FINALE NAVORSINGSVERSLAG (1999/2000) : WINETECH. Dr. J. Marais

ARC LNR ARC-FRUIT, VINE AND WINE RESEARCH INSTITUTE LNR-NAVORSINGSINSTITUUT VIR VRUGTE, WINGERD EN WYN INFRUITEC Private Bag / Privaatsak X5013, Stellenbosch 7599, South Africa /Suid-Afrika Tel: (021) 809 3100. Fax (021) 809 3400. (Int: +27 21) E-mail: netmanager@infruit.agric.za, Website: www.arc.agric.za NIETVOORBIJ Private Bag / Privaatsak X5026, Stellenbosch 7599, South Africa /Suid- Afrika Tel: (021) 809 3100. Fax (021) 809 3002, (Int: +27 21) E-mail: supervisor@nietvoor.agric.za.web site: www.arc.agric.za FINALE NAVORSINGSVERSLAG (1999/2000) : WINETECH PROJEKNOMMER : WW 13/04 PROJEKTITEL Ondersoek na die gebruik van mikroklimaatsfaktore in die bepaling van druif- en wynkwaliteit. PROJEKLEIER Dr. J. Marais PROGRAMBESTUURDER: Dr. O. P. H. Augustyn AANVANGSDATUM : 1993 EINDDATUM : 2000 Geen gedeelte van hierdie dokument mag sonder skriftelike toestemming van die ARC Infruitec-Nietvoorbij gereproduseer word, in enige vorm, of andersins weergee word nie. AN INSTITUTE OF THE AGRICULTURAL RESEARCH COUNCIL 'N ÍNSTITUUT VAN DIE LANDBOUNAVORSINGSRAAD

FINALE NAVORSINGSVERSLAG (1999/2000) PROJEKNOMMER PROJEKLEIER MEDEWERKER(S) WW13/04 Dr. J. Marais Dr. J.J. Hunter Mnr.P. Haasbroek PROJEKTITEL Ondersoek na die gebruik van mikroklimaatsfaktore in die bepaling van druif- en wynkwaliteit. ERKENNINGS Die gedeeltelike befondsing van hierdie projek deur Winetech word waardeer. PROBLEEM AANGESPREEK/DOEL OF HIPOTESESTELLING Wêreldwyd bestaan daar 'n behoefte vir die effektiewe evaluasie van druifkwaliteit. Die doel van hierdie projek is die identifisering van maklik meetbare klimaatsparameters wat gebruik kan word in die uiteindelike definiëring en voorspelling van druif- en wynkwaliteit. NAVORSINGSBEHOEFTE AANGESPREEK Winetech Tegniese Komitee 1999 : Prioriteit A DOELWITTE VAN DIE VERSLAGJAAR 1. Voorlegging van 1998/99 navorsingsverslag aan Winetech vir gedeeltelike befondsing van projek. 2. Ontleding en sintuiglike evaluasie van 1999 Sauvignon blanc wyne. 3. Gaschromatografiese ontleding van Olifantsrivier druifmonsters vir monoterpene en ibmp. 4. Statistiese verwerking van 1996 tot 1999 data en opstel van moontlike kwaliteitsvoorspellingsmodel. 5. Opstel van 1999/2000 finale navorsingsverslag. RESULTATE EN BESPREKING Die projek is suksesvol afgehandel en al die gestelde doelwitte is behaal. Die resultate is saamgevat in 10 publikasies. (Sien LYS VAN PUBLIKASIES). Samevattend kom dit op die volgende neer: Vir die eerste keer is mikroklimaatsparameters,

geurstofdata en wynkwaliteit op so 'n uitgebreide vlak, naamlik oor rypheid, streke en oesjare, ondersoek. Daar is bewys dat daar 'n verband tussen mikroklimaatsdata (ligstraling, temperatuur), geurstofdata (monoterpene, norisoprenoïede, 2-metoksi-3-isobutielpirasien) en wynkwaliteit (vrugtigheid, groenrissie/aspersie intensiteite) van Sauvignon blanc bestaan. Hierdie beginsel is in verskillende klimaatstreke (Robertson, Stellenbosch, Elgin) neergelê. Dit is moontlik om wynkwaliteit ten opsigte van die tipiese groenrissie/aspersie karakter aan die hand van ligintensiteit te voorspel. Dit is nie 'n kitsresep vir algemene kwaliteitsdefiniëring nie en moet beskou word as 'n bydrae tot hierdie ingewikkelde probleem. Effektiewe loofbestuur wat die korrekte balans tussen ligblootstelling en beskaduing van druiwe bewerkstellig, verhoog die kultivartipisiteit en wynkwaliteit van Sauvignon blanc. Die voorwaarde is egter dat Sauvignon blanc aangeplant word in geskikte, "koel" lokaliteite. Vir jare lank is daar gespekuleer oor die effek van klimaat en loofbestuur op wynkwaliteit. Daar is met hierdie ondersoek wetenskaplike bewyse verkry van die veranderings wat in Sauvignon blanc druiwe onder verskillende mikro- en makroklimaatstoestande plaasvind en dat spesifieke wynstyle in die wingerd bestuur kan word. Die resultate van hierdie studie gee duidelike riglyne vir die suksesvolle verbouing van Sauvignon blanc in feitlik al die wynproduserende streke van Suid-Afrika. In hierdie verband is uitstekende terugvoering van beide Stellenbosch (Villiera) en Robertson (Graham Beck Wines) areas verkry wat daarop dui dat die toepassing van die resultate wynkwaliteit reeds in hierdie areas verhoog het.

BEGROTING : 2000 / 2001 PROJEK : WW 13/04 TYDPERK Lopend Direkte Salarisse Kapitaal Oorhoofs Totaal 1999/2000 171 186 189 382 68 617 142 036 571 221 2000/2001 30 813 52 239 2 200 34 051 119 303 EKSTERNE BEFONDSING (a) Winetech: 2000/2001 - Aangevra : R 40 986 (BTW uitgesluit) - Ontvang : R 35 131 (BTW uitgesluit)

BEGROTING VIR KAPITALE UITGAWE PROJEK: WW 13/04 2000/2001 ITEM 1. Kleurdrukker 2. Gaschromatograaf 3. 4. 5. 6. 7. 8. 9. 10. ITEM 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. TOTAAL 2001/2002 TOTAAL KOSTE (R) 2 200 58 333 60 533 KOSTE (R) Slegs 'n gedeelte van die koste is aan die projek toegeken. Die ander gedeelte is aan projekte WW 8/18 en WW 08/19 toegeken. Vanes/Tik/NavlgOO/JM/WW13/04/11-05-00

OPSOMMING VAN FINALE NAVORSINGSVERSLAG : 1999/2000 PROJEKNOMMER PROJEKLEIER MEDEWERKER(S) WW 13/04 Dr. J. Marais Dr. J.J. Hunter Mnr. P. Haasbroek PROJEKTITEL Ondersoek na die gebruik van mikroklimaatsfaktore in die bepaling van druif- en wynkwaliteit. Die doel van die projek was die identifisering van maklik meetbare klimaatsparameters wat gebruik kan word in die uiteindelike definiëring en voorspelling van druif- en wynkwaliteit. Twee loofbestuurbehandelings wat mikroklimaat op 'n natuurlike wyse verander (kontrole/normale sonligblootstelling en oorskaduing) is op Sauvignon blanc wingerde te Robertson, Stellenbosch en Elgin oor ses seisoene (1994 tot 1999) toegepas. Geurstofkonsentrasies (monoterpene, C 13 -norisoprenoïede en 2-metoksi-3-isobutielpirasien [ibmp]) is weekliks tussen deurslaan en rypheid in die druiwe bepaal. Ligstraling bo en binne die stokke en temperature in die omgewing van die trosse is kontinueerlik tydens Desember, Januarie en Februarie van elke seisoen gemeet. Die hoogste geurstofkonsentrasies (monoterpene en C 13 -norisoprenoïede), ligstraling binne stokke en gemiddelde maksimum temperature in die omgewing van die trosse is by die kontrole behandelings waargeneem, gevolg deur die oorskadu behandelings. Daarteenoor was ibmp konsentrasies deurgaans hoër in skadu- as in sonblootgestelde druiwe. Oor die algemeen het monoterpeen- en C 13 -norisoprenoïed konsentrasies toegeneem, terwyl ibmp konsentrasie afgeneem het tydens rypwording. In 'n vergelyking tussen oesjare en streke is gevind dat koeler oesjare en streke oor die algemeen hoër aroma konsentrasies as die warmer oesjare en streke gelewer het. 'n Verband tussen chemiese- en mikroklimaatsdata is verkry en dit is ook statisties bewys in 'n model wat dit moontlik maak om ibmp vlakke en Sauvignon blanc wynkwaliteit (in tertne van groenrissie/aspersie intensiteit) te voorspel. Die resultate van hierdie studie gee duidelike wetenskaplike riglyne vir Sauvignon blanc verbouing in verskillende klimaatstreke. Dit word aanbeveel dat Sauvignon blanc slegs in geselekteerde "koel" lokaliteite aangeplant word en dat met behulp van loofbestuur die korrekte lig/skadu balans bewerkstellig word om optimale geurstofbalans, kultivartipisiteit en wynkwaliteit te verseker. Dit is basies moontlik om twee Sauvignon blanc wynstyle, naamlik die groenrissie/aspersie-agtige styl en die vrugtige/tropiese styl te produseer. Binne perke is dit die keuse van die wingerdbouer en wynmaker watter styl hulle wil produseer. Al vereiste wat geld is dat die wyn steeds as Sauvignon blanc herkenbaar moet wees.

Die navorsingsresultate is plaaslik by vyf Winetech terugvoervergaderings aan die Bedryf oorgedra, asook by 'n Sauvignon blanc bemarking simposium in Durbanville en 'n SAWWV Sauvignon blanc Inligtingsdag. Verder is die resultate ook op internasionale vlak by ses simposia aangebied. Die resultate is ook saamgevat in tien publikasies (sien LYS VAN PUBLIKASIES). Vanes/Tik:/Navlg00/JM/WW13/04 Opsom/12-05-2000

SUMMARY OF FINAL RESEARCH REPORT : 1999/2000 PROJECT NUMBER PROJECT LEADER CO-WORKER(S) WW 13/04 Dr. J. Marais Dr. J.J. Hunter Mr. P. Haasbroek PROJECT TITLE Investigation into the use of microclimatic parameters in the determination of grape and wine quality. The aim of the project was to identify easily measurable climatic parameters that can be used in the eventual definition and prediction of grape and wine quality. Two canopy manipulations altering microclimate in a natural way (control/normal sunlight exposure and shading) were applied to Sauvignon blanc vineyards at Robertson, Stellenbosch and Elgin over six seasons (1994 to 1999). Aroma compound concentrations in the grapes (monoterpenes, C 13 - norisoprenoids and 2-methoxy-3-isobutylpyrazine [ibmp]) were determined weekly between véraison and ripening. Light radiation above and within canopies and temperatures in the vicinity of clusters were also continually measured during December, January and February of each season. Highest canopy light radiation, average maximum temperatures in the vicinity of clusters, and concentrations of certain aroma compounds (monoterpenes and C 13 - norisoprenoids) were found for control treatments, followed by the shading treatments. On the other hand, ibmp concentrations were constantly higher in the shaded grapes than in the sunexposed grapes. Generally, monoterpene and C 13 -norisoprenoid concentrations increased, while ibmp concentration decreased during ripening. When seasons and regions were compared, generally higher aroma concentrations were observed in the cooler seasons and regions than in the warmer seasons and regions. A relationship was found between chemical and microclimatic data, and it was also statistically confirmed in a model which is able to predict ibmp levels and Sauvignon blanc wine quality (in terms of green pepper/asparagus intensity). The results of this study present clear scientific guidelines for Sauvignon blanc cultivation in different climatic regions. It is recommended that Sauvignon blanc should only be cultivated in selected "cool" localities and that canopy management be applied to obtain the correct light/shade balance that would secure optimum aroma balance, cultivar typicity and wine quality. Basically, it is possible to produce two Sauvignon blanc wine styles, namely the green pepper/asparagus-like style and the fruity/tropical style. The choice lies with the viticulturist and winemaker to produce, within limits, the style they prefer. The only requirement is that the wine should still be recognisable as Sauvignon blanc.

The research results were presented locally to the wine industry at five Winetech meetings, as well as at a Sauvignon blanc symposium in Durbanville, and at a SASEV Sauvignon blanc Information day. The results were further presented on an international level at six symposia. Results were also used to prepare ten publications (See LYS VAN PUBLIKASIES/PUBLICATION LIST). Vanes/Tík:/Navlg00/JM/VWV13/04/2O00

LYS VAN PUBLIKASIES/PUBLICATION LIST 1. MARAIS, J., HUNTER, J.J., HAASBROEK, P. & AUGUSTYN, O.P.H., 1995. Effect of cluster exposure on Sauvignon blanc grape aroma composition. In GOUSSARD, P.G., ARCHER, E., SAAYMAN, D., TROMP, A & VAN WYK, J. (eds.). Proc. 1ste SASEV Int. Congr., 8-10 November 1995, Cape Town, South Africa. pp. 32-34. 2. MARAIS, J., 1996. Sauvignon blanc: Metoksipirasiene en mikroklimaat. Wynboer Tegnies 80, 7-8. 3. MARAIS, J., HUNTER, J.J., HAASBROEK, P. & AUGUSTYN, O. P. H., 1996. Effect of canopy microclimate on Sauvignon blanc grape composition. In: STOCKLEY, C.S., SAS, A.N., JOHNSTONE, R. S. & LEE, T.H. (eds.). Proc. 9th Aust. Wine Ind. Tech. Conf., 16-19 July 1995, Adelaide, Australia. pp. 72-77. 4. MARAIS, J., 1996. Fruit environment and prefermentation practices for manipulation of monoterpene, norisoprenoid and pyrazine flavorants. In: HENICK-KLING, T., WOLF, T.E. & HARKNESS, E.M. (eds.). Proc. 4th Int. Symp. Cool Climate Vitic. & Enol., 16-20 July 1996, Rochester, USA. pp. V40-48. 5. MARAIS, J., 1997. Optimum rypheid: 'n Faktor in druif- en wynkwaliteit. Wynboer Tegnies 99, 10-12. 6. MARAIS, J., 1998. Temperatuur: 'n Faktor in druifverbouing en wynbereiding. Wynboer Tegnies 103, 8-12. 7. MARAIS, J., 1998. Aroma compounds and precursors in Gewurztraminer, Riesling and Sauvignon blanc grapes: The role of clone, grape maturity, canopy microclimate and region. Proc. Int. Aroma Symp., 16 April 1998, Verona, Italy. In press. 8. MARAIS, J., HUNTER, J. J. & HAASBROEK, P.D., 1999. Effect of canopy microclimate, season and region on Sauvignon blanc grape composition and wine quality. S. Afr. J. Enol. Vitic. 20, 19-30. 9. MARAIS, J., 2000. Die effek van mikroklimaat op Sauvignon blanc druif- en wynkwaliteit. Wynboer 128, 90-92.

2 10. MARAIS, J., CALITZ, F. & HAASBROEK, P.D., 2000. Relationship between microclimatic data, aroma component concentrations and wine quality parameters in the prediction of Sauvignon blanc wine quality. In press. Vanes/Tik:/NavlgOO/JM/Lys van publikasies/ww13/04

Effect of Canopy Microclimate, Season and Region on Sauvignon blanc Grape Composition and Wine Quality J. Marais, JJ. Hunter and P.D. Haasbroek ARC-Frait, Vine and Wine Research Institute, Nietvoorbij Centre for Vine and Wine, Private Bag X5026, 7599 Stellenbosch, South Africa Submitted for publication: February 1999 Accepted for publication: May 1999 Key words: Microclimate, season, region, Sauvignon blanc, aroma, wine quality The effect of canopy microclimate on the grape aroma composition and wine quality of Sauvignon blanc was investigated in three climatically-different regions, i.e. in the Stellenbosch (1996 and 1997 season), Robertson and Elgin regions (1997 and 1998 season). A canopy shade treatment altering microclimate in a natural way, was applied to Sauvignon blanc vineyards in the three regions. Control vines were not manipulated. The concentrations of aroma compounds in the grapes, namely monoterpenes, C 13 -norisoprenoids and 2-methoxy-3-isobutylpyrazine, were determined weeklv during the respective ripening periods. Solar radiation above and within the canopies as well as temperature within the canopies were also measured continually during the ripening periods. The highest canopy solar radiation, temperature, and monoterpene and C 13 -norisoprenoid concentrations were found for the control treatments, followed by the shaded treatments. An opposite tendency was found for 2-methoxy-3-isobutylpyrazine, which is one of the most important components responsible for the typical green pepper/asparagus aroma of Sauvignon blanc. There appears to be a relationship between chemical and microclimatic data in each region and over seasons. Marked temperature and aroma component concentration differences were observed among the three regions during the cool 1997 season, which manifested in wine aroma parameters such as fruitiness and vegetative/asparagus/green pepper nuances. Two defmite wine styles emerged, namcly the green pepper/asparagus "cool climate" style and the "warm climate" fruity/tropical style. However, differences in 2-methoxy-3-isobutylpyrazine and C 13 -norisoprenoid levels and wine characteristics between regions were not as pronounced during the warm 1998 season. The data contribute to establishing guidelines for canopy manipulation for obtaining a specific wine character and quality. The choice lies with the viticulturist and winemaker to strive for and obtain, within limits, the style that they prefer. It is well known that climate has a pronounced effect on grape and wine composition and quality. Considering the fact that climatic conditions between wine regions in South Africa differ greatly from regions II to IV (1389 to 2222 degree days) (Le Roux, 1974), it can be expected that wine composition, style and quality will differ between these regions. Furthermore, mesoclimate can differ between vineyards within a region and depends on factors such as aspect, slope, altitude, surrounding vegetation, etc. The effect of canopy microclimate on grape and wine composition is equally important (Smart et al., 1985a; 1985b; Smart et al., 1990). In this regard, climatic parameters such as temperature and solar radiation seem to be of special significance (Iland, 1989a; 1989b; Dokoozlian & Kliewer, 1996; Marais et al., 1996; Versini et al., 1996). Intervention by means of canopy management practices may affect grape composition to a great extent and different wine styles can be produced from the same vineyard (Hunter & Visser, 1988a; 1988b; Morrison & Noble, 1990; Hunter, De Villiers & Watts, 1991; Iacono & Scienza, 1995). Sauvignon blanc has become one of the most important white wine cultivars in South Africa. The area planted to this cultivar has increased by 99% from 2255 (1985) to 4479 (1996) hectares (Booysen & Truter, 1997). Although high quality wines with the typical cultivar aroma characteristics are produced, a high proportion of wines show a neutral character. In view of greater international and local competition, it is of utmost importance to enhance the cultivar typicity and quality of Sauvignon blanc wine in general. To achieve this goal, specific guidelines for the cultivation of this cultivar in different climatic regions are needed. Local studies (De Villiers et al, 1995; WJ. Conradie, 1997. Personal communication; Hunter & Le Roux, 1997) aimed at identifying regions and canopy management practices best suited to the cultivation of Sauvignon blanc are continuing. Knowledge about the relationship between microclimate, aroma development and wine quality of Sauvignon blanc is, however, limited (Allen &Lacey, 1993; Hunter & Ruffher, 1998). The purpose of this investigation was therefore to determine the effect of canopy microclimate on Sauvignon blanc grape composition and wine quality in different climatic regions over different seasons. MATERIALS AND METHODS Canopy manipulations: Vitis vinifera L. cv. Sauvignon blanc in three climatically-different regions, namely Robertson, Stellenbosch and Elgin, was used. According to a local degree day model (Le Roux, 1974), the three regions are classified as regions IV (1945-2222 degree days), III (1668-1944 degree days) and 11(1389-1667 degree days), respectively. Grapes were obtained over two seasons, namely 1996 and 1997 in the case of the Stellenbosch region, and 1997 and 1998 for the remaining two regions. One canopy manipulation treatment which altered microclimate in a natural way, was applied in each region (based on the method of Archer & Stxauss, 1989). During winter, two one-year Acknowledgements: Technical contributions by JJ. Strydom, DJ. le Roux, C.W. Fouché, C.G. Volschenk, LM. Adams and E. Swart, as well as financial support by the South African Vine and Wine Industry are appreciated. S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999 19

Microclimate, Season and Region Effects on Sauvignon blanc Quality old canes from each adjacent vine were trained to the middle of the cordon of the treated vine, allowing a natural shading of the vine in summer. Control vines were not manipulated. Three replications, consisting of 15 vines per replicate, were used. Suckering of shoots not located on spurs and shoot positioning were applied to all treatments. Shoots on canes used for shading were defruited to stimulate growth. The long-term annual rainfall for the Stellenbosch, Elgin and Robertson regions is : 737 mm, 1042 mm and 327 mm, respectively. The vineyards in the Stellenbosch and Elgin regions were therefore supplementarily irrigated (two irrigations per season), while the Robertson vineyard was irrigated intensively to evapotranspiration demand. Microclimatic parameters: MCS data loggers were installed and continually measured temperature within the canopies (in the vicinity of clusters) and within the clusters, as well as mainly visible light radiation (300-1200 nm) within the canopies of selected vines over the whole ripening period on an hourly basis. Each radiation sensor consisted of 10 individual sensors, fitted in parallel onto a metre-long rod, thus giving a more reliable, average reading in the canopy. Sampling and harvesting of grapes: Whole bunches (approximately 2 kg per sample) were collected weekly at random from each treatment between approximately 16 B (close to véraison) and 21 B (ripeness). In the Elgin region, sampling started at approximately 17 B. Grapes were harvested at ripeness for wine production. Wines were produced according to standard Nietvoorbij practices for small-scale white wine production. Aroma component analyses: Techniques for the extraction and analyses of free monoterpenes, glycosidically bound monoterpenes and C 13 -norisoprenoids, and 2-methoxy-3- isobutylpyrazine (ibmp) were given in a previous publication (Marais et al., 1996). The relative concentrations of linalool, hotrienol, alpha-terpineol, trans- and cis-pyran linalool oxide, citronellol, nerol and diendiol-1 were summed and expressed as total monoterpene concentration. The relative concentrations of trans- and cis-furan linalool oxide, cis-8-hydroxy linalool, trans- 1,8-terpin, trans- and cw-vitispirane, l,l,6-trimethyl-l,2-dihydronaphthalene, beta-damascenone, and actinidol 1 and 2 were summed and expressed as total acid-released monoterpene and C1 3 -norisoprenoid concentrations. During 1998, ibmp analyses were done on a new GCMS (Finnigan MAT GCQ ) instrument. The same conditions applied as with the former instrument (Finnigan 4600 Quadrupole MS). Sensory analyses: Wines were sensorially evaluated for fruitiness and vegetative/asparagus/green pepper intensities by a panel of six experienced judges. A line-method was used, i.e. evaluating the intensity of each characteristic by making a mark on an unstructured, straight 10 cm line. The left-hand and right-hand ends of the line were indicated by the terms, "undetectable" and "prominent", respectively. The wines of the 1996 season (Stellenbosch region) were not evaluated, due to faulty characters. Statistical analysis: Statistical differences between treatments were determined by applying standard analysis of variance methods to the data, obtained from each respective region and season. Least significant differences (LSD) were used to separate the means of the treatments. RESULTS AND DISCUSSION The concentrations of total monoterpenes and C13-norisoprenoids, generally considered as important components of Sauvignon blanc (Lacey et al., 1991; Sefton, Francis & Williams, 1994), decreased with an increase in canopy density, and increased during ripening between approximately 16 B and 21 B (Figs. 1 to 6). Similar tendencies were found in all three regions and during all seasons, confírming previous results from the Stellenbosch region (Marais et al., 1996). In some cases, increases in compound concentrations were followed by slight decreases close to ripeness, which could be ascribed to transformations to other compounds (Marais & Van Wyk, 1986; and references therein). The effect of canopy microclimate on the concentration of ibmp is shown in Figures 7 to 9. This methoxypyrazine is considered one of the most important cultivar-typical components of Sauvignon blanc (Lacey et al., 1991; Allen &Lacey, 1993). Contrary to monoterpenes and C 13 -norisoprenoids, the concentration of ibmp decreased with increasing grape exposure to sunlight as well as during ripening (Figs. 7 to 9). Again, similar tendencies were found in all three regions and during all seasons, and confirmed previous results (Marais et al., 1996). Average temperatures within the canopies, measured continually during the respective grape ripening periods, are shown in Figure 10. Temperature differences between "within clusters" (between berries) and "in the vicinity of clusters" were minimal and can be ascribed to air movement or wind, which had an equalising effect on temperature (Rojas-Lara & Morrison, 1989; Hunter & Visser, 1990). Therefore these two temperature measurements were considered as repetitions and referred to as "temperature within the canopy" for the purpose of this investigation. However, temperature of berry juice can differ greatly firom that of the environment, an aspect which should also be considered when temperature effects on grape composition and quality are investigated (Smart & Sinclair, 1976; Marais & Fourie, 1997). Generally, average maximum temperatures were only slightly lowered by denser canopies (Fig. 10). Therefore canopy manipulation may, to a certain degree, be successful in obtaining cooler grapes in warm locations. The benefits of cool grapes for the production of high quality wines are widely recognised (Marais, 1998; and references therein). The average temperatures (Fig. 10) are in agreement with the degree day model of Le Roux (1974), indicating Robertson as the warmer, Elgin as the cooler and Stellenbosch as the intermediate region. The fact that the 1997 season was cooler than both the 1996 and 1998 seasons (meteorological data not shown), is not completely evident in Figure 10, apart from high average minimum temperatures in the Elgin region in 1998, probably due to overcast nights. However, when variation in temperature, expressed as the number of hours below 12 C and above 30 C, is considered, it is evident that 1997 was the cooler season of the three (Fig. 11). Outlier temperatures or variation in temperature may therefore be a more important parameter than average temperatures. The apparent contradiction between the 1998 Elgin average minimum temperatures in Figures 10 and 11 may be explained by the fact that the data in Figure 10 represent only about three weeks, while those in Figure 11 represent three months. Incidentally, no <12 C hours occurred during February, the month in which ripening took place. S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microclimate, Season and Region Effects on Sauvignon blanc Quality 21 18 16 Control Shaded 14 11/2 18/2 25/2 1997 Season Sampling date FIGURE 1 20/1 27/1 1998 Season Effect of canopy microclimate on total monoterpene concentrations (average of three replicates) in Robertson Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ significantly (p<0.05). 3/2 23/1 30/1 6/2 1996 Season 13/2 10/2 Sampling date FIGURE2 17/2 24/2 1997 Season Effect of canopy microclimate on total monoterpene concentrations (average of three replicates) in Stellenbosch Sauvignon blanc grapes over two seasons (1996 and 1997). Treatments at each sampling date designated by the same letter do not differ significantly (p<0,05). 3/3 S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

22 Microclimate, Season and Region Effects on Sauvignon blanc Quality 40 35-30 - 10-5 - 19/2 26/2 5/3 1997 Season Sampling date FIGURE 3 28/1 4/2 11/2 1998 Season Effect of canopy microclimate on total monoterpene concentrations (average of three replicates) in Elgin Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ significantly (p<o,o5). Control Shaded 11/2 18/2 1997 Season 25/2 13/1 20/1 27/1 1998 Season Sampling date FIGURE 4 Effect of canopy microclimate on total acid-released monoterpene and C 13 -norisoprenoid concentrations (average of three replicates) in Robertson Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ significantly (p<0,05). 3/2 S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microclímate, Season and Region Effects on Sauvignon blanc Quality 23 50 40 30 20 23/1 30/1 6/2 1996 Season 13/2 10/2 Sampling date FIGURE 5 17/2 24/2 1997 Season Effect of canopy microclimate on total acid-released monoterpene and C 13 -norisoprenoid concentrations (average of three replicates) in Stellenbosch Sauvignon blanc grapes over two seasons (1996 and 1997). Treatments at each sampling date designated by the same letter do not differ signifícantly (p<o,o5). 19/2 26/2 1997 Seaso 5/3 28/1 Sampling date FIGURE 6 4/2 1998 Season Effect of canopy microclimate on total acid-released monoterpene and C 13 -norisoprenoid concentrations (average of three replicates) in Elgin Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ signifícantly (p<0,05). 11/2 S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

4/2 11/2 18/2 1997 Season 25/2 13/1 Sampling date 20/1 27/1 1998 Season 3/2 FIGURE 7 Effect of canopy microclimate on 2-methoxy-3-isobutylpyrazine concentration (average of three replicates) in Robertson Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ significantly (p<0,05). 23/1 30/1 6/2 1996 Season 13/2 10/2 Sampling date FIGURE 8 17/2 24/2 1997 Season Effect of canopy microclimate on 2-methoxy-3-isobutylpyrazine concentration (average of three replicates) in Stellenbosch Sauvignon blanc grapes over two seasons (1996 and 1997). Treatments at each sampling date designated by the same letter do not differ significantly (p<0,05). 3/3 S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microdimate, Season and Region Effects on Sauvignon blanc Quality 25 19/2 26/2 5/3 1997 Season Sampling date 28/1 4/2 11/2 1998 Season FIGURE 9 Effect of canopy microclimate on 2-methoxy-3-isobutylpyrazine concentration (average of three replicates) in Elgin Sauvignon blanc grapes over two seasons (1997 and 1998). Treatments at each sampling date designated by the same letter do not differ significantly (p<0,05). 1997 1998 Elgin Average maximum, average minimum and average Ki temperatures within differently manipulated Sauvignon blanc canopies in three regions and over three seasons. Temperatures represent averages during each ripening period between véraison (+15 B) and harvest (±22 B). C = control, S = shaded. Treatments in each respective region and season designated by the same letter do not differ significantly (p<0,05). S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

26 Microclimate, Season and Region Effects on Sauvignon blanc Quality It can be expected that, as a result of temperature differences between regions and seasons, respective ripening periods will differ to some extent. This in fact happened, since the 1997 season was approximately three weeks later than the other two seasons, and ripening in the Elgin region is normally about one and two weeks later than in the Stellenbosch and Robertson regions, respectively. These differences in temperature conditions may strongly affect grape composition, although other factors such as clone, soil water supply and solar radiation naturally also play a role in aroma development. As expected, solar radiation within the canopy decreased with increasing density in the vineyards in all three regions and during all three seasons (Fig. 12). Comparison between regions and seasons gives varying tendencies, which may be ascribed to the light intensity above the canopy, sunlight hours per day, row direction and canopy density of which the latter is affected by various cultivation practices. Different climatic conditions manifest differently in aroma component concentrations and grape and wine quality. Generally, total monoterpene and C 13 -norisoprenoid levels were higher in the cooler 1997 season, compared to the two warmer seasons (Figs. 1 to 6). Within one season, i.e. 1997, total monoterpene concentrations were also higher in the cooler Elgin region than in the warmer Stellenbosch and Robertson regions (Figs. 1 to 3). This tendency was, however, not repeated in the total bound aroma concentrations during the 1997 season. In fact, the Robertson levels were apparently slightly higher than the other values (Figs. 4 to 6). During the 1998 season, total monoterpene levels were also much higher in the cooler Elgin than in the warmer Robertson region (Figs. 1,3,4 and 6). Generally, the ibmp levels were higher during the cooler 1997season, compared to the other season in each region, as well as higher in the cooler Elgin than in the two warmer regions, i.e. during the 1997 season (Figs. 7 to 9). This is in general agreement with monoterpene levels in the same grapes (Figs. 1 to 3). Similarly, ibmp levels were shown to be highest in grapes from cooler climatic regions (Lacey et al., 1991; Allen & Lacey, 1993). However, during the warm 1998 season, small differences in ibmp levels occurred between the warmer Robertson and cooler Elgin regions, although the values of the latter region tended to be higher at ripeness. The sugar concentration of the Elgin grapes at ripeness was about 1 B higher than that of the Robertson grapes, which could also explain the relatively small difference in ibmp levels. In both cases, the end ibmp levels were still above the threshold value of 2 ng// (Buttery et al., 1969). The relatively high ibmp levels at ripeness in the Stellenbosch region (1996 season), and the gradual decrease in concentrations thereof during ripening (Fig. 8), in contrast to relatively prominent decreases in the other cases, is difficult to explain, but could probably be ascribed to dense canopies, caused by rainfall prior to veraison. On the other hand, solar radiation values in the same vineyard and season (Fig. 12) were much higher than all the other values concemed, which seems contradictory in the light of the extreme light-sensitivity of ibmp (Heymann, Noble & Boulton, 1986). It 100 1997 1998 Robertson 1996 1997 Stellenbosch Season and region FIGURE 11 1997 1998 Elgin Average outlier temperatures (hours) in three regions and over two seasons. Values represent averages for three months (December, January and February) for each respective region and season. S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microdimate, Season and Region Effects on Sauvignon blanc Quality 27 Above canopy Wrthin canopy (control) Within canopy (shaded) 1,5 1997 1998 Robertson 1996 1997 Stellenbosch Season and region 1997 1998 Elgin FIGURE 12 Average total radiation above and within differently manipulated Sauvignon blanc canopies in three regions and over two seasons. Values represent averages for three months (December, January and February) for each respective region and season. is evident that the interaction between solar radiation and temperature and the effect thereof on grape aroma development, is a complex phenomenon. Furthermore, other factors such as soil water supply and nutrient status may also affect vine vigour, which in turn affects light exposure of grapes and subsequently wine aroma composition (Noble, Elliot-Fisk & Allen, 1995). It appears that there is a relationship between solar radiation, temperature, monoterpene, C 13 -norisoprenoid and ibmp concentrations within each region and season. Higher monoterpene and norisoprenoid, and lower ibmp concentrations coincided with higher light intensity within the canopy, while higher monoterpene, norisoprenoid and ibmp concentrations coincided with lower environmental temperatures. It also appears that within a region and season, solar radiation, as manipulated by canopy management, has a more prominent effect on aroma component levels than temperature (Figs. 1 to 9). Morrison & Noble (1990) also suggested that changes in the composition of the berry as a result of shading of the leaf and cluster were affected more by light than by temperature. However, when regions or seasons are compared, it appears that the effect of temperature on aroma component levels becomes much more prominent. Although it is not possible to single out solar radiation and temperature as the most inportant parameters affecting aroma component concentrations, it is well known that chemical and, within limits, enzymatic reaction rates, are temperature and light dependent. Therefore the role of temperature and light in the development and/or degradation of aroma components appears to be of particular importance. In general, it can be stated that cooler conditions, i.e. a cooler region and/or a cooler season, benefitted the development and retention of monoterpenes and ibmp in the local study. It is generally accepted that monoterpenes and C 13 -noriso prenoids contribute to the fruity and tropical nuances and ibmp to the vegetative/green pepper/asparagus aroma of Sauvignon blanc wines (Sefton, Francis & Williams, 1994; Lacey et al., 1991). These contributions to wine quality are affected by, amongst others, temperature and solar radiation in each region and season. The variation in green pepper/asparagus aroma intensity of the 1997 wines between warm and cool regions (Fig. 13) corresponds to the respective grape ibmp levels (Figs. 7 to 9). A highly signifïcant correlation between ibmp concentration and vegetative aroma intensity of Sauvignon blanc was found by Allen et al, (1991). Differences in 1997 ibmp levels at harvest between regions (Figs. 7 to 9) were, however, not of the same magnitude as differences between the perceived "green" nuances (Fig. 13). When the monoterpene levels at harvest (Figs. 1 to 3) and respective fruity aroma intensities are compared, the Robertson values do not follow the expected trend, although the bound component levels tended to be higher in this region (Figs. 4 to 6). When the sensory data of the wines of the relatively warm and short 1998 season (Fig. 14), and the respective aroma component levels at harvest (Figs. 1 to 9) were considered, a number of trends were. observed. Generally, differences in fruitiness and green pepper/asparagus intensities between the cooler Elgin and the warmer Robertson region were small, which coincide with the S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microclimate, Season and Region Effects on Sauvignon blanc Quality 70 r Control Shaded a I I I I I 20-10 - lb b í Robertson Stellenbosch Bgin Robertson Stellenbosch Bgin Region FIGURE 13 Effect of canopy microclimate on aroma intensities of Sauvignon blanc wines from the Robertson, Stellenbosch and Elgin regions (1997 season). Treatments in each region designated by the same letter do not differ significantly (p<0,05). 70 Control 60 Shaded Robertson Elgin Robertson Elgin Region FIGURE 14 Effect of canopy microclimate on aroma intensities of Sauvignon blanc wines from the Robertson and Elgin regions (1998 season). Treatments in each region designated by the same letter do not differ significantliy (p<0,05). S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Microclimate, Season and Region Effects on Sauvignon blanc Quality 29 respective grape norisoprenoid and ibmp levels at ripeness. The 1998 Sauvignon blanc wines lacked prominent cultivar-typical characteristics and virtually no differences in green pepper/- asparagus intensities between treatments occurred, which is in agreement with the relatively low ibmp concentrations of the respective grapes at ripeness (Figs. 7 and 9). In a ranking evaluation of the 1998 wines, however, the panel indicated that the wines produced from the shaded treatments had more intense green pepper/grassy aromas and higher quality than the control wines in 100% of the cases. Wine composition can differ markedly between wines from climatically-different regions (Marais et al., 1992), and a large number of aroma components are involved. Therefore it is not suggested that the measured components are the only ones involved in Sauvignon blanc characteristics. Futhermore, it is common knowledge that aroma components manifest differently in different media, depending on the presence of other aroma-enhancing components. Therefore the perceived aroma nuances in this study could have been masked or enhanced by the synergistic action of different aromas present. Considering the above-mentioned results, it can be expected that prominent differences in aroma component concentrations and wine sensory characteristics may be possible between climatically-different regions during a cool season, but not necessarily in a warm season. It is clear that basically two different wine styles, i.e. the typical green pepper/asparagus style and the more tropical/fruity style were produced in the three regions in the cooler 1997 season. The results indicate that cooler regions and/or more shade are needed for the fírst, and warmer regions and/or more sunlight exposure for the second style. Wines of both styles can be of high quality and it remains the choice of the winegrower to manipulate, within the limits of his region, light exposure of the grapes (and to a certain extent also temperature), in order to obtain the style that is preferred. Whatever the choice, the following basic principle should be taken into account. Optimum ripeness, yielding maximum Sauvignon blanc wine quality, would be that stage where a sufficient level of ibmp is still present in the grapes, i.e. above its threshold value (2 ng//) to obtain the typical green pepper/asparagus nuances, and where it is complemented by a sufficiently developed level of other herbaceous, fruity and tropical aromas. CONCLUSIONS There appears to be a defïnite relationship between the concentrations of aroma components in grapes, such as monoterpenes, C13-norisoprenoids and ibmp, and microclimatic parameters such as within canopy temperature and solar radiation. Macro- and microclimatic differences between regions and seasons manifested in Sauvignon blanc grape aroma composition and could explain most of the observed tendencies. It is not suggested that solar radiation and temperature are the only parameters that affect the development or degradation of the above-mentioned components, since too many factors are involved. Nevertheless, from the present and previous studies they appear to be of particular significance. Canopy microclimate can, to a certain degree, be manipulated by viticultural practices to obtain grapes that will produce wine with the desired aroma composition, character and quality. This appears to be easier in a cooler than in a warmer season. Two Sauvignon blanc wine styles, i.e. the green pepper/asparagus style and the tropical/fruity style, can be obtained by the selection of appropriate cultivation localities and canopy management practices. The ideal is that each style should have nuances of the other. Generally, grapes in cooler regions would benefit from more sunlight exposure, while those in warmer regions would benefit from more shade. Negative effects of too high temperatures and too much sunlight exposure, or too low temperatures and lack of sunlight exposure on grape and wine quality should be borne in mind. Depending on the macro- and mesoclimate, certain localities would naturally not be suited for the cultivation of a climatically-sensitive cultivar such as Sauvignon blanc. It is, however, possible that most wine regions in South Africa have specific locations, where Sauvignon blanc could be cultivated successfully and investigations are in progress to locate such "cooler" areas in the warmer regions. The possibility exists of using microclimatic parameters as indicators or predictors of grape and wine quality, instead of lesseasily measurable aroma components, and is the subject of ongoing research. It would also make more sense to apply parameters that would give more representative information of the whole ripening process, rather than monitoring quality at harvest only. Additional data have to be collected during forthcoming vintages from different regions in order to better understand the complex interaction between light, temperature and aroma development, and to further refine existing guidelines for optimum canopy management for the production of maximum quality grapes and wines in all suitable regions. LITERATURE CITED ALLEN, M.S. & LACEY, M.I., 1993. Methoxypyrazine grape flavour: Influence of climate, cultivar and viticulture. Wein-Wiss. 48, 211-213. ALLEN, M.S., LACEY, M.J., HARRIS, R.L.N. & BROWN, W.V., 1991. Contribution of methoxypyrazines to Sauvignon blanc wine aroma. Am. J. Enol. Vitic. 42, 109-112. ARCHER, E. & STRAUSS, H.C., 1989. Effect of shading on the performance of Vitis vinifera L. cv. Cabernet Sauvignon. S. Afr. J. Enol. Vitic. 10, 74-76. BOOYSEN, J.H. & TRUTER, J., 1997. Belangrikste wyndruiftipes in die RSA. Wynboer Tegnies9S, 11-14. BUTTERY, R.G., SEIFERT, R.M., GUADAGNl, D.G. & LING, L.C., 1969. Characterization of some volatile constituents of bell peppers. J. Agric. Food Chem. 17, 1322-1327. DE VILLIERS, F.S., SCHMIDT, A., THERON, J.C.D. & TALJAARD, R., 1995. Onderverdeling van die Wes-Kaapse wynbougebiede volgens bestaande klimaatskriteria. Wynboer Tegnies 78, 10-12. DOKOOZLIAN, N.K. & KUEWER, W.M., 1996. Influence of light on grape berry growth and composition varies during fruit development. J. Amer. Soc. Hort. Sci. 121, 869-874. HEYMANN, H., NOBLE, A.C. & BOULTON, R.B., 1986. Analysis of methoxypyrazines in wines. I. Development of a quantitative procedure. J. Agric. Food Chem. 34, 268-271. HUNTER, J.J. & VISSER, J.H., 1988a. Distribution of 14 C-photosynthetate in the shoot of Vitis vinifera L. cv. Cabernet Sauvignon. I. The effect of leaf position and developmental stage of the vine. S. Afr. J. Enol. Vitic. 9, 3-9. HUNTER, JJ. & VISSER, J.H., 1988b. Distribution of 14 C-photosynthetate in the shoot of Vitis vinifera L.cv. Cabemet Sauvignon. II. The effect of partial defoliation. S. Afr. J. Enol. Vitic. 9, 10-15. HUNTER, J.J. & VISSER, J.H., 1990. The effect of partial defoliation on growth characteristics of Vitis vinifera L.cv. Cabernet Sauvignon. I. Vegetative Growth. S. Afr. J. Enol. Vitic. 11,18-25. HUNTER, J.J., DE VILLIERS, O.T. & WATTS, J.E., 1991. The effect of partial defoliation on quality oharacteristics of Vitis vinifera L. cv. Cabernet Sauvignon grapes. II. Skin color, skin sugar, and wine quality. Am. J. Enol. Vitic. 42, 13-18. HUNTER, JJ. & LE ROUX, D.J., 1997. Canopy management effects on yield, labour input, and growth compensation - new canopy composition perspectives. Proc. 5lh Int. Symp. on Grapevine Physiology, 25-30 May 1997, Jerusalem, Israel. Acta Hort. In press. S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

30 Microclimate, Season and Region Effects on Sauvignon blanc Quality HUNTER, J.J. & RUFFNER, H.P., 1998. Produktorientierte Laubarbeit in sudafrikanischen Weinbau. Schweiz. Z. Obst-Weinbau 12, 306-308. IACONO, E & SCIENZA, A., 1995. Differential effects of canopy manipulation and shading of Vitis vinifera L.cv. Cabernet Sauvignon.II. Wine sensory properties. Vitic. Enol. Sci. 50,9-13. ELAND, R, 1989a. Grape berry composition - the influence of environmental and viticultural factors. Part 1. Temperature. Aust. Grapegrower & Winemaker 326, 13-15. ILAND, P., 1989b. Grape berry composition - the influence of environmental and viticultural factors. Part 2. Solar radiation. Aust. Grapegower & Winemaker 328, 74-76. LACEY, M.J., ALLEN, M.S., HARRIS, R.L.N. & BROWN, W.V., 1991. Methoxypyrazines in Sauvignon blanc grapes and wines. Am. J. Enol. Vitic. 42, 103-108. LE ROUX, E.G., 1974. 'n Klimaatsindeling van die Suidwes-Kaaplandse wynbougebiede. M.Sc. Thesis, University of Stellenbosch, Stellenbosch, South Africa. MARAIS, J., 1998. Effect of grape temperature, oxidation and skin contact on Sauvignon blanc juice and wine composition and wine quality. S. Afr. J. Enol. Vitic. 19, 10-16. MARAIS, J. & FOURIE, E., 1997. Dniiftemperatuur : 'n Belangrike kwaliteitsparameter. Wynboer Tegnies 99, 8-9. MARAIS, 3. & VAN WYK, C.J., 1986. Effect of grape maturity and juice treatments on terpene concentrations and wine quality of Vitis vinifera L.cv. Weisser Riesling and Bukettraube. S. Afr. J. Enol. Vitic. 7, 26-35. MARAIS, J., HUNTER, J.J., HAASBROEK, P.D. & AUGUSTYN, O.P.H., 1996. Effect of canopy microclimate on Sauvignon blanc grape compositíon. In : STOCKLEY, C.S., SAS, A.N., JOHNSTONE, R.S. & LEE, T.H. (eds.). Proc. 9th Aust. Wine Ind. Tech. Conf. 16-19 July 1995, Adelaide, Australia. pp. 72-77. MARAIS, J., VERSINI, G., VAN WYK, C.J. & RAPP, A., 1992. Effect of region on free and bound monoterpene and C13-norisoprenoid concentrations in Weisser Riesling wines. S. Afr. J. Enol. Vitic. 13, 71-77. MORRISON, J.C. & NOBLE, A.C., 1990. The effect of leaf and cluster shading on the composition of Cabemet Sauvignon grapes and on fruit and wine sensoty properties. Am. J. Enol. Vitic. 41,193-200. NOBLE, A.C., ELLIOT-FISK, D.L. & ALLEN, M.S., 1995. Vegetative flavor and methoxypyrazines in Cabemet Sáuvignon. In: ROUSEFF, R.L. & LEAHY, M.M. (eds.). ACS Symposium Series 596. Fruit Flavors. Biogenesis, Characterization and Authentication. 22-27 August 1993, Chicago, Illinois, USA. pp. 226-234. ROJAS-LARA, B.A. & MORRISON, J.C., 1989. Differential effects of shading fruit or foliage on the development and composition of grape berries. Vitis 28, 199-208. SEFTON, M.A., FRANCIS, I.L. & WHXIAMS, P.J., 1994. Free and bound volatile secondary metabolites of Vitis vinifera grape cv. Sauvignon blanc. J. Food Sci. 59, 142-147. SMART, R.E. & SINCLAIR, T.R., 1976. Solar heating of grape benies and other spherical fruits. Agricultural Meteoroíogy 19, 241-259. SMART, R.E., DICK, J.K., GRAVETT, I.M. & FISHER, B.M., 1990. Canopy management to improve grape yield and wine quality - Principles and practices. S. Afr. J. Enol. Vitic. 11, 3-17. SMART, R.E., ROBINSON, J.B., DUE, G.R. & BRIEN, C.J., 1985a. Canopy microclimate modification for the cultivar Shiraz. I. Definition of canopy microclimate. Vitis 24, 17-31. SMART, R.E., ROBINSON, J.B., DUE, G.R. & BRIEN, C.J., 1985b. Canopy microclimate modification for the cultivar Shiraz. II. Effects on rnust and wine compositíon. Vitis 24, 119-128. VERSINI, G., RAPP, A., DALLA SERRA, A., NICOLINI, G. & BARCHETTI, P., 1996. Aroma profile differences among grape products from different geographic areas. In : LEM- PERLE, E., TROGUS, H. & FIGLESTAHLER, P. (eds.). llth Int. Oenol. Symp. 3-5 June 1996, Sopron, Hungary, pp. 402-424. S. Afr. J. Enol. Vitic, Vol. 20, No. 1,1999

Relationsbip between microclimatic data, aroma component concentrations and wine quality parameters io the prediction of Sauvignon blanc wine quaiity. J. Marais 1, F. Calitz 2 and P. D. Haasbroek 3 'ARC Infruitec-Nietvoorbij, Private Bag X5026, 7599 Stellenbosch, South Africa. 2 ARC Biometry Unit, Private Bag X5013, 7599 Stellenbosch, South Africa. 3 ARC Agromet, Private Bag X5O13, 7599 Stellenbosch, South Afirica. Submitted for publication : Accepted for publication : Key words : Microclimate, temperature, radiation, Sauvignon blanc quality, prediction model. ABSTRACT Sauvignon blanc grape chemical and wine sensory data, as well as meteorological data (temperature and visible light radiation), collected in three climatically-different wine regions in South Africa, over three seasons and from two different canopy treatments, were statistically analysed. A model for the prediction and/or definition of specifically Sauvignon blanc wine quality was developed. The model utilises above and within canopy radiation and can explain 68,8% of the variation in the cultivar-typical vegetative/asparagus/green pepper intensity of Sauvignon blanc wine. Other signifícant correlations, e.g. between temperature and monoterpene concentrations were also obtained. Further research is necessary to test and refíne this model for application under different environmental conditions. INTRODUCTION Research over many years was aimed at developing a simple model or recipe that could define and predict wine quality. Examples are sugar/acid ratio ( B/TTA) (Du Plessis & Van Rooyen, 1982), the glycosyl-glucose (G-G) assay (Francis et al, 1996) and the red wine colour index (Holgate, 2000). Such a model should also be useful as a method to compensate the grape producer according to grape quality. The aim further is that the