiovese (2007 to 2009): ECT, early cluster thinning; LCT, late cluster thinning; D, defoliation, C, non-defoliated, non-thinned control vines. Vertical bars show SE for each treatment combination. Interactive SE = 17.6, n = 15. 19

Figure 3 Variation over years of yield per vine (A), berry weight (B), cluster weight (C) and clusters per vine (D) for the different canopy management treatments in Sangiovese (2007 to 2009): ECT, early cluster thinning; LCT, late cluster thinning; D, defoliation, C, non-defoliated, non-thinned control vines. Vertical bars show SE for each treatment combination. Interactive SE = 0.36 (A), 0.136 (B), 13.8 (C) and 1.34 (D), n = 15. 20

Figure 4 Variation over years of skin-to-berry ratio (%) for the different canopy management treatments in Sangiovese (2008 and 2009): ECT, early cluster thinning; LCT, late cluster thinning; D, defoliation, C, non-defoliated, non-thinned control vines. Vertical bars show SE for each treatment combination. Interactive SE = 0.78, n = 15. Table 1 Vegetative growth parameters, fruit-set components and cluster compactness in response to the three canopy management treatments: early cluster thinning (ECT), late cluster thinning (LCT) and defoliation (D). Data are means of three years (2007-2009). LA = leaf area. Treatment a Final LA/shoot (cm 2 ) Final main LA/shoot (cm 2 ) Final LA /vine (m 2 ) Berries/ cluster Fruit-set (%) Cluster compactness (OIV rating) Control 4015 ab 2552 a 7.6 a 115 b 37 a 7.1 ab ECT 4550 a 2702 a 7.7 a 124 ab 37 a 6.8 b LCT 4625 a 2767 a 7.4 a 132 a 40 a 7.4 a D 3537 b 2146 b 6.3 b 93 c 30 b 5.7 c Sig. b * * * ** * ** Treat. x year ns ns ns ** ns ns a Means separated within columns by Student-Newman-Keuls test. b *, **, and ns indicate significant at p 0.05, 0.01 and not significant, respectively. 21

AJEV PAPERS IN PRESS AJEV PAPERS IN PRESS Table 2 Yield components, must composition and leaf-to-fruit ratio (shoot basis) in response to the three canopy management treatments: early cluster thinning (ECT), late cluster thinning (LCT) and defoliation (D). Data are means of three years (2007-2009). ph Treatment a Berry weight (g) Cluster weight (g) Clusters/ vine Yield/ vine (kg) Soluble solids ( Brix) TA (g/l) Total phenols (mg/g) Total anth. (mg/g) Leaf-tofruit ratio (m 2 /kg) Control 2.33 a 265 b 29 a 7.5 a 21.2 b 3.18 c 6.8 b 3.43 c 1.33 b 1.03 c ECT 2.29 a 284 ab 15 b 4.2 c 23.3 a 3.28 a 6.5 c 4.20 b 1.57 a 1.69 b LCT 2.26 a 298 a 14 b 4.1 c 23.2 a 3.30 a 6.4 c 4.01 b 1.60 a 1.65 b D 2.02 b 191 c 26 a 5.0 b 23.4 a 3.22 b 7.0 a 5.03 a 1.64 a 1.21 a Sig. b ** ** ** ** ** ** ** ** ** ** Treat. x year ** ** ** * ns ns ns ns ns ns a Means separated within columns by Student-Newman-Keuls test. b *, **, and ns indicate significant at p 0.05, 0.01 and not significant, respectively. 22

AJEV PAPERS IN PRESS AJEV PAPERS IN PRESS Table 3 Growth of berry components (absolute and relative basis ) and seed number/berry in response to the three canopy management treatments: early cluster thinning (ECT), late cluster thinning (LCT) and defoliation (D). Data are means of two years (2008-2009). Treatment a Berry weight (g) Flesh weight (g) Skin weight (g) Seed weight (g) Skin-toberry ratio (%) Flesh-toberry ratio (%) Seed to berry ratio (%) Control 2.34 c 1.97 c 0.27 0.100 b 11.9 a 83.9 b 4.2 a ECT 2.44 d 2.06 d 0.27 0.104 b 11.5 a 84.2 b 4.3 a LCT 2.25 b 1.90 b 0.26 0.090 a 12.0 a 84.1 b 4.0 a D 2.11 a 1.73 a 0.27 0.104 b 13.4 b 81.7 a 4.9 b Sig. b ** ** ns ** ** ** ** Treat. x year ns ns ** ns ** ** ns a Means separated within columns by Student-Newman-Keuls test. b *, **, and ns indicate significant at p 0.05, 0.01 and not significant, respectively. 23