canopy management to improve grape yield and …canopy management to improve grape yield and wine...

Canopy Management to Improve Grape Yield and Wine QualityPrinciples and Practices*

R.E. Smart1, Joy K. Dick2, Isabella M. Gravett2 and B.M. Fisher3

1) Cassiro Pty. (Ltd.), RMB 67, Pacific Highway, Wauchope, New South Wales 2446, Australia2) MAF Ruakura Agricultural Centre, Private Bag, Hamilton, New Zealand3) DSIR Viticultural Research Station, Te Kauwhata, New Zealand

Submitted for publication: December 1989Accepted for publication: April 1990Key Words: Canopy management, solar radiation, shading, yield, wine quality

This paper reviews the subject of canopy management with an attempt to develop principles. These principles provideguidelines for canopy surface area amount; spacing between canopies; within canopy shade, especially for the fruit-ing/renewal zone; balance between fruit and shoot growth; and uniformity of location of fruit/renewal zones, shoot tipsand cane bases. Field techniques of point quadrat analysis and canopy scoring are introduced as an aid to defining problemcanopies. These techniques are cheap, quick and effective. A set of twenty-one numeric indices and descriptors to assesswinegrape canopies is then presented as a winegrape canopy ideotype, which can be further used as managementguidelines. Recent publications are reviewed from various aspects of canopy management. These include vigour control,shoot trimming, leaf removal in the fruit zone and training system responses. The paper concludes with presentation ofthe authors' unpublished data on the effects of canopy microclimate on yield and wine quality. The trial was conductedwith the cultivar Cabernet franc on a deep, fertile soil in a cool, high rainfall region. Canopy division using the RuakuraTwin Two Tier doubled yield compared to dense, vertical shoot positioned canopies which are common in New Zealand.Shade caused reduction in all yield components, and also delayed fruit ripening and reduced wine quality. Similar resultswere obtained by comparing fruit production at different heights with the Te Kauwhata Three Tier trellis system, wherelower tiers were shaded at the canopy exterior. The results confirm that grape yield and wine quality can be simultaneouslyincreased by improved canopy management of shaded vineyards.

Canopy management is now an active area of research, espe-cially in the New World winegrowing countries. It is general-ly accepted that Dr Nelson Shaulis of New York Statepioneered canopy studies, especially with publication of theGeneva Double Curtain trellis (Shaulis et al., 1966). How-ever, many of the principles that have emerged from recentresearch on canopy management are consistent with empiri-cal observations and practices long since employed in OldWorld vineyards.

The adoption of technological advances, especially postWorld War Two has created a common situation of excessivevineyard vigour. This is particularly the case in the NewWorld, where the benefits of improved soil preparation, ir-rigation, nutrition, fertilization and pest, disease and weedcontrol practices have resulted in increased vigour. Further,there are many New World situations where vineyards havebeen planted on deep and fertile soils. The result of increasedvine vigour is typically increased within-canopy shade.Recent research into canopy management has provided tech-

niques to avoid shade.This paper aims to condense existing knowledge by

developing principles of canopy management. These prin-ciples are presented along with a series of quantitative indiceswhich permit diagnosis of problem canopies. Field techni-ques to assess canopies are also presented in detail. The paperconcludes with presentation of recent experimental evidencewhich further supports the principles proposed. This papercan be viewed as an extension of recent reviews on the samesubject (Smart, 1985a; Smart 1987c).DEFINITIONS

A Canopy, is defined as the leaf and shoot system of thevine (Shaulis & Smart, 1974). It is described by dimensionsof the boundaries in space (i.e. width, height, length etc), andalso by the amount of shoot system within this volume (typi-cally leaf area). Canopies are continuous where the foliagefrom adjacent vines down the row intermingle, and wherethere are no large gaps. If canopies are separated from vine tovine they are discontinuous. Where canopies of one vine (or

* Presented at the 13th congress of the SASEV, Cape Town, November 2,1989.Acknowledgements: The authors acknowledge assistance of field staff, especially Dave Thomsen and Maurice Roper atRukuhiaHorticultural Research Station for maintaining field experiments, and staff at Te Kauwhata Research Station for berry, must andwine analysis. Sensory data were generated by G. Kelly, B. Campbell, J. Babich, B. Morris, J. Healy, J. Hancock, M. Compton.Kay McMath helped arrange the tasting. David Jordan made helpful comments on the manuscript, and Glennis Steiner typed it.South African colleagues E. Archer, J. Steenkamp, J.J. Swanepoel, JJ. Hunter and J.M. Southey also made helpful suggestionson the manuscript.

S. Afr. J. Enol. Vitic., Vol. 11, No. 1,1990

3

Canopy Management to Improve Yield and Quality

adjacent vines) are divided into discrete foliage walls thecanopy is termed divided. Canopies are crowded or densewhere there is much leaf area within the volume bounded bycanopy surfaces - for example a high value of the ratio leafarea: canopy surface area (LA/SA, Smart, 1985a), or of leaflayer number (LLN, Smart, 1985a) or shoot density(shoots/m canopy, Smart, 1988).

The term canopy management includes a range of tech-niques which can be applied to a vineyard to alter the positionor amount of leaves, shoots and fruit in space, and so as toachieve some desired arrangement (i.e. canopy micro-climate). These techniques include winter and summer prun-ing, shoot positioning, leaf removal, shoot vigour control anduse of improved training system. Canopy management tech-niques can be used to improve production and/or winequality, reduce disease incidence, and facilitate mechanisa-tion. Open canopies also lead to more efficient distribution ofapplied agricultural chemicals (Travis, 1987).

CANOPY MICROCLIMATE

Grapevine leaves as for other plants are strong absorbersof solar radiation, especially in the waveband 400-700 nm ofphotosynthetically active radiation (Smart, 1987a). Sinceonly about 6% of light in this waveband is transmitted by aleaf (Smart, 1987b), light levels at the centre of densecanopies are very low, less often than 1% of above canopyvalues (Smart, 1985a).

With high yields, fruit may also cause shade. For examplePalmer & Jackson (1977) have quantified this effect for applecanopies. Similar results for grapevine canopies will bepresented here. Light encountered in shade conditions hasaltered quality as well as quantity (Smart, 1987a; Smart,1987b) with important physiological implications for leavesand fruit in shade conditions (Smart, 1987a). Studies to bedetailed later have indicated that shade causes reductions inyield and quality. Therefore an aim of canopy management isto produce a canopy with minimal shade.

When the canopy microclimate is altered by canopymanagement techniques, it is not only sunlight levels thatchange. Temperature, humidity, wind speed and evaporationare also modified (Smart, 1985a). Altered evaporation ratesare of significance to fungal disease incidence, (English etal.,1989), and exposure can alter thermal relations of grapeberries with implications for their composition (Kliewer &Lider, 1968; Smart & Sinclair, 1976; Crippen & Morrison1986a, 1986b). However, the most significant feature ofaltered microclimate is for light quantity and quality levelswithin the canopy, from both a yield and fruit compositionviewpoint (Champagnol, 1984; Smart, 1985a; Smart,1987a).

Shade is avoided by reducing leaf area and increasing theproportion of canopy gaps (Smart, 1988). However, too higha proportion of canopy gaps allows sunlight to be 'lost',falling on the vineyard floor. What is required is a balancebetween excessive gaps (and hence waste of sunlight energy)and insufficient gaps (and the creation of shade).

PRINCIPLES OF CANOPY MANAGEMENTListing of principles: Five principles will be briefly

stated and justified by literature reference. Recent results arepresented at the end of the paper to further support the firstfour of these principles.

Principle 1. A large canopy surface area well exposed tosunlight is desirable, and this surface area should develop asquickly as possible in spring: Sunlight interception is in-creased (Smart, 1973), and canopy density reduced (Shaulis& Smart, 1974; Smart, 1985a) with large exposed canopysurface area. Biomass production and yield potential is in-creased as more solar energy is trapped by foliage. Rapiddevelopment of canopy surface area in spring promotes solarenergy interception. Clingeleffer (1989) argues that rapidcanopy development is part of the reason for increased yieldof minimal pruned vines. Tall, thin, relatively closely spaced,vertical, north-south foliage walls provide maximum ex-posure for a canopy surface (Smart, 1973; Jackson & Palmer,1972). Note however that very high values of canopy surfacearea are not possible without violating principle 2. Overheadcanopies, and wide row, low vigour vineyards have lowvalues.

Principle 2. Canopies should not be so close together asto cause excessive shade at the base of adjacent canopies.Vertical canopies are preferred and the ratio of canopyheight to alley width should not exceed about 1:1: Wherecanopies are close together penetration of both direct anddiffused sunlight into the enclosure formed by canopy wallsand the ground surface is impaired, as is demonstrated bycalculation (Jackson & Palmer, 1972; Smart, 1973) and byfield measurement (Smart, 1985b; Intrieri, 1987). Photosyn-thesis is inhibited for exterior leaves at the base of closely-spaced canopies. Figure 1 shows calculated values of sunlightflux densities over summer down the side of vertical canopy

50-

40 -

30-

20-

10-

• Direct sunlight 51° latitude» 45° -

34"• Standard overcast sky

0 1.0 2.0 3.0Ratio height: alley width

FIGURE 1

Solar radiation flux density estimated at various positionsdown the vertical walls of canopies, expressed as a percent-age of that received by a horizontal surface above the canopy.Position on the wall indicated by the ratio of canopy height:alley width. Data are for seven summer months, direct sun-light on north south walls, 34°, 45° and 51° latitude, andstandard overcast sky (data from Jackson & Palmer, 1972).



walls for direct sunlight (north-south rows) at 34°, 45 ° and 51 °latitude and for a standard overcast sky (from Jackson &Palmer, 1972). Note the similarity between all curves, andthat values corresponding to height: alley width ratio of 1:1are about 15% of above canopy values.

Principle 3. Canopy shade should be avoided, especiallyfor the cluster/renewal zone. Leaves and fruit should have asuniform a microclimate as possible: Adequate fruit exposureto sunlight promotes wine quality, though effects of leaf andfruit exposure are not often separated experimentally. Shadedcanopies produce fruit of increased K, pH, malic acid andBotrytis bunch rot incidence, and reduced sugar, tartaric acid,phenol and anthocyanin levels (reviewed by Smart, 1985a;Shaulis, 1982; Smart et al., 1988; Kliewer & Smart, 1988;Kliewer et al., 1988). Morrison (1988) has shown that leafshading affects berry size, sugar, K and pH while clustershading lowers fruit anthocyanin levels and phenols. Shadingreduces fruit monoterpenes (Reynolds & Wardle, 1988,1989a, 1989c) and induces herbaceous characters in wine(Pszczolkowski et al, 1985). Also, Carbonneau et al. (1978)note that excessive fruit exposure may increase phenol con-centrations beyond desirable levels. It is likely that highquality wine results from processing fruit of relativelyhomogeneous composition. Preliminary studies indicate thatshaded fruit shows more variation in composition (S. Smith,unpublished data). Since fruit composition responses tomicroclimate are already evident by veraison (Smart, 1982),it is likely that pre-veraison canopy microclimate has a sig-nificant impact on wine quality. Botrytis incidence is in-creased by high canopy density as fruit exposure is reduced(Gubler et al., 1987; Savage & Sail, 1984). Similar effects ofshading are noted for Oidium incidence (Pearson & Goheen,1988). Improved spray penetration also assists disease con-trol (Travis, 1987).

The light microclimate of the renewal zone (base of theshoot which is retained at winter pruning) is important foryield expression (Shaulis 1982; Shaulis & Smart, 1974).Located in this zone are the nodes retained at winter pruning,as well as grape clusters. Shading of this zone causes reducedcluster initiation, bud break, fruit set and berry size (Shaulis& Smart, 1974; Shaulis, 1982; Smart et al., 1982b). Basalleaves are also known to be important for fruit ripeningprocesses (Hunter & Visser, 1988). The effect of shade onfruit set is little studied but poor fruit set in the centre of densecanopies is commonly observed. Recently Jackson &Coombe (1988) suggested this could be due to a physiologi-cal disorder Early Bunch Stem Necrosis (EBSN). Shadeleaves contribute little to canopy photosynthesis (Smart,1974; Smart, 1985b) and in time turn yellow and abscise.

Principle 4. Photosynthate partitioning between shootand fruit growth should be appropriate to avoid either excessor deficient leaf area relative to the weight of fruit. Thenumber of active vegetative growing points per shoot shouldbe limited: Grapevine shoots are potentially indeterminate,and excessively large shoots are a common feature ofvigorous vineyards. Vigorous shoots have relatively largediameter, long internodes and large leaves, and there is amarked propensity for active lateral growth; such features are

undesirable and indicate an imbalance between vegetativeand fruit growth (Smart et al., 1989). Vineyards of high yielddemonstrate a balance between shoot and fruit production(e.g. Lavee & Haskal, 1982), an indication that photosynthateis partitioned appropriately between shoot and fruit produc-tion. Shoot devigoration can be achieved by a combination oflight pruning and improved canopy microclimate, the so-called 'big vine' effect (Smart et al., 1989). Long shootsrepresent photosynthate diversion into superfluous leaf area,which in turn generally contributes to canopy shading. Longshoots cause high must and wine pH (Smart, 1982). Partialdefoliation improves photosynthetic efficiency of remainingleaves (Hunter & Visser, 1988). Shoots that are too short mayhave insufficient leaf area to adequately ripen fruit (i.e. Peter-son & Smart, 1975; Koblet, 1987a). Shoot growth may beregulated by summer trimming (Reynolds & Wardle, 1989b),though for vigorous shoots the lateral buds near the shootapex burst and continue extension growth (Solari et al.,1988). The balance between shoot and fruit production canbe assessed by the ratio yield: pruning mass, sometimestermed 'crop load'. Bravdo et al. (1985) found in Israel thatvalues of this ratio greater than 10 with Cabernet Sauvignonled to reduced quality, with no quality effect for values lessthan 10. Studies by Reynolds (1989) with Riesling in coolerBritish Columbia found values less than 10 desirable.

There have been few studies of shoot length heterogeneityand its effect, though it is likely that uniform populations aredesirable (Smart et al., 1989).

Principle 5. Arranging locations of individual organs inrestricted zones in space facilitates mechanisation i.e. ofshoot tips for summer pruning, of cane bases for winterpruning and of fruit for mechanical harvesting. Trainingsystem design should as much as possible create fruit-ing/renewal zones at a similar height for any one vine: Forvines trained with renewal zones at various heights, thehighest buds burst preferentially before those of lower vineparts. This effect is presumably gravimorphic though in manysituations low renewal zones are also shaded, which effectreduces bud break. Thus Van den Ende (1984) found most ofthe fruit production of Sultana vines on the Tatura trellis wasat the top of the canopy after a few years cropping, thoughoriginally vines had replacement zones at six heights.DIAGNOSIS OF PROBLEM CANOPIES

This section comprises two parts. The first presents indetail field methods for evaluating winegrape canopies. In thesecond, quantitaive indices are presented as guidelines tocanopy assessment.

The techniques we have developed for diagnosing prob-lem canopies are designed for practical field use by re-searchers and growers. The techniques are easy to learn andinterpret, quick to use and also inexpensive.

Point quadrat: This technique was first applied tovineyard canopy studies in 1980 (Smart, 1982) though thepresentation of results is now modified. This simple methoddescribes the distribution of leaves and fruit in space, andprovides quantitative canopy description. A sharpened thinmetal rod is passed at fixed inclination into the canopy (nor-mally in the fruit zone) and contacts with leaves, clusters and



TABLE 1Sample point quadrat analysis sheet for two contrasting canopies (ex Smart & Sharp, 1989).Key: L = Leaf contact C = Cluster contact G = Canopy gap

Low density canopyCabernet franc, TR2T

1 L2 LLC3 LLC4 LC5 L6 L7 LCL8 G9 G10 LLC11 G12 L

13 LC14 CL15 L16 L17 CL18 LL19 L20 L21 L22 L23 L24 G25 G

26 LCC27 LCL28 G29 G30 L31 C32 LL33 G34 G35 LCL36 LLL37 LL38 L

39 C40 G41 C42 CLCC43 G44 CC45 C46 LC47 G48 G49 L50 LL

High density canopyGewiirztraminer, standard

1 LLCCL2 LLL3 LL4 LL5 LLCL6 LLCLLL7 LLCL8 CLLL9 LCLL10 LL11 LL12 LLL

13 LLL14 LCLCL15 LLLL16 LLL17 LLL18 LL19 LCL20 LLLL21 LCLLLL22 LLL23 LLL24 CLL25 LCC

26 LLLLL27 LLLLLL28 LLLLLCLL29 LLLLL30 LLLL31 LL32 LLCCLL33 LCLL34 LLL35 LCCLL36 LL37 LLLLLLC38 LCLL

39 LCCCCLL40 CLL41 LC42 CCLLL43 LLL44 LLCLL45 LLCLL46 LLCCL47 LLCCL48 LLLCL49 LLLCLLLL50 LLCLCL

Percent gaps:LLN:Percent interior leaves:Percent interior clusters:

13/50 = 26%43/50 = 0,864/43 = 9%6/23 = 26%

0/50 = 0%167/50 = 3,3472/167 = 43%36/39 = 92%

canopy gaps noted. Stems are normally ignored. Each inser-tion takes about 10 seconds. Typically 50-100 passes aremade for each canopy to be analysed. The following canopydescriptors can be generated from the data - percent gaps,LLN (leaf layer number), percent interior leaves and percentinterior clusters (Smart & Smith, 1988; Smart & Sharp,1989). Table 1 shows typical data for high and low densitycanopies, and demonstrates the method to calculate canopydensity indices.

Vineyard scoring: The concept of visual canopy assess-ment to predict winegrape quality was first described bySmart et al., (1985a). Canopy assessment takes place be-tween veraison and harvest using eight characters as present-ly proposed (Smart & Smith, 1988; Smart & Sharp, 1989).Three of these characters describe microclimate (canopygaps, canopy density and fruit exposure) and five describeprior growth or physiological status (Table 2). Each characteris assessed out of 10 points, giving a total of 80. Note that thescorecard should not be used for diseased, unhealthy or ex-cessively stressed vines.

The scorecard as presented is based on observation andmeasurement of high quality vineyards in Australia, NewZealand, USA, France and Germany. In particular, detailedstudies of Cabernet Sauvignon vineyards in the Gravesregion of France were most instructive (Smart, Carbonneau& de Loth, unpublished). Development of the scorecard wassupported by extensive vineyard microclimate measurements(Smart, 1982; Smart etal., 1982a; Smarter al., 1985a; Smartet al., 1985b; Smart 1987b; Smart 1988; Smart & Smith,1988). For all of this, some arbitrariness is recognised in thescorecard construction. For example it is difficult to justifyonly eight characters, and that they should all have equalweighting. The scorecard is suggested as first approxima-tion only, and modification to accommodate different cul-

tivars in different regions is encouraged. For Mediterraneanclimates, the addition of a descriptor for vine water status isconsidered desirable.

Scoring a canopy takes less than two minutes. Results arepresented here to illustrate that little judge training is re-quired. In February 1987 we compared the scores from fourjudges. Judge A had five years experience with the scorecard,judge B had two years experience, and judges C and D wereusing the scorecard for the first time. Thirty experimentalplots of Cabernet franc were assessed, comprising tworootstocks by five training systems with three replicates. Thefour judges initially discussed scores for several canopies,then worked separately. Results were analysed to investigate"judge" effects for each character scored.

Analysis of variance showed that trellis treatment was thevariable with most significance (8 out of 9), followed byjudge effects (7 out of 9), and no rootstock effect. Interactionof judge (with rootstock or trellis) was less frequent, occur-ring 6 times out of a possible 27. Treatment effects were ofless significance or no significance where the range in scoreswas small i.e. for the characters shoot length and presence ofgrowing tips. Similarly, these same characters correlatedpoorly when other judges scores were correlated with judgeA, and the character fruit exposure had the highest correla-tion coefficient. There was little apparent effect of prior judgeexperience. The limited frequency of interactions with judgessupports the contention of using several judges and averagingtheir scores.

Vine growth and yield measurements: Simple measure-ments at winter pruning of cane number and retained nodenumber and total cane mass, and of yield and bunch numberat harvest allows useful indices of vine balance to be deter-mined. There are:



TABLE 2Vineyard scorecard for assessing vineyard potential to produce quality winegrapes.

NOTE: If majority of shoots are less than 30 cm long, or if these vines are clearly diseased or chlorotic or necrotic, or excessivelystressed, DO NOT SCORE VINEYARD.

A. Standing away from canopy1 CANOPY GAPS (from side to side of canopy, within

area contained by 90% of canopy boundary)*about40% 10*about 50% or more 8*about 30% 6*about 20% 4*about 10% or less 0

2 LEAF SIZE (basal-mid leaves on shoot, exterior).For this variety are the leaves relatively:*slightly small 10*average 8*slightly large 6*very large 2*very small 2

3 LEAF COLOUR (basal leaves in fruit zone)* leaves green, healthy, slightly dull and pale 10*leaves dark green, shiny, healthy 6*leaves yellowish green, healthy 6*leaves with mild nutrient deficiency symptoms 6*unhealthy leaves, with marked necrosis or chlorosis 2

B. Standing at Canopy4 CANOPY DENSITY (from side to side in fruit zone),

mean leaf layer number*about 1 or less 10*aboutl,5 8*about 2 4*more than 2 2

FRUIT EXPOSURE (remember that the canopy has twosides normally - that fruit which is not exposed on yourside may be exposed to the other side)*about 60% or more exposed*about 50%*about 40%*about 30%*about 20% or lessSHOOT LENGTH*about 10-20 nodesabout 8-10 nodesabout 20-25 nodesless than about 8 nodesmore than about 25 nodesLATERAL GROWTH (normally from about pointwhere shoots trimmed. If laterals have been trimmed,look at diameter of stubs)limited or zero lateral growthmoderate vigour lateral growthvery vigorous growthGROWING TIPS (of all shoots, the proportion withactively growing tips - make due allowance fortrimming)about 5% or lessabout 10%about 20%about 30%about 40%about 50% or more

108642

106622

1062

10

6440

Total point score

TABLE 3Summary of anova results with four judges of differing experience.

Characterassessed

Canopy gapsLeaf sizeLeaf colourCanopy densityFruit exposureShoot lengthLateral growthGrowing tipsTotal

Rangeof values3

0,6-9,27,3-9,77,1-9,82,2-8,52,8-10,09,6-10,05,7-9,8

9,9-10,046-76

Rootstock

R

---------

Trellis

T

***b*************

***-

***

SignificanceJudge

J

*********--

****

***

RXJ

*

TXJ

********

*

RXTXJ

**

Correlation with Judge A°Judge

B C D

,66 ,78 ,79,43 ,67 ,67,33 ,60 ,67,64 ,71 ,62,94 ,83 ,97-,11 -,14 ,07,69 ,62 ,710 0 0

,91 ,86 ,931 highest and lowest mean score for trellis treatments, an indication of data spread.significance level

' correlation coefficients

* P<,05**P<,01*** P<,001r>0,35 P<,05r>0,45 P<,01r>0,56 P<,001



- average cane mass (total cane mass/cane number) withhigh values indicating excessive vigour,

- total cane mass (kg/m row or per m canopy) with highvalue indicating high canopy density,

- yield/cane mass ratio, with low values indicating im-balance due to excessive vigour (Bravdo et al, 1984),

- bud burst (shoots per node) and f rui t fulness(clusters/shoot) with low values indicating among otherthings effects of shaded canopies.Other measurements: Light microclimate may be

measured with electronic sensors, with due caution exercisedfor sampling problems. Similarly leaf area may be deter-mined and the vigour index y calculated (see Smart, 1985a)as well as the leaf area/crop mass ratio. However both ofthese techniques are considered more complex than is re-quired for most practical vineyard assessment situations(Smart & Sharp, 1989).PROPOSING A WINEGRAPE CANOPY IDEOTYPE

Following on the ideotype concept of Donald (1968) foran ideal wheat plant description, we now present a grapevinecanopy ideotype. The ideotype proposed is a series ofnumeric indices and characters which can be used to assesswinegrape canopies. The values presented are believed op-timal for winegrape yield and quality with currentknowledge, but may be found to require modification fordifferent cultivars or environments. These values should beespecially useful as guidelines for canopy management inhigh vigour situations. Some of the indices presented havenot been previously introduced and references from theliterature that support the values are included.

Of the indices listed in Table 4, we have found the follow-ing to be the most useful for vineyard diagnosis and trellisevaluation; surface area, shoot spacing, shoot length, lateraldevelopment, ratios of yield: canopy surface area and yield topruning mass, and LLN.YIELD AND QUALITY RESPONSES TO CANOPYMANAGEMENT

This section will briefly review published responses tocanopy management techniques, and the next Section willpresent data for trellis system effects on yield and quality. Itshould be emphasised that a range of canopy managementtechniques are available, and the suitability of various tech-niques depends on vineyard vigour, when considerations ofeconomics and practicability are not considered. As a generalprinciple, the desirability of adoption of a more complextraining system is increased as vineyard vigour increases. Forlow to moderate vigour vineyards, summer trimming (hedg-ing) or fruit zone leaf removal may be sufficient to improvemicroclimate.

Vigour control: Low shoot vigour helps create opencanopies. Examples of this are given by Smart (1985a) andSmart et al., (1989) who reviewed management techniques toachieve devigoration.

Shoot trimming: Unless shoot growth is inhibited, forexample, by water stress (Smart & Coombe, 1983) or by lightpruning (Clingeleffer, 1989; Smarter al, 1989) shoot growth

normally continues beyond optimum length. Summer trim-ming (hedging) to contain shoot growth is widely used inhigh humidity environments like Europe and New Zealand,but less frequently in drier climates like Australia andCalifornia. As long as trimming is done early, fruit ripeningis encouraged (Solari et al., 1988; Koblet, 1987a; Koblet,1988). Kliewer & Bledsoe (1987) have however found thattrimming delayed fruit maturity. Perhaps this result was dueto removing exterior, photo synthetically active leaf area andexposing less efficient leaves.

Leaf removal in the fruit zone: Fruit exposure in densecanopies can be enhanced by preveraison leaf removal in thecluster zone. Fruit composition is improved and herbaceouswine characters reduced (Kliewer and Bledsoe, 1987; Smithetal, 1988; Freese, 1988; Kliewer et al., 1988; Hand, 1988).Leaf removal also facilitates control of Botrytis bunch rot(Gutter era/., 1987; Smith etal., 1988).

Training system: The pioneering studies of Shaulis et al(1966) demonstrated that within canopy shade was a prin-ciple cause for yield and quality reductions, and that thiscould be overcome by canopy division. Subsequently, a largenumber of studies with a range of cultivars and environmentshave confirmed the same principles. Smart (1985a) reviewedearlier work and to this list can now be added Carbonneau(1985), Carbonneau & Casteran (1987), Casteran & Carbon-neau (1987), Intrieri (1987), Cullen (1988), Kliewer et al,(1988), and Smart & Smith (1988). More complex trainingsystems typically have more old wood per vine as trunks andcordons, and this also enhances yield (Murisier & Spring,1987; Koblet, 1987b). The ability of canopy division toimprove vine microclimate and performance is dependant onvineyard vigour. This was emphasised by Smart et al.(1985a), Smart et al (1985b), Carbonneau & Casteran(1987a), and Casteran & Carbonneau (1987).RECENT RESULTS FROM CANOPY MANAGEMENTEXPERIMENTS

The paper concludes by presenting recent experimentalresults from a trellis trail which support principles developedabove. Results presented will be limited to three of the fivetreatments, and to two seasons. The full results of the trial willbe the subject of a further report.

A trail with the cultivar Cabernet franc was established in1983 at Rukuhia, near Hamilton to investigate yield andquality responses to training systems. The soil is deep andfertile (Horotiu silt loam, 80 cm of silt loam over coarse sandand gravel extending for several meters). The climate can bedescribed using the Smart-Dry index (Smart & Dry, 1980) ascool (MJT = 17,3C), maritime (CTL = 9.2C), overcast (SSH= 6,5 hrs/day), not arid (deficit 60 mm), and very humid (RH= 73%). Such conditions are favourable to high vine vigourand restrict maturity for this late season cultivar. The trialcompares five training systems and two rootstocks (1202Couderc (1202C) and Aramon Rupestris Ganzin 1 (ARGI)).There are three replicates, each plot consisting of six vines.Generally, rootstocks have only small effects and these arenot discussed here. The training systems are Te KauwhataThree Tier (TK3T) and Te Kauwhata Two Tier (TK2T) atrow spacing 1,8m, and Tatura trellis, Ruakura Twin Two Tier



TABLE 4Winegrape canopy ideotype to promote high yields and improved fruit composition.

Character assessed Optimal Value Justification of optimal value

Canopy characters:Row orientation

Ratio canopy height: alley width

Foliage walls inclination

Renewal/Fruiting area location

Canopy surface area (SA)

Ratio leaf area/surface area (LA/SA)

Shoot spacing

Canopy width

Shoot and fruit characters:

Shoot length

Lateral development

Ratio leaf area: fruit mass

Ratio yield: canopy surface area

Ratio yield: total cane mass

Growing tip presence after veraison

Cane mass (g) (in winter)

Internode length

Total cane mass: m canopy length

Microclimate characters:

Proportion canopy gaps

Leaf layer number (LLN)

Proportion exterior fruit

Proportion exterior leaves

north-south

-1:1

vertical or nearly so

near canopy top

~21,000m2/ha

—15 shoots/m

300-400mm

10-15 nodes, about 600-900mm length

limited, say less than 5-10 lateral nodes totalper shoot

about 10 cm /g (range6-15cm2/g)

1-1,5 kg fruit/m can-opy surface

6-10

nil

20-40g

60-80 mm

0,3-0,6 kg/m

20-40%

1-1,5

50-100%

80-100%

Promotes radiation interception Smart (1973) though Champagnol (1984) argues thathourly interception should be integrated with other environmental conditions i.e. temper-atures which affect photosynthesis to evaluate optimal row orientation for a site. Windeffects can also be important (Weiss & Alien, 1976a; Weiss & Alien, 1976b).

High values lead to shading at canopy bases, and low values lead to inefficiency of radia-tion interception. (See literature cited for Principle 2; data in this paper.)

Underside of inclined canopies is shaded (Smart & Smith, 1988).

A well exposed renewal/fruiting area promotes yield and, generally, wine quality, al-though phenols may be increased above desirable levels. (See literature cited for princi-ple 3; data in this paper.)

Lower values generally indicate incomplete sunlight interception, higher values are asso-ciated with excessive cross row shading. (See literature cited for principle 1; data in thispaper.)

An indication of low canopy density especially useful for vertical canopy walls. (Smart1982; Smart et al, 1985a and literature cited for principle 3).

Lower values associated with incomplete sunlight interception, higher values withshade. Optimal value is for vertical shoot orientation and varies with vigour (Smart,1988).Canopies should be as thin as possible. Values quoted are minimum likely width. Actualvalue will depend on petiole and lamina lengths and orientation.

These values are normally attained by shoot trimming. Short shoots have insufficientleaf area to ripen fruit; long shoots contribute to canopy shade and cause elevated mustand wine pH (see literature cited for principle 4).

Excessive lateral growth is associated, with high vigour (Smart et.al., 1985a; Smart &Smith, 1988; Smart, 1988, Smart et al., 1989).

Smaller values cause inadequate ripening, higher values lead to increased pH (Shaulis &Smart, 1974; Peterson & Smart, 1975; Smart 1982; Koblet, 1987a). Value around 10 op-timal.

This is value of exposed canopy surface area required to ripen grapes (Shaulis & Smart,1974). Values of 2,0 kg/m2 have been found to be associated with ripening delays inNew Zealand, but higher values may be possible in warmer and more sunny climates.

Low values associated with low yields and excessive shoot vigour. Higher values associ-ated with ripening delays and quality reduction. (See literature cited for principle 4.)

Encourages fruit ripening since actively growing shoot tips are an important alternatesink to the cluster (Koblet, 1987a).

Indicates desirable vigour level. Leaf area is related to cane mass, with 50-100cm2 leafarea/g cane mass. Values will vary with variety, shoot length (Smart & Smith, 1988;Smart et al., 1989; also data in this paper).Indicates desirable vigour level (Smart et al, 1989). Values will vary with variety.

Lower values indicate canopy is too sparse, higher values indicate shading. Values willvary with variety, shoot length (Shaulis & Smart, 1974; Shaulis, 1982; Smart, 1988; datapresented this paper).

Higher values lead to sunlight loss, lower values can be associated with shading (Smart& Smith, 1988; Smart 1988).

Higher values associated with shading, lower values with incomplete sunlight intercep-tion (Smart, 1988 and literature cited for principle 3).

Interior fruit has composition defects (literature cited for principle 3).

Shaded leaves cause yield and fruit composition defects (literature cited for principle 3).


10 Canopy Management to Improve Yield and Quality

TK3T

5,555 m row/haSA = 26,944 m2/ha

Standard

2,778 m row/haSA = 10,556 mz/ha

kssl

RT2T

2m

2,778 m row/haSA = 21,668m2/ha

FIGURE 2Cross-sectional dimensions of three training systems drawnto scale, along with calculations of row length/ha and canopysurface area. TK3T is Te Kauwhata Three Tier; Standard isvertically shoot positioned, non divided canopy, and RT2T isRuakura Twin Two Tier.

(RT2T) and 'Standard' vertically shoot positioned (STD)with 3,6m row spacing. The RT2T system was designed withthe before listed principles 1 to 5 in mind. Within row spacingis 2,0m. The first harvest was 1985, and since the 1987vintage yields have been relatively stable as canopies havefilled their allotted space.

Results will be presented here for three treatments only -TK3T, RT2T and STD, (see Figure 2) with fruit compositionand wine quality assessments for the 1988 vintage and yieldcomponents for the 1989 vintage. In general, yields for the1989 vintage were higher than for 1988, due to higher clusternumbers for the top tier of RT2T and TK3T.

Experimental wine was made from each replicate using20-40 kg fruit lots fermented on skins to dryness and sterilefiltered to avoid confounding effects of malolactic fermenta-tion. Wine sensory assessment was carried out when thewines were 7 months old, using 7 experienced industryjudges. The judges were evaluated for consistency and dis-crimination ability. The scorecard used in the sensory evalua-tion used six characters (colour density and hue, fruit charac-ter on nose and palate, palate structure and overallacceptability), each assessed out of 7 points. Results from thiscard correlated well with the three character card assessing

colour (ex 3), nose (7), palate (10), and total (20) normallyused in New Zealand and Australia (Gravett & Smart, un-published data). Standard wine analyses were made and in-cluded spectral analysis by Somers & Evans (1977) method.

Figure 2 shows the dimensions of the three systems to bediscussed, along with the calculated canopy surface area. TheTK3T system has a SA greater than the optimum cited inTable 3, while the RT2T was at the optimum and the non-divided STD canopy with wide row spacing was about halfthe optimum value.

Effect of tier height on yield, growth and quality: TheTK3T trellis provides unique insight into effects of shade onyield, growth, fruit composition and wine quality. The TK3Twas included in this trial as a test for principle 2, in that theratio of canopy height to alley width is 1,8:1, in obviousviolation of the 1:1 guideline. Thus lower tiers are shaded atthe canopy exterior. It is possible to see the effect of varyingthe height to alley width ratio on vine performance by com-paring different tier heights. Individual vines with 6m cordonlength were trained to each height and pruned to 80 nodes(13,3 nodes/m) to minimise within-canopy shading (seeTable 4).

The top tier produced five times the yield and four timesthe pruning mass and three times cane mass, of the lowest tier(Table 5). The middle tier was intermediate. These differen-ces are of similar order to the calculated mean light fluxdensity at the midpoint of the fruiting zone, taken from Figure1. Also shown in Table 5 are the height: alley width ratioscalculated at the fruit zone midpoint. All the yield com-ponents bud break, fruitfulness (clusters/shoot), berry num-ber and berry mass were reduced for lower tiers, with the lasttwo being the most sensitive to shading.

Trends in fruit composition were similar to yield (Table 6)with top tier showing advanced maturity. Sugar accumulationwas delayed for the lower tiers with the effect evident alreadyat veraison and persisting till harvest. Maximum differencesfrom top to bottom were 2,9° Brix at veraison and 4,2° Brixat harvest. There were smaller differences with acidity andpH, with maximum differences of 1,1 ° Brix and 0,07 pH unitsat harvest. Berry mass differences were substantial betweentiers, a maximum of 0,68 g at harvest. As denoted by theproportion of red berries at veraison sampling, the top tierfruit commenced colouring earlier than both middle andlower tiers.

Fruit composition trends between tiers were furtherreflected in wine analyses (Table 7). Lower tiers had higherwine K and pH, and lower titratable acidity and tartaric acid.The difference from top to bottom tier respectively was 300mg/1 K, 0,15 pH units, 1,3 g/1 titratable acidity and 0,4 g/1tartaric acid. Wine colour density was higher for the top tieras was also the concentration of anthocyanins, ionised an-thocyanins and phenols. In turn these differences were ex-pressed in the sensory scores, with all components except'fruit on the nose' showing significant effect of tier position.Wines from the top .tier scored 4,9 ex 7 for overall accept-ability, 4,6 for the middle tier and 3,8 for the bottom tier.Lower and more shaded tiers produced wine with less colourand fruit character, and less full palate.


Canopy Management to Improve Yield and Quality 11

TABLE 5Effect of tier height of TK3T on yield and yield components and growth. Cabernet franc, Rukuhia, 1988-1989 season.

Character assessed

Cordon length/vineYield (kg/vine)Clusters/vineBerry mass (g)Nodes retained/vineShoots/vineTotal cane mass (kg vine)

Percent bud breakClusters/shootBunch mass (g)Berry numberYield (g)/node retainedYield (g)/shootYield/pruning ratioCane mass (g)

Nodes retained/m cordonShoots/m cordonClusters/m cordonYield (kg)/m cordonTotal cane mass (kg)/m cordonBerries/m cordon

Calculated light (relative)3

Height: alley ratio at fruit zone midpoint

Top

5,6821,81691,4080874,11091,94129922732495,4

47,514,115,429,83,850,732740

1000,40

Middle

5,728,21200,9980721,391

1,6768691021146,817,313,912,621,01,430,221440

570,88

Bottom

5,694,389

0,8879701,089

1,29485455634,614,113,812,315,70,760,1786031

1,43

LSD

-b3,89

0,11—3

0,34

0,12191147421,34,6_

0,61,8

0,700,05430NANA

a calculated from Fig. 1 for midpoint of fruit zone.not significant indicated by -

NA = not applicableNote: Data for ARG 1 and 1202 C rootstocks combined.

TABLE 6Effect of tier height of TK3T on fruit composition.Cabernet franc, Rukuhia, 1987-1988 season.

Character assessed

Veraison 3 March 19881

Sugar (°Brix)Titratable acidity (g/1)PH% red berriesBerry mass (g)Harvest 20 April 1988 'Sugar fBrix)Titratable acidity (g/1)PHBerry mass (g)Yield (kg/vine)%rot

Must compositionSugar (°Brix)Titratable acidity (g/1)PHMalic acid (g/1)

Top

12,124,92,7295

1,39

19,310,42,951,6713,25,2

19,510,03,024,5

Middle

9,524,02,6569

1,07

15,99,62,951,2210,40,2

17,09,72,983,9

Bottom

9,224,02,6738

0,86

15,19,33,020,997,30,4

16,79,53,043,9

LSD

0,8-a

0,0520

0,19

1,60,90,050,171,81,9

1,4--

0,4

not significantly different indicated by -Data for ARG1 and 1202C rootstocks combined.

: Data for 1202C rootstock only.



Effect of RT2T training: The trial allowed comparisonsbetween RT2T and STD trellis systems, but also comparisonswere made within the RT2T plots. For each RT2T plot of sixvines, there were two 'big' vines and four 'small' vinesarranged in a 2 x 2 factorial combination with tier position('up' and 'down'). Any one vine was trained to only oneheight. 'Big' vines had 12m of cordon and 160 nodesretained, while 'small' vines had 6 m of cordon and 80 nodesretained, both at 13,3 nodes/m cordon. Shoot growth wasdevigorated on big vines because of high node numberretained at winter pruning (Smart et al., 1989). Thus despitesimilar shoot spacing, the canopy of big vines was less densethan for small vines (Smart & Smith, 1988). Data for thesecomparisons are shown in Tables 8, 9 and 10, and whereheight by vine size interactions are significant these data areshown in Table 11.

Table 8 presents the effect of vine size and tier height onyield and growth and also compares RT2T plots includingboth big and small vines with STD. RT2T more than doublesyield of STD due to greater cordon (canopy) length, morenodes retained (by 28), higher bud break (by 24%), clustersper shoot (by 0,1), and berry number per bunch (by 21). Largedifferences in yield per retained node (by 90g) and per shoot(by 50g) between RT2T and STD result. These yield respon-ses are in accord with responses noted previously (Smart &Smith, 1988), and correspond to the STD vines having a

dense shaded canopy with the RT2T canopy being open. STDvines are 'unbalanced' with excessive vegetative growth asindicated by low yield/ pruning ratios and high cane mass.

The yield, yield component and growth differences be-tween the two tier heights for the RT2T are similar to thoserecorded for the top two tiers of TK3T and require no furtherdiscussion. There was less effect on yield components of'big' versus 'small' vines. Percent bud break of small vineswas slightly higher (by 17%), as is expected with fewerretained nodes per vine (Smart & Smith, 1988). 'Big' vinesshowed better balance between vegetative growth and fruitthan 'small' vines, as indicated by higher yield per shoot (by21 g), lower yield to pruning ratio (by 2,7), and lower canemass (by 7 g). There was little interaction between level andsize. Lower tier big vines had more clusters per shoot, andlower total cane mass per m of cordon (Table 11).

Fruit composition was similar for RT2T and STD vines atveraison (Table 9). Fruit from small vines had higher aciditythan big vines at veraison by 1,3 g/1. Interactions between sizeand tier position in the RT2T are explained by lower tiersmall vines having higher pH, and berry size differences wereless for the bottom tier than the top (Table 11). At harvestthere were significant differences in fruit composition. STDvines had more sugar (by 0,9° Brix), higher pH (by 0,13) andhigher bunch rot (by 18%). Top tier vines had higher sugar

TABLE 7Effect of tier height of TK3T on wine analysis and sensory scores. Cabernet franc, Rukuhia 1987-1988 season.

Character assessed

Wine analysis:% alcohol (v/v)Titratable acidity (g/1)pHTartaric acid (g/1)Potassium (mg/1)Malic acid (g/1)Acetic acid (g/1)Extract (g/1)Sugar free extract (g/1)Wine colour densityWine hueccb

a'Total anthocyanins (mg/1)Ionised anthocyanins (mg/1)Phenols (a.u.)

Sensory scores: (Ex 7 maximum)Colour densityColour hueFruit on noseFruit on palatePalate structureOverall acceptability

Top

10,88,63,222,55703,00,1820,218,97,70,5140,337,415060

27,7

5,85,94,95,35,24,9

Middle

9,77,93,232,96302,8

0,1520,018,46,5

0,4937,236,413549

21,7

5,25,65,05,04,74,6

Bottom

8,97,33,372,18703,2

0,1918,817,15,0

0,5632,030,910333

18,7

4,34,84,44,33,93,8

LSD

-a0,40,040,4170----

2,70,05--45142,9

0,50,50,50,40,60,6

a not significantly different indicated by -a proportion ionised anthocyanins

a' proportion ionised anthocyanins without SO2Note: Data for 1202C rootstock only.



TABLE 8Effect of tier height and vine size of RT2T compared with standard trellis on yield, yield components and growth. Cabernet franc,Rukuhia, 1988-1989 season.

Characterassessed

Cordon length/vine (m)Yield (kg/vine)Clusters/vineBerry mass (g)Nodes retained/vineShoots/vineTotal cane mass (kg/vine)Percent bud breakClusters/shootBunch mass (g)Berry numberYield (g)/node retainedYield (g)/shootYield/cane mass ratioCane mass (g)Nodes retained/m cordonShoots/m cordonClusters/m cordonYield (kg)/m cordonTotal cane mass (kg)/m cordonBerries/m cordon

Trellis

STD

1,918,91161,2479653,783

1,7977601131372,558

41,434,160,74,671,973666

RT2T

7,8221,52141,241071113,11071,899781

2031876,828

13,914,928,32,840,432259

LSD0,841,812-

NA5

0,34

0,0771318150,33

1,21,13,1

0,520,13406

Tier height

UPPER

7,6329,62541,371071203,91152,1111787

2802447,634

14,116,234,33,950,562879

LOWER8,013,41731,041071022,299

1,6778751261306,022

13,413,522,21,720,311638

LSD-

2,022

0,07NA850,25

0,1691018130,83-

0,82,7

0,250,03307

Vine size

BIG12,1031,53061,231601533,796

2,00100801972018,624

13,412,825,72,660,312102

SMALL

5,6816,51671,1780902,81131,849681

2061805,931

14,215,929,62,920,492415

LSD0,692,123

0,07NA5

0,35----150,93-

0,92,8-

0,03-

Interaction(height + size)

-a0,001--

NA0,001--

,04----------

0,004-

a not significantly different indicated by -NA = not applicableNote: Data for ARG1 and 1202C rootstocks combined.

TABLE 9Effect of tier height and vine size on fruit composition of RT2T compared with standard. Cabernet franc, Rukuhia, 1987-1988season.

Characterassessed

Veraison 3 March 19881

Sugar (Brix)Titratable acidity (g/1)pH% red berriesBerry mass (g)Harvest 20 April 19881

Sugar (Brix)Titratable acidity (g/1)pHBerry mass (g)Yield/vine (kg)%rot

,,Must compositionSugar (Brix)Titratable acidity (g/1)pHMalic acid (g/1)

TrellisSTD

10,623,62,5288

1,17

18,59,63,081,2117,720

19,59,82,985,0

RT2T

10,423,62,7882

1,24

17,69,5

2,951,439,52

18,69,53,214,5,

LSD

-----

1,2-

0,060,081,59

--

0,140,5

Tier height

UPPER

10,922,52,1996

1,39

18,79,52,951,6019,14

19,19,52,994,6

LOWER

9,322,92,6766

1,08

16,59,52,961,2616,20

16,59,42,974,1

LSD

0,8-

0,04-

0,14

1,9--

0,101,71

1,4--

0,4

Vine size

BIG

10,222,02,4484

1,27

17,39,52,941,3825,6

2

17,29,5

2,954,3

SMALL

9,923,32,4278

1,19

17,99,42,971,4813,7

2

18,39,43,014,5

LSD

-1,1---

----1,8-

--

,03-


-a-

0,03-

0,03

----

0,02-

-0,03-

0,04

1 not significantly different indicated by -( Data for ARG1 and 1202C rootstocks combined.' Data for STD vs RT2T for ARG1 rootstock only, remainder for 1202C



TABLE 10Effect of tier height and vine size of RT2T compared with standard on wine analysis and sensory scores. Cabernet franc, Rukuhia,1987-1988 season.

Characterassessed

Wine analysis:% alcohol (V/V)Titratable acidity (g/1)pHTartaric acid (g/1)K(mgMalic acid (g/1)Acetic acid (g/1)Extract (g/1)Sugar free extract (g/1)Wine colour densityWine hueccb

a'Total anthocyanins (mg/1)Ionised anthocyanins (mg/1)Phenols (a.u.)Sensory scores:Colour densityColour hueFruit on noseFruit on palatePalate structureOverall acceptability

TrellisSTD

10,77,73,401,79203,3

0,2323,822,13,9

0,7717251612822

3,83,63,73,93,83,5

RT2T

10,28,43,192,97202,70,1820,319,07,00,4937371656024

4,65,95,15,55,25,1

LSD

--

0,130,71300,3---

2,7-115-29-

0,60,70,50,60,40,6

Tier heightUPPER

10,88,5

3,182,76803,1

0,1720,819,37,0

0,5338381585925

5,55,85,25,45,25,0

LOWER

8,98,1

3,202,86603,0

0,1718,817,45,5

0,5240391054117

5,25,55,25,14,84,6

LSD

0,90,3-----1,61,51,0_--1252

-0,2-

0,20,20,3

Vine sizeBIG

9,58,43,133,05702,9

0,1818,717,36,10,4942401255021

5,15,75,05,04,74,5

SMAAL

10,28,2

3,252,47703,2

0,1720,919,46,5

0,5636361385022

5,65,65,45,65,25,0

LSD

--

0,050,3100--1,61,5-

0,02--12--

--

0,10,20,20,3


a

0,03--------

0,003-

0,040,05--

-0,010,010,04--

a a proportion ionised anthocyanins.a' proportion ionised anthocyanins after removing SO2 effect.b not significantly different indicated by -Note: Data for standard vs RT2T for ARG1 rootstock, remainder for 1202C rootstock.

TABLE 11Details of significant tier hight X vine size interactions, RT2T. Cabernet franc, Rukuhia, 1987-1988 and 1988-1989 seasons.

Character assessedFrom Table 8:Yield (kg/vine) 1989Shoots/vineClusters/shootTotal cane mass (kg/m cordon)From Table 9:pH, veraisonBerry mass (g), veraisonYield/vine (kg) 1988Must titratable acidity (g/1)Must malic acid (g/1)From Table 10:Wine tritratable acidity (g/1)Wine huea'Total anthocyanins (mg/1)Sensory score - hueSensory score - fruit on noseSensory score - fruit on palate

Upper - big

43,01682,100,40

2,231,5128,79,34,3

8,40,4937,41575,95,25,3

Upper - small

22,996

2,120,64

2,161,2614,39,74,9

8,60,5838,21585,55,35,7

Lower big

20,01371,890,22

2,651,0322,69,74,2

8,40,5043,0925,44,64,6

Lower - small

10,184

1,560,35

2,681,1213,09,04,0

7,80,5334,81185,65,45,4

LSD

3,07,20,240,05

0,070,202,60,60,6

0,50,025,818

0,40,50,4



(by 1,2° Brix) and berry weight (by 0,34 g), and there was noeffect of vine size. Interactions are due to lower tier largevines having higher titratable acidity, while there was littleeffect of vine size for the bottom tier in malic acid comparedto the top tier (Table 11).

The fruit composition effects carried through to the wine,with STD wines having higher pH (by 0,21), K (by 200 m//)and malic acid (by 0,6 g//) and lower tartaric acid (by 1,2 g//),see Table 10. Wines made from fruit from the top tier hadmore acidity and extract, while wines from big vine fruit hadlower pH, K and extract and higher tartaric acid than for winefrom small vines. Interactions were caused by low values oftitratable acidity for lower tier small vines (Table 11). Winespectral analysis showed RT2T wines had more colour den-sity (by 3,1 units), and ionised anthocyanins (by 32 mg//) thanwines from the STD. Similar patterns were evident whenwines from the top and bottom tiers were compared. Bigvines produced wine with lower hue values and lower an-thocyanins. There was little effect from vine size in thebottom tier on colour hue, however small vines in the bottomtier had lower a' and higher total anthocyanins (Table 11).

Judges recorded a clear preference for RT2T wines overSTD in all characters assessed. These wines had bettercolour, more fruit character and were fuller on the palate.Overall acceptability rating was 5,1 ex 7 for RT2T and 3,5 forSTD. Wines from the top tier were preferred to those from thebottom tier, and from small vines to big vines. These differen-ces were however small. Interactions were due to lowervalues for bottom tier big vines in colour hue, and fruit onnose and palate (Table 11).

Yield and quality relationship: The relationship be-tween yield and quality for RT2T and TK2T vines ispresented in Figures 3a and b. Figure 3a shows the relation-ship between yield and sensory score for overall acceptabilityfor different tier heights for TK3T and up-down and big-small comparisons for RT2T all using 1202C rootstock. Fig-

Yield and wine quality

4 •£

. * *

4RT2T•TK3T

3 1 2Yield (kg/m)

6-

(a8 4-

o</)

I2 '

TK2T o "Z21^

TK3T* STD

—— , 0 • —————— i ——— • ——— . ——— • ——— .3 0 1 0 2 0 3 0

Yield (l/ha)

FIGURE 3

Relationship between yield and sensory evaluation score ex7 for overall acceptability. Figure on left is data for three tiersof TK3T, and four combinations of vine size and tier heightfor RT2T. Yield results expressed per m canopy length.Figure on right is for STD, TK2T, RT2T, TK3T with com-posite fruit samples taken over two heights for TK2T, threeheights for TK3T, and four combinations of vine size and tierheight for RT2T. Cabernet franc, Rukuhia, 1988 vintage.

ure 3b shows the relationship for wines made from all vinesin the plot (i.e. composite over tier height and vine size, allARG1 rootstock). Note that as shading is decreased, yieldand sensory score are simultaneously improved.

Shading caused by fruit: For vertically shoot positionedcanopies where the fruit is within a constricted region thereis a possibility that clusters may contribute to shading of therenewal zone. For the productive RT2T trellis for example,while the canopy is of low density (15 shoots/m) the shootsare very fruitful (190 g/shoot) and the basal shoot zone isliterally a wall of fruit. Measurements were made in April1989 on the Cabernet franc trial to determine clusterprojected area in relation to canopy surface area. Clustershape was approximated as a truncated triangle, and charac-teristic dimensions measured on a sample of 20 clusters. Theaverage cluster projected area was 102 cm , and so foraverage cluster mass of 114 g, this corresponds to 0,89 cmprojected area/g fruit mass. The fruit zone height was onaverage 26 cm high, or 2620 cm /canopy surface area per mcordon. The area sampled had 31 clusters per m cordon, or3190 cm projected cluster area per m cordon. Therefore, theaverage cluster layer number (CLN, ratio cluster area: canopyfruit zone area) was 1,21. Since clusters are opaque, theclusters themselves can contribute significantly to shade inthe cluster region. For the value of 1,5 kg fruit/m canopysurface area of Table 4, the cluster surface area would be

2 2about 1340 cm /m canopy surface area, or 13%.CONCLUSIONS

This paper has condensed results of recent studies intoguidelines for canopy management. These are presented asfive 'principles'. Optimal values of 21 performance indicesand character shave been incorporated into a winegrapecanopy ideotype. Along with the field techniques of pointquadrat and canopy scoring which are described, these meas-urements will assist in diagnosing problem canopies.

Recent research results are also presented which serve toconfirm some of the principles that introduced this paper.Results presented from the Te Kauwhata Three Tier trellissystem reinforce principle 2, which deals with the spacing ofvertical foliage walls. The recommendation is that the heightto alley width ratio should not exceed about 1:1. Provided thisvalue is not exceeded, light levels at the canopy exteriorshould not drop below about 15% of ambient on a horizontalplane (Figure 1). Results presented for the TK3T comparisonshow that as the height: alley width ratio increases, so doesyield and wine quality decrease. All yield components areaffected by shade, with berry number and berry mass mostsensitive. Fruit maturation was delayed by shade, and judgesscored a clear preference for wines produced from top tierfruit which had least shading. These data gave further clearevidence for the negative effects of shade on yield and winequality.

The Ruakura Twin Two Tier system was designed withthe principles and canopy ideotype listed above in mind.Results presented here confirmed the importance of theseprinciples. Yield is doubled, fruit composition is improvedand wine quality increased. The bottom tier however, has lessyield and lower quality than the top tier. We are now evaluat-



ing a modified form of the RT2T with the bottom tier ofshoots growing downwards. In this configuration, the twofruiting/renewal zones will be closer together, and at midcanopy height, which will reduce yield, fruit composition andwine quality differences between the two tiers (Smart et al.,unpublished data).

The RT2T trial also permitted an evaluation of the effectof vine size. That is, vines with large retained node numbers(160) were compared with those pruned to 80 nodes, butwhere canopy length was proportional to retained node num-ber. Use of large vines caused desirable shoot devigoration,assisting in achieving a desirable canopy microclimate forthese vigorous vines. The results suggest that a preferredplanting arrangement would be to have large vines on the toptier and small vines on the bottom tier, which will help reducefruit composition and wine quality differences between tiers.The large vine will tend to counter the tendency of the top tierto promote vigour and fruit ripening as will the small vinescounter the tendency of bottom tier to reduce vigour anddelay ripening. The results are encouraging further researchinto the use of low vine density and complex trellis systemson high fertility sites.

Adoption of the canopy management techniques outlinedhere will have a major impact on winegrape production, fruitcomposition and wine quality from vigorous vineyards. Thisviewpoint is supported by literature citation and recent ex-perimental data presented here. While these results are con-tradictory to the common opinion that high yield causesreduced wine quality, they demonstrate that improvingcanopy microclimate for dense canopies can simultaneouslyimprove yield and quality.

LITERATURE CITED

BRAVDO, B., HEPNER, Y., LOINGER, C., COHEN, S. & TABACKMAN, H., 1985.Effect of crop level and crop load on growth, yield, must and wine composition andquality of Cabemet SauvignonMm./. Enol. Vitic. 36,125-131.

CARBONNEAU, A., 1985. Trellising and canopy management for cool climate viticul-ture. Proc. First Int. Symp. Cool Climate Viticulture and Enology, June, 1984Eugene, Ore. Oregon State University, 158-174.

CARBONNEAU, A., CASTERAN, P. & LE CLAIR, P., 1978. Essai de determinationen biologic de la plante entiere, de relations essentielles entre le bioclimat natural,la physiologic de la vigne et la composition du raisin. Ann. Amelior, Plant 28,195-221.

CARBONNEAU, A. & CASTERAN, P., 1987. Interactions training system x soil xrootstock with regard to vine ecophysiology, vigour, yield and red wine quality inthe Bordeaux area. Actta Honicuhurae 206,119-140.

CASTERAN, P. & CARBONNEAU, A., 1987. Etude de la fertilite des bourgeons, dela production et de la maturation en fonction du systems de conduite de la vigne.Proc. 3rd Int. Symp. Physiologic de la Vigne, June 1986, Bordeaux. OIV., 405-412.

CHAMPAGNOL, F., 1984. Elements de physiologic de la vigne et du viticulturegenerale. F. Champagnol.

CRIPPEN, D. & MORRISON, J., 1986a. The effects of sun exposure on the composi-tional development of Cabernet Sauvignon berries. Am. J. Enol. Vitic. 37,235-242.

CRIPPEN, D., & MORRISON, J., 1986b. The effects of sun exposure on the phenoliccontent of Cabernet Sauvignon berries during development. Am. J. Enol. Vitic. 37,243-247.

CLINGELEFFER, P., 1989. Update: minimal pruning of cordon trained vines (MPCT).Aust. Grapegrower and Winemaker 304, 78-83.

CULLEN, V., 1988. Canopy management for Sauvignon Blanc vines at Margaret River.Proc. Second Int. Symp. Cool Climate Viticulture and Oenology, January, 1988.Auckland, New Zealand. NZ Society for Vitic. and Oenol., 139-143.

DONALD, C., 1968. The design of a wheat ideotype. Proc. Third Int. Wheat GeneticsSymp. Canberra. Aust. Acad. Sciences.

ENGLISH, J., BLEDSOE, A. & MAROIS, J., 1989. Influence of leaf removal from thecluster zone on the components of evaporative potential within grapevine canopies.Am. J. Enol. Vitic. 40, (In press).

FREESE, P., 1988. Canopy modification and fruit composition. Proc. Second Int. Symp.

Cool Climate Viticulture and Oenology, January, 1988, Auckland, New Zealand.NZ Soc. for Vitic. and Oenol., 134-136.

GUBLER, W., MAROIS, J., BLEDSOE, A. & BETTIGA, L., 1987. Control of Botrytisbunch rot of grapes with canopy management. Plant Disease'11, 599-601.

HUNTER, J. & VISSER, J., 1988. Distribution of 14C- photosynthetate in the shoot ofVitis Vinifera L. cv. Cabernet Sauvignon. II. The effect of partial defoliation. S. Afr.J.Enol. Viiic9,10-15. -

ILAND, P., 1988. Leaf removal effects on fruit composition. Proc. Second Int Symp.Cool climate Viticulture and Oenology, January, 1988, Auckland, New Zealand.NZ Soc. for Vitic, and Oeribl., 137-138.

INTRIERI, C., 1987. Experiences on the effect of vine spacing and trellis trainingsystem on canopy microclimate, vine performance and grape quality >ActaHorticut-turae 206,69'88.

JACKSON, J. & PALMER, J., 1972. Interception of light by model hedgerow orchardsin relation to latitude, time of year and hedgerow configuration and orientation. J.Appl. Ecol. 9, 341-357.

JACKSON, D. & COOMBE, B., 1988. Early bunch stem necrosis - a cause of poor set.Proc. Second Int. Symp. Cool Climate Viticulture and Oenology, January, 1988,Auckland, New Zealand. NZ Soc. for Vitic. and Oenol., 72-75.

KLIEWER, M. & LIDER, L., 1968. Influence of cluster exposure to the sun oncomposition of Thompson seedless fruit. Am. J. Enol. Vitic. 19,175-184.

KLIEWER, M. & BLEDSOE, A., 1987. Influence of hedging and leaf removal oncanopy microclimate, grape composition, and wine quality under California condi-tions. Acta Honicuhurae 206, 157-168.

KLIEWER, M., MAROIS, J., BLEDSOE, A., SMITH, S., BENZ, M. & SIL-VESTRONI, O., 1988. Relative effectiveness of leaf removal, shoot positioning,and trellising for improving winegrape composition. Proc. Second Int. Symp. CoolClimate Viticulture and Oenology, January, 1988, Auckland, New Zealand. NZSoc. for Vitic, and Oenol, 123-126.

KLIEWER, M. & SMART, R., 1988. Canopy manipulation for optimising vinemicroclimate, crop yield and composition of grapes. In: Wright, C. (Ed.). Manipula-tion of Fruiting. Proc. 47th Easter School in Agric. Sci. Symp. April 1988, Univ.Nottingham. Buttenvorths, London, 275-291.

KOBLET, W., 1987a. Effectiveness of shoot topping and leaf removal as a means ofimproving quality. Acta Horticulturae-206, 141-156.

KOBLET, W., 1987b. Vieux bois et performance de la vigne. Proc. 3rd Int. Symp.Physiologic de la Vigne, June 1986, Bordeaux. OIV., 418^22.

KOBLET, W., 1988. Canopy management in Swiss vineyards. Proc. Second Int. Symp.Cool Climate Viticulture and Oenology, January 1988, Auckland, New Zealand.NZ Soc. for Vitic. and Oenol., 161-164.

LAVEE, S. & HASKAL, A., 1982. An integrated high density intensification systemfor tablegrapes. Proc. Grape and Wine Centennial Symp., June 1980, Davis, Calif.Univ. of Calif., 390-398.

MORRISON, J., 1988. The effects of shading on the composition of Cabernet Sauvig-non grape berries. Proc. Second Int. Symp. Cool Climate Viticulture and Oenology,January 1988, Auckland, New Zealand. NZ Soc. Vitic. and Oenol., 144-146.

MURISIER, F. & SPRING, J., 1987. Influence de la hauteur du tronc et de la densite deplantation sur le compartement de la vigne. Proc. 3rd Int. Symp. Physiologic de laVigne, June 1986, Bordeaux, OIV 412^17.

PALMER J. & JACKSON, J., 1977. Seasonal light interception and canopy develop-ment in hedgerow and bed system apple orchards. J. Appl. Ecol. 14,539-549.

PEARSON, R. & GOHEEN, A., 1988. Compendium of grape diseases. AmericanPhytopathological Society. Minn. 93 pp.

PETERSON, J. & SMART, R., 1975. Foliage removal effects on Shiraz grapevines. Am.J.Enol. Vitic. 26,119-124.

PSZCZOLKOWSKI, P., MARALES, A. & CARA, S., 1985. Chemical composition ofmusts and wines from clusters exposed to different levels of sunlight. Ciencia andInvestigation Agraria 12, 181-188.

REYNOLDS, A., 1989. Riesling grapes respond to cluster thinning and shoot densitymanipulation./. Amer. Soc. Hon. Sci. 114, 364-368.

REYNOLDS, A. & WARDLE, D., 1988. Canopy microclimate of Gewiirztraminer andmonoterpene levels. Proc. Second Intnl Symp. Cool Climate Viticulture and Oenol-ogy, January 1988; Auckland, New Zealand. NZ Soc. for Vitic. and Oenol.,116-122.

REYNOLDS, A., & WARDLE, D., 1989a. Impact of various canopy manipulationtechniques on growth, yield, fruit composition, and wine quality ofGewilrztraminer./lm. /. Enol. Vitic. 40,121-129.

REYNOLDS, A. & WARDLE, D., 1989b. Effects of timing and severity of summerhedging on growth, yield, fruit composition, and canopy characteristics of deChaunac. I. Canopy characteristics and growth parameters. Am. J. Enol. Vitic. 40,109-129.

REYNOLDS, A., & WARDLE, D., 1989c. Influence of fruit microclimate on monoter-pene levels of Gewiirztrammer. Am. J. Enol. Vitic. 40,149-154.

SAVAGE, S. & SALL, M., 1984. Botrytis bunch rot of grapes, the influence of selectedcultural practices on infection under Californian conditions. Plant Disease 67,771-774.

SOLARI, C., SILVESTRONI, O., GUIDICI, P. & INTRIERI, C., 1988. Influence oftopping on juice composition of Sangiovese grapevines (V. Vinifera L.). Proc.



Second Int. Symp. Cool Climate Viticulture and Oenology, January 1988, Auck-land, New Zealand. NZ Soc. for Vitic. and Oenol., 147-151.

SHAULIS, N. 1982. Responses of grapevines and grapes to spacing of and withincanopies. Proc. Grape and Wine Centennial Symp., June 1980, Davis, Calif. Univ.of Calif., 353-361.

SHAULIS, N., AMBERG, H. & CROWE, D., 1966. Response of Concord grapes tolight, exposure and Geneva Double Curtain training. Proc. Am. Soc. Hortic. Sci. 89,268-280.

SHAULIS, N. & SMART, R., 1974. Grapevine canopies: management, microclimateand yield responses. Proc. XlXth [nt. Hort. Congress, September 1974, Warsaw.254-265.

SMART.R., 1973. Sunlight interception by vineyards. Am. J.Enol. Vitic. 24,141-147.SMART, R., 1974. Photosynthesis by grapevine canopies. J. Appl. Ecol. 11,997-1000.SMART, R., 1982. Vine manipulation to improve winegrape quality. Proc. Grape and

Wine Centennial Symp., June 1980, Davis, Calif. Univ. of Calif. 362-375.SMART, R., 1985a. Principles of grapevine canopy microclimate manipulation with

implications for yield and quality: a review. Am. J. Enol. Vitic, 35,230-239.SMART, R., 1985b. Some aspects of climate, canopy microclimate, vine phhysiology

and wine quality. Proc. First Symp. Cool Climate Viticulture and Enology, June1984, Eugene, Oregon. Oregon State University, 1-19.

SMART, R., 1987a. The influence of light on composition and quality of grapes. AdaHorticulturae 206, 37-47.

SMART, R., 1987b. The light quality environment of vineyards. Proc. 3rd Int. Symp.Physiologic de la Vigne, June 1986, Bordeaux. OIV. 378-385.

SMART, R.,-1987c. Canopy management to improve yield, fruit composition andvineyard mechanisation. Proc. Sixth Australian Wine Industry Technical Conf.,July, 1986, Adelaide, SA. Australian Wine Research Institute, 205-211.

SMART, R., 1988. Shoot spacing and canopy light microclimate. Am. J. Enol. Vitic. 39,325-333.

SMART, R. & SINCLAIR, T., 1976. Solar heating of grape berries. Agric. Meteorol 17,241-259.

SMART, R. & DRY, P., 1980. A climate classification for Australia viticultural regions.Aust. Grapegrower and-Winemaker 196, 8-16.

SMART, R., SHAULIS, N. & LEMON, E., 1982a. The effect of Concord vineyardmicroclimate on yield. I. The effects of pruning, training, and shoot positioning onradiation microclimate. Am. J. Enol. Vitic. 33, 99-109.

SMART. R., SHAULIS, N. & LEMON, E., 1982b. The effect of Concord vineyard

microclimate on yield. II. The interactions between microclimate and yield expres-sion. Am. J. Enol. Vitic. 33,109-116.

SMART, R. & COOMBE, B., 1983. Water relations of grapevines. In: Kozlowski, T.,(Ed.). Water deficits and plant growth. Vol. VII. Academic Press, NY, 137-196.

SMART, R., ROBINSON, J. DUE, G. & BRIEN, C, 1985a. Canopy microclimatemodification for the cultivar Shiraz. I. Definition of canopy microclimate. Vitis 24,17-31.

SMART, R., ROBINSON, J., DUE, G. & BRIEN, C., 1985b. Canopy microclimatemodification for the cultivar Shiraz. II. Effects on must and wine composition. Vitis,24,119-128.

SMART, R. & SMITH, S., 1988. Canopy management: identifying the problems andpractical solutions. Proc. Second Int. Symp. Cool Climate Viticulture and Oenol-ogy, January 1988, Auckland, New Zealand. NZ Soc. for Vitic. and Oenol. 109-115.

SMART, R., SMITH, S. & WINCHESTER, R., 1988. Light quality and quantity effectson fruit ripening of Cabemet Sauvignon. Am. J. Enol. Vitic. 39,250-258.

SMART, R. & SHARP, K., 1989. Handbook for canopy management workshop.Viticultural Bulletin 54. MAP Ruakura Agricultural Centre, Hamilton, NewZealand, 45pp.

SMART, R., DICK, J., & GRAVETT, I., 1989. Proc. Seventh Australian Wine IndustryTechnical Conf., August 1988, Adelaide, SA. Australian Wine Research Institute(In press).

SMITH, S., CODRINGTON, I., ROBERTSON, M., & SMART, R., 1988. Viticulturaland oenological implications of leaf removal for New Zealand vineyards. Proc.Second Int. Symp. Cool Climate Viticulture and Oenology, January 1988, Auck-land, New Zealand. NZ Soc. Vitic. and Oenol. 127-133.

SOMERS, T. & EVANS, M. 1977. Spectral evaluation of young red wines, anthocyanequilibria, total phenolics, free and molecular SO2, chemical age. J. Sci. Food Agric.28,279-287.

TRAVIS, J., 1987. Effect of canopy density on pesticide deposition and distribution inapple trees. Plant Disease 71, 613-615.

VAN DEN ENDE, B., 1984. The Tatura trellis - a system of growing grapevines forearly and high production. Am. J. Enol. Vitic. 35,82-87.

WEISS, A. & ALLEN, L., 1976a. Air-flow patterns in vineyard rows. Agric. Meteorol.16,329-342.

WEISS, A. & ALLEN, L., 1976b. Vertical and horizontal air flow above rows of avineyard. Agric. Meteorol. 17, 433-452.


canopy management to improve grape yield and …canopy management to improve grape yield and wine...

Documents