Differences in volatile composition and sexual morphs in rambutan (Nephelium lappaceum L.) flowers and their effect in the Apis mellifera L. (Hymenoptera, Apidae) attraction

(1)

REVISTA

BRASILEIRA

DE

Entomologia

AJournalonInsectDiversityandEvolution

w w w . r b e n t o m o l o g i a . c o m

Systematics,

Morphology

and

Biogeography

Differences

in

volatile

composition

and

sexual

morphs

in

rambutan

(Nephelium

lappaceum

L.)

ﬂowers

and

their

effect

in

the

Apis

mellifera

L. (Hymenoptera,

Apidae)

attraction

Lorena

Aceves-Chong

a

_,

_Leopoldo

_Cruz-López

b

_,

_Daniel

_{Sánchez-Guillén}

b

_,

_Julieta

_{Grajales-Conesa}

a,∗

a_Universidad_Autónoma_de_Chiapas,_Instituto_de_Biociencias,_Tapachula,_Chiapas,_Mexico b_Colegio_de_la_frontera_sur_(ECOSUR),_Tapachula,_Chiapas,_Mexico

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received12June2017 Accepted27September2017 Availableonline14October2017 AssociateEditor:LucasKaminski Keywords: Floralvolatiles Hermaphroditesex Morphologicaltraits Orchards Plant–insectinteractions.

a

b

s

t

r

a

c

t

WestudiedthevolatilecompositionandsexualmorphsofNepheliumlappaceumflowersfromtwo orchards,andinvestigatedthechoicebehaviorofthehoneybee,Apismelliferatowardthefloralextracts frombothlocations.Ourresultsshowedsignificantdifferencesinchemicalcompositionandsexual morphs;onlythehermaphroditeflowersfromtheHerraderoorchardproducedlimoneneand␣-pinene andhadlongerpeduncleandsepalthanflowersfromtheMetapaorchard;ontheotherhand,the hermaphroditeflowersfromtheMetapaorchardhadlongergynoecium.Inthebehavioralexperiment theextractsfromtheHerraderoorchardseemedtogiveA.melliferaforagersbettercuesfororientation tofoodsources,perhapsduetothepresenceoflimoneneand␣-pinene,whichareabsentinthesamples fromMetapa.Suchdifferencesinbothorchardscouldaffectpollinatorattractionandultimatelyseedset andproductivity.

Introduction

Pollinationisanecologicalinteractionmediatedbyplantsthat produceflowersasameanstoattractpollinatorstoreproductive organs.Actuallyplantsexploitthelearning capabilitiesofmany insectpollinatorsbyshowingspecificmorphologiesand produc-ingparticularfloralvolatilesthatdifferentiatethemfromflowers ofotherspecies,triggeringtheso-calledfloralconstancybehavior describedinmanyinsectspecies(Andrewsetal.,2007;Whitney and Federle, 2012). For that reason, visually perceived charac-teristics areof utmostimportance for pollination;for example, studiesonthefloralstructureofGeoffroeadecorticans(Gilliesex Hook.et.Arn)Burk,1949showedthatvariationinflowersizeand corollalengthcouldaffectpollinators’rateofvisitation(Eynard andGaletto,1999).Otherstudieshavecharacterizedfloralvolatile compoundsineconomicallyimportantcrops,suchasalfalfa (Med-icagosativaL.,1753)(Pecettietal.,2002;PecettiandTava,2000), coffee(Coffea arabicaL.,1753) (Vázquez etal.,2003), rambutan (NepheliumlappaceumL.,1767)(Méridaetal.,2003),onion(Allium cepaL., 1753)(Silva etal., 2003), pumpkin (Cucurbitamoschata Duchesne ex Poiret, 1768) (Andrews et al., 2007) and canola

∗ Correspondingauthor.

E-mail:jugrajco79@gmail.com(J.Grajales-Conesa).

(BrassicanapusL.,1753)(Wrightetal.,2002).Floralvolatileshave veryimportantrolesinattractingbeestoreproductiveorgans,in whichrewardslikepollen,nectaroroilsarepresentandstrength floralconstancy;therefore,anychangeinflower’sscent,andalso shape,couldpotentiallyaffecttherateofvisitationofpollinators, andhinderreproductivefitness(Byersetal.,2013;Jürgensetal., 2000;Nunesetal.,2016;Schäffleretal.,2015).

Mostflowering cropplants dependoninsect pollinationfor fruitandseedset;beespollinatealmost90%ofwildvegetation and35%ofcrops(Klattetal.,2013;Kleinetal.,2007;Slaaetal., 2006).Therefore, anyknowledgeontheeffectof floralcueson pollinatorswould allowustoincrease fruit setand cropyield; unfortunately,fewstudieshaveaimedtoexploresuchissue(Byers etal.,2013).Forexample,studieswithstrawberriesanditsmain pollinatorthesolitarybeeOsmiabicornisL.,1758haveshownthat foragingbehaviorisdeeplyinfluencedbyfloralscent,though flo-raldisplayalsoplaysarole;moreinterestinglyyet,differencesin volatileemissionratesbetweenvarietiesofstrawberriesinfluence beevisitationratesunderfieldconditions(Chagnonetal.,1993; HowellandAlarcon,2007;Klattetal.,2013).Studiesperformed withalfalfa(MedicagosativaL.,1753)floralvolatilesandApis mel-liferaL.,1758demonstratedthatlinaloolwastheonlycompound attractivetohoneybees,amongfiveotherantennallyperceived compounds,attheoptimizedconcentration(Henningetal.,1992). Thus,togatherabetterunderstandingabouttheimportanceof

https://doi.org/10.1016/j.rbe.2017.09.002

(2)

ﬂoralvolatilesonpollinationofeconomicallyimportantcropsitis necessarytoexplorethechemical-morphologicalandbehavioral interactionsofpollinatorsandtheirhosts.

Inthetropics,rambutan,whichisnativetoSoutheastAsia,is acropwithmajoreconomicimpactthatisrapidlyexpandinginto SouthernAsia,Australia,theCaribbean,India,SriLanka,Florida, HawaiiandSouthandCentralAmerica (Rincón-Rabanalesetal., 2015;Pohlanetal.,2007,).Rambutanplantsaretrioeeciouswith about50%maleand 50%hermaphroditeplantsandin commer-cialorchardsonlyhermaphroditeplantsaregrown,wheremost ofthemarefunctionallyfemale.RecentstudiesinIndiafoundthe stinglessbeeTetragonulairidipennisSmith,1854and theIndian honeybeeA.ceranaFabricius,1793tobethedominantvisitorsof N.lappaceum.IntheSoconuscoregion,Chiapas,Mexico,rambutan wasintroducedduring1950–1970asaneconomicalternativeto coffee,mangoandbanana,butithasbeensuccessfullyintroduced intomanymoreregions.Currently,rambutanisgrownmainlyin threemunicipalities:Cacahoatan,MetapadeDomínguezand Hue-huetan(Rincón-Rabanalesetal.,2015;Vanderlindenetal.,2004), inwhichScaptotrigonamexicanaGuérin-Meneville,1845,anative bee,isconsideredasthemainpollinator,otherwisebeekeepingof A.melliferaalsooccurs.Nonetheless,thereisalackofinformation onfloralvolatiles,morphologicaltraitsandtheirinteractionwith honeybees.Therefore,ourstudyaimedtoexploreandunderstand suchinteractionby:(a)determiningthevolatilecompositionof rambutanflowersfromtwodifferentorchards,(b)measuring mor-phologicaltraitsinflowersfrombothlocationsand(c)evaluating honeybeeattractiontorambutanfloralvolatiles.

Materialsandmethods

Studysiteandvolatilecollection

Thisstudywascarriedoutintwoorchardslocated1.5kmfrom each other in Metapa de Dominguez in the Soconusco region: Metapa (14◦50N, 92◦11W, 100masl) and Herradero orchards (14◦49N,92◦11W,100masl).Inbothorchardstheoriginofplants canbetracedbacktoonesingledonorplant;nonethelessdespite theclonal origin,evidentdifferencesinmorphologicaltraitsare present,suchasflowershapeandsize.FromFebruarytoMarch 2015ineachorchardwecollectedonepaniclewithhermaphrodite flowersper tree(N=6); wewantedtoinvestigateanypotential differencebetweenmaleandhermaphroditeflowers,thusin Her-raderoorchardwecollectedtwopanicleswithmaleflowersper tree(N=2),unfortunatelywedidnotfindmaleflowersinMetapa orchard.Allsampledpanicleswerenewlyopenedandrandomly chosen bya blind(collector didnot know thefinal fate ofthe flowers,eitherfor morphometricorvolatileanalysis) procedure fromthemiddlepartofthecanopy.Allpanicleswerecarefully removed,separatelytransferredinlabeledpolyethylenebagsand transportedinacoolerat8◦CtothelaboratoryofChemical Ecol-ogyin ElColegiodelaFronteraSur (ECOSUR).Twentyto30min aftersampling,floralvolatileswerecollectedwitha solidphase microextraction(SPME)technique;theSPMEsyringeconsistedin apolydimethylsiloxane-divinylbenzenefiberof65␮mthick,the fiberwasinsertedfor30mininaglassvial(7mL)whereflowers couldnottouchit.Next,theSPMEfiberwereimmediately ana-lyzedinagaschromatographcoupledwithamassspectrometer VarianSaturn2200(Sandovaletal.,2007)withaDB5-MSnonpolar capillarycolumncontaining95%ofdimethylpolysiloxaneand5% ofdiphenylsiloxane,withalengthof30mlongby0.25mminside diameterand25␮mthickfilm,usingheliumasacarriergas. Sam-pleswereanalyzedbyaninitialprogramof50◦Cfor2minwith anincreaseof15◦C/mintoatemperatureof280◦Cfor10min.The injectortemperaturewas250◦C.

VolatileswereidentiﬁedbycomparingKovatsindex,mass spec-traandretentiontimesofsyntheticstandardsinconjunctionwith theNationalInstituteofStandardsandTechnologylibraryversion 2.5(NIST).AllchemicalswereacquiredatSigma-AldrichChemical, ParqueIndustrialToluca2000,street6northN◦ 107,P.C.50200 Toluca,Mexico.

Morphologicalanalysis

Forthisanalysiswerandomlychoseonenewlyopened pani-clepereachofthreeplantsineachorchardandweretransported to thelaboratory as described lines above; next, we randomly chose50 ﬂowersper site and removedthem fromthepanicle. Immediatelyafterseparationfromthepanicle,sepal,gynoecium, pedunclelength,calyxdiameterandcalyxwidthweremeasured usingacalibratedscaleandadigitalcamera(CarlZeissAXIOCAM MRC,E206588,Zensoftware2011,LLC,UnitedStates)mounted onastereomicroscope(CarlZeissmodelSTEMI2000c,LLC,United States).

Behavioraltest

ForagersfromtwohealthyqueenrightA.melliferacolonieswere evaluatedfromNovember2015toJanuary2016at9–14hinthe campusofElColegiodelaFronteraSur.Theflowerextractsusedin thisexperimentwereobtainedbytaking800hermaphrodite flow-ersofeachorchardandseparatelywashedfor30minin250mLof hexaneandtakingittoafinalvolumeof100␮Lusingnitrogen; theseextractsshoweda chemicalprofilesimilartotheextracts obtainedbySPME. Theexperiment consisted oftwo sequential phases.Trainingphase:Fivetotenforagersweretrainedtocollect 2Msucrosesolutionfromafeederlocatedat10mfromthecolony understudy(Sánchezetal.,2008).Thefeederconsistedofa1.5cm diametercotton ballsoaked inthesucrose solutionand placed onaplasticPetridish,withafilterpaperinside,whichwasused toreleasethefloralvolatiles.Trainedforagersweredistinctively paint-markedontheirthoraxwithwater-basedpaintto distin-guishthemfromnewcomers.Sincetherewereferalcoloniesinthe vicinityofoursetup,wepaint-markedthethoraxoftheforagersof ourcolonywithadevicethatallowedustodistinguishthemfrom thoseforagerscomingfromothercoloniesandfromthetrained ones(Mikery-Pachecoetal.,2013).Testphase:Onceatthe exper-imentaldistance,thetrainingfeederwasremovedandtwoclean feedersseparatedfromeachotherby70cmwereshowedtothe foragers.Onefeedercontainedapieceoffilterpaperwith100␮L oftheextractofoneoftheorchardsandtheotheronewithapiece offilterpaperwith100␮Lofhexane.For20minanyunmarkedor device-markedforagerthatstartedcollectingthesucrosesolution wastrappedusinganentomologicalaspiratoruntiltheendofthe experimentstoavoidpseudoreplication,anditschoicewas regis-tered;feedersexchangedplacesevery5mintoavoidsite-learning. Besides,weregisteredthechoicesofthetrainedforagers,butthey werenottrappedtokeepbringingnewcomerstooursetup.Overall, wecarriedoutthreereplicatespercolonypereachextractorchard; eachreplicatewasperformedonadailybasis.

Dataanalysis

Webuiltclassificationtreesbyanalyzingthedatafromvolatile composition, flower sex and study sites using the randomFor-estmethod fromtherandomForest package (Liawand Wiener, 2002) implemented in R software; this analysis finds associa-tionsbetweenindividualsthroughregressionusingapermutation approach,whichallowstheanalysisofsmallsamplesets.

MorphometricdatawereanalyzedbyaKruskal–Wallistestto comparesepal,petallengthandgynoeciumheightbetweenthe

(3)

Table1

CompoundsidentiﬁedinrambutanhermaphroditesandmaleﬂowersfromHerraderoandMetapaorchards.

No. RT RI Compounds Hermaphrodites Males

Herradero Metapa Herradero

1 5.787 895.524 ␣-pinene 0.98±0.71 2 6.080 1012.609 d-Limonene 4.93±4 3 6.335 1037.342 Trans-␤-ocimene 12.83±8.03 17.87±6.88 22.15±7.15 4 7.924 1189.809 Cis-3-hexenylbutyrate 1.46±0.83 3.39±1.20 15.69±4.05 5 8.152 1212.525 Decanal 5.90±2.36 2.38±0.94 6 10.426 1462.573 Trans-geranylacetone 7.97±2.64 1.49±0.51 7 10.855 1513.982 ␣-farnesene 57.33±6.44 50.02±5.39 62.15±3.10

RT,retentiontime;RI,retentionindex;Tracesamount(Mean±StandardError).

twoorchards,respectively(<0.05)withSASsoftwarev9.1.3(SAS Institute,Carry,NC,USA).Allstatisticaltestswereperformedin Statisticaversion7.1(StatSoft,2005).Behavioraldatawere ana-lyzedwithanexactWilcoxonsignedranktestforpaireddatausing packageexactRankTestsv0.8(HothornandHornik,2015)with soft-wareRv3.02(RCoreTeam,2015).

Results

Chemicalanalysis

Overalla total of seven volatilecompoundswere identified. ThecompoundsidentifiedinmaleflowerswereTrans-␤-ocimene, Cis-3-hexenylbutyrateand␣-farnesene,whichwerepresentinEl Herraderoorchards;wedidnotfindmaleflowersintheMetapa orchard.HermaphroditeflowersfromHerraderohadd-limonene and␣-pinene,whichwereabsentinMetapa.Themajorcompound was␣-farnesene,whichwaspresentinbothsexesandorchards (Table1).Fig.1showsthegroupingofthepaniclesaccordingtothe compoundswedetected.

Morphologicalanalysis

Floral characteristics showed differences between both orchards.Allthreepeduncle,gynoeciumandsepallengthwere sig-niﬁcantlydifferent,whereascalyxdiameterandcalyxwidthwere

0.4 0.2 0.0 -0.2 -0.4 -0.2 0.0 0.2 0.4 Dim 1 Dim 2 Her_RH Her_RM Male RH

Fig.1. Relationbetweencompoundsinrambutanhermaphroditeﬂowersfrom Her-radero(HerRH)andMetapaorchard(HerRM),andmaleﬂowers(MaleRH).

not.Herraderoorchardﬂowerswerelongerinpeduncle(p<0.001, F48) and sepal length (p<0.033, F48), nevertheless in Metapa

orchard,ﬂowerswerelongeringynoeciumlength(p<0.001,F48). Table2showsdescriptivestatisticsonmorphometricdata.

Behavioraltests

We found that signiﬁcantly more newcomers chose the Herradero extract (36±2.3 foragers) than the control-hexane (24.16±2.3 foragers) (V=0, p=0.031); on the other hand, the Metapaorchardextractandthecontrol-hexanedidnotshowany signiﬁcantdifference(12.71_±2.3foragersextract;15.14_±2.3 for-agerscontrol;V=12.5,p=0.718).Fig.2showsthepreferenceofthe beesinthebehavioraltests.Inthecaseoftrained,experienced for-agersnodifferencewasfoundinchoicebehaviorforbothorchards.

Discussion

Mostofthefloralscentrelatedstudieshavefocusedon inter-specificvariationofthesesignalssofar,butverylittleisknown aboutintraspecificvariation related togeographic distributions. Our studyrevealed that rambutan flowersfrom Herraderoand

Table2

MorphologicaltraitsoframbutanhermaphroditesﬂowersfromHerraderoand Metapaorchard(Mean±StandardError).

Morphologicaltraits Orchard

Herradero(mm) Metapa(mm) Sepallength 1.59±0.03a _1.50_±_0.04b Gynoeciumlength 2.27±0.04b _2.54_±_0.09a Pedunclelength 1.32±0.04a _1.04_±_0.03b Calyxdiameter 2.35±0.04a _2.30_±_0.05a Calyxwidth 2.54±0.03a _2.58_±_0.06a

Differentlettersinthesamelinemeanstatisticaldifferences(p<0.05).

250 200 150 100 50 0 Herradero Orchards Metapa Control Extract

Number of bees (20 min)

a

a a b

Fig.2. Behavioraltestswithorchardsextracts(HerraderoandMetapa;dottedbar) againstcontrol(whitebar).TheY-axisrepresentsthetotalnumberofbeesover 20minofobservations(sixreplicatesperorchard).Differentlettersaboveeachpair ofcolumnsmeanasigniﬁcantdifference(p<0.05).

(4)

Metapadifferedbothqualitativelyandquantitativelyintheirscent profiles.Such intraspecific flowerscentvariation is observedin alimitednumberofexamples,e.g.inspecimensofSilene latifo-liaPoiret,1789wherequantitativeandqualitativevariabilitywas observedin Europeanand North Americanpopulations(Dötterl etal.,2006).Thiswasalsoregisteredinthehighlyspecific plant-insect pollination interaction, fig- figwasp mutualism, where significantdifferenceswerefoundbetweenEastAsianandIndian populations(Soleretal.,2011).Inregardtoqualitativedifferences, ourresultsshowed that two ofthe volatiles produced by Her-raderoflowers;limonene and␣-pinene,wereabsentinMetapa flowers.ThisintraspecificvariationwasalsoreportedinthreeSalix species,which couldberelated tosexdifferences(Füsseletal., 2007).Inthissense,ourdatashowedthatrambutanmale flow-ersdifferedwithhermaphroditeones,by registeringonlythree compounds:ocimene,butanoicacidand␣-farnesene.Mostofthe compoundsfoundinhermaphroditeandmalerambutanflowers havealsobeenreportedaspartofthefloralodorbouquetinother angiosperms thatare important in a varietyof plant-pollinator interactions, e.g. strawberry flowers (Fragaria×ananassa Duch-esne, 1788) emit (E,E)-␣-farnesene, limonene, p-anisaldehyde, (Z)-3-hexenylacetate,methylsalicylate,benzaldehyde,whichones areknowntobeattractantstohoneybees(Klattetal.,2013).In MimuluslewisiiPursch,1774flowersD-limonene,␤-myrceneand ␤-ocimene,arethemostimportantvolatilesdrivingbumblebee flowervisitationrate(Byerset al.,2013).Inlemon flowers,the majorcompoundwaslimonene(Grajales-Conesaetal.,2012). Fra-garia virginiana Miller, 1768 and Euphorbia cyparissiasL., 1753 emitfloralscentsmadeupofsimilarlywidespreadcompounds, includinglinalool,␤-caryophyllene,␣-terpineol,␣-pineneand (E)-␤-ocimene,which are crucial atattracting potentialpollinators duringdaytime,mostly ants(Vegaetal.,2014).Hence,scentis oneofthefloraltraitsparticularlyimportantindrivingpollinator behaviorandmediating reproductioninfloweringplants(Byers etal.,2013).

The intraspecific variation of floral scent compounds could result of relaxed selective pressure, genetic drift, introgression ofscenttraitsthroughhybridization,pleiotropiceffectsofplant defenseonscentbiosynthesis,orphenotypicplasticityresulting fromedaphicorclimaticdifferences(Raguso,2008).Floralscent oftenoperatesynergisticallywithfloralvisualdisplay,and mor-phologicaltraitscontributetoeffectiveresourcepartitioning(Byers etal.,2013);forexample,differencesinproboscislengthinbumble beesallowforagerswithlongerproboscistocollectfrom flow-erswithlongercorollas(Johnson,1986),anddifferencesinbody colorationintheMeliponagenusallowsomespeciestoforagein sunnier,warmerconditions(Biesmeijeretal.,1999).Inourstudy, wefoundthathermaphodriteflowersformtheHerraderoorchard hadlongerpeduncleandsepalthanflowersfromMetapaorchard, whichwerelongeringynoecium.Innaturalpopulationsof Echinop-sisancistrophora Spegazzini,1905 flowerlength wasassociated withfloraltraitsofvisualdisplay,rewardpresentationand pol-linatorfit(Schlumpbergeretal.,2009).

Pollinatorssuchashoneybeesarecapableoflearningthescent ofaplantspeciesandforagealmostexclusivelyonitprovidedit isdeliveringnectarand pollen(Wrightetal.,2005).Our results showedthatHerraderofloralextractwasmoreattractivethan con-trol;however,Metapafloralextractwasnot.Fidelityofhoneybee foragerstoMetapaorchardthuscouldbebetterexplainedbysite constancy(Grüteretal.,2011)thanfloralconstancyitself,asaresult thatthebeesdidnotdiscriminatebetweenextractandhexaneand theyjustlearnedtheresourcelocation.Honeybeescan discrimi-nateamongthescentsofdifferentcultivarsandrecognizesubtle differencesinscent,e.g.withthreevarietiesofsnapdragon (Antir-rhinummajusL.,1753)beesdifferentiatedtheodorofonevariety fromtheodorofothervarieties,whichsuggeststhatscentintensity,

thetypesofcompoundspresentandthenumberofcommonodors thatabeehaspreviouslyexperiencedaffecttheabilityof honey-beestorecognizeanddifferentiateamongcultivarsandvarieties (Wrightetal.,2005;Wrightetal.,2002).Thisdifferencemightbe relatedtopresenceandconcentrationofodorsignals,whichvary fromflowertoflower,thereforeslightchangesinvolatile composi-tioncouldcausetherejectionofcertainflowersonceisperceivedby thepollinator.Duringthelifespanofaflower,changesinscent com-positionmaybecausedbybothenvironmentalfactorssuchaslight intensityandtemperature,orbiologicaltraitssuchasgenetic dif-ferentiationorcircadianrhythmsofscentemission,floralage,and pollinationstatuswhichmayinfluencepollinatordiscrimination (Wrightetal.,2005).Ourstudyprovidesafirststeptounderstand howfloralfragrances,morphologicaltraitsandsexualitymediate beeattractiontorambutanflowers.Nonetheless,thereisaneed toevaluateidentifiedcompoundsorthemixoftheminside the orchardswithrambutanpollinators,toprovewhichonesaremore attractiveforthem.

Conﬂictsofinterest

Theauthorsdeclarenoconﬂictsofinterest.

Acknowledgments

Wewouldliketothanktherambutanproducersthatallowedus tocollectrambutanﬂowers,ElColegiodelaFronteraSur(ECOSUR) fortheirsupportthroughoutthisprojectand M.SC.JavierValle Moraforhisassistanceduringthedataanalysis.

References

Andrews,E.S.,Theis,N.,Adler,L.S.,2007.Pollinatorandherbivoreattractionto Cur-cubitaﬂoralvolatiles.J.Chem.Ecol.33,1682–1691.

Biesmeijer,J.C.,Richter,J.A.P.,Smeets,M.A.J.P.,Sommeijer,M.J.,1999.Niche differ-entiationinnectar-collectingstinglessbees:theinﬂuenceofmorphology,ﬂoral choiceandinterferencecompetition.Ecol.Entomol.24,380–388.

Byers,K.J.,BradshawJr.,H.D.,Riffell,J.A.,2013.Threeﬂoralvolatilescontribute todifferentialpollinatorattractioninmonkeyﬂowers(Mimulus).J.Exp.Biol., http://dx.doi.org/10.1242/jeb.092213.

Chagnon,M.,Gingras,J.,Oliveira,D.d.,1993.Complementaryaspectsofstrawberry pollinationbyhoneyandindigenousbees(Hymenoptera).J.Econ.Entomol.86, 416–420.

Dötterl,S.,Jürgens,A.,Seifert,K.,Laube,T.,Weißbecker,B.,Schütz,S.,2006.Nursery pollinationbyamothinSilenelatifolia:theroleofodoursinelicitingantennal andbehaviouralresponses.NewPhytol.169,707–718.

Eynard,C.,Galetto,L.,1999.Estructuraﬂoralyvariabilidadintraespecíﬁcade Geof-froeadecorticans(Fabaceae).Darwiniana37,219–228.

Füssel,U.,Dötterl,S.,Jürgens,A.,Aas,G.,2007.Inter-andintraspeciﬁcvariationin ﬂroalscentinthegenusSalixanditsimplicationforpollination.J.Chem.Ecol. 33,749–765.

Grajales-Conesa,J.,Meléndez-Meléndez,V.,Cruz-López,L.,Sánchez-Guillén,D., 2012.EffectofCitrusﬂoralextractsontheforagingbehaviorofthestingless beeScaptotrigonapectoralis(DallaTorre).Rev.Bras.Entomol.56,76–80. Grüter,C.,Moore,H.,Firmin,N.,Helantera,H.,Ratnieks,F.L.W.,2011.Flower

con-stancyinhoneybeeworkers(Apismellifera)dependsonecologicallyrealistic rewards.J.Exp.Biol.,214.

Henning, J.A., Peng, Y.-S., Montague, M.A., Teuber, L.R., 1992. Honey bee (Hymenoptera: Apidae) behavioral response to primary alfalfa (Rosales: Fabaceae)ﬂoralvolatiles.J.Econ.Entomol.85,233–239.

Hothorn,T.,Hornik,K.,2015.exactRankTests:ExactDistributionsforRankand Per-mutationTests.Rpackageversion0.,pp.8–28.

Howell,A.D.,Alarcon,R.,2007.Osmiabees(Hymenoptera:Megachilidae)candetect nectar-rewardingﬂowersusingolfactorycues.Anim.Behav.74,199–205. Johnson,R.A.,1986.IntraspeciﬁcresourcepartitioninginthebumblebeesBombus

TernariusandB.Pennsylvanicus.Ecology67,133–138.

Jürgens,A.,Webber,A.C.,Gottsberger,G.,2000.Floralscentcompoundsof Amazo-nianAnnonaceaespeciespollinatedbysmallbeetlesandthrips.Phytochemistry 55,551–558.

Klatt,B.K.,Burmeister,C.,Westphal,C.,Tscharntke,T.,Fragstein,M.v.,2013.Flower volatiles,cropvarietiesandbeeresponses.PLoSONE8,e72724.

Klein,A.-M.,Vaissière,B.E.,Cane,J.H.,Steffan-Dewenter,I.,Cunningham,S.A., Kre-men,C.,Tscharntke,T.,2007.Importanceofpollinatorsinchanginglandscapes forworldcrops.Proc.R.Soc.B274,303–313.

Liaw,A.,Wiener,M.,2002.ClassiﬁcationandregressionbyrandomForest.RNews 2,ISSN1609-3631.

(5)

Mérida,J.,Rincón,M.,Malo,E.,Rojas,J.,Cruz-López,L.,2003.Volátilesemitidospor lasﬂoresdelrambután(NepheliumlappaceumL.)ysuefectosobrelaabejasin aguijónTetragoniscaangustula(Hymenoptera:Meloponinae).In:IIISeminario Mesoamericanosobreabejassinaguijón.,pp.83–85.

Mikery-Pacheco,O.,Solórzano-Gordillo,E.,Sánchez-Guillén,D.,2013.Métodode marcajemasivodeabejasApismellifera(Hymenoptera:Apidae)paraestudios ecoetológicos.ActaZool.Mex.29,248–251.

Nunes,C.,Pe ñaflor,M.F.,Bento,J.M.,Salvador,M.J.,Sazima,M.,2016.Thedilemma ofbeingafragrantflower:themajorfloralvolatileattractspollinatorsand florivoresintheeuglossine-pollinatedorchidDichaeapendula.Oecologia182, 933–946.

Pecetti,L.,Tava,A.,2000.Effectofﬂowercolorandsamplingtimeonvolatile ema-nationinalfalfaﬂowers.CropSci.40,126–130.

Pecetti,L.,Tava,A.,Felicioli,A.,Pinzauti,M.,Piano,E.,2002.Effectofthreevolatile compoundsfromlucerneﬂowersontheirattractivenesstowardspollinatorsB. Insectology55,21–27.

Pohlan,H.A.J.,Moya,W.G.G.,Centeno,D.J.S.,Agreda,F.M.,Janssens,M.J.J.,Leyva Galán,Á.,Guzman,E.,Toledo,E.T.,Álvarez,R.G.,2007.Organichorticultureinthe tropics:situationandexamplesformMeso-America.J.Agric.RuralDev.Trop. 108,123–148.

Raguso,R.A.,2008.Wakeupandsmelltheroses:theecologyandevolutionofﬂoral scent.Annu.Rev.Ecol.Evol.Syst.39,549–569.

RCoreTeam,2015.R:ALanguageandEnvironmentforStatisticalComputing.R FoundationforStatisticalComputing,Vienna,Austria.

Rincón-Rabanales,M.,Roubik,D.W.,Guzmán,M.A.,Salvador-Figueroa,M., Adriano-Anaya,L.,Ovando,I.,2015.Highyieldsandbeepollonationofhermaphoditic rambutan(NepheliumlappaceumL.)inChiapas,Mexico.Fruits70,23–27. Sánchez,D.,Nieh,J.C.,Vandame,R.,2008.Experience-basedinterpretationofvisual

andchemicalinformationatfoodsourcesinthestinglessbeeScpatotrigona mexicana.Anim.Behav.76,407–414.

Sandoval,I.S.,Hernández,A.S.,Valdez,G.V.,Cruz-López,L.,2007.Volatilesofmango var.ataulfocharacterizedbySPMEandcapillaryGC/MSspectroscopy.J.Mex. Chem.Soc.51,145–147.

Schlumpberger, B.O., Cocucci, A.A.,Moré, M.,Sérsic, A.N., Raguso,R.A., 2009. Extreme variation in ﬂoral characters and its consequences for pollina-tor attraction among populations of an Andean cactus. Ann. Bot. 103, 1489–1500.

Schäfﬂer,I.,Steiner,K.E.,Haid,M.,vanBerkel,S.,Gerlach,G.,Jhonson,S.,Wessjohann, L.,Dötterl,S.,2015.Diacetin,areliablecueandprivatecommunicationchannel inaspecializedpollinationsystem.Sci.Rep.5,12779.

Silva,E.M.,Dean,B.B.,Hiller,L.K.,2003.Honeybee(Hymenoptera:Apidae)foraging inresponsetopreconditioningwithonionﬂowerscentcompounds.J.Econ. Entomol.96,1510–1513.

Slaa,E.J.,Chaves,L.A.S.,Malagodi-Braga,K.S.,Hofstede,F.E.,2006.Stinglessbeesin appliedpollination:practiceandperspectives.Apidologie37,293–315. Soler,C.,Hossaert-McKey,M.,Buatois,B.,Bessière,J.-M.,Schatz,B.,Profﬁt,M.,2011.

Geographicvariationofﬂoralscentinahighlyspecializedpollination mutual-ism.Phytochemistry72,74–81.

Vanderlinden,E.J.M.,Pohlan,H.A.J.,Janssens,M.J.J.,2004.Cultureandfruitqualityof rambutan(NepheliumlappaceumL.)intheSoconuscoregion,Chiapas,Mexico. Fruits59,339–350.

Vázquez,G.R.,Rincón,M.,Rojas,J.,Malo,E.,Cruz,L.,2003.Respuesta electroﬁsiológ-icayconductualdelaabejaScaptotrigonamexicanaalosvolátilesﬂoralesde Coffeacanephoravar.robusta.In:IIISeminarioMesoamericanosobreabejassin aguijón.,pp.79–82.

Vega,C.d.,Herrera,C.M.,Dötterl,S.,2014.Floralvolatilesplayakeyroleinspecialized antpollination.Perspect.PlantEcol.Evol.Syst.16,32–42.

Whitney,H.M.,Federle,W.,2012.Biomechanicsofplant-insectinteractions.Curr. Opin.PlantBiol.16,1–7.

Wright,G.A.,Lutmerding,A.,Dudareva,N.,Smith,B.H.,2005.Intensityandtheratios ofcompoundsinthescentofsnapdragonﬂowersaffectscentdiscriminationby honeybees(Apismellifera).J.Comp.Physiol.A191,105–114.

Wright,G.A.,Skinner,B.D.,Smith, B.H.,2002.Abilityofhoneybee,Apis mellif-era,todetectanddiscriminateodorsofvarietiesofcanola(Brassicarapaand Brassicanapus)andsnapdragonﬂowers(Antirrhinummajus).J.Chem.Ecol.28, 721–740.