Investigating the metabolic potential of Merlot in a characterised vineyard

Thesis (MScAgric)--Stellenbosch University, 2020.ENGLISH ABSTRACT: This study focused on characterising the variability in a Merlot vineyard after which the effect of aspect and ripeness on the composition of grape skins and seed, as well as wines made from these berries, was investigated. The North-South orientated vineyard rows resulted in the two canopy sides (East and West) being exposed to the sun at different stages during the day. There was a strong indication that this related to variable microclimatic conditions between the two-canopy sides which could in turn affect the metabolic composition of the berries. Merlot is one of the red wine cultivars which is susceptible to the influence of environmental conditions, resulting in heterogeneity in the characteristics of the grapes. In order to focus our investigation solely on the effect of aspect, it was important to exclude other variable factors which might also impact metabolite composition. Thus, a workflow was created to incorporate multiple sources of information describing the vineyard variability at different spatial scales. First, remote sensing was conducted on a vineyard-scale, gathering data-rich images including a digital elevation model, soil electrical conductivity scan and Sentinel 2 satellite images. The images were combined as data layers in a 3-D data cube which was statistically analysed using principle component analysis. This showed that about 60 % of the variability in the vineyard was caused by a combination of the soil properties and the elevation profile of the plot (thus fixed and stable factors) while another 10 % related to the phenological stages of the berries. To investigate aspect and ripeness we excluded the variability related to the stable factors and created a data-driven experimental design to identify and avoid the areas of highest variability when the experimental layout (subplot) was chosen. Pruning weights and trunk circumference measurements of the vines in this subplot showed homogeneity with regard to vigour. Subsequently, grapes were sampled separately for the two canopy sides and at two harvest dates at each sampling point (11 x positions) and analysed for phenolic compound composition. Climate analysis concluded that aspect did indeed influence the microclimate. The East side canopy experienced a sudden increase in the rate of temperature incline during the hours when this canopy was directly exposed to the sun (higher light exposure) as well as a higher temperature at 8 a.m., both of which resulted in a higher phenolic content of the grapes. Further investigations focused on the different berry tissues (skins and seeds) and the final Merlot wines and showed that the skin and wine phenolic profile were similar with regard to the patterns seen for each canopy side, as well as the two harvest dates. The East canopy side wines had a higher sugar and phenolic compound and a lower methoxypyrazine concentration in comparison to those of the West canopy. This indicates an increased rate of grape ripening on the East canopy side probably due to higher temperatures and light exposure. These differences were clear in the first harvest but less dominant in the second which might indicate synchronization of grape maturity. The metabolite differences seen in the grapes persisted to the wine but were not as prominent which is probably the result of metabolite bioconversions and interactions during wine making. The sensory analysis revealed that the higher methoxypyrazine concentrations (above the perception threshold) in wines of the West canopy side was likely masked by the high presence of organic acids, C13-norisoprenoids and monoterpenes in the first harvest date, as well as high alcohol concentrations in wines made from grapes harvested at the second harvest date. This study illustrated the design and implementation of an integrated data-driven experimental design to characterise vineyard-scale variability and to make informed choices on sampling points for detailed investigation on a subplot-scale (or bunch/berries scale) depending on your desire to include or exclude different variability causing factors.AFRIKAANSE OPSOMMING: Hierdie studie fokus op die karakterisering van veranderlikheid in ‘n Merlot wingerd en die impak van aspek en rypheidsgraad op die samestelling van die druifdop en saad, sowel as die ooreenkomstige wyn. Die Noord-Suid oriëntasie van die wingerdrye het daartoe gelei dat die twee lowerkante (Oos en Wes) tydens verskillende tye gedurende die dag aan die son blootgestel is. Daar was vervolgens ‘n sterk aanduiding dat dit variasie in die twee lowerkante se mikroklimaat kan veroorsaak en gevolglik ook in die metaboliese samestelling van die korrels. Merlot staan bekend as ‘n rooiwyn kultivar wat relatief vatbaar is vir variasies in omgewings kondisies wat kan lei tot heterogeniteit in druifkenmerke. Om te verseker dat hierdie studie gefokus bly op die impak van aspek, was dit belangrik om die invloed van ander veranderlike faktore wat ook ‘n moontlike effek op metaboliese samestelling kan hê, uit te sluit. ‘n Werkstroom is dus ontwikkel om die integrasie van verkillende vlakke van inligting rondom wingerd veranderlikheid op verskeie ruimtelike skale, te bewerkstelling. Eerstens is afstandswaarneming op die wingerd, as geheel, toegepas. Dit behels die verkryging van dataryke foto’s/beelde insluitende ’n hoogte bo seevlak model, grond elektriese geleidings model, sowel as Sentinel 2 satellietbeelde. Hierdie beelde is gekombineer as data lae in ‘n 3-D datakubus en statisties ontleed met behulp van hoofkomponentanalise (PCA). Die tegniek het gewys dat die kombinasie van grond elektriese geleiding en hoogte bo seevlak (stabiele faktore) ongeveer 60% van die variasie in die wingerd verduidelik, terwyl die fenologiese groei stadia rondom 10% verduidelik. Vir die bestudering van aspek en rypheid, is die variasie wat veroorsaak is deur die stabiele faktore, uitgesluit. Vervolgens is ‘n sub-plot, data gedrewe eksperimentele uitleg ontwikkel met die doel om areas van ekstreme variasie te identifiseer en te vermy. Die snoeigewigte en stam omtrekke van die wynstokke in die sub-plot het bevestig dat groeikagtigheid relatief homogeen was. Druiwe monsters is geneem van elke steekproefpunt (x11) vir beide lowerkante tydens twee rypheidsgrade en getoets vir fenoliese komponente. Klimaatsontleding het bevestig dat aspek die mikroklimaat beïnvloed. Die Oostelike lowerkant het ‘n skielike verhoging in die tempo van temperatuur toename ondervind tydens die ure wat hierdie lowerkant direk aan sonlig blootgestel was, asook hoër temperature teen 8 vm. Dit het gelei het tot ‘n hoër fenoliese inhoud van die druiwe. Verdere ondersoeke het gefokus op veskillende druifweefsels (doppe en sade) sowel as die finale Merlot wyn gemaak vanaf die druiwe. Resultate het gewys dat die druifdop en wyn fenoliese samestelling vergelykbaar is in terme van die patrone wat waargeneem is vir elke lowerkant en oesdatum. Die wyne afkomstig vanaf die Oostelike lowerkant het hoër suiker en fenoliese kosentrasies gehad, maar laer metoksipirasiene in vergelyking met wyne gemaak vanaf die Westelike lowerkant. Dit dui op ‘n verhoogde rypwordings tempo van druiwe vanaf die Oostelike kant, moontlik as gevolg van hoër temperature en sonlig blootstelling. Hierdie verskille was duidelik tydens die eerste oesdatum, maar minder so tydens die tweede oes, moontlik as gevolg van die sinchronisering van druif rypwording. Die metaboliet verskille wat in die druiwe bespeur is, was ook aanwesig, maar minder prominent in die finale wyne, moontlik as ‘gevolg van metaboliet bio-omskakelings en interaksies tydens die wynmaak proses. Sensories ontleding wys dat die hoër vlakke van metoksipirasiene (bo die persepsievlakke) teenwoordig in wyne vanaf die Westelike lowerkant nie sensories bespeur word nie. Hierdie maskeringseffek kan toegeskruif word aan die hoër konsentrasie organiese sure, C13-norisoprenoïede en monoterpene tydens die eerste oesdatum en hoër alkohol vlakke teenwoordig in wyne vanaf die tweede oesdatum. In hierdie studie is ‘n geïntegreerde data gedrewe eksperimentele uitleg ontwerp en geïmplimenteer waarmee die veranderlikheid van ‘n wingerd gekararteriseer is. Daarmee is dit moontlik om ingeligte keuses te maak oor steekproefpunte vir hoër vlak ondersoeke op ‘n sub-plot skaal (of tros/korrel skaal) afhangende van die navorser se keuse om veranderlikheid afkomende van verskillende faktore, in of uit te sluit.Master