Characterising sensory interactions between volatile phenols and other taint-causing compounds in South African red wines

[摘要] ENGLISH ABSTRACT: South African red wine competes for limited market space, and must maintain consistent,excellent quality. One of the important modalities in assessing the quality and typicality of wineis odour perception (Hopfer et al., 2015). Most investigations quantify aroma compounds bychemical/analytical means and compare levels to odour detection thresholds (ODTs) providedby scientific literature. If malodourous compounds are present at concentrations above theirODTs, they may be considered a threat to wine quality as they exhibit odour activity values >1(Louw et al., 2010; Prida & Chatonnet, 2010). A review of the literature reveals that studies inwine can use inappropriate ODTs for work carried out in a new wine matrix, and matrix effectson the activity and perception of a compound are often ignored. To properly scientificallyevaluate the effect of any compound to wine aroma, formal sensory evaluation in the studymatrix is essential (Villamor & Ross, 2013; Perry & Hayes, 2016).Certain volatile phenols (VPs) can contribute to a continuum of smoke-taint related off-flavoursincluding 'burnt', 'bretty', 'smoky', and 'ashy' attributes in wine (Jiranek, 2011; Kennison et al.,2011) at higher levels but are generally accepted as being benign to wine aroma at subthresholdlevels (Boidron et al., 1988; Prida & Chatonnet, 2010). Useful sensory studies on VPsin wine have been carried out (Simpson et al., 1986; Boidron et al., 1988; Chatonnet et al.,1992; Kennison et al., 2008; Petrozziello et al., 2014), but at supra-threshold levels, and theeffect of combinations of subthreshold levels of VPs is not taken into account.The main aims of this research were to characterise the sensory contribution of specific VPs atlow levels to red wine odour, and to assess their effect on the perception of attributes throughinteractions with themselves and other compounds known to be involved in specific wine offflavours.Chapter 2 addresses key knowledge and gaps within the literature around origin,perception and odour detection threshold of VPs associated with specific off-flavours (including'smoke' and 'ashiness'), and previous studies concerning interaction of aroma compoundswhich may have relevance to the current study. Chapter 3 investigated trends within thechemical (GC-MS) and sensory (Descriptive Analysis) results for twelve commercial samples ofsmoke-affected wine. Associations between negative attributes and bushfire events prior toharvest were found. Results also showed that certain sensory effects could have resulted fromcombinations of subthreshold levels of VPs. The need arose to formally test perceptualinteractions at subthreshold levels to see if various effects could be explained.It was thus decided to investigate effects of three VPs: guaiacol, ortho-cresol, 4-ethylphenol,and two compounds associated with certain off-flavours in wine, viz. 3- isobutyl-2-methoxypyrazine (IBMP) and 2,4,6-trichloroanisole (TCA). The work was carried out in partially de-aromatised Shiraz, and it was necessary to establish if ODTs provided in the literature wereappropriate for this matrix. Formal sensory work is known to be complex, time-consuming andexpensive, particularly in establishing ODTs, and thus a pragmatic sensory approach to thework is outlined in Chapter 4.Chapters 5 and 6 address the perceptual effects of combinations of two, three, four and five offflavourcompounds on red wine aroma, which has not been conducted on this scale before. Theresults of this DA sensory study showed olfactory opposition between clean controls, winesspiked with single compounds (generally fruity and sweet-associated), and wines spiked withcomplex combinations of VPs and IBMP (linked to negative attributes). Chapter 7 demonstratesanother pragmatic approach, using projective mapping (PM) with a large sample size (n=18).Comparable results to the DA interaction study for four compounds in red wine were shown.Chapter 8 investigates effects of combinations of two VPs on four cultivars in order to establishwhether there were perceptual olfactory effects that were common to all cultivars, or whetherthe matrices responded differently from an aroma perception perspective, and shows thatsamples spiked with combinations of VPs and IBMP show consistently negative olfactoryattributes that are independent of cultivar.This research contributes to the sensorial and chemical characterization of selected VPs in redwines, and shows that subthreshold levels of VPs in combination with very low levels of IBMPand TCA can lead to olfactory interactions that cause various olfactory effects, some of themnegative. This may help inform winemaking decisions, particularly when dealing with smokeaffectedgrapes, and/or cultivars that naturally have higher levels of methoxypyrazines, likeMerlot and Cabernet Sauvignon. This study also emphasises the importance of understandingeffects of VPs on wine aroma, and escalating awareness and sensitivity to these issues in thewine industry.

[发布机构] Stellenbosch University