Session 4: Managing seasonal production challenges Relationships between harvest time and wine composition in Cabernet Sauvignon Keren Bindon Cristian Varela, Helen Holt, Patricia Williamson, Leigh Francis, James Kennedy, Markus Herderich, Stella Kassara, Christina Linke, Paul Smith
Alcohol in Australian Wine 18 Alcohol content (% v/v) 17 16 15 14 13 12 11 10 Average 14.2% Ethanol 9 8 1984 1988 1992 1996 2000 2004 2008 Year
Why reduce alcohol? Too much! 15% 14% 13% 12% Ethanol content Alcohol abuse Stuck Normal Sluggish Wine & society
Why is alcohol increasing? Climate change? Cultural practices? Hypotheses regarding extending hang time : Decreases the unripe green and vegetal flavours? Consumer preference for fruit-driven, high colour, full-bodied wines? Grape maturity enhances ripe fruit flavour, fullness and colour? Despite the simplicity of the question, data is lacking
Concepts
Palate balance Ethanol Sugar Polysaccharides Tannin Acid Component Physiological Response Body Sweetness Bitterness Astringency Sourness Wine Descriptor Fat Thick Silky Velvety Ripe Hard Green Coarse Unripe
The phenolics sweet spot Various studies have looked at commercial allocation gradings for principal red varieties Anthocyanin and colour density Tannin concentration Skin tannin proportion (% epigallocatechin) Tannin mdp Consumer testing Tells a slightly different story Tannin: 1.3 1.9 g/l Astringency: 2.9 to 3.9 Mean liking score 7.5 6.5 5.5 1.43 1.58 1.18 1.33 1.32 1.92 1.98 2.30 1.18 2.38 1.18 1.84 2.5 3.0 3.5 4.0 4.5 5.0 Astringency (adhesive) intensity Tannin sweet spot (1.3
Factors contributing to variability in phenolics Negative correlation of berry anthocyanins with maximum temperatures in both October and January (Barnuud et al. 2014). Apart from degradation, it could be due to: compression of ripening early season growth high vigour Positive relationship between berry anthocyanins and rainfall between September and October (Barnuud et al. 2014). This confers both strong regional and seasonal effects Seasonal effect appears to outweigh regional or site-specific effects Skin tannin and anthocyanin are correlated
Large seasonal variability exists 2010 2011 2012 Continuous increase Continuous increase Continuous decrease Target wine tannin and colour at 25 Brix Increase to plateau Increase to plateau, then decrease No significant decrease Target wine tannin and colour not reached Accumulation slowed Plateau at 22 Brix then constant Sharp decrease to 22 Brix, then constant Target wine tannin and colour not reached
Flavour and aroma balance Vegetative Fruity Cabernet Sauvignon is described as having: a dichotomy of vegetative and fruity attributes To achieve balance in these attributes is desirable However, in practice, perception of green is considered a negative
Sensory perception of quality S. African study on wine sensory attributes of Cabernet S. wines: 4 vineyard locations three vintages multiple ripeness stages vegetative no correlation with wine quality poor relationship with ripeness grade astringency colour Strong correlation with wine quality by vintage (high variability) Window identified = quality unchanged either side of colour maximum berry fruit acidity fullness spicy Strong correlation with both wine quality and ripeness grade Due to seasonal variability no clear objective measure could be identified (Botes, 2009, Oberholster et al., 2008)
Aroma/flavour profile changes during ripening Observed changes in sensory profile during ripening Cabernet Sauvignon Site/Season 1 Site/Season 2 Site/Season 3 Ripening Fresh Green/ Vegetal Red fruit/ Fresh Green Red fruit Dark fruit Over-ripe Fruit Alcohol concentration from: 11.8 % v/v to 15.5 % v/v
Sensory analysis results 2010 aroma variables 5 harvests, all significant 11.8 to 15.5 v/v alcohol Sewage/Drain Aroma Pungent Overall fruit Aroma 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Dark fruit Aroma Red fruit Aroma H1 H2 H3 H4 H5 Earthy Aroma Fresh green Aroma Cooked vegetable Aroma Bindon et al., 2014, Food Chemistry
Sensory analysis results 2010 flavour variables 5 harvests, all significant 11.8 to 15.5 v/v alcohol Acidity Overall fruit palate 5.0 4.0 3.0 Dark fruit palate H1 H2 H3 H4 H5 2.0 1.0 0.0 Fruit aftertaste Red fruit palate Fresh Green palate Vanilla palate Bindon et al., 2014, Food Chemistry
Sensory analysis results 2010 texture and colour variables 5 harvests, all significant 11.8 to 15.5 v/v alcohol Opacity 5.0 H1 H2 H3 H4 Viscosity 4.0 Red colour H5 3.0 2.0 1.0 Hotness 0.0 Purple colour Viscosity Astringency Bitter Bindon et al., 2014, Food Chemistry
Objective measures Was correlated with higher lower methoxypyrazine and C6 compounds only for palate red fruit red colour fresh green Was associated with higher levels of ester formation (aroma only) purple colour, opacity dark fruit anthocyanin skin tannin polymeric pigments Higher DMS (below threshold) acidity astringency low ph, high TA tannin only for H5 Was associated with lower levels of ester formation pungency hotness ethanol butanol ethanol mannoprotein viscosity
Consumer mean liking scores (n=104) 6.50 6.25 6.00 Cabernet Sauvignon Liking 5.75 5.50 5.25 5.00 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 Alcohol (% Alcohol v/v) A score of 6 represents a significant liking effect A plateau was seen over 13.5% alcohol where liking did not improve
Drivers of consumer preference F2 (24.0 %) H1 H1 H2 Bitter Sewage Cluster 2 33% Cooked veg Hotness TA H5 Earthy Pungent Acidity Ov fruit f G+FDark colour Fruit AT Alcohol Dark fruit Purple colourviscosityvanilla H4 Astringency H3 23% Cluster 3 Total sample Fresh green VA ph Red colour Red fruit Ov fruit Fresh Green f Cluster 1 44% H3 H4 H5 H2 F1 (59.9 %)
The effect of alcohol Grape phenolic measures anthocyanin, skin tannin related to higher wine colour density, polymeric pigment wine opacity, perceived purple colour and consumer liking Over 13.5% alcohol consumer liking did not improve despite increases in colour, dark fruit attributes and viscosity Liking maximum did not necessarily relate to target commercial levels of anthocyanin, wine colour or tannin Some questions raised from this study: Moderating effect of wine alcohol on perception? Dominating effect of pungency or hotness, i.e. unbalanced wine?
The effect of alcohol Heymann et al., 2013 LOW BRIX GRAPES (fortification) fresh green and sourness hotness and viscosity (consistent over seasons) Seasonal differences in effect on sweetness, bitterness, astringency HIGH BRIX GRAPES (diluted) hotness, viscosity, sweetness No effect of any treatment on fruitiness perception
Take-home messages Large seasonal variability in ripening rate and colour development Target objective measures of phenolics may not be reached A transition from fresh green and red fruit to dark fruit may be observed for Cabernet S. but is also season-dependent The transition from red fruit to dark fruit may be related to losses in methoxypyrazine and C6 compounds Viticultural management practices may be modified to reduce methoxypyrazine and enhance colour, rather than rely on hang time
Take-home messages Maximum consumer liking was not related to highest levels of dark fruit, purple colour or viscosity in wine. Alcohol-related hotness and pungency may reduce consumer preference for otherwise positive attributes Given the challenge of seasonal variability, earlier harvests may achieve lower-alcohol wines Although wine style may be affected, the evidence suggests that consumer liking will be maintained
Acknowledgments SARDI team: Mike McCarthy Matt Ayers Jo Pech Research at The AWRI is supported by Australia s Grapegrowers and winemakers through their investment agency the Australian Grape and Wine Authority (AGWA), with matching funds from the Australian Government.