Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Winemaking merlot leeuwen dubourdieu

313 views

Published on

Published in: Entertainment & Humor
  • Be the first to comment

  • Be the first to like this

Winemaking merlot leeuwen dubourdieu

  1. 1. - 27 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)CONTRIBUTION OF GRAPE SKIN AND FERMENTATIONMICROORGANISMS TO THE DEVELOPMENTOF RED- AND BLACK-BERRY AROMAIN MERLOT WINESBénédicte PINEAU1, Jean-Christophe BARBE1, 2, Cornelis Van LEEUWEN1, 2and Denis DUBOURDIEU11: Université Bordeaux Segalen, USC ŒnologieInstitut des Sciences de la Vigne et du Vin, 210 chemin de Leysotte, CS 5000833882 Villenave d’Ornon cedex, France2: Université de Bordeaux, École Nationale d’Ingénieurs des Travaux Agricoles de Bordeaux,1 cours du Général de Gaulle, CS 40201, 33175 Gradignan cedex, France*Corresponding author:jc-barbe@enitab.frAim:TheaimofthisstudywastoelucidatehowaninitiallyneutralMerlotmustresultedinawinewithcharacteristicaromasofred-andblack-berryfruit,focusingontherespectivecontributionsofyeastmetabolismtogetherwith grape juice, pulp, and skins.MethodsandResults:SensoryanalyseswereperformedonMerlotgrapeskinmacerationmodels,basedonobservationsinthewinery.Initialfindingsrevealed that strong fruity nuances appeared during pre-fermentationmaceration. In the maceration models used, the development of aromaof red- and black-berry fruit systematically paralleled the growth of theyeast population. The respective roles of grape skins and yeasts wereinvestigated throughout the alcoholic fermentation of model musts withaddition of Merlot skins or Merlot skin extract in ethanol. The aromaticnuancesrevealedbyalcoholicfermentationinamustalonehadnospecificwhite-, rosé-, or red-wine character. In contrast, wines made bymicrovinificationwithgrapeskinsand/orgrapeskinextractinethanolhadaclear,intensearomaofred-andblack-berry.MicrovinificationwithbothMerlotskinextractandgrapeskinsrevealedthemostintensefruitycharacter.Conclusions: Inodorous skin constituents produced a specific aroma ofred-andblack-berryfruitafteralcoholicfermentationbyyeast.Thephysicalpresenceofgrapeskinsduringfermentationenhancedtheintensityofthefruity nuances obtained.Significanceandimpactofthestudy:Thestudyestablished,forthefirsttime, the existence of inodorous constituents in Merlot grape skins,extractible by ethanol and transformed by yeasts to produce a specificaroma of red- and black-berry fruit in the finished wines.Keywords: red wines, grape skins, red- and black-berry aroma, aromaorigins, precursorsBut:Cetravailaétéeffectuépourtâcherdecomprendrecommentunmoûtissu du cépage Merlot et initialement neutre pouvait donner un vin avecdes arômes caractéristiques de fruits rouges et noirs.Méthodes et Résultats: Des analyses sensorielles ont été effectuées surdesmacérationsdepelliculesdeMerlotréaliséesenfonctiondobservationsau champ. Nous avons, dans un premier temps, montré que des notesfruitéesintensesapparaissaientlorsdelamacérationpréfermentaire.Dansles milieux de macération utilisés, le développement des arômes de typefruits rouges et fruits noirs était lié à laugmentation de la populationlevurienne. Les rôles des constituants pelliculaires ainsi que des levuresontétéétudiésparlebiaisdelafermentationalcooliquedemoûtsmodèlessupplémentés en pellicules de Merlot ou en extrait éthanolique de cespellicules.Lesnuancesaromatiquesrévéléesparlafermentationalcooliquedumoûtseulnétaientenriensimilairesàcellesdesvinsquilssoientblancs,rosésourouges.Aucontraire,lesvinsobtenuslorsdesmicrovinificationsen présence de pellicules ou dextraits pelliculaires présentaient des notesclairesetintensesdefruitsrougesetnoirs.Lesmicrovinificationsréaliséesavecàlafoispelliculesetextraitspelliculairesontrévélélesnotesfruitéesles plus intenses.Conclusions: Des constituants inodores des pellicules conduisent à desnotesspécifiquesdefruitsrougesetnoirsaprèslafermentationalcoolique.La présence physique de pellicules lors de la fermentation augmentenettement lintensité des notes fruitées obtenues.Impact de létude: Ce travail montre pour la première fois lexistencede constituants pelliculaires inodores et extractibles par léthanol quisont transformés par les levures pour conduire à des arômes spécifiquesde fruits rouges et noirs dans le vin obtenu.Motsclés:vinrouge,pellicules,arômesdefruitsrougesetnoirs,originesdes arômes, précurseursAbstract Résumémanuscript received 15th September 2010 - revised manuscript received 4th January 201105-barbe 27/03/11 17:18 Page 27
  2. 2. - 28 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Bénédicte PINEAU et al.INTRODUCTIONWineisproducedfromnumerousgrapevarietieswithawiderangeofaromaticexpressions.Paradoxically,wineswithveryrich,complexaromasaremainlyobtainedfromgrapes and musts with little noticeable aroma (Delfini etal., 2001; Escudero et al., 2004; Lee and Noble, 2006).Theonlyexceptionis«floral»varieties,themostfamousbeingMuscatfirststudiedbyRibéreau-Gayonetal.(1975).Itsunfermentedmusthasaspecificfloralcharacter,directlycorrelatedwithitshighterpeniccontent,whichisrelativelyunaffected by fermentation (Cordonnier and Bayonove,1974;Günataetal.,1985).Incontrast,thevarietalcharacterof Sauvignon blanc and Verdejo grapes is attributed tosome polyfunctional thiols, already present in tinyamounts in grapes, but mainly released fromS-cysteine conjugates during fermentation by yeast(Campoetal.,2005;Tominagaetal.,1996;Tominagaetal.,1998).Whitewinearomasarethusknowntobemainlyrevealed by the fermentation of odourless grape juice.Varietal aroma in red wine is a much more complexissue. Red wine exhibits highly specific fruitycharacteristics, described as red-berry and black-berryfruit, that are not detected in white wines (Pineau et al.,2010).IntheBordeauxregion,Cabernet-Sauvignonwinesoften have a strong blackcurrant aroma, while Merlotwinesaregenerallycharacterizedbycaramelandcherry.Kotseridis and Baumes (2000) identified bothfuraneol®and homofuraneol®as potentially responsiblefor the caramel aroma of Bordeaux wines. Recently,Pineau et al. (2009) demonstrated that a combination offifteen ethyl esters and alkyl acetates was responsible fortheoverallberryfruit-likearomaofredBordeauxwines.Empiricalobservationsinthewineryindicatethatthered- and black-berry characteristics of red wines resultfrom aromatic development throughout alcoholicfermentation. The intensity of these aromas may,moreover, be enhanced by pre-fermentation maceration(Girard et al., 1997; Girard et al., 2001). The productionofthesetypicalred-andblack-berryaromasoccursduringtheentirevinificationprocess,whichmainlydiffersfromwhite-winemaking by the presence of grape skinsthroughouttheprocess.Moreover,skincontacthasclearlybeen shown to enhance the varietal character of somewhite wines. In Muscat grapes, Wilson et al. (1986)showed that monoterpenes were present as both freevolatileformingrapepulpandglycoside-boundinodorousform mainly in grape skins. Traditionally, Muscat wineis made by fermenting only the juice obtained by directcrushing and pressing of harvested grapes. However,many studies showed an increase in the terpene contentof the wine following pre-fermentation maceration(Cabaroglu and Canbas, 2002; Guilloux-Benatier et al.,1998; Ho et al., 1999; Schmidt and Noble, 1983). Thepresence of grape skins throughout the red winevinification process may thus be an important factor inthe aromatic differences between red and white wines.Commontastingprofiles,evenamongwinesobtainedfromredgrapes,featuredifferentexpressions,dependingon the vinification process. Rosés, made from directlypressedgrapes,areoftenquitesimplewinesdescribedasfresh and slightly fruity. Clairet wines, produced in theBordeaux region from grape juice following a short pre-fermentationskincontact,havemoreintensearomas,butnot the complexity of a red wine. « Nouveau » wines,obtainedbycarbonicmacerationofredgrapes,arestronglycharacterized by banana and « boiled-sweet « aromas.Red-wine production, characterized by skin contactthroughout vinification, results in very strong aromareminiscent of red- and black-berry fruits ranging fromfresh to jammy (Pineau et al., 2009).Therefore,themaingoalofourstudywastoelucidatehowaninitiallyneutralmustsuchasaMerlotmustwouldresult in red wine with characteristic fruity aroma of red-and black-berries. Achieving this objective was madepossible through focus on the respective contributions ofMerlot grape juice, pulp, and skins. The specific role ofyeast metabolism was also investigated.MATERIALS AND METHODS1. Chemicals and solventsAnalytical grade chemicals and solvents were used.Solvents were purchased from VWR (Fontenay-sous-bois,France)andallotherchemicalswerepurchasedfromSIGMA-ALDRICH (St Louis, MO, USA).2. GrapesThe red grapes used were Merlot from the 2006harvest. They were sourced from the Château Luchey-Halde vineyard in Pessac-Léognan (Bordeaux, France)and manually harvested at technological maturity (sugarcontent: 176 g/L - total nitrogen content: 130 mg/L -pH: 3.35) from homogenous plots. The grapes wereimmediately frozen or used for pre-fermentationmacerations or vinifications.Italia table grapes were used as a control of a whitegrapevarietyforthemicrovinificationexperiments.Theywereselectedforbeingcommerciallyavailableatthetimewhen the experiments were conducted.a) MustsModel must (pH 3.3), prepared according to Marulloet al. (2004), contained the following components(expressed in g/L): glucose (105), fructose (105), tartaricacid(3),citricacid(0.3),L-malicacid(0.3),MgSO4 (0.2),05-barbe 27/03/11 17:18 Page 28
  3. 3. andKH2PO4 (2).Totalnitrogenwasadjustedto190mg/Lwith0.3g/L(NH4)2SO4 and6.44mL/Laminoacidsolutioncontaining (in g/L): tyrosine (1.4), tryptophan (13.7),isoleucine (2.5), aspartic acid (3.4), glutamic acid (9.2),arginine(28.6),leucine(3.7),threonine(5.8),glycine(1.4),glutamine (38.6), alanine (11.1), valine (3.4), methionine(2.4),phenylalanine(2.9),serine(6),histidine(2.5),lysine(1.3), cysteine (1), NaHCO3 (20), and proline (46.8) inbuffersolution(NaHCO3,2%:20g/L).Mineralsaltswerethen added (mg/L): MnSO4.H2O (4), ZnSO4.7H2O (4),CuSO4.5H2O (1), KI (1), CoCl2.6H2O (0.4), (NH4)6-Mo7O24.4H2O(1),andH3BO3 (1).Variousvitaminswerealsointroduced(mg/L):mesoinositol(300),biotin(0.04),thiamine(1),pyridoxine(1),nicotinicacid(1),pantothenicacid(1),andp-aminobenzoicacid(1).Thefollowingfattyacids were added (mg/L): palmitic acid (1), palmitoleicacid (0.2), stearic acid (3), oleic acid (0.5), linoleic acid(0.5),andlinolenicacid(0.2).Themediumwassterilizedby filtration (cellulose nitrate membrane, 0.45 µm,Millipore, France), supplemented with sulphur dioxide(6g/hL),accordingtostandardwinemakingpractice,andthen inoculated with yeast. Finally, a sterile solution ofanaerobic factors was added to obtain 0.5 mL/L Tween80, 15 mg/L ergosterol and 5 mg/L sodium oleate.Merlotmustwasmadebydestemming,crushing,andpressingthegrapes.Forlaboratoryexperiments,thegrapemust was supplemented with 17 g/L glucose, 17 g/Lfructose, 20 mg/L ammonium sulphate, and 1.61 mg/Laminoacidsolutiontoadjustthesugarandnitrogencontenttothatofthemodelmustdescribedabove.Incontrast,mustsfermentedunderwineryconditionswerenotsupplemented.b) SkinsFrozen Merlot and Italia grapes were defrosted andwashedusingdistilledwater.Grapeskinswereseparatedfrom pulps and seeds by manually pressing each berry,andwerethenwashedanddrained.Onaverage,200and120 g skins were obtained per kg of Merlot and Italiagrapes, respectively.3. Ethanol extraction of grape skinsConstituentswereextractedfrom200gofgrapeskinsusing200mlofdistilledwater/ethanolsolution(1/1:v/v).The extraction was performed in a sterile closedErlenmeyer flask with continuous agitation for 5 days.Theextractwasthenpaper-filteredandconcentratedusinga Rotavapor® with a bath temperature of 20 °C, until theethanol had been eliminated. The extract obtained was avery thick syrup with fresh grape aromas.4. Skin macerationsa) Winery conditionsMerlot grapes from two homogeneous plots in thevineyardweredestemmed,crushed,putintoseparatevats(1 and 2), and supplemented with 6 g/hL SO2. Dry icewas used to reduce the temperature inside the vats to 10-12 °C. This temperature was maintained throughoutmaceration. Macerations were monitored by tasting andlasted for 4 to 5 days.b) Laboratory conditionsTwosolutionswereused:distilledwaterspikedwith4g/Ltartaricacid(solution1)anddilutealcoholsolutionobtained by adding 12 % ethanol and 4 g/L tartaric acidtodistilledwater(solution2).Bothsolutionswereadjustedto pH 3.5 with KOH (10 N). The three maceration testsare presented in table 1. Each sample contained 300 mlof solution and 50 g of Merlot grape skins in a sterile,closedErlenmeyerflask.Macerationtookplaceinadarkroomwherethetemperaturewasmaintainedat11-12°C.- 29 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Table 1 - Laboratory maceration experiments using Merlot skins:solution used, inhibitor added and maceration time.1Solution 1: distilled water spiked with 4 g/L tartaric acid, pH adjusted to 3.5 with KOH (10N); solution 2: dilute alcohol solution obtainedby adding 12% ethanol and 4 g/L tartaric acid to distilled water, pH adjusted to 3.5 with KOH (10N). 2Pimaricin: specific inhibitor of yeastgrowth; penicillin: specific inhibitor of acetic acid bacteria growth; chloramphenicol: specific inhibitor of bacterial growth.05-barbe 27/03/11 17:18 Page 29
  4. 4. 5. MicrovinificationsActiveDryYeast(10g/hL,Saccharomycescerevisiae,F10, Sarco SA, France) was used in all vinificationexperimentstoavoidpossiblearomaticvariationsinducedby different yeast strains. Rojas et al. (2003) showed, inparticular, that indigenous yeasts may contribute to theexpression of a varietal or endemic character in wines.To ensure homogeneity, the non-macerated Merlot mustused for both winery and laboratory microvinificationswastakenfromthesamevat.Moreover,theMerlotskinswere taken from a grape sample from the same harvestusedforroséandredwinemicrovinificationsinthewinery.a) Winery conditionsThe Merlot grapes used for both rosé and red wines,with and without pre-fermentation maceration, weredestemmed, crushed, put into two vats (A and B), andsupplemented with 6 g/hL SO2. Vat A was immediatelyinoculated to make the non-macerated red wine. Vat Bwas cooled to 10-12 °C using dry ice. A 60 l sample ofcooled, non-macerated must was taken from vat B andinoculated immediately to make the non-macerated roséwine. Pre-fermentation maceration continued in vat B ata constant temperature of around 10 °C for 5 days, afterwhich it was inoculated to make the macerated red wine.A60lsampleofcooled,macerated,inoculatedmustfromvat B was fermented to make the macerated rosé wine.Thevattemperaturewasmaintainedat20-22°Cand28-30 °C for the rosé and the red wines, respectively, andalcoholic fermentation was monitored by optical densitymeasurements Once fermentation was completed, i. e.whendensityhaddroppedto0.990,thewineswerefrozenuntil required for sensory analysis.b) Laboratory modelsThe F10 yeast strain was pre-cultured for 24 heuresin the model must or in the non-macerated Merlot mustdiluted with distilled water (1/1: v/v). Fermentation tookplace in 375 ml glass bottles containing 300 ml of mustinoculated with 3.5*106 yeast cells/mL. Adding culturedyeast in exponential growth phase ensured goodfermentation kinetics and prevented indigenous strainsfromtakingovertheselectedyeaststrainduringlaboratorymicrovinifications.Cultureswereincubatedat20°Cwithagitation (50 rpm), and fermentation was monitored bymeasuringweightloss.Whenfermentationwascompleted(stable weight for 48 heures, after 8 to 10 days), thesamples were filtered and immediately tasted or frozenpriortotasting.Themicrovinificationsamplescontainedmust,eitheraloneorsupplementedwithgrapeskinsand/orgrape skin extract prior to inoculation. The initialcomposition of each sample is presented in table 2.6. Microorganism countsTheyeastcellcountwasdeterminedusingthemethoddescribedbyZott etal. (2008).Briefly,thesampleswereplated on YPG-based medium prepared by adding yeastextract (10 g/L), tryptone (pancreatic digest of casein,10 g/L), glucose (20 g/L) and agar (30 g/L) to distilledwater,andadjustingpHto4.5withorthophosphoricacid.- 30 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Bénédicte PINEAU et al.Table 2 - Laboratory microvinification tests:must used and elements added prior to inoculation and fermentation.* Skin-to-juice ratio based on that of Merlot grapes, where skins represent approximately 20% of total weight. A ratio of 1:1 indicates thatthe grape skins added to the must represented 20% of the total weight of the sample or that the grape skin extract added corresponded to aninitial quantity of grape skins equivalent to 20% of the total weight.05-barbe 27/03/11 17:18 Page 30
  5. 5. Sterilized YPG-based medium was supplemented with0.15g/Lbiphenyland0.1g/Lchloramphenicoltoinhibitmould and bacterial growth.Acetic and lactic acid bacteria were counted on aspecific medium prepared by adding grape juice(500ml/L),yeastextract(5g/L),Tween80(1ml/L),andagar (30 g/L) to distilled water. The pH was adjusted to4.5 with orthophosphoric acid and the medium wassterilized.Forspecificlacticacidbacteriacounts,0.1g/Lpimaricin and 12.5 mg/L penicillin were also added tothe medium.Four successive 10-fold dilutions of macerated andnon-macerated must samples were plated (100 µL) induplicate.CellcountsaregivenasCFU(ColonyFormingUnit)/ml and were calculated by averaging the CFUs ofthetwoplatescorrespondingtothelowestdilutionwheresingle colonies were observed.7. Sensory evaluationsThe sensory evaluations were performed undercontrolledroomtemperature(20°C),inindividualbooths,using black Afnor (Association Française des Normes)glasses containing about 40 ml liquid and covered withpetri dishes.a) PanelsThereweretwopanelsofassessors.Panel1consistedof 14 trained and experienced judges – winemakers andresearchers - all members of research laboratories at theBordeaux Faculty of Oenology (France). During regulartastings of wine samples, they showed a good aptitudefor recognising specific red- and black-berry aromas,described as « redcurrant - raspberry - strawberry » and« blackcurrant - blackberry - cherry », respectively. Fruitliqueurs were used as reference samples for thesedescriptors.The three assessors on panel 2 were selected asrepresentative members of panel 1, following the tastingdescribed above. All three were researchers workingspecifically on fruity aromas in wine and were thusconsidered experts in this field.b) Tasting winery must and wineFor the must tasting, the assessors on panel 1 wereasked to describe the dominant aromas they perceivedand to rate them on a 6-point scale labelled very light,light, medium, strong, intense, and very intense. Onlydescriptorsgivenbyamajorityof9judgesoutof12wereretained.For the wine tasting, the assessors on panel 1 wereasked to rate the intensity of the overall red- and black-berry aroma they perceived in the wines on a 10-pointscale ranging from 0 = absent to 10 = very intense.Significant differences among wines and in the overallratingbytheassessorswerestatisticallydeterminedusingone-way analysis of variance (« wine treatment » usedas single factor and aall pair-wise comparisons adjustedfor multiple testing using Tukey’s Honest SignificantDifference at α = 0.05) and Kruskal-Wallis tests,respectively.c) Tasting must and wine from laboratorymicrovinification experimentsThe assessors on panel 2 were asked to describe thedominantaromasinbothmustsandwinesandtoratetheintensity of each aroma perceived on a 6-point scalelabelled very light, light, medium, strong, intense, andvery intense. Assessors were asked to focus on the red-and/or black-berry fruit characteristics of each sample,but were free to use their own descriptors. Responseswerecomparedandonlydescriptorscommontoallthreeassessors were retained.RESULTS AND DISCUSSION1. Reliability of the sensory evaluation resultsThis investigation was based on sensory analyses,which can involve interactions between visual andolfactorystimuli.Thiswasavoidedbyusingblackglassesfor each tasting session. One key finding of Morrotet al.(2001)wasthatwhenawhitewinewascolouredred,thesensorydescriptorsappliedtotheodourofthewinewereconsistentlythosenormallyusedforredratherthanwhitewine.Theassessorsabilitytoperceiveandassessanoverallaroma of red- and black-berry fruit was evaluated byapplyingtheKruskal-Wallistesttothewinetastingresultsobtainedfrompanel1.Thisrevealednodifferenceamongtheassessorsscoring(Hvalue=2.49,associatedP-value= 1, compared to α = 0.05 confidence limit). Threeassessors were selected to form panel 2 based on thesewinetastingresults:theirratingsreflectedtheaverageofthe whole panel, so they were considered representativeassessors of the specific fruity aromas under study.The assessors on panel 2 were also asked to performdescription tasks. As shown by Gawel and Godden(2008),thetasterrepeatabilitycannotbeguaranteed,whenassessing wines for quality. Nevertheless, using thecombined scores of a small team of tasters (three in thatpublication)resultsinmoreconsistentqualityassessments.Thus,thethreeexperiencedassessors,chosenamongthemembers of a laboratory specialized in wine aromaresearch,ensuredthereliabilityofthesampledescriptionsobtained.Moreover,theattributesgeneratedbyconsumers- 31 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)05-barbe 27/03/11 17:18 Page 31
  6. 6. or untrained panellists are known to be more ambiguousand repetitive, but less specific, than those proposed bytrained assessors (Chollet and Valentin, 2001; Cholletand Valentin, 2006; Chollet et al., 2005). This point wasconfirmedbycomparingthedescriptorsgivenbypanel2:all three noted a « fermentation aroma ». When askedtodefinethisterm,theassessorsagreedthatthisdescriptorcovered a blend of « banana », « cider », « cucumber »,« fatty acids », and « higher alcohols ». None of themmentioned berry fruit aromas.2. Fruity character revealed during pre-fermentation macerationVariationsinthearomaticdescriptionoftwodifferentMerlot musts during pre-fermentation maceration in thewinery are presented in table 3. When first put into vats(day 0), the two musts were consistently described asexhibiting light herbaceous characteristics. This wasprobably due to C6 aldehydes and alcohols, well-knownto develop in early stages of the winemaking process(destemming, crushing, maceration, and pressing)(Baumes et al., 1988). These compounds, produced byenzymatic cleavage of polyunsaturated linoleic andlinolenic fatty acids concentrated in the grape skins, arecommonly regarded as pre-fermentation aromacompounds (Cordonnier and Bayonove, 1981; Crouzet,1986;Cordonnier,1989).Ferreiraetal.(1995)foundthatlarge quantities were produced during pre-fermentationskin contact, using Chardonnay.On the contrary, at the end of maceration (day 5), themusts were characterized by fruity aromas, reminiscentof strawberry, raspberry, redcurrant, and blackberry,varying from fresh to jammy and perceived as intense orveryintense.YeastcellcountsintheMerlotmustinvat1ispresentedinfigure1.Atday0ofthemacerationperiod,themusthadatotalyeastpopulationof6.3x103CFU/ml.- 32 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Bénédicte PINEAU et al.Table 3 - Aroma development in two different Merlot musts (vats 1 and 2)during cool pre-fermentation maceration in the winery.Note: sensory assessments were performed by a panel of 14 experienced assessors who were asked to describe the dominant aromas theyperceived in the musts. Only descriptors given by a majority of 9 assessors out of 12 were retained. Perceived intensities were evaluated usinga 6-point scale (very light, light, medium, strong, intense, and very intense).Figure 1 - Total yeast population in a Merlot must during cool pre-fermentation macerationin the winery.05-barbe 27/03/11 17:18 Page 32
  7. 7. There was very little yeast activity during the first twodays,asindicatedbythedecreasingpopulation.Incontrast,startingonthethirddayofmaceration,theyeastpopulationgradually increased to approximately ten times its initialvalue. The pre-fermentation macerations in the winerythus suggested that the fruity aromas and the total yeastpopulation developed simultaneously.Laboratory experiments modelling pre-fermentationmacerationinthewineryinitiallyconsistedofmaceratingMerlotskinsindistilledwater.Theyshowedconsiderablearomadevelopment,aspresentedintable4(Merlotskinsmacerated in distilled water). Initially perceived asherbaceous, reminiscent of grape juice, by the 3rd day,themacerationstartedtorevealraspberryandblackberryaromas. The fruity character increased markedly inintensity on the 6th day, accompanied by gas release. Onthe 7th and 8th days, very intense cider and banananuances were also perceived, described by the panel asfermentation aromas. At the end of maceration, analysisof the microorganisms revealed the presence of yeastsand acetic acid bacteria, with populations of 103-104CFU/mLand10-102CFU/mL,respectively,whereasthey were absent from the initial distilled water (data notshown). The aromatic development and increase in thetotalyeastpopulationwereverysimilartothoseobservedin winery samples. This laboratory model was thusconsideredrepresentativeofpre-fermentationmaceration.Samples macerated with a specific inhibitor of yeastgrowth(Table4,Merlotskinsmaceratedindistilledwaterwith 100 mg/l pimaricin) produced only herbaceousaromas,initiallydescribedasfreshgrapes,thencucumberand hay nuances at the end of maceration time. Thisaromatic development was similar to findings reportedby Delfini et al. (2001). Analysis of the microorganismsconfirmed the absence of yeasts. In contrast, the aceticacid bacteria population was 10 to 100 times larger thanin the models without pimaricin (data not shown). Thiswas attributable to the inhibition of acetic acid bacteriagrowth by yeasts, widely described since the 1980s(Lonvaud-Funeletal.,1988aandb;Wibowoetal.,1985).Moreover,Beelmanetal.(1982)showedthataminoacidassimilationbyyeastinfermentingmustwasveryrapid,leading to nutritional deficiencies that hinder the growthof acetic acid bacteria.- 33 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Table 4 - Daily evolution of the nature and the intensity of the aroma nuances perceivedin Merlot skin macerations in two media: distilled water and distilled water supplementedwith pimaricin to inhibit yeast growth.1Sensory assessments were performed by a panel of three experienced assessors who were asked to describe the dominant aromas theyperceived in the musts. Only descriptors common to all three assessors were retained.2 Perceived intensities were evaluated using a 6-point scale (very light, light, medium, strong, intense, and very intense).05-barbe 27/03/11 17:18 Page 33
  8. 8. Bénédicte PINEAU et al.J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France) - 34 -3. Alcoholic fermentation and the development offruity aromasAsshowninfigure2,roséandredwinesfermentedinthewineryfrombothdirectlypressedandpre-fermentation-macerated must exhibited significant differences in theperceivedintensityoftheoverallaromaofred-andblack-berry.Roséwinemadefromnon-maceratedmusthadtheleastintensearoma,followedbytheroséwinemadewithpre-fermentation-macerated must. Both red wines had amuchmoreintensered-andblack-berryaroma,irrespectiveof any pre-fermentation maceration.There was a remarkable difference in the intensity offruity aromas between the rosé and red wines. The onlydifferencebetweenthewinemakingmethodsusedfortherosé and red wines was that the grape skins remained inthe must throughout alcoholic fermentation for the reds,maintaining continuous maceration. In these four wines,the perceived intensity of the fruity aroma apparentlyreflected the length of maceration before and duringfermentation.AsalreadydiscussedbyDelfinietal.(2001),themajordifficultywastomaintainareproducibleskin-to-juiceratiounderbothwineryandlaboratoryconditions.Severalskin-to-juiceratiosweretestedinthelaboratory(table2),with1:1 corresponding to the ratio in the winery. However, itisonlypossibletocomparemodelswithidenticalskin-to-juice ratios. Nevertheless, the extraction process duringalcoholicfermentationofredgrapesinthewineryismuchmoreintensethaninmicrovinificationmodels.Whilevatsarecommonlypumpedover1to3timesperday,withoutdeterioratingthearomasoftheresultingwine,applyingasimilarprocedureto300mlmicrovinificationsintroducesFinally, no lactic acid bacteria were found in sampleswith or without inhibitor (data not shown), which wasquite surprising, as the literature reports averageconcentrationsaround103 to104 CFU/mLinmust,beforeandduringtheearlypartofalcoholicfermentation.Lacticacid bacteria are nevertheless known to be very sensitivetomediumconditions,particularlypH(Davisetal.,1986;Lafon-Fourcadeetal.,1983).Theirtotalabsencemaybeduetotherestrictiveconditionsinthismacerationmedium.Aspresentedintable5,maceratingskinswithspecificinhibitors of yeast and bacterial growth (test 3A) did notproduceanyfruityaromas.Theabsenceofmicroorganismsattheendofthemacerationwasconfirmedbycellcounts.Consequently, the aromas of red- and black-berry fruitperceived cannot be explained by the presence of aromacompoundsextractedinthemacerationmediumfromthegrapes, as is the case with Muscat (Cordonnier andBayonove, 1974; Günata et al., 1985).Tests 3B and 3D confirmed the results presented intable4.Lacticacidbacteriadidnotgrowduringmaceration,and acetic acid bacteria were inhibited in the presence ofyeast. Furthermore, bacterial populations were onlyassociatedwithherbaceousaromasinmacerationmedia.Inviewofthesefindings,thefruityaromasthatdevelopedin the winery samples cannot be of bacterial origin.Overall, yeasts were the only microorganismssystematicallyassociatedwithfruityaromasinmacerationmedia under both laboratory and winery conditions. Inlaboratory experiments, a considerable increase in theintensity of the fruity aromas was also systematicallycorrelatedwithgasreleaseinthemacerationmedia.Thismayprovideindirectevidenceofyeastmetabolicactivity.Table 5 - Aroma evolution of Merlot skin maceration in distilled water spikedwith 4 g/L tartaric acid, pH adjusted to 3.5 and supplemented with specific microbial growth inhibitors.1 Pimaricin: specific inhibitor of yeast growth; penicillin: specific inhibitor of acetic acid bacteria growth; chloramphenicol: specific inhibitorof bacterial growth.2 Sensory assessments were performed by a panel of three experienced assessors who were asked to describe the dominant aromas theyperceived in the musts. Only descriptors common to all three assessors were retained.3 Perceived intensities were evaluated using a 6-point scale labelled very light, light, medium, strong, intense, and very intense.4 -: absence +: little gas released ++: large amounts of gas released05-barbe 27/03/11 17:18 Page 34
  9. 9. too much oxygen, considerably modifying the aromaticprofileofthewine.Microvinificationmodelswerepartiallyclosedafterinoculationinordertoreleasethegasproducedby fermenting yeasts while preventing oxidation. Tocompensate for the less intense extraction, skins werearbitrarily added to the must at a 2:1 skin-to-juice ratio.Thearomanuancesrevealedbyalcoholicfermentationinamustalonearepresentedintable6.Winesmadefrombothmodelmustandnon-maceratedMerlotmust(tests4Aand4B)exhibitednocharacteristicsofred-orblack-berry.According to the assessors, they generally had« fermentation » or « vinous » aromas, with no specificwhite, rosé-, or red-wine character.In contrast, wines made by microvinification withMerlotskinshadclear,intensered-andblack-berryaromas,reminiscent of « strawberries », « raspberries », and« cherries » (table 6, tests 4C and 4D). The overall aromawas described by the assessors as similar to that of a« nouveau » red wine. Consequently, fruity aromas onlydeveloped when red grape skins were present duringalcoholic fermentation, irrespective of the macerationmedium.Furthermore, as reported above, macerating Merlotskinsindilutealcoholsolutiononlyproducedherbaceousaromas, described as « hay » and « crushed leaves». Theincreaseinethanol,characteristicofalcoholicfermentation,cannotthereforeberesponsibleforrevealingthesespecificfruity aromas.Both physical and chemical effects of skins contactwerefurtherinvestigated.Theresultsintable6showthatwines produced by adding either Italia or Merlot skinspreviously macerated in ethanol to must did not developany red- or black-berry aromas (tests 4E, 4F and 4G).Thisindicatedthatthefruityaromadidnotsimplydevelopwhen yeast fermented in the physical presence of redgrape skin solids.Merlot skin extract in ethanol had only a slight grapejuice-like aroma, with no specific character of red- orblack-berry.Nevertheless,asshownintable6,winesmadefromthisextractbymicrovinificationrevealedfreshred-berry fruit (test 4H). These wines were generallyreminiscent of « Clairet » or « nouveau » red wines. Onthe contrary, wine made from Italia grape skin extract inethanolexhibitednoaromaofred-orblack-berry(table6,test 4I). Finally, microvinification with both Merlot skinextractandgrapeskins,irrespectiveofwhethertheywereMerlotorItaliaskins,revealedmoreintensefruitycharacter(tests4Jand4K).Thesewineswereperceivedbythepanelas very similar to the winery-fermented red wines.The latter results clearly showed, for the first time,that fruity aromas can develop when an aromatically-neutral ethanol extract, made from specific red grapeskins, is fermented. This indirectly demonstrates thatinodorous skin constituents are responsible for thedevelopment of a specific aroma of red- and black-berryafter alcoholic fermentation by yeast. The most intensefruityaromasdevelopedunderconditionssimilartothosein the winery, i. e. the presence of both red grape skinextract and grape skins.CONCLUSIONThe alcoholic fermentation of red grapes was shownto produce characteristic fruity aromas, reminiscent ofred-andblack-berries,parallelingthegrowthoftheyeastpopulation. These fruity aromas were, for the very firsttime, shown to be directly linked to inodorous red grapeskin constituents, which were successfully isolated froman ethanol extract. The yeast activity characteristic ofalcoholic fermentation was then shown to transformthe inodorous precursors into aromatic compounds,revealing clearly perceived red- and black-berry aromasinthewine.Moreover,theintensityofthesefruityaromaswas enhanced by the physical presence of grape skinsduring alcoholic fermentation.- 35 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Figure 2 - Perceived intensity of a specific red- and/or black-berry aroma in Merlot rosé and red winesat the end of alcoholic fermentation.05-barbe 27/03/11 17:18 Page 35
  10. 10. For the first time, these findings shed light on theextremelyimportantquestionoftheoriginsoffruityaromasin red wines. However, the identity of the inodorousconstituents involved remains unknown and furtherexperimentsarerequiredtodeterminetheirchemicalnature.Microvinification provides an excellent tool for indirectobservation of the development of red- and black-berryaromas, coupled with analysis of grape composition.Acknowledgements:TheresearchwasfundedbytheBordeauxWineCouncil. We thank the Château Luchey-Halde in Pessac-Léognan(Bordeaux, France) for supply of the grape samples and for havingmade it possible to conduct the experiments in the winery. We thankKatarina Zott and Isabelle Masneuf-Pomarède for assistance inmicroorganism counts. Finally, we express our gratitude to thewinemakers and the members of the Bordeaux Faculty of Oenologywho performed all of the sensory assessments.REFERENCESBaumes R., Bayonove C., Barillère J.M., Escudier J.L. andCordonnier R., 1988. La macération pelliculaire dans lavinification en blanc. Incidence sur la composante volatiledes moûts. Connaissance Vigne Vin, 32 (3), 209-223.Beelman R.B., Keen R.M. Banner M.J. and King S.W., 1982.Interactions between wine yeast and malolactic bacteriaunder wine conditions. Developments in IndustrialMicrobiol., 23, 107-121.Cabaroglu T. and Canbas A., 2002. Effects of skin-contact onaromatic composition of the white wine of V. vinifera L.cv.MuscatofAlexandriagrowninSouthernAnatolia.ActaAlimentaria, 31, 45-55.CampoE.,CachoJ.andFerreiraV.,2005.Predictionofthewinesensory properties related to grape variety from dynamic-headspacegaschromatography–olfactometrydata.J.Agric.Food Chem., 53, 5682-5690.CholletS.andValentinD.,2001.Impactoftrainingonbeerflavorperception and description: Are trained and untrainedsubjectsreallydifferent?J.SensoryStudies,16,601–618.CholletS.andValentinD.,2006.Impactoftrainingonbeerflavourperception.Cerevisia,BelgianJ.BrewingBiotechnol.,31,189-195.- 36 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)Bénédicte PINEAU et al.Table 6 - Aroma description of wine obtained by laboratory microvinification of model or Merlot musts,with or without grape skins and/or grape skin extract added before inoculation.1Sensory assessments were performed by a panel of three experienced assessors who were asked to describe the dominant aromas theyperceived in the musts. Only descriptors common to all three assessors were retained.2Perceived intensities were evaluated using a 6-point scale (very light, light, medium, strong, intense, and very intense).* Aroma described as banana, cider, cucumber, fatty acids, and higher alcohols.05-barbe 27/03/11 17:18 Page 36
  11. 11. Chollet S., Valentin D. and Abdi H., 2005. Do trained assessorsgeneralize their knowledge to new stimuli? Food QualityPreference, 16, 13-23.Cordonnier R., 1989. Principaux facteurs de la formation decomposés à flaveurs herbacées. Rev. Œnol., 53, 17-20.CordonnierR.andBayonoveC.,1974.C.R.Acad.Sci.Paris,278,387-395.Crouzet J., 1986. Les enzymes et l’arôme du vin. Rev. Fr. Œnol.,102, 42-49.DavisC.R.,WibowoD.,LeeT.H.,andFleetG.H.,1986.Growthandmetabolismoflacticacidbacteriaduringandaftermalo-lactic fermentation of wines at different pH. AppliedEnvironmental Microbiol., 51, 539-545.Delfini C., Cocito C., Bonino M., Schellino R., Gaiaj P., andBaiocchi C., 2001. Definitive evidence for the actualcontribution of yeast in the transformation of neutralprecursorsofgrapearomas.J.Agric. FoodChem.,49(11),5397-5408.Escudero A., Gogorza B., Melus M.A., Ortin N., Cacho J., andFerreira V., 2004. Characterization of the aroma of a winefrom Maccabeo. Key role played by compounds with lowodoractivityvalues.J. Agric.FoodChem.,52,3516-3524.FerreiraB.,HoryC.,BardM.H.,TaisantC.,OlssonA.,andLeFurY.,1995.EffectsofskincontactandsettlingonthelevelofC18: 2, C18: 3 and C6 compounds on burgundyChardonnaymustsandwines.FoodQualityPreference,6,35-41.Gawel R. and Godden P.W., 2008. Evaluation of the consistencyofwinequalityassessmentsfromexpertwinetasters.Aust.J. Grape Wine Res., 14, 1-8.GirardB.,KoppG.T.ReynoldsA.G.andCliffM.,1997.Influenceofvinificationtreatmentsonaromaconstituentsandsensorydescriptors of Pinot noir wines. Am. J. Enology Vitic., 48,198-206.GirardB.,YukselD.,CliffM.A.,DelaquisP.andReynoldsA.G.,2001. Vinification effects on the sensory, colour and GCprofiles of Pinot noir wines from British Columbia. FoodRes. Int., 34, 483-499.Guilloux-BenatierM.,LeFurY.andFeuillatM.,1998.Influenceof fatty acids on the growth of wine microorganismsSaccharomycescerevisiaeandOenococcusoeni.J.Indus.Microbiol. Biotech., 20, 144-149.Günata Z., Bayonove C., Baumes R., and Cordonnier R., 1985.The aroma of grapes. Extraction and determination of freeand glycosidically bound fractions of some grape aromacompounds. J. Chromatography A., 331, 83-90.Ho P., Rogerson F.S.S., Watkins S.J., Silva M.C.M., Hogg T.A.,and Vasconcelos I., 1999. Effect of skin contact andoxygenation of musts on the composition of white portwines. Sci. Alim., 19, 687-699.Kotseridis Y., and Baumes R., 2000. Identification of impactodorants in Bordeaux red grape juice, in the commercialyeast used for its fermentation and in the produced wine.J. Agric. Food Chem., 48 (2), 400-406.Lafon-Lafourcade S., Carre E. and Ribéreau-Gayon P., 1983.Occurenceoflacticacidbacteriaduringthedifferentstagesof vinification and conservation of wines. AppliedEnvironmental Microbiol., 46, 874-880.Lee S.J., and Noble A.C., 2006. Use of partial least squareregressionandmultidimentionalscalingonaromamodelsofCaliforniaChardonnaywines.Am. J. Enol. Vitic.,57(3),363-370.Lonvaud-Funel A., Masclef J.P., Joyeux A. andParaskevopoulos Y., 1988a. Étude des interactions entrelevures et bactéries lactiques dans le moût de raisin.Connaissance Vigne Vin, 22, 11-24.Lonvaud-Funel A., Joyeux C. and Desens C., 1988b). Inhibitionof malolactic fermentation of wines by products of yeastmetabolism. J. Sci. Food Agric., 44, 183-191.Marullo P., Bely M., Masneuf-Pomarède I., Aigle M., andDubourdieu D., 2004. Inheritable nature of enologicalquantitative traits is demonstrated by meiotic segregationof industrial wine yeast strains. FEMS Yeast Research, 4,711-719.Morrot G., Brochet F. and Dubourdieu D., 2001. The color ofodors. Brain and Language, 79, 309-320.Pineau B., Barbe J.-C., Van Leeuwen C. and Dubourdieu D.(2010) Olfactory specificity of red- and black-berry fruitaromas in red wines and contribution to the red Bordeauxwine concept. J. Int. Sci. Vigne Vin 44 (1), 39-49.PineauB.,BarbeJ.-C.,VanLeeuwenC.andDubourdieuD.,2009.Examples of perceptive interactions involved in specific« red-» and « black-berry » aromas in red wines. J. Agric.Food Chem., 57, 9, 3702-3708.Ribéreau-GayonP.,BoidronJ.N.andTerrierA.,1975.AromaofMuscatgrapesvarieties.J. Agric.FoodChem.,23,6,1042-1047.Rojas V., Gil J.V., Piñaga F. and Manzanares P., 2003. Acetateesters formation in wine by mixed cultures in laboratoryfermentations. Int. J. Food Microbiol., 86, 181-188.SchmidtJ.O.andNobleA.C.,1983.Investigationoftheeffectofskin contact time on wine flavor. Am. J. Enol. Vitic., 34,135-138.TominagaT.,DarrietP.andDubourdieuD.,1996.Identificationdel’acétatede3-mercaptohexanol,composéàforteodeurde buis, intervenant dans l’arôme des vins de Sauvignon.Vitis, 35 (4), 207-210.TominagaT.,MuratM.L.andDubourdieuD.,1998.Developmentofamethodforanalyzingthevolatilethiolsinvolvedinthecharacteristic aroma of wines made from Vitis vinifera L.CVSauvignonblanc.J.Agric. FoodChem.,46(3),1044-1048.WibowoD.,EschenbruchR.,DavisC.R.,FleetG.H.andLeeT.H.,1985.Occurrenceandgrowthoflacticacidbacteriainwine.A review. Am. J. Enol. Vitic., 36, 302-313.WilsonB.,StraussC.R.andWilliamsP.J.,1986.Thedistributionoffreeandglycosidally-boundmonoterpenesamongskin,juiceandpulpfractionsofsomewhitewinesvarieties.Am.J. Enol. Vitic., 37, 107-111.Zott K., Miot-Sertier C., Claisse O., Lonvaud-Funel A. andMasneuf-Pomarède I., 2008. Dynamics and diversity ofnon-Saccharomyces yeasts during the early stages inwinemaking. Int. J. Food Microbiol., 125, 197-203.- 37 -J. Int. Sci. Vigne Vin, 2011, 45, n°1, 27-37©Vigne et Vin Publications Internationales (Bordeaux, France)05-barbe 27/03/11 17:18 Page 37

×