specific length during the growing season. Three trellising designs were used to ob- tain three distinct shoot inclinations within Type of Trellis Affects Radiation the vineyard. The positioning of shoots on trellises resulted in canopies that were 1) po- Absorption and Must Composition but sitioned vertically (vertical canopy), 2) po- sitioned in a “V” pattern with sides inclined not Yield of ‘Petite Sirah’ 60o from horizontal (inclined canopy), and 3) inclined toward the vineyard floor (stan- T.H. Morsil, A.D. Matthias2, and J.L. Stroehlein3 dard canopy). The vertical canopy was ob- Department of Soil and Water Science, University of Arizona, Tucson, tained by attaching cordons and shoots to a vertically arranged three-wire trellis. The AZ 85721 lower wire was used for cordon support, and Additional index words. radiation balance, PAR, wine, vineyard, the middle and upper wires were used to sup- port shoots. Spacing between lower and Abstract. The effects of trellising on absorption of photosynthetically active radiation middle wires was 0.46 m, and spacing be- (PAR, 400- to 700-nm wavelength) by foliage and fruits, on fruit composition, and tween middle and upper wires was 0.58 m. yield were studied in 1988 under semi-arid conditions using field-grown Vitis vinifera This canopy was designed to provide in- L. ‘Petite Sirah’ grapevines in a mature vineyard. A vertical inclination was obtained creased PAR absorption and fruit shading near o by attaching shoots to a vertically arranged three-wire trellis; 60 shoot inclination solar noon when incoming radiant energy from horizontal was obtained by attaching shoots to a “V-type” Tatura trellis; a stan- fluxes were greatest. The inclined canopy dard two-wire trellis (control) was used in which shoots attached to the upper wire was supported on a “V-type” Tatura trellis. were permitted to orient downward to the vineyard floor. PAR absorption by foliage This trellis consisted of five wires for sup- during mid-morning to mid-afternoon periods was highest in the standard trellis and port of cordons and shoots. A lower wire lowest in the Tatura trellis. PAR available for absorption by fruits was lowest in the was used to support cordons. Four upper wires standard trellis and highest in the Tatura trellis. Analysis of fruit composition at harvest were attached to the ends of stacked 0.38- o revealed that total dissolved solids ( Brix) was significantly higher for berries from the and 0.76-m double crossarms for support of Tatura trellis than for the vertical trellis or the control. The Tatura trellis resulted in shoots. Spacing between crossarms was 0.58 the highest alcohol content of wine. Per-vine yields did not differ significantly among m. The inclined canopy was used to study the three trellis systems. the effect of decreased PAR absorption by foliage and decreased shading during midday Trellis design is an important factor that in- We present preliminary results from a field periods. The standard canopy (control) was fluences the physiological and management study in Arizona to determine the effects of supported on a standard two-wire trellis with aspects of grapevines (Smart, 1988; Smart et three trellising systems on absorption of pho- a 0.46-m spacing between the lower and up- al., 1982, 1985; Van den Ende, 1984). De- tosynthetically active radiation (PAR) by fo- per wires. Shoots were attached to the upper signs in which excessive shading of fruits by liage and fruits, fruit composition, and yield wire and were permitted to orient downward foliage occurs have generally reduced fruit of mature ‘Petite Sirah’ grapevines. This study toward the vineyard floor. This orientation quality and yield. A two-wire (Geneva Double was motivated by the need to determine op- resulted in a relatively dense foliage config- Curtain) trellis design increased ‘Concord’ timal shoot positioning for the production of uration that nearly enveloped the fruits and yields by as much as 90%, relative to a single- wine grapes in a hot, semi-arid climate, such provided nearly total fruit shading from di- wire (Kniffin) design, as a result of improved as Tucson, because, to our knowledge, noth- rect sunlight throughout the day. The stan- leaf and shoot exposure (Shaulis et al., 1966). ing has been published about the effect of dard trellis is commonly used in vineyards The yield of ‘Crouchen’ grapes was increased shoot positioning on productivity of grapes in the southwestern United States. Cordon up to 30% due to trellis widening from 0.3 to grown in such a climate. Growth of wine height above ground for all three canopies 1.4 m (May et al., 1976). Improved fruit ex- grapes under high air temperatures and in- was » O.9 m. Establishment of vertical and posure with four- and six-wire trellising sys- tense midday solar radiant energy fluxes re- inclined canopies occurred during Mar. 1988. tems increased ‘Cabernet Sauvignon’ yields by quires efficient configuration of the vine A completely randomized design was used, 45% and 48%, respectively, relative to the yield canopy to achieve optima1 radiation absorp- consisting of the three treatments. Each can- from a standard two-wire trellis (Steinhauer tion by foliage, fruit shading, and fruit tem- opy treatment was replicated eight times, with and Bowers, 1979). Reynolds et al. (1986) perature conditions within the canopy. each replicate containing three vines. Data found that the most exposed ‘Seyval blanc’ Excessive shading by foliage or excessive from the two outer (border) plants within a fruit had the highest pH levels, in addition to exposure to intense radiant energy may re- replicate were not collected; thus, data from the highest oBrix and tartrate content, and the duce fruit quality. only the middle vine were used in the study. lowest titratable acidity. Reynolds et al. (1986), The experiment was conducted during the Components of the PAR budget of each however, reported no significant wine quality 1988 growing season at a 0.7-ha vineyard at type of canopy were measured on several differences due to differences in fruit expo- the Univ. of Arizona Campus Agricultural clear days during the season using a 80-cm- sures. Recently, Patterson and Zoecklein (1990) Center, Tucson, using lo-year-old ‘Petite long, Decagon Model SF-80 Sunfleck Cep- showed that increased radiation penetration due Sirah’ grapevines. This cultivar was chosen tometer (Decagon, Pullman, Wash.). Mea- to shoot positioning and removal of nonfruitful because it generally produces wines with ad- surements of the descending and ascending shoots resulted in superior fruit (i.e., reduced equate color and flavor in warm climates PAR (µmol·m-2·s-1) components of the malate and potassium) relative to other treat- (Mielke et al., 1980). The vines were grown budget were made at 0.5-h to 2-h intervals, ments in ‘Norton’ grapes. in Brazito sandy loam soil, which is classi- typically from mid-morning to mid-after- fied as mixed thermic Typic Torrifluvents. noon, and were used to calculate the PAR Received for publication 17 Dec. 1990. Accepted The vines were basin-irrigated at » 3-week absorbed by foliage (APAR) for each canopy for publication 28 Aug. 1991. Journal article no. intervals. Spacing was 2.1 m between vines, design. Instantaneous APAR was calculated 7308 from the Arizona Agr. Expt. Sta, The cost with rows 3.3 m apart. Rows were oriented using Eq. 1 from Gallo and Daughtry (1986) of publishing this paper was defrayed in part by north-south. Before 1988 each vine had been given here as: the payment of page charges. Under postal regu- trained to a standard two-wire trellis. All vines lations, this paper therefore must be hereby marked APAR = (PARo + RPARs)— advertisement solely to indicate this fact. were spur-pruned in Jan. 1988 to 20 buds/ (TPAR + RPARc) [1] 1Graduate Student. cordon with two cordons per plant. There 2Associate Professor. were » 20 shoots/meter of row and » 130 where PAR, is the PAR flux incident on the 3 Professor. leaves/shoot. Shoots were not pruned to a canopy, RPAR, is the PAR reflected from the soil surface to the canopy, TPAR is the of PAR available for absorption by the fruits. incoming PAR (PARo) was absorbed by the PAR transmitted through the canopy, and As done with APAR, FPAR was also con- foliage in the standard canopy, thus provid-

RPARc is the total PAR reflected from the verted to daily (integrated) values. Also, ing relatively uniform shading of fruits at the canopy and the soil. PARo and RPARc were FPAR was expressed as a fraction of PARo base of the canopy. This high percentage of measured at » l m above the canopy using on a daily basis according to: FPAR' = midday PARo absorption by foliage may in- the leveled ceptometer facing either upward FPAR/PARo. Only FPAR' and APAR' val- crease photosynthesis and growth in the or downward, respectively. RPARs was ues are presented. standard canopy relative to the others. In measured with the inverted ceptometer at The fruits were harvested 17 Aug. 1988, contrast to the standard canopy, the inclined » 0.2 m above the soil surface. TPAR was which was several days earlier than the nor- canopy absorbed, on average, only » 71% measured by positioning the upright level mal harvest of ‘Petite Sirah’ at the end of of PARo, thus allowing for increased expo- ceptometer diagonally across the canopy at August in Tucson (Mielke et al., 1980). Fruit sure of fruits, and potentially higher fruit about the height of the basal fruits. Ten rep- was harvested early to avoid yield loss due temperatures. Similarly, the vertical canopy licate measurements of each of the four PAR to possible infestation by birds and insects. absorbed on average » 74% of midday PARo. components on the right-hand side of Eq. [l] Fruit from each vine was weighed (kilo- Because of the open canopy architecture were made per plant for three plants within grams per vine) and transported to the Univ. in the inclined and vertical canopies, » 30% each canopy design. From these measure- of Arizona Winery, Tucson. All grapes were of PARo was potentially available for ab- ments, an average instantaneous APAR was crushed within » 6 h after harvest. Must from sorption by the fruits. This contrasted mark- determined from Eq. [l] for each canopy de- each replicate vine was analyzed for total edly with the average 9% value of FPAR' sign. Daily (integrated) values of APAR (units soluble solids (TSS; oBrix), pH, and titrat- for the standard canopy. In the standard can- of MJ·m-2) were then computed by numer- able acidity (grams of acid/100 ml of juice) opy, there was much less radiation penetrat- ical integration of the average instantaneous using procedures described by Amerine and ing the canopy from the sky, as well as less values of APAR determined throughout the Ough (1974). After chemical analysis of the radiation reflected from the vineyard floor day. In addition, the daily APAR values were juice from each of the eight replicates per into the vicinity of the fruits. computed as fractions of the daily PARo ac- treatment, the juices were combined to pro- Mean (± 1 SD) fruit yields were 20.1 ± cording to: APAR' = APAR/PARo. duce one batch of wine per treatment. The 10.4, 14.0 ± 4.1, and 16.9 ± 5.2 kg/vine PAR available for absorption by fruits three finished wines were analyzed » 6 months for the standard, inclined, and vertical treat- (FPAR) at the base of each canopy was cal- after harvest for pH, titratable acidity (g/100 ments, respectively. The large standard de- culated as the sum of the PAR transmitted ml), volatile acidity (g/100 ml), and alcohol viation in yields within each treatment downward to the fruits through the foliage (%) by standard procedures (Amerine and (particularly for the standard canopy) re- (TPAR) and the PAR reflected upward to the Ough, 1974). Wine color was determined sulted in nonsignificant mean yield differ- fruits from the soil surface (RPARs), i.e., calorimetrically at 420 and 520 nm accord- ences (range 14.0 to 20.1 kg/vine). Non-

FPAR = TPAR + RPARs. Actual absorp- ing to Amerine and Ough (1974). Statistical significant yield differences are generally tion of PAR by individual fruit clusters could comparisons of treatment means (LSD at P consistent with most previous studies of the not be measured directly by the ceptometer; = 0.05) for measured must attributes are effects of trellising on yield. Baldwin et al. thus, FPAR is only an indicator of the amount reported. However, chemical characteristics (1979) found no differences in ‘Sultana’ yields for wine produced from each treatment were due to six T-trellis treatments. They found, not tested statistically because juice from the however, that three pruning levels did result replicates had been combined. in significant yield differences. Couvillon and PAR absorption by foliage (APAR and Nakayama (1970) reported that yields of APAR') during six midday periods was, on ‘Concord’ grapes grown on a two-wire average, highest in the standard canopy crossarm (horizontal) trellis were higher than treatment and lowest in the inclined canopy in a two-wire vertical trellis. Weaver et al.

treatment (Table 1). PARo for those days was (1984) found that the highest (2 m tall) trellis » 30 MJ·m-2. During midday, when heat and used resulted in the highest yield of ‘Thomp- radiation levels were highest, » 85% of the son Seedless’ grapes, but the use of cros- sarms did not significantly increase yields, which is consistent with our results. Weaver and Kasimatis (1975), however, found that ‘Thompson Seedless’ vines trained on trel- lises with crossarms gave the highest yields. Buttrose (1970) reported that use of crossarm trellises did not increase yields of several cultivars because a lack of light penetration in the interior of the vines limited fruit bud differentiation and the light exposure of shoots, especially the basal portions. Shaulis and May (1971) suggested that the low amount ever, found oBrix unaffected by trellis treat- photosynthetically active radiation in corn can- of solar radiation penetrating the canopy on ment (two-, four-, and six-wire) or yield opies. Agron. J. 78:752-756. a standard T-trellis is a major factor limiting differences of ‘Cabernet Sauvignon’. Also, Hedberg. P.R. and J. Raison. 1982. The effect of production of ‘Thompson Seedless’ grapes. the concentration of sugars and acids and the vine spacing and trellising on yield and fruit quality of Shiraz grapevines. Amer. J. Enol. They achieved higher yields by reducing shoot mean berry weight in ‘Crouchen’ were un- Viticult. 33:20-30. crowding through extension of canopy length affected by trellis width (0.3 to 1.4 m) and Kasimatis, A.N., L.A. Lider, and W.M. Kliewer. and by reducing the number of nodes per unit pruning treatments (May et al., 1976). Hed- 1976. Increasing growth and yield of Thompson canopy length. berg and Raison (1982) concluded that the Seedless vines by trellising. Calif. Agr. 30(5):14- The oBrix for all three trellises (Table 2) tendency for ‘Shiraz’ vines with close rows 15. was somewhat low compared to a typical to have lower oBrix than wider-spaced and Kliewer, W.M. and L.A. Lider. 1968. Influence oBrix » 21 for ‘Petite Sirah’ in Tucson trellised vines is related to the shaded hori- of cluster exposure to the sun on the composi- (Mielke et al., 1980), which may have been zontal canopies that developed. Kliewer and tion of Thompson Seedless fruit. Amer. J. Enol. due in part to the early harvest. The oBrix Lider (1968) found that increased fruit clus- Viticult. 19:174-181. was significantly higher for the inclined can- ter exposure to sunlight reduced total acidity May, P., P.R. Clingeleffer, and P.B. Scholfield. 1976. The response of the grape cultivar opy (by 14%) than that for the standard or and malate concentration, and pH. The higher o Crouchen ( syn. Clare ) to var- vertical (by 9%) canopies. The Brix for pH in the musts produced from the fruits of ious trellis and pruning treatments. Austral. J. grapes from the vertical and standard cano- the standard two-wire trellis in our study is Agr. Res. 27:845-856. pies was similar. The pH values were con- consistent with data obtained by Smart et al. Mielke, E.A., G.R. Dutt, S.K. Hughes, W.H. sistent with pH » 4 for ‘Petite Sirah’ in (1982, 1985). May et al. (1976) found that Wolfe, G.J. Loeffler, R. Gomez, M.D. Bryant, Tucson (Mielke et al., 1980). The mean pH widening the trellis did not significantly af- J. Watson, and S.H. Schick. 1980. Grape and for the vertical canopy was significantly lower fect oBrix of the juice, pH, and acidity of wine production in the four comers region: Univ. than that for the standard canopy, but there the wine. However, they found that acidity of Arizona, Agr. Expt. Sta., Tech. Bul. 239. was no significant difference for berries from tended to decrease and sugar concentration Patterson, W.K. and B.W. Zoecklein. 1990. Veg- inclined and standard canopies. The titrata- tended to increase with increasing yield. etative and berry chemistry response to canopy manipulation and ethephon in ‘Norton’ grapes. ble acidity levels were somewhat higher than The PAR absorbed by foliage and avail- HortScience 25:905-908. expected (» 0.48 g/100 ml; Mielke et al., able for absorption by fruit differed among Reynolds, A.G., R.M. Pool, and L.R. Mattick. 1980), but they were not affected by the dif- the three trellising treatments. Highest PAR 1986. Influence of cluster exposure on fruit ferent positioning of shoots on the trellis sys- absorbance by the standard canopy did not, composition and wine quality of Seyval blanc tems (range 0.842-0.950 g-100 ml-l). however, significantly affect yield. Simi- grapes. Vitis 25:85-95. However, the oBrix : acid ratio was signifi- larly, the inclined canopy treatment in- Shaulis, N.J., H. Amberg, and D. Crowe. 1966. cantly higher for the inclined than for the creased the sugar content of fruits at harvest Response of Concord grapes to light, exposure standard canopy. and appeared to improve the wine quality and Geneva Double Curtain training. Proc. Amer. The relatively low oBrix for berries in each characteristics relative to the other two trel- Sot. Hort. Sci. 89:268-280. treatment resulted in an alcohol content be- lises. Differences in must characteristics can Shaulis, N.J. and P. May. 1971. Response of ‘Sultana’ vines to training on a divided canopy low the expected level of » 10% (Mielke et be achieved through modification of the fruit and to shoot crowding. Amer. J. Enol. Viticult. al., 1980) (Table 3). The wine from berries microclimate by canopy management, but 22:215-222. of the inclined canopy had a higher alcohol long-term studies are needed to determine Smart, R.E. 1988. Shoot spacing and canopy light content than that from either the standard or the magnitudes of these differences and their microclimate. Amer. J. Enol. Viticult. 39:325- the vertical canopies. The color of wine from effect on the quality of the wine. 333. the inclined and vertical canopy treatments Smart, R.E., J.B. Robinson, G.R. Due, and C.J. was excellent and consistent with the results Literature Cited Brien. 1985. Canopy microclimate modifica- of Mielke et al. (1980) (absorbances of » 0.19 Amerine, M.A. and C.S. Ough. 1974. Wine and tion for the cultivar Shiraz. II. Effects on must and 0.22 at 420 and 520 nm, respectively). must analysis. Wiley, New York. and wine composition. Vitis 24:119-128. However, the wine from the standard canopy Baldwin, J.G., G.E. Bleasdale, R.S., Cadman, and Smart, R.E., N.J. Shaulis, and E.R. Lemon. 1982. The effect of Concord vineyard microclimate on treatment had 37% and 27% lower absor- I.L. Keens. 1979. Comparison of trellises and pruning levels for Sultana vines in the Murray yield. II. The interrelations between microcli- bances at 420 and 520 nm, respectively, than Valley. Austral. J. Expt. Agr. Animal Hus- mate and yield expression. Amer. J. Enol. Vi- Mielke et al. (1980) reported. Titratable bandry 19:634-640. ticult. 33:109-116. acidity of wine from the standard canopy was Buttrose, M.S. 1970. Fruitfulness in grapevines: Steinhauer, R.E. and K.W. Bowers. 1979. Influ- » 25% above the typical level of » 0.5 g·100 The response of different cultivars to light, tem- ence of trellis on yield and quality of Cabernet ml-1. The pH values of wine from the three perature and daylength. Vitis 9:121-125. Sauvignon in Napa Valley. Amer. J. Enol. Vi- canopies were » 10% higher than the results Couvillon, G.A. and T.O.M. Nakayama. 1970. ticult. 30:236-240. of Mielke et al. (1980). The wine chemistry The effect of the modified Munson training sys- Weaver, R.J. and A.N. Kasimatis. 1975. Effect results were consistent with those for the must tem on uneven ripening, soluble solids, and yield of trellis height with and without crossarms on (Table 2). of ‘Concord’ grapes. J. Amer. Sot. Hort. Sci. yield of Thompson Seedless grapes. J. Amer. o 95:158-162. Sot. Hart. Sci. 100:252-253. The values for Brix and titratable acidity Ende, Van den B. 1984. The Tatura trellis-A Weaver, R.J., A.N. Kasimatis, J.O. Johnson, and are consistent with results of Kasimatis et al. system of growing grapevines for early and high N. Vilas. 1984. Effect of trellis height and (1976), whose four-wire trellis produced fruit production. Amer. J. Enol. Viticult. 35:82-87. crossarm width and angle on yield of Thompson o with a higher Brix than did the single-wire Gallo, K.P. and C.S.T. Daughtry. 1986. Tech- Seedless grapes. Amer. J. Enol. Viticult. 35:94- trellis. Steinhauer and Bowers (1979), how- niques for measuring intercepted and absorbed 96.