1IntroductionAmongallcrops,grapevineisoneofthosewhichinspiredthehighestnumberofbioclimaticanalysisrelatedtoclimatechange(HallandJones,2009;Andersonetal.,2012;Molitoretal.,2014;Koufosetal.,

2014;Dunnetal.,2015).Theterm“terroir” itselfstemsfromthe ideathatwineproduction issteadilyentangledwiththeenvironmental features– includingclimate–whichcharacterizeoriginandgrowingofa

variety.Climatechangeobviouslyposesurgentquestionsaboutthefutureofpremiumwineproductioninmanywine-growingregions.Morethanformostcrops,themaintenanceofthelinkofmanywineswiththeir

typicalproductionareasisanaddedvaluewhichimpactsontheculturaland–ultimately–marketvalueoftheproducts.Hence,itisofconcernforwinegrowers–aswellasforallthewineproductionspin-offs–to

knowwhichclimaticconditionsanareaisgoingtofaceinachangingclimatescenario.

Generallyspeaking,warmeranddrierconditionshavemostlycontributedtoanincreaseofwinequalityinmanyregionsoftheworld(Jonesetal.,2005;Daluetal.,2013).Ifthelatterauthorshighlighted

above all the improvements ofwinequality in the last 50 years, attributing an important role towarmer climate conditions, on the otherhand, the sameauthors claimed that a further increase could take the

bioclimatic physical driver beyond the optimum, leading to a worsening of ratings. For warm vine-growing regions, a temperature increase would probably bring with itself the undesired effect of a dramatic

Originalpapers

Simulationsofquantitativeshiftinbio-climaticindicesintheviticulturalareasofTrentino(ItalianAlps)byanopensourceRpackage

EmanueleEccela,⁎

emanuele.eccel@fmach.it

AlessandraLuciaZollob,c

PaolaMercoglianob,c

RobertoZorerd

aDepartmentofSustainableAgro-ecosystemsandBioresources,ResearchandInnovationCentre,FondazioneEdmundMach(FEM),ViaE.Mach1,38010SanMicheleall’Adige,Italy

bMeteorologyLaboratory,CentroItalianoRicercheAerospaziali(CIRA),ViaMaiorises.n.c.,81043Capua,Italy

cRegionalModelsandGeo-HydrogeologicalImpactsDivision,CentroEuro-MediterraneosuiCambiamentiClimatici,ViaMaiorises.n.c.,81043Capua,Italy

dDepartmentofBiodiversityandMolecularEcology,ResearchandInnovationCentre,FondazioneEdmundMach,ViaE.Mach1,38010SanMicheleall’Adige,Italy

⁎Correspondingauthor.

Abstract

Inconsiderationofthesteadyentanglementsofviticulturewiththeenvironmentalfeatures– includingclimate– it isofconcerntoprojectwhichclimaticconditionsanareaisexpectedtofaceina

changingclimatescenario.Aquantitativeapproachhelps inassessingclassshift inclimateclassification;both“generic”climaticandbioclimatic indiceswereconsidered inthisstudy,namely:Köppen –

Geigertypesandsubtypes,aridity(6indices),and“Internationalvineandwineorganization”(OIV)classification(10indices).AllindiceswereeasilycalculatedbyanopensourceRlibrary(ClimClass),which

alsoincludestoolsforbasepre-processingofweatherseries.FutureclimatescenariosforthisstudywereobtainedusingtheRegionalClimateModel(RCM)COSMO-CLM,employingtwoIPCC’sgreenhouse

gasconcentrations(RCP4.5andRCP8.5),statisticallydownscaledto39weatherstationsinTrentino,aregionintheItalianAlps.Thesimulationsenvisagethenewclimaticprofileofthearea,withashift

towardswarmerandsomewhatdrierconditions.Whilenolimitationtowinegrowingisexpectedintheloweraltitudeareas,newclimaticsuitabilityisprojectedformountainareas,presentlydevotedtoother

soiluses.Thelatteranalysiswastheresultofacalibrationofathermalrequirementindextothepresentsoiluseconditionintheregion.

Keywords:Grapevine;Climatechange;Models;Indices;OIV;Italy

restrictionofthesuitableviticulturalareas(Whiteetal.,2006;HallandJones,2009).Moriondoetal.(2011)suggestedpossiblenegativeeffectsoftemperatureincreaseonwinequalityintheMediterraneanarea

(particularlyItaly).Therecordedadvanceinphenologicaltiming(CaffarraandEccel,2011;Tomasietal.,2011;Fragaetal.,2015)isaclearwarningagainstachangeinthematchbetweenavarietyanditslong-lived

growingcontext,questioningthesubsistenceofthebalancebetweentheripeningprocessandthemeteorologicalconditionsoccurringintherelevantperiod(Mullinsetal.,1992).Theactionofmeteorologicaldrivers

onphysiologicaldevelopmentmayalsoaffectyield(Bindietal.,1996;DuchêneandSchneider,2005;Webbetal.,2008),duetoanexpectedhigher,unsatisfiedwaterdemandinsummer,andtoashorterripening

period.

Spreadingviticulturalareasnorthwardentailstheshortcomingofashiftofproductionareastoregionswhichmaybenewtoeitherspecificvarietiesoreventovinegrowing(Fragaetal.,2016;Jones,2006).

Moriondoetal.(2011)suggestedpossiblenegativeeffectsoftemperatureincreaseonwinequalityintheMediterraneanarea,highlightingthatspreadingthecultivationareastohigheraltitudeswithinthesame

geographiccontextoffersthechancetomeetbetterconditionsthantheonesexpectedinthetraditionalproductionareasatlowerelevations,accordingtotheclimaticscenarios.Trentino,atemperatewine-producing

areaintheItalianAlps,canbenefitofthealtitudinalrangeofmountainregionstofaceaprogressivewarmingofclimaticconditions.

Thisworkaimsatprovidinganin-depthanalysisoftheprojectedchangeinthequantitativebioclimaticindicesthathavebeendevelopedtoassessthelandscapesuitabilityforviticulture.Themodelchainfor

thisanalysiswasbuiltbyapplyingacodewhichcalculatesclimateindicestotheoutputofclimatesimulations.

2Materialsandmethods2.1Geographical,climatologicalandviticulturalfeatures

Trentino isa6212km2 region in thecentral-easternItalianAlps (Fig.1).Altituderangesbetween70ma.s.l. (lowerSarcaValley) to3769m (Cevedalepeak).Trentinoclimate ismostlyoceanic,withsomeareasshowing

featuresoftransitiontoamorecontinental-alpineclimate,coolerandoftendrier,moretypicaloftheinnermountainvalleys.Precipitationismostlydistributedovertwomaxima,intheautumn(main)andinthespring(secondary).

ThecurrentviticulturalareasinTrentinoconsistofabout10,.000ha,distributedfrom70to850ma.s.l.Thelistofthecultivatedgrapevines(Vitisviniferassp.vinifera)includesbothredandwhitevarieties,mainlyfromFrance

(white:Pinotgris,Chardonnay,Sauvignon;rosé:Gewürztraminer;red:Pinotnoir,Merlot,CabernetSauvignon)inadditiontosomelocalcultivars(white:Nosiola;rosé:Schiavagentile;red:Teroldego,Marzemino,Lagrein)andthe

SwissMüllerThurgau,whichissuccessfullycultivateduptothehighestelevations.

Themostupdatedlandusemap(August2003),includingviticulture,hasbeendrawnanditisprovidedbytheurbanplanningserviceoftheAutonomousProvinceofTrento.ThemapisavailableasEsrishapefileattheTrentino

OPENdatawebsite(http://dati.trentino.it/)underCreativeCommonsZero–CC0v1.0Universaltermsofuse.

Fig.1Mapofthe39selectedweatherstations(PATMeteoTrentino).

2.2Observationaldataset39dailymeteorologicalseriesofprecipitationandtemperatureweremadeavailablefromMeteotrentino,themeteorologicalserviceoftheAutonomousProvinceofTrento(Fig.1).Theseriescoveredallthealtitudinalrangeof

theregion,fromthe90ma.s.l.ofTorbole,intheproximityofLakeGarda,to2600ma.s.l.ofCareserdam,anartificialreservoirintheheartofcentralAlps.Theseseriesarethesame (Pleasedelete"arethesame")werechosenaccording

totheirlengthandqualitytobias-correctanddownscalethemodeloutputinOrientgateproject(http://www.orientgateproject.org/)–seefurther.

2.3ClimatesimulationsClimatedataforthisstudywereobtainedusingtheRegionalClimateModel(RCM)COSMO-CLM(Rockeletal.,2008),theclimateversionoftheoperationalnon-hydrostaticmesoscaleweatherforecastmodelCOSMO-LM

developedbytheGermanWeatherService(DWD)andoptimizedatCMCC,REMHIdivision.ThemainfeaturesofthemodelandofitssetuparereportedinBucchignanietal.(2015)andarebrieflyrecalledinTable1.Simulationswere

performedover Italy,employingaspatial resolutionof0.0715° (about8km). InitialandboundaryconditionswereprovidedbyglobalmodelCMCC-CM(Scoccimarroetal.,2011),whoseatmosphericcomponent (ECHAM5)hasa

horizontalresolutionofabout85km.Thetimeperiodconsideredwas1976–2070;specifically,theCMIP5historicalexperiment(basedonhistoricalgreenhousegasconcentrations)wasusedfortheperiod1976–2005,while,forthe

period2006–2070,twodifferentsimulationswereperformed,employingtheIPCCRCP4.5andRCP8.5greenhousegas(GHG)concentrations.RCP4.5scenarioischaracterizedbyastabilizationinGHGemissionsafteryear∼2070,

whileRCP8.5byarapidlyincreasingGHGconcentration.Numericalsimulationswereperformedemploying512coresonaclusterof30IBMP575nodes,installedatCMCCwithanaggregatepeakpowerof18TFlops.Theelapsed

timerequiredtosimulateoneclimatologicalyearwasabout20h.

Table1MainfeaturesofCOSMO-CLMsimulationsetup.

COSMO-CLMmodelversion 4.8_clm13

INT2LMmodelversion 1.10_clm2

Horizontaldiscretization Second-orderdifferencetechnique

Horizontalresolution 0.0715°

Computationalmesh 224×230gridpoints

Timestep 40s

Timeintegrationscheme Third-orderRunge–Kutta

Convectionscheme Tiedtke

Numberofverticallevels 40

Numberofsoillevels 7

Soilmodel TERRA_ML

Frequencyupdateboundaryconditions 6h

The simulations have been deeply analysed in Bucchignani et al. (2015) and – in terms of extreme events – in Zollo et al. (2015a), showing a general good agreement with observed data. Furthermore, as explained in

Bucchignanietal.(2015),theresultsofthesesimulationsareconsistent,intermsofbothmeanbiasandprojections,withthoseobtainedfromtheEURO-CORDEXensembledata(Giorgietal.,2009),whichrepresentthe‘state-of-the-

art’ofregionalclimatesimulationsoverEuropeat0.11°ofhorizontalresolution.AlthoughRCMsarepowerfultoolsfordescribingsmallscaleclimateconditions,theirdirectuseinimpactstudiesisstillchallengingsincetheyare

commonlybiased.Inordertoprovidecorrectedclimatescenarios,simulateddatawerebias-correctedconsideringastatisticaldownscalingapproachbasedonquantilemapping(Zolloetal.,2015b).Specifically,simulateddatawere

downscaledatstationlevelusingdatafromtheavailablemeteorologicalstationsdescribedinSection2.2.

2.4ClimaticandbioclimaticindicesAllindiceswerecalculatedbytheRpackageClimClass2.0.1(Ecceletal.,2015),freelyavailableatRCranrepository(https://cran.r-project.org/).Thepackagemakesuseofmonthlyordailyseries,accordingtotheclassof

indicestobecalculated:monthlyforKöppen–Geigerclimateclassification,foraridityindices,andforRiou’sdroughtindex(oneoftheOIVindices),dailyforallbioclimaticindicesexceptRiou’s.Thecodescancarryoutgapfilling

operations,ifrequired,fortheinstrumentalseries.Meandailyandmonthlytemperatureswerecalculatedasaveragesbetweencorrespondingminimaandmaxima,availableeitherasrecordsorclimateprojections,accordingly.

2.4.1Climateclassification

(i) Köppen–Geigerclimatetypesandsub-types.TheclassicalclimateclassificationfollowsthedescriptiongiveninTrewarthaandLyle(1980).Therearefivemainclimatetypes(Ato–pleasereplaceFwithEF),andmoresub-types,accordingtothegeneralprecipitation

regimeandtoitsdistributionalongtheyear.

(ii) Aridityindices.Fiveannualindicesarecalculated:Ia–DeMartonne,Im–Thornthwaite,Q–Emberger,R–Lang,Io–Rivas-MartinezandRivas-Sáenz(detailsinAppendix1–SupplementaryMaterial,andinClimClassusermanual).

2.4.2BioclimaticindicesforviticultureOIV(2012)selectedandsuggestedalistof10indicesusefulforassessingtheclimaticclassificationofwine-growingregions.Inthepresentwork,weusedareducedlistoftheproposedindices(descriptioninAppendix1)

groupedbyphysiologicalmeaning.GrowingSeasonaverageTemperature(GST),WinklerIndex(WI),orGrowingDegreeDays(GDD),BiologicallyEffectiveDegreeDays(BEDD),HuglinheliothermicIndex(HI)areindicatorsoftheheat

requirements,necessary forapropervinedevelopmentandgraperipening.These four indiceswere testedagainst theactualgrapevinepresence in thearea, toassessa roughviticulturalareaaccording to the future,expected

conditions, as simulated by the climatic model. For the assessment of exposure of grapevine to biotic and abiotic adversities we considered the minimum temperature during rest period (Tn_rest), minimum temperature during

vegetativeperiod(Tn_veg),andmaximumtemperatureduringvegetativeperiod(Tx_veg).

ToniettoandCarbonneau(2004)proposedtoclassifyworldviticulturalareasaccordingtothreebioclimaticindices,amongthoselistedabove:Huglin’sheliothermicIndex(HI),CoolnightIndex(CI),andRiou’sDroughtIndex

(DI).Alltheseindices(seeAppendix1,Supplementarymaterial)werecalculatedtoassessclimate-drivenchanges.

2.5AssessmentofchangesintheviticulturalareaThepotentialenlargementoftheviticulturalareas,withrespecttoclimatechangescenariosandgeo-morphologicalfeatures,hasbeenassessedusingopensourceGISsoftware(QGISGeographicInformationSystemver.2.8.2,

QGISDevelopmentTeam),ahighresolutiondigitalelevationmodel(10m),andfinallyalandusemap,bothprovidedbythe“P.A.T.SistemaInformativoAmbienteeTerritorio”serviceoftheProvinceofTrento.

3Results3.1SimulatedclimatechangesinTrentino

Thebias-correctedanddownscaledclimatedatadescribedinSection2.3wereanalysedintermsofclimatechangesignaloverTrentino,consideringtheanomalybetweenthemeanclimatologyinthereferencepastperiodand

intwodifferentfutureperiods:2021–2050and2041–2070.

Asconcernsprecipitation, theprojectedclimatechangesignal indicatesageneraldecreaseof total rainfallandnumberof rainydays; incontrast,an increaseofmeanprecipitationamountonwetdaysandprecipitation

maximumisprojected.Furthermore,anincreaseinthenumberofconsecutivedrydaysandadecreaseinthenumberofconsecutivewetdaysisexpected.Ingeneral,comparisonofresultsobtainedwithscenariosRCP4.5andRCP8.5

indicatesthatthesignoftheclimatechangesignalgenerallyremainsthesameforbothscenarios,buttheintensityofthesignalisgreaterforRCP8.5.

For temperature, theclimatechangesignal indicatesasignificant increase.Specifically, this increase ishigher for themoredistant futureperiod (2041–2070), ismoreaccentuated for scenarioRCP8.5, andaffectsboth

minimumandmaximumtemperature.Inlinewiththisincreaseintemperature,thenumbersoffrostandicedays(dayswithTmin<0°CandTmax<0°C,respectively)showadecreaseandthenumberofsummerdays(dayswith

Tmax>25°C)anincrease.Forthevegetativeseason(ApriltoOctober),themeanoftheexpectedincreaseonallstationsis1.4–1.5°Cforthetimewindow2021–2050,whilefor2041–2070theextentoftheincreaseandtherangeof

valuesaremuchhigher:2.5to–3.0°C.

3.2SimulatedchangesinclimateclassificationofviticulturalareasIngeneral,theareaswhichare–orwouldeventuallybecome–suitableforvinegrowinginTrentino,canbeascribedtooneclimatetypeaccordingtoKöppen–Geiger(TrewarthaandLyle,1980):“C”(mildtemperate,rainy)–

Table2.Thesub-typeisalways“f”,whichdenotesabsenceofatruedryseason.Thefurthersub-typeofCfcanbe“a”,“b”,or“c”accordingtosummertemperature,“hot”,“cool”,or“coolandshort”,respectively;theshiftfromone

sub-typetotheotheroccurswithinatransitionrangebetweenabout250and500ma.s.l.TheotherstationsinTrentinobelongto“microthermal”climatebeltD,withonealpinestationinthe“polar”typeET(tundra-likeclimate).In

thefirstfuturetimewindow(2021–2050)atrendisevidenttotheincreaseofstationsofsubtype“a”,furtherincreasinginthesecondtimestep(2041–2070).

Table2Köppen–Geigerclimateclassificationforallstationsaccordingtoseveraltimewindowsandscenarios.Tablecellscontainthenumberofstationsineachclass.

Period&RCPscenario ′76–′05 ′21–′50 ′41–′70

4.5 8.5 4.5 8.5

Cfa 6 11 11 15 17

Cfb 24 21 21 19 20

Cfc 1 0 3 1 0

Dfb 1 0 1 3 1

Dfc 6 6 2 1 1

ET 1 1 1 0 0

Theanalysisofaridityclassesismorecomplex,involvinganumberofdifferentindices(Table3).Asageneralrule,asexpected,theclimatechangeleadstoashiftofanumberofsitestowardsdrierclasses,fromthegenerally

humidorveryhumidcategoriestogenerallysub-humidclasses,withevensomecasesofdryclassesinthefarthesttimewindow,mostlylimitedtoscenarioRCP8.5.

Table3Aridityindicesclassificationforallstationsaccordingtoseveraltimewindowsandscenarios.Tablecellscontainthenumberofstationsineachclass.SeeClimClassmanualforindexformulationsand

classificationtables.

Period&RCPscenario ′76–′05 ′21–′50 ′41–′70 PeriodRCPscenario ′76–′05 ′21–′50 ′41–′70

4.5 8.5 4.5 8.5 4.5 8.5 4.5 8.5

Ia Io

Dry 0 0 0 0 0 Ultrahyperarid 0 0 0 0 0

Semidry 0 0 0 0 0 Hyperarid 0 0 0 0 0

Mediterranean 0 0 0 0 0 Arid 0 0 0 0 0

Semihumid 0 0 0 0 0 Semiarid 0 0 0 0 0

Humid 0 0 0 0 2 Dry 0 0 0 0 0

Veryhumid.a 18 21 18 22 24 Subhumid 6 12 11 14 16

Veryhumid.b 21 18 21 17 13 Humid 25 21 20 22 22

Hyperhumid 8 6 8 3 1

Q Ultrahyperhumid 0 0 0 0 0

Desert 0 0 0 0 0

Arid 0 0 0 0 0 Im

Semiarid 0 0 0 0 0 Arid(E) 0 0 0 0 0

Subhumid 3 9 1 9 9 Semiarid(D) 0 0 0 0 0

Humid 36 30 38 30 30 Drysub-humid(C1) 0 1 0 1 2

Moistsub-hum.(C2) 2 11 2 8 12

R Humid(B1) 11 4 11 8 4

Steppe 1 0 0 0 0 Humid(B2) 4 3 4 2 7

Semiarid 0 0 0 0 5 Humid(B3) 1 7 1 8 4

Temper,warm 11 16 15 19 20 Humid(B4) 8 4 8 3 3

Temper.humid 15 13 14 12 7 Perhumid(A) 13 9 13 9 7

Humid 12 10 10 8 7

Moreindetail,Thornthwaitemoistureindex(Im),whichoffersahigherdegreeofdetail,showsageneralshifttowardslesshumidclasses,sothatinthefarthesttimewindowthemostpopulatedclasswouldbe“moistsub-

humid”,insteadof“humid–B1”,mostfrequentinthepresenttime;oneortwostationsreachthe“drysubhumid”class.Rivas-MartinezandRivas-Sáenz’sombrothermicindexattainsthesamequalitativeresult,almostequallysharing

themajorityofpresentvine-growingareabetweenthehumidandsub-humidtypes.

Inscenario′41–′70,RCP8.5,thenumberofsub-humidstationsisincreasedbyafactorof2.7,withrespecttotheinitialconditions;the“hyperhumid”classisreducedtojustonesite,whereasatthebeginningthestationsin

thisclassweresix.

ThesimulationsofEmberger’sQfactoralsoleadtoashiftfromthehumidtothesub-humidtype(inthenearesttimewindowonlyforRCP4.5scenario–amorehomogeneoussignalisdetectableforthesecondtimewindow).

Lang’spluviometricfactorRshowednegligiblechangesuntil′41–′70RCP8.5scenario;inthiscase,fivesiteswouldfallintoa“semiarid”type.AninterestingfeatureisthatsimulationsforQandImindices,inthe2021–2050RCP8.5

case,yieldeithernosignificantchangesorevenastrengtheningofhumidtypes.Finally,DeMartonneindex(Ia)showsminorchangesinthesimulations,mainlyaschangeswithinthetwosubclassesof“veryhumid”type.

Aridityindicesshowedagenerallygoodcollinearity,insomecaseswiththeexceptionofthelowervalues(driersites)andelevations.Inordertocreateanindexthatsummonsupallclassifications,irrespectiveoftheunitof

eachindex,rankshavebeencalculatedforeachstationandindex,andthemedianofrankshasbeentakenasasynthesisindex.Averystrong(p<0.001)negativecorrelationofariditywithelevationcanbeinferred(Fig.2),yielding

theexpectedresultoflesshumidclassesatthelowestelevationsites.

3.3SimulatedchangesinviticulturalindicesThemostcommonviticulturalbioclimaticindiceswereconsidered–seeAppendix1,Supplementarymaterial,fordefinitions–sevenbyOIV(2012)(BEDD,GST,WI,HI,Tn_rest,Tn_veg,andTx_veg),andthreesuggestedby

ToniettoandCarbonneau(2004).ResultsaresummarizedinTable4.

Fig.2Aridity(medianofranksofsixindices)vs.elevationforstationsinviticulturalareasforperiod1976–2005.Lowervaluesreflecthigheraridity.

Table4ClassificationaccordingtoOIVbioclimaticindicesforallstationsaccordingtoseveraltimewindowsandscenarios.Tablecellscontainthenumberofstationsineachclass.SeeAppendix1,Supplementary

material,forindexformulations.

Period&RCPscenario ′76–′05 ′21–′50 ′41–′70 PeriodRCPscenario ′76–′05 ′21–′50 ′41–′70

4.5 8.5 4.5 8.5 4.5 8.5 4.5 8.5

Tm_veg WI

Toocool<13°C

18 11 11 9 8 Toocool<850

19 10 10 9 8

Cool13–15°C

10 11 11 6 3 RegionIa850–1111

8 9 9 2 2

Intermediate15–17°C

6 6 6 12 12 RegionIb1111–1389

4 8 8 11 9

Warm17–19°C

5 8 7 5 8 RegionII1389–1667

5 4 4 5 8

Hot19–21°C

0 3 4 7 5 RegionIII1667–1944

3 4 4 4 3

Veryhot21–24°C

0 0 0 0 3 RegionIV1944–2222

0 4 4 5 4

Toohot>24°C

0 0 0 0 0 RegionV2222–2700

0 0 0 3 5

Toohot>2700

0 0 0 0 0

HI

Toocool<1200

13 9 9 8 8 BEDD

Verycool1200–1500

11 4 4 2 1 <1000 27 19 19 11 11

Cool1500–1800

4 11 11 9 4 1000–1200 4 8 8 12 11

Temperate1800–2100

5 4 4 8 11 1200–1400 6 5 5 6 6

Warmtemper2100–2400

6 5 5 3 4 1400–1600 2 7 7 8 8

Warm2400–2700

0 6 6 6 5 1600–1800 0 0 0 2 3

Verywarm2700–3000

0 0 0 3 5 1800–2000 0 0 0 0 0

Toohot>3000

0 0 0 0 1 >2000 0 0 0 0 0

AccordingtoGladstones’biologicaleffectivedegreedaysindex(BEDD),alargepartoflandinTrentinodoesnotmeettheminimumthermalrequirementforwinegrowing(lessthan1000BEDD).However,roughlyhalfof

weatherstationswhicharepresently inviticulturalareasfall inthe1200–1400BEDDclass;a fewsites fall inawarmerclass,whileothers fit in thecolderclasses.Thesimulatedprogressivewarmingreducesthenumberof the

selectedsites,currentlyunsuitableforviticulturalusebecausetoocool,tolessthanhalfoftheinitialextentinthe′41–′70timewindow(Fig.3).

ThecorrelationbetweenBEDDandelevationcanbeconsideredlinearbelow1800ma.s.l.andthereforeforeachscenariotheupperlimitforviticulturewasassessedusingthelinearregression’scoefficients(Table5)anda

thresholdvalueof1000BiologicallyEffectiveDegreeDays,whichisconsideredsufficienttofulfilthethermalrequirements,neededtocompleteboththevegetativeandreproductivecycleofgrapevine.

Table5LinearregressionsbetweenBEDDandelevationforcurrentandfutureconditions,basedonclimatechangescenarios.Theupperlimitsforviticulturehavebeencalculatedusingtheregressioncoefficients

(slopeandintercept)ofeachscenarioandthelowestvalueofBEDD(1000).Pearsoncoefficientsandthenumberofweatherstations(nr.ofsites)fallingintothesuitableareaarealsoreported.

Ref.period 1976–2005 2021–2050 2041–2070

Scenario Current RCP4.5 RCP8.5 RCP4.5 RCP8.5

Slope −0.64 −0.64 −0.65 −0.63 −0.61

Intercept 1478 1640 1654 1736 1760

Pearson −0.98 −0.98 −0.98 −0.98 −0.98

Upperlimit(ma.s.l.) 750 992 1000 1161 1243

Nr.ofsites 13 22 22 26 28

Huglin’sindex(HI)assessesthemainclassshiftsfrom“temperate”or“cool”classestothewarmeronesinthefarthestperiod,thankstotheincreasedsuitablearea,butwithagrowingnumberofcasesinthewarmerclasses

andonesiteeveninthe“toowarm”class.

Asfarasvegetativeseasonindex(GST)isconcerned,the“intermediate”class(15–17°C)remainsforallscenariosthemostpopulatedinthepresentviticulturalarea,butwithatrendofaprogressivedepopulationofthecolder

classes,andtheconsequentincreaseofmembersinthewarmerones,withafewsitesreachingtheclass“veryhot”;theareawhichistoocoolforgrapevineundergoesareductionofthesamemagnitudeasthatpredictedbyBEDD.

SimilarresultsareobtainedfromtheanalysisofWinkler’sindex(WI),whichishighlycorrelatedtoGST,beingbothmeasuresofmeanthermalconditionsduringgrowingseason(seeformulationinAppendix1,Supplementary

material).Classesof“RegionIa”arestronglyreducedinthe ′41–′70RCP8.5scenario.“RegionIV”,todaynotrepresentedinthepotentiallyviticulturalareasoftheregion,beginstobepopulatedbysomesites inthe ′21–′50time

window,andeventhethermalhigherclassof“RegionV”,thelastbeforetheclassdefinedas“toohot”,hassomesitesfittingin.

Fig.3CorrelationsbetweenBiologicallyEffectiveDegreeDaysandelevation,accordingtocurrentandfutureclimaticconditions,basedonclimatechangescenarios.Toocoolareas(lessthan1000BEDD)arenotconsideredsuitableforviticulture.

Theanalysisofindicesofextremetemperature(Table6)hasentailedfixingthresholdstoassesstherecurrenceofundesirableordangerousevents;indeed,nostandardswereproposedforsuchindices.Thevalueshavebeen

chosenalsotakingintoaccounttheirmodelledoccurrenceinthepresentconditions,andconsideringtheirabilityintrackingchangesinthedifferenttimewindowsandscenarios,asfollows.

Table6ClassificationaccordingtoOIVbioclimaticindicesforextremetemperaturesforallstationsaccordingtoseveraltimewindowsandscenarios.Tablecellscontainthenumberofstationsineachclass.See

Appendix1,Supplementarymaterial,forindexformulations.

Period&RCPscenario ′76–′05 ′21–′50 ′41–′70

4.5 8.5 4.5 8.5

Tn_rest

Below−15°C 30 30 16 18 10

Above−15°C 9 9 23 21 29

Tn_veg

Below−2°C 39 37 36 36 36

Above−2°C 0 2 3 3 3

Tx_veg

Below35°C 34 26 26 21 15

Above35°C 5 13 13 18 24

Tn_rest.Grapevineduringdormancycantoleratetemperaturesaslowas−15°Cto−25°C,accordingtosources(Düring,1997;Lisek,2009).IntheviticulturalareasofTrentino,thelattertemperaturewasneverattained,

whilesomedamagescouldalreadyoccuratminimumair temperatureofabout−15°C(Zulinietal.,2010).Thiswas the thresholdchosen forTn_rest, calculatedas theminimumreached ineach30-yearperiod.Theanalysisof

simulationsstartswithalargemajorityofstationswhereextremelylowtemperaturesoccuratleastoncein30years,andthesameconditionismaintainedfor′21–′50RCP4.5.Thefollowingscenariosshowaprogressivedecreaseof

thenumberofstationswhereextremelylowtemperaturesareattained,andin′41–′70RCP8.5theratiooftheareasbelowandabovethethresholdisalmostreversed.

Tn_veg.Giventhatthe“vegetationperiod”accordingtotheindexissettostartfromApril,aratherearlyperiodforsprouting,atemperatureof−2°C,generallyacceptedasadamagethresholdforreproductivebuds,wasset

asadiscriminatingvalueforfrostepisodesduringthesensitiveperiod,calculatedastheminimumreachedineach30-yearperiod.Thetrendforthisindexshowsonlyminorchangesalongthetimelineandtheseverityofscenarios,

withamodestdecreaseinthenumberofstationswherefrostduringthevegetativeperiodisstillsimulatedtooccur.

Tx_veg.Areasonablethresholdforphysiologicalimpairmentingrapevineis35°C(HunterandBonnardot,2011);owingtoafrequentattainmentofthisthresholdinfullsummer,themedianofeach30yearmaximumvalues

wasconsideredabetterindicatorfortheoccurrenceofhotsummerconditions.Asreferencetemperaturesprogressivelygrowaslongasscenariosbecomemoresevere,thepercentageofserieswherethe35°Cthresholdistrespassed

risesprogressively,becomingthemajorityin′41–′70RCP8.5,whentheviticulturalareaismostlyexpectedtofallinthisclass.

Ingeneral,alltheselectedindicesshowastrongdependencewithelevation,whichismaintainedinfuturescenarioswithsimilarfeaturesfor“meantemperature”indices(BEDD,GST,HI,WI).Thisrelationshipisrepresented

inFig.3forBEDD,theindexchosentoassessthearealextension.

Tonietto–Carbonneau’ssyntheticclassificationmakesuseofthreeindices,asexplainedinSection2(Table7).OneisaslightsimplificationofHuglin’sindex(HI)classification,whoseresultshavealreadybeendiscussed;

accordingtothisclassification,in′41–′70RCP8.5thefullrangeofHIclassesarepresentinTrentino.Thesecondisthe“Coolnightindex”;mostoftheseriesstillbelongtothecoolestclass,butasthereferencescenariogetswarmer,

anincreasingnumberofsitesareexpectedtofallintoa“negative”coolnessclass(CI−1),highlightingtheshifttothemacro-categoryof“warm”viticulturalclimatesforthosesites,wherenighttemperatureintheproximityofharvest

aregenerallyhigherthandesirable.TheRiou’sdroughtindex(DI)showsaprogressiveappearanceofsiteswhere,onaverage,moderatelydryconditionstakeplaceinthelastmonthofripening(September),whereas,atpresent,moist

conditionsprevailforallsites.

Table7ClassificationaccordingtoToniettoandCarbonneau’sbioclimaticindicesforallstationsaccordingtoseveraltimewindowsandscenarios.Tablecellscontainthenumberofstationsineachclass.Seetextfor

explanationsonindices.

Period&RCPscenario ′76–′05 ′21–′50 ′41–′70 Period&RCPscenario ′76–′05 ′21–′50 ′41–′70

4.5 8.5 4.5 8.5 4.5 8.5 4.5 8.5

CI−2 0 0 0 0 0 HI−3 24 13 13 10 9

CI−1 1 1 3 4 6 HI−2 4 11 11 9 4

CI+1 4 5 4 5 8 HI−1 5 4 4 8 11

CI+2 34 33 32 30 25 HI+1 6 5 5 3 4

HI+2 0 6 6 9 10

DI−2 24 10 19 14 9 HI+3 0 0 0 0 1

DI−1 15 19 19 20 15

DI+1 0 10 1 5 15

DI+2 0 0 0 0 0

3.4SimulatedchangesintheviticulturalareaAccordingtotheclimatechangescenarios,thecurrentviticulturalareawillpresenthigherthermalavailabilityleadingtoearlyphenologicalevents(JonesandDavis,2000),animprovementofgrapequalityforlate-ripening

varieties,andearlyharvest,particularlyforsparklingwineproduction.Moreover,apotentialextensionofthesuitableareaforviticulturetohigherregionsisexpected,althoughtherealgrowthpotentialshouldcomplywithother

importantconstraints(Table8),suchasgentleslope,favourableorientation,largeavailabilityofpotentialdailymeansun-hours(AvgSunHursVeg)andcumulatedglobalsolarradiation(CumGloRadVeg)duringthevegetativeperiod.

Finally,newpotentialareasshouldbelargeandcontiguousenoughtojustifyandsustainthecostsforvineyardestablishment.Aminimumsurfaceof1hawasset.

Table8Criteriaappliedtospatialqueries,toassessthelandscapesuitabilityforviticulture,basedoncurrentclimaticconditions,climatechangescenarios(RCP4.5and8.5),andfavourabletopographicconditions

fortheestablishmentofvineyards.AvgSunHursVeg:potentialdailymeansun-hours.CumGloRadVeg:cumulatedglobalsolarradiation.

Ref.period Scenario Elevation(ma.s.l.) Slope(%) Aspect(DEGfromNorth=0CW) AvgSunHoursVeg(h) CumGloRadVeg(kWhm−2)

1976–2005 Current <=750 <=15 >=67.5AND<=292.5 >=10 >=850

2021–2050 RCP4.5 >750AND<=992

RCP8.5 >992AND<=1000

2041–2070 RCP4.5 >1000AND<=1161

RCP8.5 >1161AND<=1243

Accordingtothecurrentclimaticconditionsandfutureclimatechangescenarios,thesuitableareasforviticulture,asaresultofthespatialqueries(Table8,Fig.4),presenttodatethreemainlanduses:annualandperennial

crops (mainlygrapevine,apple,berries,horticulture),pasturesandmeadows,andurbanorcommercialareas.Thepercentagesofnewavailablesoil foreachscenario, fromthecurrentclimaticconditions to thewarmestclimate

changescenario(′41–′70RCP8.5)aresummarizedinFig.5.

4DiscussionThe climate simulations performed according to RCP 4.5 and RCP 8.5 greenhouse gas concentrations depict a future climate scenario characterized by higher temperature, especially if the RCP 8.5 is

considered,andageneralreductionofprecipitationandincreaseinthenumber (Pleaseinsert"of")consecutivedrydays.FutureclimatescenariosconsideredinthisstudywereobtainedbyasingleRegionalClimate

Model(COSMO-CLM).Inordertoassessuncertainty,assuggestedbyIPCC,thereliabilityofprojectionsmightbeimprovedbymulti-modelensembles,weightingsingleelementswithmodelskillmeasures.However,

thehugecomputationalcosts required toperformhigh-resolutionsimulationswithmodelensembles representa limit, leaving this solution to futureworks. Indeed,COSMO-CLMresultshavebeenproved tobe

consistentwiththe‘state-of-the-art’ofregionalclimatesimulations(Bucchignanietal.,2015;Zolloetal.,2015a).

Thewidevarietyofclimatesconsideredinthisstudy,fromthe90mlowervalleybottomtohigh-mountainsites,coversthefullrangeofpresentandfuturepossibleviticulturalareas.

Ingeneral,theexpectedeffectsofwarmeranddrierconditionswouldleadtobetterripeningandsanitarystatusofthegrapesofredvarieties,increasingtheirqualityrating(Grifonietal.,2006),andtoearlier

harvestofwhitevarieties,particularlyforsparklingwineproduction,andapossibleenlargementoftheviticultureareatohigherelevation.Onlyafewwhitegrapevinevarietiesgrownintheregion,suchasMüller

Fig.4ResultsofGISspatialqueries,usedtoassessthelandscapesuitabilityforviticultureinTrentino,accordingtocurrentclimaticconditionsandfuturescenarios.

Fig.5Classification(%)ofnew,incrementalavailablesoil(withrespecttothepreviousscenario/time),accordingtothedifferentclimaticscenariosandreferenceperiods.

Thurgau,areparticularlyaffectedbywarmconditionsandnewplantingarerecommendedrightnowinthecoolerareas(above450ma.s.l.),tofulfiloptimalripening;suchvarietiesarelessresilientandaregoingto

facesomelimitationsinthelowerareas,wheretemperatureisexpectedtoattaintoohighvalues.

Besideaprobablegeneralimprovementofthegrowingconditionsforgrapevine,asawhole,thetypicalthermalregimesoftheareaswhicharepresentlyoccupiedbyvineyardsareexpectedtoshifttowards

differentbioclimaticclassifications,butonlypartiallyisthisfeaturegoingtoaffectthechoiceofthegrapevinevarieties,thankstotheirgeneralresiliencetowarmerconditions.

Extremewarmanddryconditionsarelikelytooccurmoreofteninthefuture,leadingtoahighernumberofundesirableevents;however,inTrentinomostvineyardsarenowadaysequippedwithirrigation

plants, importanttoconfronttheoccurrenceofdroughtevents.Theexpectedhigherfrequencyofdrycondition (conditions),however,shouldbetakenintoaccountiftheresultsofthisstudyareextendedtosimilar

areas,lackingsuchfacilities,orwhereirrigationisnotenvisagedintheproductionprotocol.

Anassessmentof theexpected,plausibleexpansionof viticultural areas shouldalso consider the current landuseandexcludeunsuitable soil (i.e. urbanand industrial areas,parksandprotectedareas,

touristicplaces)fromcalculation.Pasturesandmeadowsrepresentthemostaffordableareasforanenlargementofviticultureinawarmerclimaticscenario.Infact,itisunlikelythatperennialcrops(i.e.intheapple

producingareaofNonValley)andhighqualityproductions(horticulture,berries)willbereplacedduetoclimatechange,ifnotdrivenbyadecreaseofproductionandquality;insuchconditions,changesinprofitdue

tothemarketrulescouldindeedbecomethedriversofpossiblechangesinbothgrapevarietyandcrop.

Norestrictionofviticulturalareas ispredicted to takeplaceat the lowervalleybottom,even if thechange is likely tosuggest–at least in thestrongesthypothesisof thermal increase– to consider the

convenienceofavarietalchange.

5ConclusionsInthisstudyahigh-resolutionmodellingchainwasappliedtoanAlpineareatoassessthechangeinclimaticandbioclimaticindices,withparticularreferencetotheviticulturalarea.Theresultscomefrom

theapplicationoffunctionsfromanopen-sourceRlibrary(ClimClass)totheoutputofaRegionalClimateModel(COSMO-CLM).Whilesomechangesarealreadyobservable(Ecceletal.,2015),theshiftexpectedfor

thefuturecanbequantitativelyassessedbytheuseofclimaticclassificationsandindices,specificallycreatedforviticulturalpurposes.Theimplementationoftheseindicesinanopen-sourcelibrarymayenhance

theiruseas“standards”;moreover, theR libraryClimClasssuppliesuseful tools forpre-processingclimaticseries, facilitatingtheirpreparationandanalysis inviewof theassessmentofclimaticandbio-climatic

indices.

Ageneralincreaseintheoccurrenceofwarmerand–withalesserextent–drierseasonsispredictedforeverytimewindow(2021–2050and2041–2070)andreferencescenario(RCP4.5andRCP8.5);of

course,thechangeismorepronouncedinthefarthestperiodandforthescenarioenvisagingacontinuousincreaseinGHGemission(RCP8.5).

Indicesreflecttheeffectofthisdriver:viticulturecouldbenefitfromtheexpectedeffectsofclimatechange,whichincludehigherairtemperatureanddrierconditions,withbetterripeningandsanitarystatus

ofthegrapes,andspreadingtonewsuitableareas,butattheexpenseofannualandperennialcropsormeadowsandpastures,whicharepartoftheruralmountainlandscape.Alltheseimportantdegreesoffreedom

inthevocationalchoiceofagriculturalareas,togetherwithothersarisingeitherfromagriculturalpolicyorfromtheeconomicalaspectsofwinemarketandfarmmanagement,willprobablyplaythemajorroleinthe

determinationoftherealwine-growingareas.Nevertheless,forthefirsttimeaftertheendofthe“LittleIceAge”period,ataboutthehalfof19thcentury,therearenowsomepremisesforwitnessingareturnof

climateasaco-driverinthechoiceofcrops.

6UncitedreferenceMichalet(1991).

AcknowledgmentsWork co-funded by Climate Change Funds of Provincia Autonoma di Trento, 2013, project IndiClima, and by SEE Transnational Cooperation Programme of EU, 2012, project OrientGate. Thanks to

MeteotrentinoandtheCTT-SIGunitofFond.Machfordatasupply.

AppendixA.SupplementarymaterialSupplementarydataassociatedwiththisarticlecanbefound,intheonlineversion,athttp://dx.doi.org/10.1016/j.compag.2016.05.019.

ReferencesAndersonJ.D.,JonesG.V.,TaitA.,HallA.andTroughtM.C.T.,AnalysisofviticultureregionclimatestructureandsuitabilityinNewZealand,J.Int.DesSci.DeLaVigneEtDuVin46(3),2012,149–165.

BindiM.,FibbiL.,GozziniB.,OrlandiniS.andMigliettaF.,Modelingtheimpactoffutureclimatescenariosonyieldandyieldvariabilityofgrapevine,Clim.Res.7,1996,213–224.

BucchignaniE.,MontesarchioM.,ZolloA.L.andMercoglianoP.,HighresolutionclimatesimulationswithCOSMO-CLMoverItaly:performanceevaluationandclimateprojectionsforthe21stcentury,Int.J.

Climatol.2015,http://dx.doi.org/10.1002/joc.4379.

CaffarraA.andEccelE.,Projectingtheimpactsofclimatechangeonthephenologyofgrapevineinamountainarea,Aust.J.GrapeWineRes.17(1),2011,52–61.

DaluJ.D.,BaldiM.,DallaMartaA.,OrlandiniS.,MaracchiG.,DaluG.,GrifoniD.andManciniM.,MediterraneanclimatepatternsandwinequalityinNorthandCentralItaly,Int.J.Biometeorol.57,2013,729–742,

http://dx.doi.org/10.1007/s00484-012-0600-4.

DunnM.R.,LindesayJ.A.andHowdenM.,Spatialandtemporalscalesoffutureclimateinformationforclimatechangeadaptationinviticulture:acasestudyofUserneedsintheAustralianwinegrapesector,Aust.J.

GrapeWineRes.21,2015,226–239.

Eccel,E.,Cordano,E.,Zottele,F.,Toller,G.B.,2015.ClimClassandClimClassMap:twoR-packagesforclimaticandagro-bioclimaticindices.AnapplicationtoTrentino.XVIIIConvegnoNazionalediAgrometeorologia,9–11giugno2015,SanMicheleall’Adige(TN),BookofExtendedAbstract.

FragaH.,SantosJ.A.,Moutinho-PereiraJ.,CarlosC.,SilvestreJ.,Eiras-DiasJ.,MotaT.andMaleiroA.C.,StatisticalmodellingofgrapevinephenologyinPortuguesewineregions:observedtrendsandclimatechange

projections,J.Agric.Sci.1–17,2015,http://dx.doi.org/10.1017/S0021859615000933.

FragaH.,SantosJ.A.,MalheiroA.C.,OliveiraA.A.,Moutinho-PereiraJ.andJonesG.V.,ClimaticsuitabilityofPortuguesegrapevinevarietiesandclimatechangeadaptation,Int.J.Climatol.36,2016,1–12,

http://dx.doi.org/10.1002/joc.4325.

GiorgiF.,JonesC.andAsrarG.R.,Addressingclimateinformationneedsattheregionallevel:theCORDEXframework,WMOBull.58(3),2009,175–183.

GrifoniD.,ManciniM.,MaracchiG.,OrlandiniS.andZipoliG.,AnalysisofItalianwinequalityusingfreelyavailablemeteorologicalinformation,Am.J.Enol.Viticult.57,2006,339–346.

HallA.andJonesG.V.,EffectofpotentialatmosphericwarmingontemperaturebasedindicesdescribingAustralianwinegrapegrowingconditions,Aust.J.GrapeWineRes.15(2),2009,97–119.

HunterJ.J.andBonnardotV.,SuitabilityofsomeclimaticparametersforgrapevinecultivationinSouthAfrica,withfocusonkeyphysiologicalprocesses,S.Afr.J.Enol.Viticult.32(1),2011,137–154.

JonesG.V.,Climateandterroir:impactsofclimatevariabilityandchangeonwine,In:MacqueenR.W.andMeinertL.D.,(Eds.),FineWineandTerroir–TheGeosciencePerspective.GeoscienceCanada,2006,Geological

AssociationofCanada;Newfoundland,Canada.

JonesG.V.andDavisR.E.,Climateinfluencesongrapevinephenology,grapecomposition,andwineproductionandqualityforBordeaux,France,Am.J.Viticult.Enol.51(3),2000,249–261.

JonesG.V.,WhiteM.A.,CooperO.R.andStorchmannK.,Climatechangeandglobalwinequality,Clim.Change73(3),2005,319–343.

KoufosG.,MavromatisT.,KoundourasS.,FyllasN.M.andJonesG.,Viticulture–climaterelationshipsinGreece:theimpactsofrecentclimatetrendsonharvestdatevariation,Int.J.Climatol.34,2014,1445–1459.

MichaletR.,1991Uneapprochesynthétiquebiopédoclimatiquedesmontagnesméditerranéennes:exempleduMarocseptentrional,1991,DoctoratuniversitéJosephFourier;GrenobleI,273.

MolitorD.,CaffarraA.,SinigojP.,PertotI.,HoffmannL.andJunkJ.,Latefrostdamageriskforviticultureunderfutureclimateconditions:acasestudyfortheLuxembourgishwinegrowingregion,Aust.J.GrapeWine

Res.20(1),2014,160–168.

MoriondoM.,BindiM.,FagarazziC.,FerriseR.andTrombiG.,Frameworkforhigh-resolutionclimatechangeimpactassessmentongrapevinesataregionalscale,Reg.Environ.Change11(3),2011,553–567.

MullinsM.G.,BouquetA.andWilliamsL.E.,BiologyoftheGrapevine,1992,CambridgeUniversityPress;GreatBritain,239.

OIV,2012.OIVGuidelinesforVitivinicultureZoningMethodologiesonaSoilandClimateLevel.ResolutionOIV-VITI423-2012REV1.Adoptedbythe10thGeneralAssemblyoftheOIV.June2012,Izmir,

QueriesandAnswersQuery:Yourarticleisregisteredasaregularitemandisbeingprocessedforinclusioninaregularissueofthejournal.IfthisisNOTcorrectandyourarticlebelongstoaSpecialIssue/Collectionpleasecontacts.sekar@elsevier.comimmediatelypriortoreturningyourcorrections.Answer:OK

Query:Theauthornameshavebeentaggedasgivennamesandsurnames(surnamesarehighlightedintealcolor).Pleaseconfirmiftheyhavebeenidentifiedcorrectly.

Turkey.Availableat:<http://www.oiv.int/oiv/info/enresolution2012?lang=en>.

RockelB.,WillA.andHenseA.,TheregionalclimatemodelCOSMO-CLM(CCLM),Meteorol.Z.17,2008,347–348,http://dx.doi.org/10.1127/0941-2948/2008/0309.

ScoccimarroE.,GualdiS.,BellucciA.,SannaA.,FogliP.G.,ManziniE.,VichiM.,OddoP.andNavarraA.,Effectsoftropicalcyclonesonoceanheattransportinahighresolutioncoupledgeneralcirculationmodel,J.

Clim.24,2011,4368–4384,http://dx.doi.org/10.1175/2011JCLI4104.1.

TomasiD.,JonesG.V.,GiustM.,LovatL.andGaiottiF.,Grapevinephenologyandclimatechange:relationshipsandtrendsintheVenetoregionofItalyfor1964–2009,Am.J.Enol.Viticult.62(3),2011,329–339.

ToniettoJ.andCarbonneauA.,Amulticriteriaclimaticclassificationsystemforgrape-growingregionsworldwide,Agric.For.Meteorol.124(1–2),2004,81–97.

TrewarthaG.T.andLyleH.H.,AnIntroductiontoClimate,sixthed.,1980,MacGraw-Hill,(Appendix:Koeppen’sClassificationofClimates).

WebbL.B.,WhettonP.H.andBarlowE.W.R.,ModelledimpactoffutureclimatechangeonthephenologyofwinegrapesinAustralia,Aust.J.GrapeWineRes.13,2008,165–175,http://dx.doi.org/10.3354/cr00739.

WhiteM.A.,DiffenbaughN.S.,JonesG.V.,PalJ.S.andGiorgiF.,ExtremeheatreducesandshiftsUnitedStatespremiumwineproductioninthe21stcentury,Proc.Natl.Acad.Sci.USA103(30),2006,11217–11222.

ZolloA.L.,RilloV.,BucchignaniE.,MontesarchioM.andMercoglianoP.,TemperatureandprecipitationextremeeventsoverItaly:assessmentofhighresolutionsimulationswithCOSMO-CLMandfuturescenarios,Int.

J.Climatol.2015a,http://dx.doi.org/10.1002/joc.4401.

ZolloA.L.,TurcoM.andMercoglianoP.,AssessmentofhybriddownscalingtechniquesforprecipitationoverthePoriverbasin,Eng.Geol.Soc.Territory1,2015b,193–197.

ZuliniL.,FischerC.andBertaminiM.,Chlorophyllfluorescenceasatoolforevaluationofviabilityinfreeze-stressedgrapevinebuds,Photosynthetica48(2),2010,317–319.

AppendixA.SupplementarymaterialMultimediaComponent1

Supplementarydata1

Highlights

• Awarmingclimatecanleadtonewclimaticconditionsinviticulturalareas.

• Thechangecanbeassessedbyapplyingbioclimaticindicestoclimatemodeloutput.

• Climateindiceswerecalculatedbyadedicatedopen-sourceRlibrary.

• Aspreadofwine-growingareastothemountaincanbequantified.

• Highertemperaturegenerallyincreasesopportunitiesincoolviticultureclimates.

Answer:Yes.

Query:Pleasechecktheeditsmadeintheaffiliations,andcorrectifnecessary.Answer:Yes

Query:Thecountrynamehasbeeninsertedfortheaffiliations‘b’and‘c’.Pleasecheck,andcorrectifnecessary.Answer:Yes

Query:Thecitation‘HallandJones(2008)’hasbeenchangedtomatchtheyearinthereferencelist.Pleasecheck,andcorrectifnecessary.Answer:Yes

Query:References‘DuchêneandSchneider(2005)’,‘Düring(1997),andLisek(2009)’arecitedinthetextbutnotprovidedinthereferencelist.Pleaseprovidetheminthereferencelistordeletethesecitationsfromthetext.Answer:DuchêneE.,Schneider,C.,2005.Grapevineandclimaticchanges:aglanceatthesituationinAlsace.AgronSustainDev25,93-99.doi:10.1051/agro:2004057

Düring,H.,1997.Potentialfrostresistanceofgrape:Kineticsoftemperature-inducedhardeningofRieslingandSilvanerbuds.Vitis.36(4),213-214

Lisek,J.,2009.Frostdamageofbudsonone-year-oldshootsofwineandtablegrapevinecultivarsinCentralPolandfollowingthewinterof2008/2009.JournalofFruitandOrnamentalPlantResearch.17(2),149-161.

Query:Thissectioncomprisesreferencesthatoccurinthereferencelistbutnotinthebodyofthetext.Pleasepositioneachreferenceinthetextor,alternatively,deleteit.Anyreferencenotdealtwithwillberetainedinthissection.Answer:Pleasedelete.