Sustainable Grape Productivity and the Growth-Yield Relationship: A Review

Transcription

1 Sustinble Grpe Productivity nd the Growth-Yield Reltionship 165 Honorry Reserch Lecture ASEV Annul Meeting, 1999, Reno, Nevd Sustinble Grpe Productivity nd the Growth-Yield Reltionship: A Review G. Stnley Howell* Reserch reports nd experimentl efforts during the lst century re presented with the gol to encourge discussion of blncing grpevine fruit yield nd vine growth nd lef re. Fruit nd subsequent processed qulity re eqully relevnt economic issues s we strive to crete conditions for both sustinble grpevine productivity nd vine cpcity for tolerting biotic nd biotic stress episodes. It is proposed tht methods to chieve vine blnce will vry with regrd to mcroclimte nd cultivr, but will be most criticl for those mcroclimtes commonly clled cool-climte regions. Regrdless, vine blnce is most redily understood when bsed on the principles of vine crbon blnce s medited through well-understood fctors such s cm 2 lef re/grm fresh weight of fruit t hrvest nd llmetric prctices s the Rvz Index nd the Growth- Yield Reltionship. Key words: Vine blnce, miniml pruning, lef re, crop rtio, Rvz Index, crbon blnce, photosynthesis, vine yield, vine size, fruit mturtion The chllenge of commercil grpevine culture nd production is the bility to consistently produce quntity of ripe grpes sufficient to cover ll costs of production nd return profit to the producer. There re numerous models on which this my be chieved. Vribles such s vlue of the cultivr, viticulturl mngement, perceived qulity of the crop, production costs, nd production consistency ll come together to determine whether the production is sustinble. In recent yers sustinble hs been co-opted s buzzword for vrious methods of culture, including integrted, bio-dynmic, nd/or orgnic. For purposes here the term is used in its erlier, simpler form; sustinble production hs both viticulturl nd economic dimensions. In this sense we sy tht sustinble production is collective methodology tht produces highest yields of ripe fruit per unit lnd re with no reduction in vine vegettive growth nd does so over period of yers t costs which return net profit. *Professor, Deprtment of Horticulture, Michign Stte University, Est Lnsing, MI [emil: howell@pilot.msu.edu] Acknowledgments: The efforts reported here would not hve been possible without the contributions of students nd postdoctortes: Teres Brros, Leh Clerwter, Chrles Edson, Timothy Mnsfield, Mike McLen, Dvid Miller, Pul Petrie, Bsil Stergios, Keith Streigler, Crmo Vsconcelos, nd Jmes Wolpert. The contributions of collegues nd friends re grtefully cknowledged: Peter Clingeleffer, Jmes Flore, Werner Koblet, Aln Lkso, Justin Morris, Robert Pool, Nelson Shulis, nd Robert Wmple. Encourgement nd finncil support by Ntionl Grpe Coopertive, Michign Grpe nd Wine Industry Council, nd Estern Viticulturl Consortium hve mde continued reserch possible. Commercil vineyrds tht llowed us to do reserch in their vineyrd provided gret, vlued trust. These include: Allegn County, Fenn Vlley Vineyrds; Berrien County, Dongvillow Vineyrds nd Tbor Hill Vineyrds; Grnd Trverse County, Chteu Chntl; Leelnu County, Boskydel Vineyrd, Good Hrbor Vineyrds, nd Leelnu Wine Cellrs; nd Vn Buren County, Cronenwett Vineyrds, Mohney Vineyrds, nd Bob Rogers Vineyrd. Finlly, I wish to express grtitude to those cdemic nd reserch institutions tht hve fostered my understnding of viticulture: New York Agriculturl Experiment Sttion, Genev; Lincoln University, Cnterbury, New Zelnd; Forshugsnstlt fur Obst-Wein- und Grtenbu, Wdenswil, Switzerlnd; nd my home institution for over 30 yers, Michign Stte University. Copyright 2001 by the Americn Society for Enology nd Viticulture. All rights reserved. Recently, Gldstones [10] used similr words to describe the term blnce. Blnce ws chieved when vegettive vigor nd fruiting lod re in equilibrium, nd consistent with high fruit qulity. The terms sustinble nd blnce re concepts consonnt with the mteril to follow. A few premises re worth noting. Viticulture in cool-climte portions of viticulturl production requires ccommodtion of those climtic fctors ner the limits of commercil grpe production. These environmentl limits re the bsis for the following premises: (1) for ny genotype-environment interction there is n optimum method of culture to chieve highest yields of ripe grpes of cceptble qulity over yers; (2) good viticulturl prctices must result from the ppliction of sound principles of vine growth nd development; (3) sustinble levels of highest fruit qulity t mximum yield cn occur only through the chievement of vine blnce through the ppliction of the lef re:fruit weight rtio or the Growth-Yield Reltionship. Miniml Pruning nd the Growth-Yield Reltionship The introduction of miniml pruning (MP) by Clingeleffer nd ssocites in Austrli [3-5,57,58] hs proven to be mjor brekthrough for winegrpe culture in tht region. It hs been shown to be superior to trditionl spur-pruned nd cne-pruned pproches in both sustinble yield nd fruit nd wine qulity. On the surfce, the dt would seem to chllenge the vlidity of premise 3. Clerly MP works well for winegrpes in Sunrysi, Austrli, nd the vine physiology on which the method is posited seems to suggest tht once vines equilibrte, the pproch should be brodly pplicble even in cool-climte 165

2 166 Howell regions. The following mteril is presented with two gols: (1) to shre the bsis of our concern bout pplicbility of MP technology in cool, short-seson viticulturl mcroclimtes; nd (2) to encourge crbon-budget pproch towrd finding solutions to problems involving n rry of biotic nd biotic stresses s well s vineyrd prctices. Vine Blnce Although the discussion of vine blnce hs expnded in recent yers, it is not new. Rvz [47] is the erliest source of relevnt informtion, nd the Rvz Index suggests tht the rtio of fruit to wood is the key to chieving both fruit qulity nd consistent production. He lso showed generl reltionship between lef production nd fruit production. As he ssessed the close reltionships between lef nd wood production, he chose the ltter for his Rvz Index, s he sought mens for viticulturists to put the reltionship into prctice. He chose this llometric pproch becuse he wnted growers to use it. In the erly 1920s, Prtridge [37-42] put forwrd very similr concept. He resoned tht vine produced two forms of yield ech growing seson: reproductive yield nd vegettive yield. Blnce ws chieved when yield of ripe fruit ws mximized with no detrimentl impct on vegettive growth. If the contribution by Prtridge hd stopped there, he would be of only pssing note; it ws merely modifiction of Rvz s concepts. His genius ws to tke the next step. He proposed to use the weight of cne prunings produced in yer 1 s n indictor of the upper limit of vine s cpcity to produce nd ripen crop in yer 2. While numerous fctors cn reduce yield in given yer, this upper limit ws mjor improvement in chieving vine blnce. This ws mjor step. Blnce s defined by the Rvz Index ws posthrvest evlution. It could tell the viticulturist how nerly ctul blnce hd been pproched, but only fter the fct. Prtridge clled his pproch the Growth-Yield Reltionship. I compre the contribution of Prtridge nd the subsequent prcticl refinements by Shulis [22,48-52,59] to tht of Drwin with reltion to orgnic evolution. Gould [11] rgues convincingly tht the ide of evolution hd been round for centuries prior to Drwin s time. The genius of Drwin ws his definition of mens by which it could be chieved nturl selection. Prtridge nd Shulis, nlogous to Drwin, produced prcticl methodology by which the process could be both explined nd put to prcticl use [37-39,42,48,52] to chieve blnce nd sustined production. The ppliction of vine blnce concepts is complicted by severl considertions: (1) grpevines re perennil plnts nd for tht reson the positive or negtive impct of seson s vineyrd mngement cn be mesured for one or more yers fterwrd [20]; (2) in cool-climte viticulturl regions there re strong nnul fluctutions in wether conditions during the growing seson [16]; nd (3) under conditions of high bud number reltive to vine cpcity, the weight of mture cnes reltive to lef re declines [32,33]; there is more lef re per unit weight of cnes. In ny event, prescription pproch to vineyrd mngement under such conditions is uncceptble s it limits both yield nd qulity in good vintges nd will yield unripe fruit nd reduced vine growth, mesured s vine size (Kg cne prunings/meter of row), or s ctul re of exposed folige, in poor vintges. Lef Are nd Crop Blnce As noted by Rvz, Prtridge, Shulis, nd subsequent reserchers [14,17,18,43,44], blnce my lso be considered s the mount of lef re required to ripen unit of crop weight. This is commonly expressed s cm 2 lef re /gm fresh weight of fruit. The literture reports rnge of 7 to 14 cm 2 /gm to chieve ripening. The proposl of 2 X rnge of difference immeditely ttrcts our ttention. Wht mkes it possible for cultivr to chieve vine blnce t 7 cm 2 in one culturl sitution nd require 14 cm 2 in nother will be ddressed in this discussion. Crop Blnce nd Growing Seson Length Grpevines cultured in region llowing significnt period of time posthrvest with vines retining functionl, exposed lef re will require less lef re to ripen the crop. This posthrvest period llows vine crop levels tht likely not only use the current seson s photosyntheticlly produced crbohydrtes but lso mobilize crbohydrtes stored in vegettive tissues [23,26,55]. A long folited period fter hrvest could llow the reccumultion of crbohydrtes in storge tissues tht will be necessry for the finl stges of bud nd inflorescence differentition nd support the spring growth flush in yer 2 [50]. Thus, long folited period posthrvest could potentilly ripen lrger crop per unit lef re. Crop Blnce nd Light Intensity Another likely fctor relted to crop blnce is light intensity over the growing seson [25]. Grpevine culture in Cliforni s Centrl Vlley or the Sunrysi district of Austrli is gretly fcilitted by high light intensity. Few dys in these viticulturl regions do not exceed the 800 to 1000 µ E m -2 s -1, which is sturtion for lef photosynthesis [25,53,54], nd mny dys the level is nerer 2000 µ E m -2 s -1. By contrst, cool-climte regions my be limited by growing seson length, light intensity, or both. Smrt [53] hs reported tht bout 8 to 10% of photosyntheticlly ctive rdition (PAR) striking cnopy psses through the lef, nd tht is key component of his cnopy mngement philosophy regrding lef lyer number. When the light intensity is t or bove 2,000 µ E m -2 s -1, the second lef lyer cn receive 200+ µ E m -2 s -1, well bove the lef compenstion point. When mbient PAR is t 800 µ E m -2 s -1, the resulting 80 µ E m -2 s -1 PAR is t or below the compenstion point (Howell nd Trought, 1997, unpublished dt). Further evidence suggests tht such shde leves lck the cpcity to chieve the rtes of photosynthesis of sun leves, even when plced in full sun [21]. Consistently high light intensities improve photosynthesis of interior, shded leves nd cn reduce the lef re necessry to ripen the crop. Limittions in Cool-Climte Viticulture The culture of grpevines ner the cool-climte limits of commercil production often lcks one or both of the bove-

3 Sustinble Grpe Productivity nd the Growth-Yield Reltionship 167 Tble 1 Growing seson length nd het ccumultion in different viticulturl regions. After Gldstones [10], Vn Den Brink et l. [60], Shulis et l. [50], nd Mills [personl communiction, 2001]. Viticulturl region Growing seson length Growing degree dys (10 C) Mildur, Austrli > Fresno, Cliforni > Coonwr, Austrli Np, Cliforni Bordeux (Medoc) > Centrl Wshington Long Islnd, New York > Mrlboro, New Zelnd Hobrt, Tsmni > Geisenheim, Germny > Burgundy (Dijon) Chmpgne (Reims) Genev, New York Benton Hrbor, Michign Trverse City, Michign mentioned dvntges [10]. Consider the growing seson chrcteristics of rnge of viticulturl regions (Tble 1). The long, wrm growing sesons of Mildur nd Fresno (>230 dys; 1750 growing degree dys; bse 10 C [GDD]) decline in Coonwr, Np, nd Bordeux (_200 to 210 dys; _1300 to 1400 GDD), which hve similr seson length s Mrlboro, Hobrt, nd Geisenheim (200 to 210 dys), but hve reduction in seson wrmth (_1100 to 1110 GDD). The remining cool regions include Dijon, Chmpgne, Genev, New York, Benton Hrbor, nd Trverse City, Michign (170 dys; _1050 to 1125 GDD). Further, light intensities in the Gret Lkes Region re well below the Fresno or Mildur exmples, nd lef loss s result of the first utumnl freeze very nerly coincides with hrvest dte. All of these fctors suggest viticulturl condition requiring greter lef re/ unit crop weight so tht importnt physiologicl functions bud initition nd differentition [50], crop ripening [50], crbohydrte storge [23], wood nd bud mturtion, nd cclimtion to freezing temperture nd mintennce of vine cold hrdiness [12] cn be ccomplished with the vilble exposed lef re. Vine Crbohydrte Stress nd Crop Blnce In the mid-1960s, Dr. Nelson Shulis spent sbbticl leve in Austrli t the Merbein Reserch Sttion ner Mildur. His experiences there involved efforts with vrying levels of lef removl nd vine defolition. Our discussions with him in the erly 1970s led to the initition of similr experiments s we were interested in crbohydrte stress nd grpevine cold hrdiness in Michign. Tble 2 provides synopsis of tht dt [19]. The dt resulted from fctoril experiment with different cropping levels (thinned to one cl/shoot or not thinned ), two blnced pruning formule (60+10, 30+10), nd defolition t verison (yes or no) over 3 yers. Under Michign conditions the impct of excess crop nd indequte lef re resulted in reduced vine size (weight of dormnt cne prunings). Vine yield ws reduced by indequte lef re, nd the grpes produced were unripe nd of no economic vlue. Bud hrdiness (Tble 2) nd cne hrdiness (dt not shown) were lso reduced when vines were subjected to stresses imposed by the vrious tretment combintions. In second experiment (Tble 3) we employed defolition t verison in different mnner [26]. To Hudson River Umbrell ( bilterl cordon t 1.8 m bove ground) trined vines, we creted folited (F) nd defolited (D) controls with three methods of chieving 50% defolition per vine: (1) removl of the lef t lternte node positions on the shoot yielding node F nd node D tretments; (2) removl of ll leves on lternte shoots yielding shoot F nd shoot D tretments; nd (3) removl of ll leves on one-hlf of the bilterl cordon in the 2.4 meter within row spcing, creting cordon F nd cordon D tretments. The 1.8 m trunk length, the employment of two trunks, nd the 1.2 m cordon joining ech trunk result in 6.1 m of bove-ground perennil structure. These dt re interesting on severl levels, but the key point of focus for this discussion is the dynmics of sugr ccumultion in the fruit postverison. Removl of 50% of vine leves ws insignificnt s node D, mesurble s shoot D, nd very detrimentl s cordon D (Tble 3). Of interest ws the bility of Tble 2 Influence of crop lod produced by different pruning severities (30+10 or 60+10), thinning to one cluster per shoot (T), or not thinned (NT), nd folited (F) or 100% defolition (D) t verison on Concord grpevines. After Howell et l. [19]. Tretment Vine size (Kg) Yield (Kg) % SS Primry bud hrdiness C Fll Spring b F-30-T F-60-T F-30-NT F-60-NT D-30-T D-60-T D-30-NT D-60-NT Tukey s HSD Min effects F vs D ** c ** ** ** ns 30 vs 60 * ns ns ns ns NT vs T * * ns ns ns Hrdiness ssessment mde on 21 Nov b Hrdiness ssessment mde on 15 April c *, **, nd ns indicte significnce t 0.05, 0.01, nd not significnt, respectively.

4 168 Howell Tble 3 The influence of differentil defolition t verison on yer 1 Brix response (A), yer 1 nd yer 2 yield nd bud hrdiness response (B), nd the reltive impct of these tretments on Brix in yer 1 nd bud fruitfulness in yer 2 (C). After Mnsfield nd Howell [26]. A. The influence of defolition t verison on yer 1 Brix response Tretment Verison b September 3 September 20 c Brix Control F d Cordon F b 7.6 Shoot F bc 7.2 Node F c 6.8 Node D c 6.9 Shoot D b 15.5 c 6.6 Cordon D b 14.2 d 5.1 b Control D b 11.2 e 2.5 c F-test ns e ** ** *** B. Yer 1 nd yer 2 vine yield nd bud hrdiness response. Yield/node (g) Bud hrdiness % of shoot- % of primry % of Tretment Yer 1 Yer 2 less nodes bud mortlity control D Control F 136 b e 6.6 d 7 d Cordon F 190 b bc 13.3 c 15 c Shoot F 144 b cd 11.6 cd 13 c Node F 183 b e 7.8 d 9 d Node D e 8.1 d 9 d Shoot D 155 b c 14.8 c 17 c Cordon D 149 b 153 b 45 b 41.4 b 47 b Control D 125 b 23 c F-test ** *** ** ** ** C. Reltive impct of imposed crbohydrte stress on % soluble solids in yer 1 nd bud fruitfulness in yer 2. % Soluble solids Fruitfulness Tretment % SS % of control F gm/node % of control F Control F Cordon F Shoot F Node F Node D Shoot D Cordon D 5.1 b 69 b 153 b 56 b Control D 2.5 c 34 c 23 c 9 c F-test *** *** *** *** F = folited; D = defolited; control = either fully folited (F) or fully defolited (D). Cordon, shoot, nd node tretments were 50% folited (F) or defolited (D). For cordon, ll leves on 50% of the vine cordon were removed. For shoot, ll leves from lternte were removed. For node, leves were removed from lternte nodes. b Timing of defolition: 17 August. c Hrvest dte. d Numbers within column hving the sme letter re not different by Duncn s New Multiple Rnge Test. e **, ***, nd ns indicte significnce t 0.01, 0.001, nd not significnt, respectively. the fruit sink on the cordon D tretment to mobilize crbohydrte from the cordon F tretment nd move it up to 6 m nd result in incresed sugr in fruit. Even more impressive ws the response of the fruit on the control D tretment vines. In the bsence of leves, the fruit mobilized stored crbohydrtes nd resulted in 2.0 Brix increse in the fruit. The power of the postverison fruiting sink is gret. Trining System nd Vine Crbohydrte Dynmics In the 30 yers of the 1970s through the 1990s, revolution in cultivrs used for wine hs occurred in Michign nd other portions of the Gret Lkes region. The cultivrs used for 95% of Michign wine in 1970 ccounted for less thn 5% of the wine in 1995 (Mich. Liquor Control, personl communiction). This cultivr chnge resulted in questions concerning whether pproches deemed desirble or cceptble for Vitis lbruscn Biley cultivr with procumbent, growth hbit would be pproprite for cultivrs possessing more upright growth hbit. To resolve this question, experiments were undertken involving rnge of trining systems. These hve been recently summrized [15]. Our effort sought to understnd principles, not just evlute prctices. New pproches to vine trining occur nerly every yer. Once principles re uncovered, the ppliction of those principles should be possible fter n initil ssessment of cultivr s growth hbit. We should not be required to reinvent the wheel every time new trining system is suggested. We employed four trining systems tht differed in height of the fruiting zone nd were hed or cordon trined systems: low hed, high hed, low cordon, nd high cordon. We hve conducted this kind of experiment on nine cultivrs nd conducted ech for minimum of five yers [17,18]. The mount of perennil wood vried significntly with ech trining system tested. All vines were double trunked so the length of perennil wood for ech system ws: 1.8 m for low hed; 3.6 m for high hed; 4.3 m for low cordon; nd 6.1 m for high cordon. With the exception of the cultivr Aurore, we hve invribly seen the reltionship: high cordon > low cordon > high hed > low hed. This hs been true whether we were considering vine size, vine yield, fruit composition vlues, or bud nd cne cold hrdiness. The impct of perennil vine structure on vine performnce hs lso been reported by My [27] in Austrli. Similrly, work in Switzerlnd [23,24] employing trunk modifiction yielding 12 to 15% increse in perennil wood resulted in significnt increses in fruit Brix s compred to the trditionl trunk conformtion. Collectively, these dt suggest tht choice of trining system hs considerble impct on the level of sustinble production of ripe grpes. Trining systems with more

5 Sustinble Grpe Productivity nd the Growth-Yield Reltionship 169 will be influenced by the methods used to mesure it nd the mnner in which such dt re interpreted. Single-lef mesurements do hve considerble utility. The key to their effective use is to define precisely the question sked nd to be very criticl in ny extrpoltion of lef response to cnopy response [34,43]. Crop lod nd crbohydrte prtitioning. One of the findings often reported bsed on single-lef ssessments hs been the positive influence of crop lod on vine photosynthesis [5,57]. Tht seems to be intuitively obvious; more fruit should reduce ny fruit-bsed feedbck inhibition of photosynthesis to minimum. Thus, more CO 2 should be fixed per vine nd tht should be shown s incresed dry weight per vine. Tble 4 suggests tht the ssumption is untrue. An evlution of prtitioning dt t fruit set, verison, nd hrvest shows shifts in the reltive dry weight of the vrious vine orgns but shows no difference in totl vine dry weight on ny mesurement dte [7-9]. The mount of crop per vine influences where the crbohydrtes produced ccumulte. At hrvest, fruit ccounted for over 40% of the totl vine dry weight for the most hevily cropped vine. This high percentge of dry weight ccumultes t the expense of vegettive tissues, prticulrly the roots. This dry weight dt is supported by subjective ssessment of root qulity (Tble 5). Between 60 to 80% of the grpeperennil wood show fvorble response of yield, vine size, fruit composition, nd cold hrdiness [15,17,18]. Old vines mke better wine? These experiences lso led us to conjecture: s vine trining systems with greter quntity of perennil wood resulted in fruit with superior fruit composition vlues, could the oft-expressed sentiment tht old vines mke better wine be result of greter volume of perennil wood nd concomitnt incresed crbohydrte storge re? If so, the response would be most often expressed in poor vintges. Tht, of course, would be the condition when it could be most redily detected. This specultion cn be esily subjected to criticl experimentl evlution, nd I expect it will be in the coming decde. Grpevine Photosynthesis nd Crbohydrte Prtitioning Experiments in grpevine photosynthesis nd crbohydrte prtitioning hve been conducted in coopertion with n rry of ssocites in Switzerlnd [1,20,23,24], New Zelnd [2,43,44], nd Michign [6-9,13,31,33,34]. The methods employed involved ssessment t the level of the single lef, whole potted vines, nd whole mture vines in the vineyrd. Potted vine studies hve been of two types: vines produced by the Mullins Technique [36] nd two-yer-old bering vines in 20-liter pots [6-9,31-34]. An rry of cultivrs hs lso been employed, including Chmbourcin, Chrdonny, Concord, Nigr, Pinot noir, Seyvl, nd Vignoles. The following principles re consistent with dt derived from these very different cultivrs. Tble 4 Influence of vine crop lod on the quntity nd percentges of dry mtter prtitioned to different vine structures t fruit set (A), verison (B), nd hrvest (C). After Edson et l. [9]. Percent of totl Totl vine Clusters/vine Fruit Lef Shoot Wood Root dry wt. (g) Predicting Vine Crbon Sttus Single-lef versus whole-vine ssessment of photosynthesis. Since CH 2 O is the vine s metbolic currency for growth, differentition, fruit ripening, nd host of other processes, photosynthesis becomes cndidte for ssessing circumstnce in vine culture tht my influence sustinbility. One gol in vine photosynthesis study is to produce mesurement tht cn predict whole-vine performnce. One pproch to chieve tht gol is to ssess the photosynthetic CO 2 fixtion of precise re on single lef nd then multiply tht by the lef re on the vine. An lterntive pproch involves the ssessment of CO 2 fixtion by the entire cnopy [31]. Using vine dry weight nd whole-vine photosynthesis s the bsis for ssessment, the single-lef ssessment is not predictive for either fctor; the wholevine ssessment is predictive of vine dry weight sttus [7-9,33,34]. Bsed on these dt, the perceived influence of vine photosynthesis on sustinble yield of ripe grpes A. Fruit set n n n n n n Liner regression ** b ns ns ns ns ns B. Verison n n n n n n Liner regression *** *** *** ns * ns C. Hrvest Liner regression *** *** *** ns *** ns n: not vilble. b *, **, ***, nd ns indicte sttisticl significnce t 0.05, 0.01, 0.001, nd not significnt, respectively.

6 170 Howell Tble 5 Influence of vine crop lod on roots t different phses of the growing seson. After Edson et l. [9]. Fruit set Verison Hrvest Clusters/vine Dry wt. (g) Root clss Dry wt (g) Root clss Dry wt. (g) Root clss n b n n n Liner regression ns c ns * * *** *** Rting system: 1=poor, few ctive roots; 5=good, mny vigorous roots. b n: not vilble. c *, ***, nd ns indicte significnce t 0.05, 0.001, nd not significnt, respectively. vine roots produced ech growing seson die, n ongoing process of turnover of the fibrous white roots [28]. The dt in Tble 5 support the dry weight dt nd suggest tht the observed decline in root qulity results from reduced replcement of roots s older roots die [28,60,62]. This observtion differs from the Austrlin experiences [3] nd tht of Robert Wmple in Wshington Stte [personl communiction, 2001]; no reduction in roots ws mesured. As noted bove in the discussion on light intensity nd growing seson length nd in Tble 1, ner-idel conditions for culture eliminte fctors tht commonly limit crbon ssimiltion nd ccumultion in cooler climtes. Grpevine Crop Level nd Fruit Mturtion on Vines Potted vines. The dt collected to dte suggest tht if the growing seson with dequte growing conditions is long enough, the vine will ripen the crop (Tble 6) [9]. Wht is not shown is criticl component of grpe qulity vrietl chrcter. Anecdotl experience nd micro-vins produced from grpes in the experiment reported in Tble 6 suggest tht fruit composition is generlly ssocited with vrietl chrcter (dt not shown) but is not predictive of the intensity of tht vrietl chrcter. The grpes tht chieved mture Brix erliest hd gretest vrietl chrcter for this cultivr. Importntly, this will vry with vriety nd the compounds tht collectively produce vrietl chrcter. Tble 6 Influence of vine crop lod on vine yield nd fruit composition. After Edson et l. [9]. Two weeks prehrvest Clusters/vine Yield (g) Brix ph TA Brix ph TA Liner regression *** * *** ns ns ns ns *, ***, nd ns indicte significnce t 0.05, 0.001, nd not significnt, respectively. Grpevine lef re nd verison. Under conditions of serious lef re reductions occurring prior to verison, source:sink imblnce cn result. The creted source indequcy cn hve n impct similr to tht of excess crop with resulting dely in the onset of verison. As with fruit mturtion curves [43], ll berries ultimtely pss through verison, but the tretment with most restricted lef re to fruit weight rtio ws delyed by over 30 dys. Mture vines in the vineyrd. While potted vines re convenient for prtitioning studies, we would never be comfortble putting conceptul viticulturl principle into prctice without first evluting the response of mture, bering vines in vineyrd. The dt in Tble 7 result from n experiment on Concord vines conducted with fctoril sttisticl design. There were three vine size ctegories selected with the number of nodes retined t pruning rnging from 20 to 160 per vine. The experiment ws conducted from 1991 to 2000 [35]. Similr dt hve been produced in other reserch efforts [29,30]. The key point of these vineyrd dt is their greement with those gined by experimenttion on potted vines. The bove-ground response is very similr. We therefore mke the inductive inference tht the fctor not mesured, tht is, root dry weight, lso responded in similr mnner to the potted vines. This inference should be subjected to criticl direct ssessment in the vineyrd. Erly Development of Lef Arry nd Crop Mturtion Another ide tht seems intuitively obvious is the positive impct of erly lef rry on totl vine crbon ssimiltion over the growing seson. Erly cnopy fill, it would seem, should trp sunlight tht would otherwise strike the vineyrd floor [57,58]. Like the previously mentioned cse of incresed Pn nd crop lod, the dt do not support the hypothesis. Dt in Tbles 7 nd 8 show different response. Concord vines were t equilibrium with rnge of either 15 to 91 buds (Tble 7) or 17 nd 66 buds (Tble 8) retined per meter of row. Differences in lef re t bloom nd verison coupled with the lef re:fruit weight rtio fvor the lrger bud number in every cse, but fruit mturtion (bsed on fruit composition vlues) is delyed for the lrger bud Hrvest number tretments. Erly lef re development ws not n dvntge once vines were t equilibrium with the imposed tretments. Photosynthesis nd Lef Age Dt reported by Kriedemnn [25] nd Poni nd Intrieri [46] rises one concern

7 Sustinble Grpe Productivity nd the Growth-Yield Reltionship 171 Tble 7 Influence of erly seson lef rry on yield nd fruit composition of Concord grpevines. Vines t equilibrium fter eight yers of tretment, After Howell, Miller, nd Stocking, unpublished. Nodes retined Yield (Kg) cm 2 lef re/g finl fruit wt. per meter of row per meter of row % SS Bloom Verison F-test ** ** ** ** Liner regression * ** *** ** *, **, nd *** indicte sttisticl significnce by liner regression t 0.05, 0.01, nd 0.001, respectively. Tble 8 Influence of multisesonl cropping stress on yield, lef re, nd totl vine sugr production of Concord grpevines. Vines t equilibrium fter six yers of tretment, After Miller nd Howell, unpublished. Nodes retined Fctor nlyzed F-test Vine size (Kg cne prunings) *** b Lef re t hrvest (cm 2 ) *** Yield (Kg) ** Clusters *** Clusters/node retined *** Cluster weight (g) *** Berry weight (g) ** Berries/cluster ** % Soluble solids * Lef re (cm 2 )/g fruit ** Crop lod (yield/vine size) *** Kg sugr ** Kg sugr/lef re (10-6 ) *** Per meter of row. b *, **, nd ***, indicte sttisticl significnce t 0.05, 0.01, nd levels of probbility, respectively. bout culturl methods with lrge shoot number per vine. Such vines re chrcterized by single growth flush in the spring, nd little dditionl cnopy is dded over the growing seson [3,4,33,34,57,58]. The potentil to hve n old cnopy during the criticl period from verison to hrvest ws n issue. Reserch on Pinot noir [44,45] using potted Mullins [36] vines nd on mture bering Chrdonny vines (Howell nd Trought, unpublished) suggests tht lef ge response vries with the rtio of the lef re to fruit weight. Tretments tht gretly reduced lef re (nerly 100%) per vine resulted in leves on the shoot tht were ctive photosyntheticlly well fter leves t similr positions (sme ge) were senescent on fully folited vines (Tble 9). The dt in Tble 9 pper to disgree with the erlier point tht induced crbohydrte stress vi croplod resulted in no in- crese in totl vine dry weight (Tbles 4 nd 5). This potentil misunderstnding is bsed on the fctors being mesured: single-lef photosynthesis in the first instnce nd totl vine dry weight in the ltter. The cpcity of individul leves to remin photosyntheticlly ctive nd compenste when lef re is reduced is cler (Tble 3). However, when lef re ws severely reduced (Tble 9), the incresed lef Pn rte could not compenste for the loss of the mjority of other leves in reltion to fruit mturtion or totl vine dry weight [43; dt not shown]. Miniml Pruning nd Vine Blnce Erlier, we mentioned our concern bout the direct ppliction of MP technology to cool climtes. Further, we noted tht this new pproch to vine culture hd the potentil to refute concepts of vine blnce. There re severl resons why I do not believe tht to be true. First, trditionl Austrlin cne or spur-pruning did not crete blnce. Bsed on the principle of the Rvz Index, the Growth- Yield Reltionship, nd lef re:fruit weight, trditionlly pruned vines retined indequte bud number to produce crop levels t the vine-environment potentil [3,5,57]. Lower yields, excessive vine vigor, nd cnopy shding were the results. Employment of MP under the conditions of Sunrysi, Austrli, produced bud number nd subsequent crop tht resulted in blnced lef re nd crop. As Rvz [47] noted, there is close reltionship between vine size nd lef re. MP Cbernet Suvignon vines were in blnce bsed on the ripening criteri nd the bility of those vines to produce t high levels over yers of culture. The victory of MP is victory of blnce, sustinbility, the Growth-Yield Reltionship, nd lef re:fresh weight rtio over prescription ppliction of bud number regrdless of vine growth sttus. Achieving Highest Yields in Sitution with Vrying Vintge Conditions A mjor chrcteristic of cool-climte viticulture is the nnul vrition in growing conditions. Achieving highest sustinble yields of ripe grpes must involve culturl methods tht consider this potentil limittion. Tble 9 Pre- nd postverison photosynthesis. Different levels of within-shoot lef removed two weeks postbloom, Chrdonny. After Howell nd Trought, unpublished. µ moles/cm 2 /sec % Leves removed Preverison Postverison F-test ** ** ** ** F-test ns ns Liner regression * ** *, **, nd ns indicte sttisticl significnce by liner regression or the F-test t 0.05, 0.01, nd not significnt, respectively.

8 172 Howell The 1990s produced the gretest vrition mong growing sesons in Michign since temperture recording begn. The yers 1991, 1998, nd 1999 were mong the best on record for growing degree dys. The 1992 vintge ws the worst. Such vrition cretes dilemm for the producer. Does producer crop t the level to chieve blnce in 1992, nd lose the mount of ripe crop possible in 1991, 1998, nd 1999? Alterntively, does the producer crop for vintges like 1991, 1998, nd 1999 nd risk n unsleble crop in 1992? This is relity in cool-climte viticulture, with mjor economic implictions for the viticulturist. In fct, we do not know how much crop level should be djusted downwrd in response to one or more environmentl nd/or biotic stresses. This re deserves much more ttention by viticulturists for two resons: the scenrios on globl wrming suggest more vrible climte sitution, nd we will hve fewer pest control options in the future. We must nticipte conditions tht will result in vrible bility to ripen crop. In ny event, it is cler tht prescription pproch cnnot be stisfctory. Crop djustment nd vine blnce. Crop djustment provides one solution. This pproch would llow the viticulturist to crop t the level tht would chieve blnce in the historicl best vintge nd djust the crop downwrd prior to verison bsed on the sttus of GDD ccumultion hlfwy between bloom nd verison. The vine would esily djust [35,38,55,56] nd even compenste [1]. Pest Control, Sustinble Grpe Production, nd the Growth-Yield Reltionship The future of commercil viticulture is perceived through cloudy crystl bll under the best of circumstnces, but one fct seems very cler: future grpe production will hve fewer chemicl tools to combt pest problems. A likely result will be periodic episodes of stress when vines re subjected to insect or disese ttck on vine folige. Gretly limiting our bility to predict the impct of these episodes is the lck of informtion bout economic thresholds. How much lef dmge occurs before there is n economic impct? We do not know the impct of powdery or downy mildew, lefhopper burn, or Jpnese beetle reduction of lef re on lef CO 2 ssimiltion or net photosynthesis. Nor do we know whether the impct of the biotic stress chnges with reltion to shoot nd fruit growth nd mturtion phenology of the vine. Bsed on work with biotic stresses reported bove, we expect tht timing will be importnt. Unfortuntely such dt re very scrce (2). Applying the principles of the Growth-Yield Reltionship to biotic stresses provides direction for future efforts on pest-induced stresses. In poor vintge, crop djustment cn produce the blnce pproprite for tht seson s climtic conditions. A similr pproch for pest stress should be possible once the physiologicl nd economic impct of the pest stress hs been determined. Within vineyrd vrition nd sustinble production. In ddition to sesonl vrition is the relity tht soil vrition within the vineyrd cn produce considerble rnge of vine vigor nd resulting vine size nd lef re. Agin, prescription pproch cnnot work: ech vine must be considered individully or smll vines will be overcropped nd lrge vines undercropped. This cn result in the smller vines becoming weker nd producing unripe fruit nd the lrger vines producing indequte yields of fruit ripened in the shde of n excessively vigorous vine cnopy [29,30]. At most vineyrds, crop control is done t pruning, bsed on numbers of nodes retined. Becuse this pruning is being done by hired pruners, the bility to chieve such individul ttention is very smll. However, the future does hold promise. The Future Severl fetures of the viticulturl future re visible now: 1. Site mpping vi globl positioning stellites (GPS) to determine where problem res exist. 2. On-bord hrvester yield ssessments re relity. In the future, these dt coupled with GPS dt will monitor low production res within vineyrd nd provide bsis for ttention nd culturl modifiction. 3. Vine size or estimtes of exposed cnopy t verison will be possible using existing trctor-mounted computerbsed visul technology coupled with GPS positioning. The potentil crop bsed on lef re or Prtridge s Growth-Yield blnce concepts will be determined by the computer for ech individul vine. 4. Crop lod estimtes will be mde using methods noted bove in item 3 t the time of the prebloom spry when flower clusters re esily visible. 5. Crop djustment in mid-july (for northern hemisphere) or bout hlfwy between bloom nd verison will be ccomplished mechniclly so tht the input of ech individul vine from items 3 nd 4 bove plus GDD sttus re integrted nd individul vine blnce chieved. These five fetures re now possible. The dtbses nd the reserch required to produce these dtbses re lcking. In ddition, it will become incresingly importnt for the viticulturist to employ the most dvnced methods of monitoring vineyrd growth nd pest sttus nd to dopt the principles of vine blnce. As yields pproch the upper limit for ny mcro- or mesoclimte, the buffering cpcity of photosynthesis compenstion will be reduced, nd further stresses tht negtively influence vine crbon blnce, regrdless of origin, cn produce disstrous results. Tht is the chllenge for sustinble viticulture in the twenty-first century, nd meeting tht chllenge will hve its roots in the lef re:fruit weight rtio, the Rvz Index, nd the Growth-Yield Reltionship, s understood by Newton Prtridge nd Nelson Shulis. Conclusions The concept of vine blnce is nerly 100 yers old. Rvz introduced the concept nd Prtridge nd Shulis pursued methods to use it s mens to predict vine performnce vi the Growth-Yield Reltionship. This llometric method substituted vine growth or vine size (weight of cne prunings per vine) for lef re per vine nd the lef re reltionship to fresh fruit weight (7 to 14 cm 2 per grm). Tht reltionship is tied to vine blnce nd long-term sustinble viticulture.

9 Sustinble Grpe Productivity nd the Growth-Yield Reltionship 173 Trining systems employing mximum mounts of perennil wood tht lso fcilitte sunlight penetrtion into the fruiting nd renewl zone re to be preferred. Spur systems on cordons my be uncceptble on some cultivrs due to low fruitfulness of bsl buds [15]. Photosynthesis in the period preverison is not source-limited under typicl vineyrd conditions, nd leves seldom exceed 50% of their mesured photosynthesis cpcity. This hs been demonstrted on Mullins vines, two-yer-old potted vines, nd mture bering vines using severl cultivrs nd different Vitis species. Single-lef photosynthesis mesurements re not correlted with either whole vine photosynthesis or totl vine dry weight increses. Whole vine photosynthesis is closely relted to vine dry weight increses. Miniml pruning nd/or mchine hedging in some form, coupled with cpcity for timely crop djustment, offers good potentil for future likelihood of chieving mximum sustinble yield of ripe grpes cross rnge of cultivrs. The key to the success of this effort will be vine-by-vine control of crop djustment to chieve vine blnce under conditions of vrible crop lod, previous yer s vine size, nd current seson s growth nd mturtion sttus preverison. Leves on vines with either indequte lef re or excess crop (low source:sink rtio) retin chlorophyll, dely senescence, mintin high photosynthetic rtes, nd dely the ging response chrcterized by leves on similr vines possessing fully expressed cnopies. Indequte lef re delys verison nd lengthens the time from verison to ripening. Vine blnce s understood by Rvz, Prtridge, Shulis, nd others remins key to the chieving of mximum consistent production over long yers of production. Modifictions in our pproches to vine culture nd mngement should begin with n ssessment of tht modifiction s impct on vine blnce s understood bsed on the Rvz Index, the Growth:Yield Reltionship, nd cm 2 lef re/grm fresh weight of fruit t hrvest. Literture Cited 1. Cndolfi-Vsconcelos, M. Crmo, W. Koblet, G.S. Howell, nd W. Zweifel. Influence of defolition, rootstock, trining system nd lef position on gs exchnge of Pinot noir grpevines. Am. J. Enol. Vitic. 45: (1994). 2. Clerwter, L.M., M.C.T. Trought, nd G.S. Howell. Impct of powdery mildew infection nd fungicide ppliction on grpevine photosynthesis. Submitted to Aust. J. Grpe Wine Sci. (2000). 3. Clingeleffer, P.R. Vine response to modified pruning prctices. In Proc. of the Second N.J. Shulis Grpe Symposium, Fredoni, NY. R.M. Pool (Ed.), pp Cornell University Press, Genev, NY (1993). 4. Clingeleffer, P.R., nd L.R. Krke. Responses of Cbernet frnc grpevines to miniml pruning nd virus infection. Am. J. Enol. Vitic. 43:31-37 (1992). 5. Downton, W.J.S., nd W.J.R. Grnt. Photosynthetic physiology of spur pruned nd miniml pruned grpevines. Aust. J. Plnt Physiol. 19: (1992). 6. Edson, C.E., nd G.S. Howell. A comprison of vine rchitecture systems t different crop lods: Lef photosynthesis, vine yield, nd dry mtter prtitioning. Vitic. Enol. Sci. 48:90-95 (1993). 7. Edson, C.E., G.S. Howell, nd J.A. Flore. Influence of crop lod on photosynthesis nd dry mtter prtitioning of Seyvl grpevines. I. Single lef nd whole vine response pre- nd post-hrvest. Am. J. Enol. Vitic. 44: (1993). 8. Edson, C.E., G.S. Howell, nd J.A. Flore. Influence of crop lod on photosynthesis nd dry mtter prtitioning of Seyvl grpevines. II. Sesonl chnges in single lef nd whole vine photosynthesis. Am. J. Enol. Vitic. 46: (1995). 9. Edson, C.E., G.S. Howell, nd J.A. Flore. Influence of crop lod on photosynthesis nd dry mtter prtitioning of Seyvl grpevines. III. Sesonl chnges in dry mtter prtitioning, vine morphology, yield, nd fruit composition. Am. J. Enol. Vitic. 46: (1995). 10. Gldstones, J.S. Viticulture nd Environment. Wine Titles, Hyde Prk Press, Adelide, Austrli (1992). 11. Gould, S.J. Shdes of Lmrck. In The Pnd s Thumb, pp W.W. Norton, New York (1980). 12. Howell, G.S. Culturl mnipultion of vine cold hrdiness. In Proc. of the Second Int. Cool Climte Symposium. R. Smrt et l. (Eds.), pp Aucklnd, New Zelnd (1988). 13. Howell, G.S. Crbohydrte prtitioning in grpevines. In Proc. of the Fourth Annul Romeo Brgto Conference, pp New Zelnd Grpe Growers Council (1998). 14. Howell, G.S. Wht is blnced pruning nd why should wine grpe growers cre bout it? Vintner Vineyrd 4(1):3,11 (1990). 15. Howell, G.S., D.P. Miller, D. Jckson, nd D. Stocking. Selecting the optimum trining system for grpevines in Michign. Michign Grpe & Wine Industry Council nd Michign Stte University, Est Lnsing (2000). 16. Howell, G.S. Grpevine crop control: Implictions for yield, fruit composition nd subsequent wine qulity, cold hrdiness nd sustinble production, pp Hertlnd Wine Colition, Ohio Stte University, Columbus (2000). 17. Howell, G.S., T.K. Mnsfield, nd J.A. Wolpert. Influence of trining system, pruning severity, nd thinning on yield, vine size nd fruit qulity of Vidl blnc grpevines. Am. J. Enol. Vitic. 38: (1987). 18. Howell, G.S., D.P. Miller, C.E. Edson, nd R.K. Striegler. Influence of trining system nd pruning severity on yield, vine size, nd fruit composition of Vignoles grpevines. Am. J. Enol. Vitic. 42: (1991). 19. Howell, G.S., B.G. Stergios, nd S.S. Stckhouse. Interreltion of productivity nd cold hrdiness of Concord grpevines. Am. J. Enol. Vitic. 29: (1978). 20. Howell, G.S., M. Crmo Cndolfi-Vsconcelos, nd W. Koblet. Response of Pinot noir grpevine growth, yield nd fruit composition to defolition the previous growing seson. Am. J. Enol. Vitic. 45: (1994). 21. Hunter, J.J., nd J.H. Visser. The effect of prtil defolition, lef position nd developmentl stge of the vine on the photosynthetic ctivity of Vitis vinfer L. cv. Cbernet Suvignon. S. Afr. J. Enol. Vitic. 10:67-73 (1989). 22. Kimbll, K., nd N. Shulis Pruning effects on growth, yield nd mturity of Concord grpes. Proc. Am. Soc. Hort. Sci. 71: (1958). 23. Koblet, W., M. Crmo Cndolfi-Vsconcelos, E. Aeschimnn, nd G.S. Howell. Influence of defolition, rootstock, nd trining system on Pinot noir grpevines. I. Mobiliztion nd reccumultion of ssimiltes in woody tissue. Vitic. Enol. Sci. 48: (1993). 24. Koblet, W., M. Crmo Cndolfi-Vsconcelos, W. Zweifel, nd G.S. Howell. Influence of lef removl, rootstock, nd trining system on yield nd fruit composition of Pinot noir grpevines. Am. J. Enol. Vitic. 45: (1994).

10 174 Howell 25. Kriedemnn, P.E. Photosynthesis in vine leves s function of light intensity, temperture nd lef ge. Vitis 7: (1968). 26. Mnsfield, T., nd G.S. Howell. Response of soluble solids ccumultion, fruitfulness, cold resistnce nd onset of bud growth to differentil defolition stress t verison in Concord grpevines. Am. J. Enol. Vitic. 32: (1981). 27. My, P. The grpevine s perennil, plstic nd productive plnt. In Proc. of the Sixth Austrlin Wine Industry Tech. Conf., pp Adelide, Austrli (1987). 28. McLen, M., G.S. Howell, nd A.J.M. Smucker. A minirhizotron system for in situ root observtion studies of Seyvl grpevines. Am. J. Enol. Vitic. 43:87-89 (1992). 29. Miller, D.P., nd G.S. Howell. Influence of vine cpcity nd crop lod on the yield, fruit composition nd sugr production per unit lnd re of Concord grpevines. In Proc. of the Fourth Int. Symposium on Cool Climte Enology nd Viticulture. T. Henick-Kling et l. (Eds.), pp Rochester, NY (1996). 30. Miller, D.P. nd G.S. Howell. Influence of vine cpcity nd crop lod on cnopy development, morphology, nd dry mtter prtitioning in Concord grpevines. Am. J. Enol. Vitic. 49: (1998). 31. Miller, D.P., G.S. Howell, nd J.A. Flore. A whole-plnt, open, gsexchnge system for mesuring net photosynthesis of potted woody plnts. HortScience 31(6): (1996). 32. Miller, D.P., G.S. Howell, nd J.A. Flore. Effect of shoot number on potted grpevines. I. Cnopy development nd morphology. Am. J. Enol. Vitic. 47: (1996). 33. Miller, D.P., G.S. Howell, nd J.A. Flore. Effect of shoot number on potted grpevines. II. Dry mtter ccumultion nd prtitioning. Am. J. Enol. Vitic. 47: (1996). 34. Miller, D.P., G.S. Howell, nd J.A. Flore. Influence of shoot number nd crop lod on potted Chmbourcin grpevines. II. Whole-vine vs. single-lef photosynthesis. Vitis 36: (1997). 35. Miller, D.P., G.S. Howell, nd R.K. Striegler. Reproductive nd vegettive response of mture grpevines subjected to differentil cropping stress. Am. J. Enol. Vitic. 44: (1993). 36. Mullins, M.G., nd K. Rjskekren. Fruiting cuttings: Revised method for producing test plnts of grpevine cultivrs. Am. J. Enol. Vitic. 32:35-40 (1981). 37. Prtridge, N.L. The fruiting hbits nd pruning of the Cmpbell Erly grpe. Mich. Agr. Expt. St. Tech. Bull. No. 106 (1930). 38. Prtridge, N.L. The influence of long pruning nd thinning upon the qulity of Concord grpes. Proc. Am. Soc. Hort Sci. 28: (1931). 39. Prtridge, N.L. The use of the growth-yield reltionship in field trils with grpes. Proc. Am. Soc. Hort Sci. 23: (1925). 40. Prtridge, N.L. The fruiting hbits nd pruning of the Concord grpe. Mich. Agr. Expt. St. Tech. Bull. No. 69 (1925). 41. Prtridge, N.L. Growth nd yield of Concord grpe vines. Proc. Am. Soc. Hort. Sci. 23:84-87 (1925). 42. Prtridge, N.L. Culturl methods in the bering vineyrd. Mich. Agr. Expt. St. Circulr Bull. No. 130 (1927). 43. Petrie, P.R., M.C.T. Trought, nd G.S. Howell. Growth nd dry mtter prtitioning of Pinot noir (Vitis vinifer L.) in reltion to lef re nd crop lod. Aust. J. Grpe Wine Res. 6:40-45 (2000). 44. Petrie, P.R., M.C.T. Trought, nd G.S. Howell. Fruit composition nd ripening of Pinot noir (Vitis vinifer L.) in reltion to lef re. Aust. J. Grpe Wine Res. 6:46-51 (2000). 45. Petrie, P.R., M.C.T. Trought, nd G.S. Howell. Influence of lef ging, lef re, nd crop lod on photosynthesis, stomtl conductnce, nd senesence of grpevine (Vitis vinifer L. cv. Pinot noir) leves. Vitis 39:31-36 (2000). 46. Poni, S., C. Intrieri, nd O. Silvestroni. Interctions of lef ge, fruiting, nd exogenous cytokinins in Sngiovese grpevines under non-irrigted conditions. I. Gs exchnge. Am. J. Enol. Vitic. 45:71-78 (1994). 47. Rvz, M.L. L effeuellge de l vigne. Annles d L Ecole Ntionle d griculture de Montpellier. 11: (1911). 48. Shulis, N. Culturl prctices for New York vineyrds. Cornell Ext. Bull. No. 805 (1950). 49. Shulis, N. Some effects of pruning severity nd trining on Fredoni nd Concord grpes. Proc. Am. Soc. Hort. Sci. 51: (1948). 50. Shulis, N., nd C. Prtt. Grpes: Their growth nd development. Frm Reserch. Cornell Univ. No. 401 (1965). 51. Shulis, N., nd W.B. Robinson. The effect of seson, pruning severity nd trellising on some chemicl chrcteristics of Concord nd Fredoni grpe juice. Proc. Am. Soc. Hort. Sci. 62: (1953). 52. Shulis, N., T.D. Jordn, nd J.P. Tomkins. Culturl prctices for New York vineyrds. Cornell Ext. Bull. No. 805 (1966). 53. Smrt, R.E. Principles of grpevine microclimte mnipultion with implictions for yield nd qulity. A review. Am. J. Enol. Vitic. 36: (1985). 54. Smrt, R.E., nd M.D. Robinson. Sunlight into Wine: A Hndbook for Wine- Grpe Cnopy Mngement. Wine Titles, Hyde Prk Press, Adelide, Austrli (1991). 55. Smithymn, R.P., G.S. Howell, nd D.P. Miller. Influence of cnopy configurtion on vegettive development, yield nd fruit composition of Seyvl blnc grpevines. Am. J. Enol. Vitic. 48: (1997). 56. Smithymn, R.P., G.S. Howell, nd D.P. Miller. The use of competition for crbohydrtes mong vegettive nd reproductive sinks to reduce fruit set nd Botrytis bunch rot in Seyvl blnc grpevines. Am. J. Enol. Vitic. 49: (1998). 57. Sommer, K.J., nd P.R. Clingeleffer. Comprison of lef re development, lef physiology, berry mturtion, juice qulity, nd fruit yield of miniml nd cne-pruned Cbernet Suvignon. In Proc. of the Second N.J. Shulis Grpe Symposium, Fredoni, NY. R.M. Pool (Ed.), pp Cornell University Press, Genev, NY (1993). 58. Sommer, K.J., P.R. Clingeleffer, nd Y. Shulmn. Comprtive study of vine morphology, growth nd cnopy development in cne-pruned nd miniml-pruned Sultn. Aust. J. Exp. Agric. 35: (1995). 59. Tompkins, J., nd N. Shulis. The Ctwb grpe in New York. II. Some effects of severity of pruning on the production of fruit nd wood. Proc. Am. Soc. Hort. Sci. 66: (1955). 60. Trought, M.C.T., nd G.S. Howell. Crbohydrte blnce of grpevines: How does it control fruit qulity? In Proc. Grpe nd Wine Spring School. E. Costello (Ed.), pp Lincoln University, New Zelnd (1997). 61. Vn Den Brink, C., N.D. Strommen, nd A.J. Kenworthy. Growing degree dys in Michign. Mich. Agr. Expt. St. Res. Report No. 131 (1971). 62. Vn Zyl, J.L. The grpevine root nd its environment. S. Afric Dept. Agr. Wter Supply Tech. Comm. No. 215 (1988).