The use of viticultural terroir units for demarcation of geographical indications for wine production in Stellenbosch and surrounds

Carey, Victoria Anne (2005-04)

Dissertation (PhD(Agric)) -- Stellenbosch University, 2005.

Thesis

ENGLISH ABSTRACT: Due to increased consumer demand for products labelled by origin, and the requirement that these labels are a guarantee of both quality and product character, there is an increasing global focus on delimitation of denominations of origin. The integrity of denominations of origin and their defensibility can be ensured through the use of terroirs as a basis for delimitation. The aims of this study were to establish the dominant environmental criteria that affect the viticultural behaviour and wine character of two important cultivars (Cabernet Sauvignon and Sauvignon blanc) in the Stellenbosch Wine of Origin District, to use an appropriate methodology to identify viticultural terroirs in this district based on these criteria and with the use of a geographic information system, and finally to use these viticultural terroirs to identify denominations of origin within the same area. A terroir can be defined as a grouping of homogenous environmental units, or natural terroir units, based on the typicality of the products obtained. Identification and characterisation of terroirs depends on knowledge of environmental parameters, the functioning of the grapevine and the characteristics of the final product. Field studies, resulting in point data, are necessary to investigate the functioning of the grapevine but in order for this information to be of use within zoning studies it must be placed in a spatial context. As a first phase in data acquisition, the Stellenbosch Wine of Origin District was characterised and natural terroir units were identified using existing digital data and a geographic information system. A natural terroir unit (NTU) can be defined as a unit of land that is characterised by relatively homogenous topography, climate, geological substrate and soil. A total of 1389 NTUs were identified in the Stellenbosch Wine of Origin District (84 537 ha). The identified NTUs were homogenous with respect to terrain morphological unit, altitude, aspect and soil type. Each of the identified units was further described with respect to the extent of the expected sea breeze effect and, for certain of the soil types, the associated parent material. As a second phase of data acquisition, a network of plots of Sauvignon blanc and Cabernet Sauvignon were delimited in commercial vineyards in proximity to weather stations and their viticultural and oenological response monitored for a period of seven years. Regression tree analyses were performed on the complete data set and the relative importance of the environmental and management related variables determined for each dependent variable. Excepting for scion clone, which had a high relative importance for bunch mass of Sauvignon blanc and yield to pruning mass index of Cabernet Sauvignon, no other non-environmental variable included in the analyses appeared to have a strong effect on grapevine performance and wine character. The performance of Cabernet Sauvignon was affected by the potassium content of the subsoil and the climate of the season. The performance of Sauvignon blanc appeared to be related to soil texture, wind exposure and temperature, both during the green berry growth stage and the month prior to ripening. From the results presented, it appears that environmental parameters have an overriding effect on the performance of both Cabernet Sauvignon and Sauvignon blanc, but that these two cultivars react differently to environmental stimuli. A knowledge-driven model used the rules generated in the regression tree analyses to directly classify natural terroir units with respect to expected response of Cabernet Sauvignon and Sauvignon blanc in the Stellenbosch Wine of Origin District. The natural terroir units were thus grouped into terroir units that were homogenous with respect to predicted viticultural and oenological response for each cultivar. The use of representative sites to determine the response of the grapevine to its environment is time consuming and costly and limits terroir studies to research related investigations. Vineyard managers were therefore surveyed with respect to the functioning of established Sauvignon blanc vineyards in the Stellenbosch Wine of Origin District in an attempt to obtain the necessary data. Comparison of the data generated with these questionnaires to measured data in commercial vineyards suggested that the vineyard managers were able to characterise the performance of vineyards with respect to vigour, signs of drought stress and yield. Each vineyard was mapped and the responses were linked to modelled environmental variables. Classification and regression trees were used to construct decision trees, which could be applied to environmental data in a geographic information system to determine viticultural terroirs for production of Sauvignon blanc. These terroirs, although fewer, were comparable to those generated using field data. Data gathered during terroir studies, and the identified viticultural terroirs for Cabernet Sauvignon and Sauvignon blanc, were used to revisit the boundaries of the Stellenbosch Wine of Origin District and the Simonsberg-Stellenbosch ward. Modifications were proposed based on expected wine characteristics. Boundaries for two new wards in the Helderberg basin were proposed. It was also possible to identify vineyards within a ward for the production of terroir specific wines.

AFRIKAANSE OPSOMMING: Toenemende verbruikersaanvraag vereis produkte waarvan die etikette nie net die oorsprong aandui nie, maar ook kan dien as ‘n waarborg vir die produk se kwaliteit en kenmerkendheid. Hierdie tendens verklaar die toenemende wereldwye fokus op afgebakende areas van oorsprong. Die integriteit van die afgebakende areas van oorsprong sowel as hul verdedigbaarheid kan gewaarborg word deur terroirs as basis vir afbakening te gebruik. Die doelstellings van die studie was om die oorheersende omgewingseienskappe wat wingerdprestasie en wynkarakter van twee belangrike cultivars, naamlik Cabernet Sauvignon en Sauvignon blanc, in die Stellenbosch Wyn van Oorsprongsdistrik bemvloed, te bepaal; tweedens om ‘n toepaslike metodologie te gebruik om terroirs in die die distrik te bepaal wat gebaseer is op die geidentifiseerde omgewingseienskappe met die gebruik van ‘n geografiese inligtingstelsel; en, ten slotte, om hierdie terroirs vir wingerbou te gebruik om afgebakende areas van oorsprong in dieselfde omgewing te identifiseer. ‘n Terroir kan gedefinieer word as ‘n samestelling van homogene omgewingseenhede, of natuurlike terroir-eenhede (NTE), wat gebaseer word op die kenmerkende eienskappe van die produkte wat daaruit verkry word. Identifisering en karakterisering van terroirs sal afhang van kennis van die omgewingsparameters, die funksionering van die wingerdstok en die eienskappe van die finale produk. Veldstudies waaruit puntdata verkry word, is noodsaaklik om die funksionering van die wingerdstok te ondersoek. Dit is egter noodsaaklik om eers hierdie inligting in ruimtelike konteks te plaas alvorens die inligting vir soneringstudies gebruik kan word. As 'n eerste fase van datagenerering, was die Stellenbosch Wyn van Oorsprongsdistrik gekarakteriseer en NTE’s geidentifiseer deur gebruik te maak van bestaande digitale data en ‘n geografiese inligtingstelsel. ‘n NTE kan gedefinieer word as ‘n landseenheid wat gekarakteriseer word deur ‘n relatiewe homogene topografie, klimaat, geologiese substraat en grondtipe. ‘n Totaal van 1389 NTE’s is geidentifiseer in die Stellenbosch Wyn van Oorsprongsdistrik (84 537 ha). Die geidentifiseerde NTE’s was homogeen met betrekking tot die terrein morfologiese eenheid, hoogte bo seespieSI, hellingsaspek en grondtipe. Elk van die geidentifiseerde eenhede was verder beskryf volgens die omvang van die seewindinvloed en, vir toepaslike grond tipes, die geassosieerde moedermateriaal. As ‘n tweede fase van datagenerering is ‘n netwerk van persele van Cabernet Sauvignon en Sauvignon blanc afgebaken binne bestaande kommersiele wingerde in die nabyheid van weerstasies. Hul wingerd- en wynkundige respons is vir ‘n periode van sewe jaar gemonitor. Regressieboomanalises is gebruik om die volledige stel data te analiseer en om die relatiewe belang van omgewings- en bestuurspraktykverbonde veranderlikes te bepaal. Die bostokkloon (wat 'n hoe relatiewe belang vir die trosmassa van Sauvignon blanc en die oes- tot snoeimassa verhouding van Cabernet Sauvignon het) is die enigste van die nie-omgewingsparameter wat ‘n sterk invloed op wingerdprestasie of wynkarakter blyk te he. Die prestasie van Cabernet Sauvignon is beTnvloed deur die kaliuminhoud van die ondergrond sowel as die seisoensklimaat. By Sauvignon blanc het dit voorgekom of die prestasie verband hou met grondtekstuur, windblootstelling en temperatuur tydens die groen fase van korrelgroei sowel as die maand voor rypwording. Alhoewel dit blyk uit die resultate dat omgewingsparameters 'n oorheersende invloed op die prestasie van beide Cabernet Sauvignon en Sauvignon blanc uitoefen, reageer die twee cultivars verskillend op omgewingsprikkels. ‘n Kennisgedrewe model waarvan die riglyne uit resultate van die regressieboomanalise saamgestel is, word gebruik om NTE direk te klassifiseer ten opsigte van die verwagte respons van Cabernet Sauvignon en Sauvignon blanc in die Stellenbosch Wyn van Oorsprongsdistrik. Die NTE is dus gegroepeer om terroir eenhede te vorm wat homogeen was ten opsigte van die verwagte wingerd- en wynkundige respons vir elke cultivar. Die gebruik van verwysingspersele om die respons van die wingerdstok teenoor sy onmiddelike omgewing te bepaal, is tydrowend en duur en beperk sodoende terroir studies tot navorsing. Gevolglik is ‘n opname onder wingerdbestuurders gemaak om inligting oor die prestasie van Sauvignon blanc in die Stellenbosch Wyn van Oorsprongsdistrik in te win en sodoende die nodige data te verkry. Na vergelyking van die ingewinde data, wat uit die opname verkry is, met gemete data vanaf kommersiele wingerde, kon afgelei word dat wingerdbestuurders by magte is om die prestasie van wingerde ten opsigte van groeikrag, tekens van droogtestres en opbrengs te karakteriseer. Elke wingerd is gekarteer en die respons is gekoppel aan die gemodeleerde omgewingsparamters. Klassifikasie en regressiebome is gebruik om besluitnemingsmodelle saam te stel wat toegepas kon word op omgewingsdata in ‘n geografiese inligtingstelsel om terroirs vir die produksie van Sauvignon blanc te bepaal. Hierdie terroirs, alhoewel minder, was vergelykbaar met die wat gegenereer was met behulp van veldstudies. Data verkry met terroirstudies, sowel as gei'dentifiseerde terroir-eenhede, was gebruik om die grense van die Stellenbosch Wyn van Oorsrongsdistrik en die Simonsberg-Stellenbosch wyk te herbepaal. Voorgestelde modifiserings was gebaseer op verwagte wyneienskappe. Grense vir twee nuwe wyke in die Helderbergkom is voorgestel. Dit was ook moontlik om wingerde binne ‘n wyk te identifiseer vir die produksie van terroir-spesifieke wyne.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/50533
This item appears in the following collections: