Fruit Ripening in Vitis vinifera L.: Responses to Seasonal Water Deficits

Transcription

1 Fruit Ripening in Vitis vinifera L.: Respnses t Seasnal Water Deficits MARK A. MATTHWS ~* and MICHAL M. ANDRSON 2 The respnse f fruit ripening t vine water status was investigated in a hillside Cabernet franc vineyard in the Nrth Cast regin f Califrnia. Treatments were impsed by drip irrigatin at 2 X the standard practice rate (cntinual) t maintain high water status, by withhlding water befre (early deficit) r after (late deficit) veraisn, r by withhlding water thrughut mst f the seasn (full deficit). Midday leaf water ptential f cntinual vines decreased frm apprximately-.3 MPa befre blm t-1.13 MPa at veraisn and t MPa at harvest. Leaf water ptentials f early deficit and late deficit vines were apprximately.3 MPa mre negative than cntinual vines at veraisn and harvest, respectively. After veraisn, water status f early deficit vines recvered t the level f cntinual vines. These mderate differences in water status at different phenlgical stages altered fruit cmpsitin at harvest. The cncentratins f phenlics in juice and dermal extracts and f anthcyanins in dermal extracts were increased by all treatments which withheld water. Malate cncentratins were significantly lwer in treatments which impsed lw vine water status befre veraisn. Lw vine water status after veraisn increased prline cncentratin significantly. There were n treatment effects n the nset f veraisn, the duratin f ripening, juice ph, r ptassium levels, and little difference in Brix r titratable acidity. Thus, irrigatin t btain seasnal water deficits may ffer a cultural cntrl f winegrape cmpsitin withut significant effects n the time required t reach maturity. KY WORDS: water stress, irrigatin, malate, prline, phenlics, ptassium The Nrth Cast regin f Califrnia is recgnized fr the prductin f premium winegrapes. In this regin, many vineyards are irrigated weekly, while thers are ttally dependent upn stred sil water. Plantings have expanded frm the valley flrs, where sils are ften f adequate depth and sme grwers have ample irrigatin water, t the hillsides where sils are shallw and reservirs limited. Hence, the water status f vines is likely t vary amng vineyards and during the seasn. It is clear that water status affects a myriad f plant functins (4). The imprtance f understanding physilgical respnses t water status is magnified in wine grapes, where the cmpsitin f fruit challenges yield as the primary parameter f prductivity. Hwever, there are n reprts f vine respnses ( e.g., grwth r fruit cmpsitin) t irrigatin r t vine water status in Nrth Cast vineyards. Indeed, the rle f vine water status in determining the reprductive develpment and cmpsitin fwinegrapes is, in general, nt knwn (27,33). Therefre, this study was cnducted t determine the extent t which reprductive develpment, including the ripening prcess, is sensitive t vine water status. In this paper, we reprt that the levels f selected juice slutes f ptential imprtance in winemaking respnd differentially t seasnal water deficits. 1.2Department f Viticulture and nlgy, University f Califrnia, Davis, CA "Authr t whm crrespndence shuld be addressed. This research supprted in partby grants frm the Winegrwers f Califrnia and the Nrth Cast Viticultural Research Grup. The authrs wish t thank Jhn Zaya and Steve Lagier fr technical assistance, and Vernn Singletn fr use f his phenlic assay system. This research was cnducted at the UniversitY f Califrnia, Davis. Manuscript submitted fr publicatin 11 January Cpyright 1988 by the American Sciety fr nlgy and Viticulture. All rights reserved. Materials and Methds The site f the study was selected fr its shallw, light sil (gravelly lam), suthwestern aspect, premium winegrape variety, and vine unifrmity. Sixyear-ld vines (Vitis vinifera L., cv. Cabernet franc n Ganzin (A X R1) rtstck) in a cmmercial hillside (apprx. 2% grade) vineyard near Saint Helena, Califrnia, were cultured and irrigated as previusly described (15). Briefly, irrigatin was supplied weekly by a drip system at apprximately 45 L/vine in the standard practice treatment (SP) and at 9 L/vine/week in all ther treatments. In the SP and cntinual (C) treatments, water was supplied thrughut the seasn. Water was withheld befre veraisn in the early deficit (D) treatment, withheld after veraisn in the late deficit (LD) treatment, and applied twice (2 wk) befre veraisn and twice (2 wk) befre harvest in the full deficit (FD) treatment. The ttal vlume f water applied per vine was apprximately 32,64,73,82, and 15 L in the FD, D, SP, LD, and C treatments, respectively. N measureable rain ccurred during the treatment perid in any seasn. (Ttal evaptranspiratin during the grwing seasn fr a full-canpied vineyard can be estimated rughly frm the wrk f Pruitt et al. (23) as 3 L/vine during treatment applicatin). Treatments were applied t three-rw X sevenvine plts. Data were cllected nly frm the middle five vines f the middle rw. ach treatment was replicated five times. Midday leaf water ptential was determined with a pressure chamber as previusly described (15). Tw r three leaves per replicate plt were sampled; a ttal f 1 t 15 leaves were used t estimate treatment water ptential fr each sample date. The air temperature at a central, internal site in 313 Am. J. nl. Vitic., Vl. 39, N. 4, 1988

2 314-- WATR DFICITS AND FRUIT RIPNING individual clusters was determined by carefully inserting a lng (1 cm) thermistr prbe int representative basal clusters withut damage t berries. Ambient temperature within the canpy was determined with similar sensrs suspended near the cluster with an internal prbe. Sensrs were read several times thrughut the day (day 2, apprx, veraisn). Berry samples (apprx. 125), cmprised f berries representing each vine and all psitins n the shts and within clusters, were btained biweekly frm each plt. Subsamples (1 berries) were wrapped in dublelayer cheeseclth and crushed with a small hand press. The resulting juice was centrifuged at 3 g fr tw minutes, t remve debris. Aliquts f the supernatant were retained fr immediate analysis f ph, titratable acidity (TA) by titratin with NaOH (9), and sluble slids by refractmetry (1). Ptassium was determined by emissin spectrscpy. The remaining juice was immediately frzen and stred fr further analysis. After thawing t lab temperature, aliquts f juice were taken fr determinatin f malate, prline, and sluble phenlics. Malate cntent was determined enzymatically accrding t Hhrst (1). Prline was determined with ninhydrin frm the A52 as described by Ough (19). Ttal phenlics were estimated using the methds f Singletn and Rssi (25) as mdified by Slinkard and Singletn (26). Gallic acid standards included glucse and fructse (1:1) at cncentratins equivalent t Brix f the juice samples t avid verestimates f phenls caused by the psitive reactin f sugars with the alkaline Flin-Cicalteu reagent (25). Disks f dermal tissue (.2 cm 2) were remved frm 1-berry subsamples with a crk brer. Anthcyanins were extracted with acidified methanl and estimated frm the A535 accrding t Kliewer (13). Ttal phenlics f skin extracts were estimated frm the A2s using gallic acid standards. Precipitatin with trichlracetic acid had n measureable effect n absrbance; n crrectin was made fr chlrphyll absrptin. Fruit grwth was determined by repeated determinatins f berry diameter with a hand-held micrmeter as previusly described (15). Berry vlume was calculated assuming a spherical berry. In rder t determine berry water cntent, five-berry subsamples were weighed, frzen, and lyphilized until n further decrease in weight was bserved. Results Midday leaf water ptential (W) was apprximately -.3 MPa in all treatments at the nset f the experiment (Table 1). Vine water status declined until veraisn in all treatments, but mre in D vines than in ther treatments. At veraisn, W was greater than -1.2 MPa in C and LD vines, but was MPa in D vines (Table 1). Thereafter, W was relatively stable in C vines, but decreased in FD and LD vines (which had water withheld after veraisn) and increased in D vines (which had water supplied after veraisn). W declined mre after veraisn in LD vines than in FD vines (Table 1). At Table 1. Midday leaf water ptential at early, mid-, and late seasn f vines receiving varius irrigatin treatments (described in Materials and Methds). Budbreak ccurred at apprximately day 75. Data are means f 1 t 15 samples per treatment. The standard errr f the mean never exceeded.1 MPa. Treatment Juliandate (veraisn) (harvest) Midday leaf water ptential Standard practice Cntinual arly deficit Late deficit Full deficit harvest, W was highest in D vines and lwest in LD vines (Table 1). Thus, withhlding water during different phenlgical stages (i.e., befre and after the nset f ripening) created significant deficits in plant water status in cmparisn with C vines. The ripening cnsequences f these differences were the fcus f the fllwing measurements. Diurnal measurements f the cluster interir air temperatures were cnducted during fruit ripening. N differences amng treatments in cluster temperatures r between ambient air temperatures and cluster temperatures were bserved until afternn. Readings at 13 (Table 2) and 16 hurs indicated slightly higher temperatures in the vines which had been expsed t preveraisn water deficits, but n differences were greater than 1.5 C at any sample time. Cluster temperatures did nt increase measurably when clusters were expsed t direct sunlight by restraining the fliage thrughut the diurnal measurement perid. Table 2. Midday (13 h) cluster temperature and internal canpy ambient temperature in plts receiving different irrigatin treatments. Values are the mean f tw samples (clusters) in each treatment. Treatment Interir air temperature ( C) Cluster Canpy Cntinual Late deficit arly deficit Full deficit Fruit grwth respnded t plant water status. At harvest, berry vlume was significantly less in vines frm which water was withheld than in C vines (inset table, Fig. 1). The vlume f C berries was apprximately 31% and 39% greater than LD and D berries, respectively. In all treatments, berry water cntent decreased similarly thrughut the seasn, reaching apprximately 72% f the fresh weight at harvest (Fig. 1). The duble-sigmid grwth habit f the berry creates a minr cmplicatin in interpretatin f differences in slute cncentratins, since fruit ripening and grwth ccur simultaneusly. A dilutin f inrganic cnstituents (decrease in cncentratin) withut a net lss theref may ccur in berries during Stage III grwth. On the ther hand, slute accumulatin may Am. J. nl. Vitic., Vl. 39, N. 4, 1988

3 WATR DFICITS AND FRUIT RIPNING ~ c- c- O {._) L_ (D >,, k._. k._ m ~ ~ ~ ~ ~ ~ ~ Treatment Final Vlume (cm 3 1 berries-i) Cntinual Late Deficit arly Deficit Full Deficit Fig. 1. Water cntent (% fresh wt) f Cabernet franc berries during ripening. Cntinual treatment shwn nly; ther treatments exhibited the same pattern and were mitted fr clarity. Final (harvest sample) water cntent was within 1.5% f 72% f berry fresh weight fr all treatments. Inset table shws the final berry vlume fr vines receiving different irrigatin treatments. Data shwn are means +standard errr (n = 5). 5O 15, I, I Julian day ccur during expansive grwth despite a cnstant (r even decreasing) cncentratin. Cnsequently, juice cmpsitin was analyzed n the bases f cncentratin and cntent per berry. In mst cases, the final juice cncentratin is the parameter f significance fr winemaking and sensry attributes. The rate f increase in sugar cncentratin was initially similar in all treatments but slwed near harvest in LD vines (Fig. 2). At harvest, the cncentratin f sluble slids was 22. _+.1, 21.9 _+.1, and 2.8 _+.3 Brix (x +_ s.e.m., n = 5) in C, D, and LD treatments, respectively (Fig. 2). The final cncentratins f sluble slids in FD and SP treatments (21.5 and 21.7 Brix, respectively) were intermediate t the D and LD treatments. The amunt f sugar in each berry was always greatest in the C vines, and this difference increased near harvest when the rate f grwth (14) and accumu- I, I latin f sugar per berry (inset, Fig. 2) slwed in D and LD vines. Juice TA increased slightly befre and then declined apprximately 1X after veraisn in all treatments (Fig. 3). At veraisn, the TA f D vines was slightly greater than in treatments irrigated weekly (Fig. 3), althugh the difference was nt statistically significant. In all treatments, the TA decreased significantly befre an increase in sluble slids was detectable (cf. Fig. 2, 3). At harvest, mean TA was slightly lwer in D juice (.37 mg tartrate equiv/1 ml, respectively). Final TA levels f FD and SP vines were intermediate and did nt differ frm ther treatments. TA per berry fllwed a pattern similar t that fjuice TA (inset, Fig. 3). All berries lst apprximately 75% f the TA present befre veraisn, but TA per berry at harvest was significantly greater in C berries than in D berries (inset, Fig. 3). Fig. 2. Sluble slids (expressed as Brix)in the juice at varius times f the seasn fr Cabernet franc vines which received different irrigatin treatments during fruit develpment. There were n significant differences until the final sample date. Final Brix given in text, except full deficit and standard practice treatments (nt shwn), which were 21.5 and 21.7 Brix, respectively. Inset figure shws accumulatin f sluble slids n a per berry basis. All data are means fr the juice f 1-berry samples taken frm each f 5 replicate plts, x c m._. -- _ ~ _ I b L_ 225 ~ 15 m iot m Julian day y Cntinual Late Deficit " ~"--- ~ ~ " arly Deficit e ~ ~ ~ S veraisn I, I, I Julian day Am, J. nl. Vitic., Vl. 39, N. 4, 1988

4 _ ~ WATR DFICITS AND FRUIT RIPNING i, C) CT (1) 4 3 ca L_ 2 L_ CD < k--. ~ 8~~-~--.,~ ~ Cntinual _ Late Deficit ~'~-/~ % ~ ~ arly Deficit _ f-= l :~'-"--" el..... :--'7't - ;6s lss 2s 22s 2,;s ~ ~ ~, ~ Julian day. Z ~ ~ m ~ veralsn, I, I, I, Fig. 3. Titratable acidity (TA) f juice at varius times f the seasn frm Cabernet franc vines which received different irrigatin treatments during fruit develpment. There were n significant differences amng treatments until the final sample date (see text). Inset table shws TA n a per berry basis at varius times f the seasn. All data are means fr the juice f 1- berry samples taken frm each f 5 replicate plts. JuliGn day Table 3. The cncentratin f malate, prline, and ttal sluble phenlics in juice and ttal sluble phenlics in dermal extracts fr fruit f vines which received different irrigatin treatments. Data are means + standard errr, n = 5. Phenlics (gallic acid equiv.) Treatment Malate Prline Juice Dermal (g/1ml) (mm) (pm) (~ml/cm 2) Cntinual Late deficit arly deficit Full deficit Fig. 4. The ph f juice at varius times f the seasn frm Cabernet franc vines. Cntinual treatment shwn nly; ther treatments exhibited the same pattern and were mitted fr clarity. The inset table shws the final juice ph f vines which received different irrigatin treatments during fruit develpment (C, cntinual; LD, late deficit; D, early deficit; and FD, full deficit). All data are means fr the juice f 1-berry samples taken frm each f 5 replicate plts. T L ~ ~ ~ ~ Final ph (_+s.e.m. n=5) C LD D ~ ~ ~ Cntinual ~ / FD i I ~ veraisn I i I Julian day Am. J. nl. Vitic., Vl. 39, N. 4, 1988

5 WATR DFICITS AND FRUIT RIPNING O m 4 rq ,5 ± b L Cntinual arly Deficit verisn ~ ~ 1 "~ 5 l 13_ 165 T {, i i Julian day Fig. 5. The cncentratin f ptassium in the juice at varius times f the seasn frm Cabernet franc vines which received different irrigatin treatments during fruit develpment. Data frm the late deficit, full deficit, and standard practice treatments were similar t the cntinual treatment and were mitted fr clarity. Inset figure indicates the ptassium cntent n a per berry basis. All data are means + standard errr (n = 5)., I, I Julin dy In cntrast t TA, malate cncentratin was markedly dependent upn vine water status. At harvest, malate levels in juice were.19,.16, and.13 g/1 ml in C, LD, and D vines, respectively (Table 3). Malate level was als lw in FD vines (.11 g/1 ml) and significantly less in D and FD vines than in LD and C vines (Table 3). The cncentratin f the primary catins in grape juice (H and K ) differed little amng treatments. The ph f C juice at varius times during ripening is shwn in Figure 4 t indicate the seasnal pattern. The ph f juice was virtually identical and increased after veraisn in cncert in all treatments. Final juice ph did nt differ by mre than.6 ph units amng the treatments (inset table, Fig. 4). There were n significant differ- ences nr apparent trends amng treatments. Ptassium cncentratins in juice during ripening were als similar amng treatments, increasing befre veraisn and decreasing after veraisn until reaching a stable value ca 15 days befre harvest (Fig. 5). There were n significant differences in K cncentratin at harvest. Hwever, K cncentratin cnsistently increased mre rapidly and peaked earlier in D vines than in vines irrigated weekly until veraisn, i.e., C and LD treatments (Fig. 5). K per berry increased thrughut ripening in all treatments (inset, Fig. 5). The rate f increase was similar and relatively cnstant in D and LD vines but was significantly greater in C vines after veraisn (inset, Fig. 5). The cncentratin f anthcyanins extractable frm.8 Fig. 6. The cncentratin (area basis) f anthcyanins extracted frm dermal tissue f berries at varius times f the seasn fr Cabernet franc vines which received different irrigatin treatments during fruit develpment. All data are means + standard errr (n = 5). O4 I (.9 T~ O c-- _ (- (D.6.4 Cntinual "/ $ Lre Deficit ~ /' arly Deficit Z ~ T (D (- (- < I, I, I, I, Julian day 245 Am. J. nl. Vitic., Vl. 39, N. 4, 1988

6 318 m WATR DFICITS AND FRUIT RIPNING dermal tissue began t increase rapidly at apprximately day 212 in all treatments (Fig. 6). During the subsequent 3 days, anthcyanins increased mre than 1X (Fig. 6). The rate f increase during the initial 15 days was apprximately 2X that f the final 15 days in all treatments (Fig. 6). Treatment differences were established during the initial 15 days f rapid increase in anthcyanin cntent when the increase in cncentratin was mst rapid in D vines and slwest in C vines (Fig. 6). Cncentratins f anthcyanins at harvest were.51,.59, and.64 mg/cm 2 in C, LD, and D vines, respectively (Fig. 6). The cncentratin f ttal phenlics in juice and dermal extracts were increased significantly by withhlding water at different times f the seasn. When water deficits were impsed befre r after veraisn, phenlic cncentratin at harvest was apprximately 1.15 ~M (Table 3), whereas in C vines, phenlic cncentratin Was ca.88 ~M (Table 3). Similarly, cncentratin f phenlics in dermal extracts f C vines was significantly less than in vines which experienced water deficits (Table 3). Phenlics in FD juice and dermal extracts were virtually identical t D levels (Table 3). The prline cncentratin in juice at harvest was least in C vines (3.95 mm), intermediate in D vines (4.66 mm), and greatest in LD vines (5.76 mm) (Table 3). FD vines als had a significantly higher prline level (4.79 mm) than C vines. An exceptin t this ccurred in 1985 when the prline level f C vines was greater than thse f D and FD vines (data nt shwn). Discussin The results shw that vine water status was readily manipulated by altering the amunt and timing f water applicatins t a drip-irrigated, hillside vineyard in Napa Valley, Califrnia. Differences in vine water status, established befre and after veraisn, led directly t differences in the size and cmpsitin f Cabernet franc winegrapes. Berry slutes sensitive t vine water status included rganic acids, amin acids, anthcyanins, and ttal sluble phenlics. Brix, TA, and K were slightly higher in D vines than in C r LD vines befre veraisn. This may have been due t the mderate preveraisn water deficit having a greater effect n fruit grwth than n fruit metablism. Since TA was slightly lwer in D vines at harvest, the rate f acid lss was prbably greater in D vines (and smewhat slwer in C vines) than in ther treatments, althugh differences in fruit grwth may cnfund this interpretatin. The high malate cncentratin at harvest in C juice and lw cncentratin in D juice supprt this cnclusin, since mst f the acid lst during fruit ripening is malate (11). The malate cncentratin at harvest was als lw in FD juice, which indicates that preveraisn water deftcits decreased the final malate cncentratin independent f vine water status during fruit ripening. The pattern f decline in TA after veraisn suggests that the differences in malate may have been due t differences in catablism after veraisn rather than t the malate level at veraisn. These bservatins are relevant t vineyard management and winemaking decisins, since the large effect n malate and relatively small effect n TA suggest that early seasn water deficits result in increased tartrate t malate ratis. Van Zyl (29) clearly shwed that the tartrate t malate rati increased when water was withheld frm drip- irrigated Clmbar in Suth Africa. This may be imprtant in determining the methd fdeacidificatin f musts with high TA (3) and in the stability f ph and TA during mallactic fermentatins. Althugh K uptake cntinued thrughut fruit ripening treatments, K cncentratin increased befre veraisn and decreased slightly after veraisn in all treatments. Thus, juice ph increased as K was decreasing. Cmbe (5) als shwed a decrease in K cncentratin frm 17 t 26 Brix in Muscat Grd. The respnses fjuice K and ph t seasnal water deficits were similar in that there were n treatment effects evident at harvest, althugh early-seasn water deficits caused a slight decrease in juice TA. Thus, under the cnditins f this study, there appears t be limited ptential t manipulate juice ph status with irrigatin scheduling, whether via K uptake frm the sil, K transprt t fruit, r ther mechanisms. Similar results have been bserved with ther cultivars and experimental prtcls (16,29,32). Althugh irrigatin studies have shwn increased (8,28) and decreased (17)juice ph as a results f supplemental irrigatin, the effects have almst always been marginal. Fr example, significant differences were bserved in ne ut f tw years (27) r in ne ut f three years (8). Therefre, the general sensitivity f juice ph t vine water status is nt high and may be site- and variety-specific. The cncentratin f phenlics was dependent upn vine water status. Bth early- and late-seasn water deficits resulted in phenlic cncentratins in the juice and dermal extracts which were mre than 3% and 15% greater, respectively, than in vines maintained at a higher water status thrughut the seasn. At present, n assay f sluble phenlics can discriminate between phenls which impart bitterness and astringency and thse that d nt (31). Changes in phenlic cncentratin in the juice suggest that the nnflavnid (24) phenlics, present predminantly in the vacules f the mescarp cells (31), were particularly sensitive t vine water status. Sensry research has shwn that the nnflavnids cntribute little t wine flavr (18,3). Hence, the differences in phenlics f dermals extracts, albeit less than in the juice, may be f greater imprtance t the sensry characteristics f the wine, since there is a high prprtin f flavnids in the phenlics f the dermal cells (2). The increase in phenlic cncentratin in the juice was similar t the decrease in fruit vlume caused by lw water status. Hwever, the phenlic cncentratin in dermal extracts als increased when expressed n a surface area basis. These differences are f clear imprtance due t the prminent rle Am. J. nl. Vitic., Vl. 39, N. 4, 1988

7 WATR DFICITS AND FRUIT RIPNING f phenlics in determining the clr, bitterness, and astringency f table wines (24,31). The bservatin f increased anthcyanin cntent in juice r wine in treatments which decrease (r were likely t have decreased) vine water status can nw be cnsidered cmmnplace (33). The results here indicated that clr develpment was mst rapid during the first tw weeks after the nset f anthcyanin synthesis and that clr develpment was mre sensitive t vine water status in the early rather than late stages f the ripening prcess. The general respnse t water deficits and recgnitin f the imprtance f the early phase f fruit ripening in the synthesis f anthcyanins may facilitate imprved winegrape prductin fr cultivars and envirnments in which clr prductin is a cncern. Prline is the primary free amin acid in the juices f many winegrape cultivars, including Cabernet Sauvignn, Merlt, Petit Sirah, and Zinfandel (12). Althugh there are n reprts fr Cabernet franc, it is likely that prline is als the primary free amin acid fr this variety, since the amin acid prfiles f the abve varieties (which are similar t Cabernet franc) were very similar (12). Kliewer (12) shwed that the cncentratin f prline in juice increases during ripening f many cultivars. In Cabernet franc, the prline cncentratin at harvest was higher in vines which were at lw water status and lwer in vines at high water status. Hwever, withhlding irrigatin decreased the accumulatin f prline in Carignane at Davis, Califrnia (8). The cause f the increase in prline cncentratin during ripening and f the different respnses t lw vine water status is nt clear. Prline is unlikely t play a direct rle in fermentatin r wine flavr, since prline is flavrless and is utilized nly when ther amin acids have becme limiting (21). Hwever, amin nitrgen clearly plays an imprtant rle in yeast grwth during fermentatin (2), and prline levels have been crrelated psitively with summed amin acid cncentratins in ripening grapes (6). There are several indirect mechanisms by which seasnal perids f lw water status culd alter fruit cmpsitin. The ptential cnfundment f waterstatus-induced differences in "crp lad" (7,21) was effectively avided by cnsidering data frm the initial seasn f the study nly. It was nly in the initial seasn, in which yields differed but due nly t differences in berry size (15), that the cumulative effects f water deficits n reprductive develpment (14) culd be avided. Hwever, the reprted treatment differences in vine water status and fruit cmpsitin were bserved in each f three seasns (with ne exceptin, nted in Results). Fruit size may be imprtant in determining the extractin (dilutin) f dermal cell cntents, which are clearly the primary site f several imprtant slutes fr winemaking (5). Large diameter fruit wuld have a greater slvent (mescarp cell sap) t slute (dermal cell sap) rati as a result f the lwer surface t vlume rati cmpared t smaller fruit. Hence, the inhibitin f fruit expansin by water deficits may diminish the dilutin f dermal slutes in the must. The differences in fruit cmpsitin reprted here are unlikely t be attributable t the simple inhibitin f fruit expansin (decrease in slvent) r t an indirect effect f increased fruit temperature (due hypthetically t increased expsure f clusters t slar radiatin). First, the levels f berry slutes did nt change in ne directin which wuld have been cnsistent with simple altered vlume r temperature hyptheses. Cmpnents exhibited unchanged ( Brix, K), decreased (malate), r increased cncentratins (phenlics, prline) as a result f lw water status. Secnd, the directin f the change in slute cncentratins was nt always cnsistent with that predicted by these hyptheses. Fr example, althugh water deficits increased the cncentratin f prline, the prline cncentratin was nt greatest in the D r FD treatments, which had the smallest fruit and inhibited canpy grwth (15). Als, the increase in phenlics in the treatments which experienced water deficits is incnsistent with an increased fruit temperature (18). Finally, we btained little evidence f increased fruit temperature (at least at the cluster interir) as a result f water deficits, althugh small differences may have ccurred. Water deficits might als alter the nset r duratin f the ripening perid. Differences in vine water status befre veraisn had n effect upn the nset f veraisn (15; Fig. 6). Withhlding water after veraisn eventually slwed the increase in sluble slids, but this effect was nt evident until water deficits were maximal (shrtly befre harvest in the LD treatment). Alltreatments were harvested n the same date and were within 1.2 Brix. Smart and Cmbe (27), citing studies with yield differences f up t 131% (irrigated cmpared t nnirrigated), suggested that the time required fr ripening was inversely related t yield increases brught n by increased water status. In this study, the water status and yield f C vines were greater than thse f ther treatments (15), but there was n delay f fruit maturatin in C vines cmpared t any treatment. Cnsequently, the differences in cmpsitin discussed belw cannt be attributed t any general effect n the duratin f the ripening perid. There are evidently direct effects f vine water status n berry metablism during ripening. The water deficits encuntered in the LD treatment were at the margin f that which wuld significantly delay fruit maturatin. Hwever, it may be imprtant t nte that the increase in sluble slids was nt delayed in FD vines which received much less water than LD r ther vines. Cnsistent with this bservatin was the lwer water status f LD vines cmpared t FD vines at harvest (Table 1). This suggests that the water deficits created by supplying water at a relatively high rate befre veraisn and withhlding water after veraisn Am. J. nl. Vitic., Vl. 39, N. 4, 1988

8 32-- WATR DFICITS AND FRUIT RIPNING were near the maximum achievable in that vineyard. This als demnstrates the imprtance f quantitating vine water status in investigatins f vine respnses t water deficits, since LD vines received mre than twice the ttal amunt f water applied t FD vines. The develpmental perid during which water deficits were impsed selectively determined the slutes affected. Therefre, in vineyards in which vine water status is sensitive t irrigatin, irrigatin scheduling ffers the grwer an pprtunity t cntrl the cmpsitin f the raw material and, hence, f the prduct wine. Hwever, these results relate fruit cmpsitin t vine water status nly. The irrigatin scheduling required t btain varius vine water statuses in different mesclimates has nt been addressed. It may als be imprtant t nte that althugh sme f the cmpsitinal respnses t water deficits shwn here may be perceived as psitive, inhibitin f flral develpment may be a simultaneus cnsequence f lw water status. Literature Cited 1. Amerine, M. A., and C. S. Ough. Methds fr Analysis f Musts and Wines. 341 pp. Jhn Wiley and Sns, New Yrk (198). 2. Arnld, R. A., A. C. Nble, and V. L. Singletn. Bitterness and astringency f phenlic fractins in wine. J. Agric. Fd Chem. 28:675-8 (198). 3. Bultn, R. B. The general relatinship between ptassium, sdium and ph in grape juice and wine. Am. J. nl. Vitic. 31:182-6 (198). 4. Bradfrd, K. J., and T. C. Hsia. Physilgical respnses t mderate water stress. In: ncyclpedia f Plant Physilgy, Vl. 12B. O. L. Lange, P. S. Nbel, C. B. Osmnd, and H. Ziegler (ds.). pp Springer-Verlag, New Yrk (1982). 5. Cmbe, B. G. Distributin f slutes within the develping grape berry in relatin t its mrphlgy. Am. J. nl. Vitic. 38:12-7 (1987). 6. Du Plessis, C. S. Optimum maturity and quality parameters in grapes: a review. S. Afr. J. nl. Vitic. 5:35-42 (1984). 7. Freeman, B. M. ffects f irrigatin and pruning f Shiraz grapevines n subsequent red wine pigments. Am. J. nl. Vitic. 34:23-6 (1983). 8. Freeman, B. M., and W. M. Kliewer. ffect f irrigatin, crp level and ptassium fertilizatin n Carignane vines. I1. Grape and wine quality. Am. J. nl. Vitic. 34: (1983). 9. Guymn, J. F., and C. S. Ough. A unifrm methd fr ttal acid determinatin in wines. Am. J. nl. Vitic. 13:4-5 (1962). 1. Hhrst, H. J. L-(-)-Malate determinatin with malicdehydrgenase and DPN. In: Methds f nzymatic Analysis. H.-U. Bergmeyer (d.). pp Academic Press, New Yrk (1963). 11. Kliewer, W. M. Changes in the cncentratin f malates, tartrates, and ttal free acids in the flwers and berries f Vitis vinifera. Am. J. nl. Vitic. 16:92-1 (1965). 12. Kliewer, W. M. Free amin acids and ther nitrgenus fractins in wine grapes. J. Fd Sci. 35:17-21 (197). 13. Kliewer, W. M. Influence f temperature, slar radiatin and nitrgen n clratin and cmpsitin f mperr grapes. Am. J. nl. Vitic. 28:96-13(1977). 14. Matthews, M. A., and M. M. Andersn. Reprductive develpment in grape (Vitis vinifera L.): Respnses t seasnal water deficits. Am. J. nl. Vitic. (In press, 1989). 15. Matthews, M. A., M. M. Andersn, and H. R. Schultz. Phenlgical and grwth respnses t early and late seasn water deficits in Cabernet franc. Vitis 26:147-6 (1987). 16. Mrris, J. R., and D. L. Cawthn. ffect f irrigatin, fruit lad, and ptassium fertilizatin n yield, quality, and petile analysis f Cncrd (Vitis labrusca L.) grapes. Am. J. nl. Vitic. 33:145-8 (1982). 17. Neja, R. A., W.. Wildman, R. S. Ayers, and A. N. Kasimatis. Grapevine respnse t irrigatin and trellis treatments in the Salinas Valley. Am. J. nl. Vitic. 28:16-26 (1977). 18. Nble, A. C. Bitterness and astringency in wine. In: Bitterness in Fds and Beverages. R. Russef (d.). lsevier Science Publishers, Amsterdam (1987). 19. Ough, C. S. Rapid determinatin f prline in grapes and wines. J. Fd Sci. 34:228-3 (1969). 2. Ough, C. S., and M. A. Amerine. Fermentatin rates f grape juice. IV. Cmpsitinal changes affecting predictin equatins. Am. J. nl. Vitic. 17: (1966). 21. Ough, C. S., and R. Nagaka. ffect f cluster thinning and vineyard yields n grape and wine cmpsitin and wine quality f Cabernet Sauvignn. Am. J. nl. Vitic. 35:3-4 (1984). 22. Peynaud,., and S. Lafn-Lafurcade. Cmpsitin azt6e des vins (r vignes) en fnctin des cnditins de vinificatins. Ann. Technl. Agric. 1:143-6 (1961 ). 23. Pruitt, W. O.,. Ferreres, K. Kaita, and R. L. Snyder. Reference vaptranspiratin fr Califrnia. Univ. Calif. Div. Agric. Natur. Res. Bull. N (1987). 24. Singletn, V. L., and A. C. Nble. Wine flavr and phenlic substances. In:Phenlic, Sulfur and Nitrgen Cmpunds in Fd Flavrs. G. Charalambus and I. Katz (ds.). pp American Chemical Sciety Sympsium Series, N. 26 (1976). 25. Singletn, V. L., and J. A. Rssi, Jr. Clrimetry f ttal phenlics with phsphmlybdic-phsphtungstic aid reagents. Am. J. nl. Vitic. 16: (1965). 26. Slinkard, K., and V. L. Singletn. Ttal phenl analysis: autmatin and cmparisn with manual methds. Am. J. nl. Vitic. 28:49-55 (1977). 27. Smart, R.., and B. G. Cmbe. Water relatins f grapevines. In: Water Deficits and Plant Grwth, Vl. VII. T. T. Kzlwski (d.). pp Academic Press, New Yrk (1983). 28. Vaadia, Y., and A. M. Kasimatis. Vineyard irrigatin trials. Am. J. nl. Vitic. 12:88-98 (1961 ). 29. Van Zyl, J. L. Respnse f Clmbard grapevines t irrigatin as regards quality aspects and grwth. S. Afr. J. nl. Vitic. 5:19-28 (1984). 3. Ver~tte,., A. C. Nble, and T. C. Smmers. Hydrxycinnamates f Vitis vinifera: sensry assessment in relatin t bitterness in white wines. J. Sci. Fd Agric. (In press, 1988). 31. Webb, A. D. Quality factrs in Califrnia grapes. In: Quality f Selected Fruits and Vegetables f Nrth America. R. Teranishi and H. Barrera-Benitez (ds.). Am. Chem. Sc. Symp. Ser. 17:1-9 (1981 ). 32. Wildman, W.., R. A. Neja, and A. N. Kasimatis. Imprving grape yield and quality with depth-cntrlled irrigatin. Am. J. nl. Vitic. 27:168-75(1976). 33. Williams, L.., and M. A. Matthews. Grapevines. In:Irrigatin f Agricultural Crps. B. A. Stewart and D. R. Nielsen (ds.). Am. Sc. Agrn. Mngr. (In press, 1988). Am. J. nl. Vitic., Vl. 39, N. 4, 1988