Ambrosia Beetle

12 Pages·2013·0.23 MB·English

CHEMICALECOLOGY
Ambrosia Beetle (Coleoptera: Curculionidae) Responses to Volatile
Emissions Associated With Ethanol-Injected Magnolia virginiana
CHRISTOPHERM.RANGER,1,2,3MICHAELE.REDING,1,2PETERB.SCHULTZ,4
ANDJASONB.OLIVER5
Environ.Entomol.41(3):636Ð647(2012);DOI:http://dx.doi.org/10.1603/EN11299
ABSTRACT Xylosandrusgermanus(Blandford)andotherspeciesofambrosiabeetlesarekeypests
ofornamentalnurserytrees.Avarietyoflaboratory-andÞeld-basedexperimentswereconductedin
pursuitofimprovedmonitoringstrategiesandtodevelopatraptreestrategyforambrosiabeetles.
TrapsbaitedwithboltspreparedfromMagnoliavirginianaL.injectedwithethanolcaughtÞvetimes
moreX.germanusthanethanol-baitedtraps.BasalsteminjectionsofethanolintoM.virginianainduced
more ambrosia beetle attacks than irrigating or baiting with ethanol, and no attacks occurred on
water-injected trees. A positive correlation was also detected between concentration of injected
ethanolandcumulativeattacks.Solidphasemicroextraction-gaschromatography-massspectrometry
characterizedbarkemissionsfromethanol-andwater-injectedM.virginianaat1,2,10,and16dafter
treatment.Ethanolemissionfrominjectedtreessteadilydeclinedfrom1to16daftertreatment,but
was not emitted from water-injected trees. A variety of monoterpenes were also emitted in trace
amounts from the ethanol- and water-injected trees. Antennal responses of X. germanus via gas
chromatography-electroantennographic detection to volatiles from ethanol-injected M. virginiana
occurredforethanol,butnotthevariousmonoterpenes.X.germanusandotherambrosiabeetleswere
alsoequallyattractedtotrapsbaitedwithethanolalonecomparedwithasyntheticmixtureofethanol
plus various monoterpenes formulated to mimic ethanol-injected M. virginiana. Injecting concen-
tratedsolutionsofethanolintotreesmaybeusefulforestablishingodor-basedtraptrees,whichcould
aidwithmonitoringprogramsand/orpotentiallydeßectambrosiabeetlesawayfromvaluablenursery
stock.
KEYWORDS ambrosiabeetles,Xylosandrusgermanus,ethanol,solidphasemicroextraction
Exoticambrosiabeetles(Coleoptera:Curculionidae: germanushassincebecomeestablishedinthenorth-
Scolytinae) are increasingly being recognized as eastern,southern,midwestern,anduppernorthwest-
emerging pests of ornamental nursery stock (Oliver ernUnitedStates(Mudgeetal.2001,Rabagliaetal.
and Mannion 2001, Reding et al. 2010). Symptoms 2006).Itisknowntoattack(cid:2)200species,butdecid-
associated with colonization by these fungivorous uoushostsarepreferredoverconiferousspecies(We-
wood-boringbeetlescanincludedieback,wiltingfo- berandMcPherson1983).Nurserystockcommonly
liage,profusebasalsprouts,andestheticdamagetothe attacked in Ohio include dogwood (Cornus spp.),
bark. Xylosandrus germanus (Blandford) is the most Magnolia spp., redbud (Cercis spp.), and Styrax spp.
commonspeciesofambrosiabeetleattackingnursery (C.M.Ranger,personalobservation).
stockinOhio(Rangeretal.2010,Redingetal.2010). AnefÞcientolfactorymechanismaidsX.germanus
Since Þrst being detected in New York in 1932, X. in distinguishing among host qualities, with physio-
logically stressed trees being preferentially attacked
Mentionofproprietaryproductsorcompaniesdoesnotconstitute (WeberandMcPherson1984,Solomon1995,Ranger
anendorsementorrecommendationbytheUSDA-AgriculturalRe- etal.2010).Ethanolisemittedfromstressed,dying,
searchService,TheOhioStateUniversity,TennesseeStateUniver- anddecayinghosts(KimmererandKozlowski1982,
sity,orVirginiaTech.
Joseph and Kelsey 1997, Kelsey and Joseph 2001),
1USDA-Agricultural Research Service, Application Technology
ResearchUnit,HorticulturalInsectsResearchLaboratory,1680Mad- and represents the most attractive semiochemical
isonAve.,Wooster,OH44691. known for X. germanus (Montgomery and Wargo
2Department of Entomology, The Ohio State University, Ohio 1983; Ranger et al. 2010, 2011a). As such, ethanol-
AgriculturalResearchandDevelopmentCenter,1680MadisonAve.,
baited traps are used for detecting X. germanus and
Wooster,OH44691.
3Correspondingauthor,e-mail:[email protected]. other ambrosia beetles and monitoring their ßight
4Hampton Roads Agricultural Research and Extension Center, activity(Gandhietal.2010,Redingetal.2010,Ranger
VirginiaTech.,VirginiaBeach,VA23455. etal.2011a).
5OtisL.FloydNurseryResearchCenter,CollegeofAgriculture,
Buchanan (1941) induced attacks by X. germanus
HumanandNaturalSciences,TennesseeStateUniversity,McMin-
nville,TN37110. underlaboratoryconditionsbyinjecting50%ethanol
0046-225X/12/0636Ð0647$04.00/0(cid:2)2012EntomologicalSocietyofAmerica
June2012 RANGERETAL.:AMBROSIABEETLERESPONSESTOVOLATILEEMISSIONS 637
intoboltsofelm(Ulmussp.).Basalsteminjectionsof MA)wasusedtoinjectM.virginianawith75mlof90%
ethanolintolivingtreesalsoinducedambrosiabeetle ethanolorwater.Treeswere(cid:3)4yrold,1mtall,and
attacks under Þeld conditions, and X. germanus was growingin13litercontainerswithamixtureofaged
themostpredominantspeciesemergingfromethanol- pine bark, peat, and coarse sand (60:30:10 vol:vol).
injectedtreesinOhio(Rangeretal.2010).Injecting Injection sites were initiated by drilling a single 9.5
ethanolintoM.virginianaalsoinducedmoreattacks mm hole (cid:3)16 mm deep into the base of the trees,
thanotherlowmolecularweightvolatilesknowntobe whichhadabasecaliperof40Ð50mm.Theholewas
emitted in response to physiological stress, namely, immediatelypluggedwithanArborjetinjectionport
acetaldehyde, acetone, ethyl acetate, and methanol (9.5mmdiameter)and75mlof90%aqueousethanol
(KimmererandMacDonald1987,KimmererandKoz- solutionswereinjectedatadeliverypressureof4.14
lowski 1982, Ranger et al. 2010). Using the ethanol- bar(60psi).
injection technique to ensure ambrosia beetle pres- Stemsectionsof10cminlengthweresubsequently
sure on experimental trees subsequently made it taken24hafterinjectionat5cmabovetheinjection
possibletoassesstheefÞcacyofbotanicalinsecticides portfromtheethanol-orwater-injectedtrees.Asingle
underÞeldconditionsinOhio(Rangeretal.2011b). boltfromeitherethanol-orwater-injectedtreeswas
ThetechniquealsoallowedfordeterminingtheefÞ- thenpositionedandsecuredverticallywithinbottle
cacyofconventionalinsecticidesagainstXylosandrus trapspreparedaccordingtoRangeretal.(2010)and
crassiusculus(Motschulsky)andotherambrosiabee- picturedinRedingetal.(2011).Inshort,aTornado
tles in North Carolina, along with providing insight Tube(SteveSpanglerScience,CO)wasusedtocon-
intosecondarypestoutbreaksresultingfromthepre- nectthemouthsofa1literplasticbottleanda0.5liter
ventative insecticide applications (Frank and Sadof plastic bottle. Two rectangular openings (cid:3)12.5 cm
2011).Field-basedefÞcacydataforambrosiabeetles long(cid:4)6cmwidewerecutintothesidesofthe1liter
was previously difÞcult to obtain because of the in- bottletoallowtheentranceofambrosiabeetles.In-
abilitytoensurepressureontreatedandcontroltrees dividualboltswerehungverticallywithinthe1liter
(MizellandRiddle2004). bottle.Additionaltrapswerealsobaitedwithanin-
InducingambrosiabeetleattacksonspeciÞctrees dividual heat-sealed, permeable membrane pouch
couldaidwithmonitoringßightandcolonizationac- Þlled with 10 ml of 95% ethanol (65 mg/d at 30(cid:5)C;
tivity for properly timing insecticide applications AgBio, Westminster, CO). Empty traps served as a
withinornamentalnurseries.Attacksbywood-boring blankcontrol.The0.5litercollectingbottlewasÞlled
beetlesotherthantheScolytinaehavenotbeenob- withpropyleneglycolasakillingandpreservingagent.
served to occur as a result of ethanol-injection into Traps were positioned 0.6 m above ground level by
living trees (Ranger et al. 2010). In addition, unin- attaching the inverted end of the 1 liter bottle to a
jectedtreesadjacenttoethanol-injectedtreesarenot metalrod.Trapswerearrangedinarandomizedcom-
attacked by X. germanus or other ambrosia beetles pleteblockdesignalongtheedgeofawoodlotatthe
(Ranger et al. 2010). Thus, strategic placement of OhioAgriculturalResearchandDevelopmentCenter
ethanol-injected trees may also be useful for inter- (OARDC),Wooster,OH(WayneCo.),with6mbe-
cepting ambrosia beetles and deßecting them away tweentrapsand15mbetweenblocks.Intotal,eight
fromvaluablenurserystock,particularlyaspartofa completeblockswereused,thuseightreplicatesfor
Ôpush-pullÕ management strategy (Pyke et al. 1987, each treatment. Traps were held under Þeld condi-
MillerandCowles1990). tionsfrom4May2009through18May2009.
Aspartofourcurrentstudy,avarietyoflaboratory- Inducing Attacks on Specific Trees. To aid in de-
andÞeld-basedstudieswereconductedtoaidinmoni- velopinganodor-basedtraptreefordetecting,mon-
toringanddevelopingatraptreestrategyforambrosia itoring, and potentially mass trapping purposes, the
beetles.Ourobjectiveswereto:1)assesstheattractive- effectiveness of injecting, irrigating, or baiting with
nessoftrapsbaitedwithboltsfromethanol-injectedM. ethanolforinducingambrosiabeetleattacksonspe-
virginiana;2)compareinjecting,irrigating,orbaitingM. ciÞctreeswascompared.Theaforementionedtech-
virginianawithethanolforinducingattacksonspeciÞc niqueswereusedtoinjectpottedM.virginianawith
trees;3)examinetherelationshipbetweenconcentra- 75mlof90%ethanolor75mlofdeionized/distilled
tionofinjectedethanolandattacks;4)characterizevol- water. Trees were also irrigated with 75 ml of 90%
atileemissionsfromethanol-injectedtreesandthecor- ethanol,orbaitedwithoneofthepreviouslydescribed
responding antennal response by X. germanus; and 5) ethanolluresat40Ð50cmabovethetreebase.
determine the attractiveness of a complex mixture of Immediately after injection, experimental trees
ethanolandmonoterpenesbasedonemissionsfrometh- werearrangedinrandomizedcompleteblocksalong
anol-injectedtrees. theedgeofaserviceroadpassingthroughawoodlot
at the OARDC. Trees within each block were 3 m
apartandreplicatedblockswereseparatedby6m.In
MaterialsandMethods
total,Þvecompleteblockswereused,thusÞverepli-
Bolt Attractiveness. To aid with developing im- cates for each treatment. Trees were injected, irri-
proved detection and monitoring tactics, the attrac- gated,orbaitedon21July2009andmaintainedunder
tivenessofboltsfromethanol-injectedM.virginiana Þeldconditionsuntil6August2009duringwhichtime
werecomparedwithanethanollure.BasedonRanger theywerethoroughlyexaminedfornewgalleriesat1,
etal.(2010),anArborjetTreeI.V.system(Woburn, 4,7,10,13,and16daftertreatment(DAT).Gallery
638 ENVIRONMENTALENTOMOLOGY Vol.41,no.3
croboretubing(0.568mminnerdiameter(cid:4)1.07mm
outerdiameter,ColeParmer)wasalsoÞttedtopass
throughoneofthesepta,whichwouldlaterallowfor
aSPMEsyringetobeinsertedintothechamber(Fig.
1A). The free end of the microbore tubing that ex-
tendedoutsideofthechamberwastemporarilysealed
by folding multiple times and compressing with a
microcabletietopreventthelossofvolatilesduring
equilibration.Cablestieswereusedtosnuglysecure
the PTFE chamber against the stem in parallel,
therebyconÞning6cm2ofbarktissue.Theparticu-
larly smooth bark associated with M. virginiana al-
lowedforclosecontactbetweenthetissueandsam-
plingchamber,whicheffectivelysealedthechamber
frompassiveairßow.Thelowerendofthechambers
werepositioned(cid:3)12Ð15cmaboveandin-linewiththe
Fig.1. (A)Volatilesamplingchamber(1)detailingmi-
injectionport.
croboretubing(2)passingthroughasemicircularseptum.
Chamberswereheldinplacefor30mintoallowfor
(B)ThemicroboretubingallowedforaSPMEsyringetobe
volatile equilibration. After sufÞcient equilibration,
positionedwithinthechamberduringsampling.
thesealedmicroboretubingextendingoutsideofthe
chamberwascutwithin0.5cmoftheseptumanda
entrances were circled with a waterproof marker syringecontainingaretractedSPMEÞberwasimme-
(Sharpie,OakBrook,IL)toquantifyattacksateach diately inserted through the microbore tubing and
timepointandsubsequentlytotaledtocalculatecu- intothesealedchamber.Cabletieswereusedtose-
mulativeattacks. cure the syringe holder against the stem in parallel,
Concentration Response. To optimize the attrac- andthentheSPMEÞberwasextrudedfromthesy-
tivenessofethanol-injectedtrees,theassociationbe- ringetipandexposedfor10minwithinthechamber
tweenconcentrationofinjectedethanolandambrosia (Fig.1B).TheÞberwasretractedintothesyringeafter
beetle attacks was examined. Potted M. virginiana samplingandimmediatelycappedwithasealedsec-
wereinjectedwith75mlof90,45,22.5,11.25,or5.6% tion of PTFE microbore tubing to prevent contami-
solutions of ethanol based on the aforementioned nation.FiberswerethenwrappedinaTeßonbagand
techniques.Controltreeswereinjectedwith75mlof brießy held at (cid:6)40(cid:5)C until analysis. A Þber coating
water. Immediately after injection, trees were ar- of divinylbenzene-carboxen-polydimethylsiloxane
rangedinrandomizedcompleteblocksalongtheedge (DVB/CAR/PDMS; 50/30 (cid:2)m coating; Sigma-
of a service road passing through a woodlot at the Aldrich,St.Louis,MO)wasused,whichisidealfora
OARDC.Treeswithineachblockwere3mapartand range of volatile and semivolatile compounds (MW
replicatedblockswereseparatedby6m.Intotal,Þve 40Ð275)(Sigma-Aldrich2011).
complete blocks were used, thus Þve replicates for Volatileswerethermallydesorbedbyexposingthe
each treatment. Trees were injected on 7 May 2009 Þberwithintheinjectionportofagaschromatograph
and maintained under Þeld conditions until 28 May (VarianCP-3800)equippedwithaMerlinMicroseal
2009duringwhichtimetheywerethoroughlyexam- septum system (Sigma-Aldrich). The injection port
inedfornewgalleriesat1,3,7,14,and21DAT. consistedofadeactivatedglasslinerwitha0.75mmID
Volatile Emissions. The time-course emission of and a single tapered end speciÞc for SPME applica-
volatiles from M. virginiana at 1, 2, 10, and 16 DAT tions.Fiberswereheldintheinjectionportfor2min
wereassessedusingtreesinjectedwith75mlof90% at250(cid:5)C,whichwasoperatedinsplitlessmodefrom0
ethanol or water. Four ethanol-injected and water- to 2.5 min and then split at a ratio of 1:20 for the
injectedtreeswereexaminedwhilebeingmaintained remainderoftherun.AcapillarynonpolarDB-5col-
inagreenhouseundernaturallightconditions.Solid umn(0.25mm(cid:4)30m(cid:4)0.25(cid:2)m;cross-linked/surface
phase microextraction (SPME) was used to sample bonded5%phenyl,95%methylpolysiloxane;Agilent
volatilesemanatingfromlowerwoodystemregions, J&W,SantaClara,CA)wasusedforanalysisaccording
whichtypicallysustainsthelargestnumberofambro- to the following program: 40Ð150(cid:5)C at 3(cid:5)C/min and
sia beetle attacks (C. M. Ranger, personal observa- 150Ð246(cid:5)Cat15(cid:5)C/min.Arampof15(cid:5)C/minwasused
tion).Volatilesamplingchambers(6(cid:4)1(cid:4)7cm;l(cid:4) afterreaching150(cid:5)Cbecauseoftheconsistentabsence
w(cid:4)h)withaninternalvolumeof3cm3(Fig.1A,B) ofhost-derivedvolatilesdetectedabovethistemper-
wereconstructedbyÞrstusingarazorbladetolon- ature.Heliumwasusedasacarriergasat1ml/min.A
gitudinallysliceasectionofpolytetraßuoroethylene Varian 2200 mass spectral detector was operated in
(PTFE)tubinginhalf(10mminternaldiameter;2mm electron impact mode with a scan range of 14Ð415
tubingthickness;ColeParmer,VernonHills,IL).An m/z. System control was accomplished with Star
11 mm diameter circular septum (Molded Ther- ChromatographyWorkstationsoftware(StarToolbar,
mogreenLB-2Septa;Supelco,Bellefonte,PA),cutin version6.8;Varian,PaloAlto,CA).Fiberswerecon-
halftoformasemicircle,wasthennestledwithineach ditionedbeforeeachanalysisbyexposurewithinthe
endofthePTFEtubing.A6cmpieceofPTFEmi- injectionportfor20minat250(cid:5)C.
June2012 RANGERETAL.:AMBROSIABEETLERESPONSESTOVOLATILEEMISSIONS 639
All identiÞcations were conÞrmed by comparing moistpapertowels.Volatileswerecollectedfromthe
mass spectral fragmentation patterns and retention bark of M. virginiana using SPME according to the
timeswithauthenticstandards,alongwithreferencing aforementioned procedures. The GC-EAD system
Adams(2007)andtheNationalInstituteofStandards wascomprisedofaVarianCP-3800gaschromatograph
andTechnology(NIST)MSdatabase.Fordetermin- equippedwithaßameionizationdetector(FID)and
ing relative quantities, serial dilutions of ethanol an electroantennographic detector (EAD) system
(99.5%purity,AcrosOrganics,Geel,Belgium)ranging (Syntech,Hilversum,TheNetherlands).TheGCwas
from 100 to 0.0001 g/L were made in water. Serial operated in splitless mode with the aforementioned
dilutions(100to0.0001g/L)ofthefollowingmono- columnandprogram.Efßuentfromthecolumnwas
terpene standards were also prepared: (-)-(cid:3)-pinene split1:1betweentheFIDandEAD.Nitrogenwasused
((cid:4)99.0%; Sigma-Aldrich), ((cid:7))-(cid:3)-pinene ((cid:4)99.0%; asamakeupgas(30ml/min)atthejunctionofthe
Sigma-Aldrich), ((cid:7))-camphene ((cid:4)99.0%; Sigma- splitter(SGEAnalyticalScience,Austin,TX).Deac-
Aldrich), (-)-(cid:5)-pinene ((cid:4)99.0%; Sigma-Aldrich), tivated,fusedsilicatubing(0.32mmdiameter)deliv-
((cid:7))-(cid:5)-pinene ((cid:4)98.5%; Sigma-Aldrich), myrcene eredhalfoftheefßuenttotheFIDwhiletheotherhalf
((cid:4)95.0%;Sigma-Aldrich),(cid:6)-cymene((cid:4)99.5%;Sigma- passed through a heated transfer line maintained at
Aldrich),R-((cid:7))-limonene((cid:4)99.0%;Sigma-Aldrich), 246(cid:5)C. The terminal column end for the EAD ex-
and eucalyptol ((cid:4)99.0%; Sigma-Aldrich). Monoter- tended0.5mmintoastainlesssteelodordeliverytube
penestandardswereÞrstdilutedinethanolbecauseof thatwaspositioned11cmupwindfromtheantennal
theirinsolubilityinwater,butsubsequentserialdilu- preparation.HumidiÞed,charcoalÞlteredairpassed
tionsweremadeinwater.Standardsofvariouscon- throughtheodordeliverytubeat150ml/minandover
centrationswereanalyzedbyÞrstusingcabletiesto theantennalpreparationat(cid:3)2.5cm/s.Antennalprep-
looselysecurethepreviouslydescribedvolatilesam- arationsconsistedofaglasspipettewithanAg/AgCl
plingchamberinparallelwithaglasstesttube(2.5cm indifferent electrode containing Beadle-Ephrussi sa-
diameter). A 20 (cid:2)l aliquot containing a known con- linethatwasinsertedintotheforamenofX.germanus
centration of an individual compound was then ap- Õhead.Thesaline-ÞlledAg/AgClrecordingelectrode
plied to a piece of Þlter paper (2 (cid:4) 1 cm; l (cid:4) w) was micromanipulated to superÞcially puncture the
positionedwithinthevolatilesamplingchamber.The centeroftheleftantennalclubasitwasbracedagainst
cable ties were then immediately pulled tight to ef- thehead.
fectivelysealthechamber.SPMEÞberswereexposed Recording electrode signals were ampliÞed by a
for10minwithinthesamplingchamberaccordingto highimpedanceprobeandfurtherampliÞed,Þltered,
the previously described methods and analyzed im- and optimized using a two-channel data acquisition
mediatelyafterwards. controller (IDAC-2; Syntech). Data were processed
Peak areas associated with standard compounds usingGC-EAD2010software(Syntech).Amplitudes
weremeasuredusingtheStarChromatographyWork- were calculated for all EAD peaks that occurred
stationsoftware.Standardconcentrationcurveswere withina15minwindowthatencompassedallofthe
then calculated and used to determine the relative compounds identiÞed from ethanol-injected M. vir-
quantities of individual compounds associated with giniana. Spikes were only characterized as antennal
volatile emissions from M. virginiana. Under the responsesiftheyoccurredinatleast4ofthe10runs.
properconditions,SPMEcanbeusedforcomparing LureFormulation.Amountsofthefollowingstan-
therelativequantitiesofanindividualcompoundin dards at the aforementioned purities were used to
differentsamples(Bartelt1997,Romeo2009).How- prepare a synthetic mixture of ethanol and selected
ever,becauseofdifferencesinafÞnitiesofSPMEÞ- monoterpenes approximating ratios associated with
bersfordifferentcompounds,relativequantitationby ethanol-injected M. virginiana at 2 DAT: ethanol
SPMEdoesnotallowonetocomparethequantitiesof (118.35 g), (-)-(cid:3)-pinene (22.31 mg), ((cid:7))-(cid:3)-pinene
differentcompoundswithinthesamesample(Bartelt (22.31mg),((cid:7))-camphene(43.78mg),(-)-(cid:5)-pinene
1997,Romeo2009).Absolutequantitationofdifferent (22.67mg),((cid:7))-(cid:5)-pinene(22.67mg),myrcene(15.88
compoundswithinthesamesamplecanbeachieved mg),(cid:6)-cymene(44.56mg),limonene(43.74mg),and
usingSPME(Bartelt1997,Romeo2009),butwasnot eucalyptol(239.85mg).AvolumeofparafÞnoilcom-
conductedaspartofourcurrentstudybecauseofthe parable to the total volume of monoterpenes was
lack of an internal standard during initial sampling. addedtoethanolasacontrolformulation.Testmix-
Thus,therelativequantitiesofindividualcompounds tureswerethenloadedintoseparatePher-Emitdis-
emittedfromM.virginianaoverseveralpointsintime pensers(15mlvolume;Med-E-Cell,SanDiego,CA)
wereonlycomparedwithin(i.e.,ethanol-injectedor manufacturedwithaproprietaryreleasemechanism
water-injected) and between (e.g., ethanol-injected of394mg/dat23(cid:5)Cforethanol.
versuswater-injected)samples. AttractivenessofthetestmixturesinthePher-Emit
Electrophysiological Response. Gas chromatogra- dispensers was assessed using the aforementioned
phy-electroantennographicdetection(GC-EAD)was traps.Trapswerearrangedincompletelyrandomized
usedtoassesstheantennalresponseof10adultfemale blocksalongtheedgeofawoodlotinWayneCo.,OH,
X. germanus to volatile emissions from ethanol-in- with6mbetweentrapsand15mbetweenblocks.In
jected M. virginiana at 2 DAT. Adult females were total,eightblockswereused,thuseightreplicatesfor
reared from maple bolts previously infested with X. eachtreatment.Fieldtrappingwasconductedfrom12
germanusthatwereconÞnedtoplasticcontainerswith May2011through2June2011.
640 ENVIRONMENTALENTOMOLOGY Vol.41,no.3
Statistics.One-wayanalysisofvariance(ANOVA)
and Duncan multiple range test (MRT) at (cid:3)(cid:8) 0.05
wereusedtocomparetotalambrosiabeetlecountsin
Þeldtrappingstudies(SASInstitute2001).Datasets
involving cumulative ambrosia beetle attacks over
time were Þrst analyzed by repeated measures
ANOVAtotestforwithin-subjectandbetween-sub-
ject treatment (cid:4) time effects (SAS Institute 2001).
One-wayANOVAandDuncanÕsMRTat(cid:3)(cid:8)0.05were
subsequentlyusedtocomparemeansatspeciÞctime
pointswhensigniÞcant(P(cid:9)0.05)within-subjector
between-subject treatment (cid:4) time effects were de-
tected.PearsonÕscorrelationcoefÞcientanalysiswas
alsousedtodeterminethedegreeofcorrelationbe- Fig. 2. Cumulative mean ((cid:10)SE) number of ambrosia
beetleattacksassociatedwithinjectingorirrigatingM.vir-
tween cumulative ambrosia beetle attacks and DAT
ginianawithethanol,orbaitingwithanethanollure.Control
(SAS Institute 2001). PearsonÕs correlation coefÞ-
trees were injected with water. Different letters indicate
cientscharacterizetherelationshipbetweentwoin- signiÞcantdifferences16DAT(DuncanÕsMRT;(cid:3)(cid:8)0.05;see
dependentvariablesandrangefromanabsoluteneg- Resultssectionforstatisticalanalyses).
ative correlation of (cid:6)1 to an absolute positive
correlation of 1 (SAS Institute 2001). The general
linearmodelsprocedure(SASInstitute2001)wasalso or baiting M. virginiana with ethanol was compared
used to Þrst compare the overall equality of slopes underÞeldconditions(Fig.2).SigniÞcantdifferences
associatedwithcumulativeattacksandDAT.Thegen- weredetectedamongallfourtreatmentsat16DAT
erallinearmodelsprocedureandpredeterminedcon- withamean((cid:10)SE)of116(cid:10)19.7cumulativeattacks
traststatementswerethenusedtomakespeciÞccom- onethanol-injectedtrees,followedby22.5(cid:10)7.9at-
parisons among and between slopes (SAS Institute tacksontreesirrigatedwithethanol,6.25(cid:10)1.46at-
2001). tacksontreesbaitedwithethanol,and0.0(cid:10)0.0attacks
The time-course emission of an individual com- ontreesinjectedwithwater(F(cid:8)77.79;df(cid:8)3,16;P(cid:9)
pound within and between treatments was also ex- 0.0001)(Fig.2).SigniÞcantdifferenceswerealsode-
amined using repeated measures ANOVA. One-way tectedintheoverallequalityofslopesassociatedwith
ANOVAandDuncanÕsMRTweresubsequentlyused cumulativeattacksandDATfortreesinjected(0.49),
tocomparetherelativeconcentrationsofanindivid- irrigated(0.16),orbaitedwithethanol(0.17),orin-
ualvolatilewithinaparticulartreatmentwhenasig- jectedwithwater(0.0)(F(cid:8)23.35;df(cid:8)3;P(cid:9)0.0001).
niÞcant within-subject treatment (cid:4) time effect was Subsequent speciÞc contrasts determined the slope
detected. PearsonÕs correlation coefÞcient analysis associated with trees injected with ethanol was sig-
was also used to examine the correlation between niÞcantlylargerthantheremainingtreatments(F(cid:8)
emissionofanindividualvolatilecompoundandDAT. 61.77;df(cid:8)1;P(cid:9)0.0001).
WhensigniÞcantbetween-subjecttreatment(cid:4)time Repeated measures ANOVA indicated signiÞcant
effects were detected for an individual compound, within-subject treatment (cid:4) time effects (F (cid:8) 43.55;
relative concentrations associated with ethanol-in- df(cid:8)15;P(cid:9)0.0001).SigniÞcantdifferencesinmean
jected and water-injected trees were compared at ((cid:10)SE) cumulative attacks on ethanol-injected trees
eachtimepointusinganunpairedt-test((cid:3)(cid:8)0.05;SAS weredetectedbetween1DAT(7.3(cid:10)1.7)and4DAT
Institute 2001). In all cases, data were square root (48.0 (cid:10) 11.4), which were both signiÞcantly lower
transformedbeforeanalysis,butuntransformeddata than the remaining dates, but differences were not
arepresented. detectedamong7DAT(85.5(cid:10)12.6),10DAT(103.0(cid:10)
15.7),13DAT(106.8(cid:10)17.0),and16DAT(116(cid:10)19.7)
(F(cid:8)23.18;df(cid:8)5,24;P(cid:9)0.0001)(Fig.2).PearsonÕs
Results
correlation coefÞcient analysis detected a positive
BoltAttractiveness.Amean((cid:10)SE)of23.88(cid:10)3.28 correlation between cumulative ambrosia beetle at-
X. germanus were collected in traps baited with an tacks and DAT for trees injected with ethanol (r (cid:8)
individual bolt from ethanol-injected M. virginiana, 0.81;P(cid:9)0.0001).
whichwassigniÞcantlyhighercomparedwith4.75(cid:10) SigniÞcantlyfewermean((cid:10)SE)cumulativeattacks
1.16 specimens collected from traps baited with an occurred on trees irrigated with ethanol at 1 DAT
ethanollure,0.13(cid:10)0.13fromtrapsbaitedaboltfrom (3.0(cid:10)0.41)comparedwiththeremainingdates,but
awater-injectedtree,and0.0(cid:10)0.0fortheblanktrap differenceswerenotdetectedamong4DAT(15.3(cid:10)
(F(cid:8)80.98;df(cid:8)3,28;P(cid:9)0.0001).Theethanollure 4.8),7DAT(20.8(cid:10)7.5),10DAT(22.3(cid:10)7.7),13DAT
was more attractive to X. germanus than the water- (22.3(cid:10)7.7),and16DAT(22.5(cid:10)7.9)(F(cid:8)4.30;df(cid:8)
injectedboltandblankcontrol,butsigniÞcantdiffer- 5,24;P(cid:8)0.006)(Fig.2).Apositivecorrelationwas
ences did not occur between these latter two treat- detected between cumulative attacks and DAT for
ments. treesirrigatedwithethanol(r(cid:8)0.53;P(cid:8)0.0024).
InducingAttacksonSpecificTrees.Theabilityto SigniÞcantdifferencesinmean((cid:10)SE)cumulative
induceambrosiabeetleattacksbyinjecting,irrigating, attacks on trees baited with ethanol were detected
June2012 RANGERETAL.:AMBROSIABEETLERESPONSESTOVOLATILEEMISSIONS 641
tween14DATand21DAT(82.8(cid:10)11.0)(F(cid:8)43.12;
df(cid:8)4,20;P(cid:9)0.0001)(Fig.3).PearsonÕscorrelation
coefÞcientanalysisdetectedapositivecorrelationbe-
tween cumulative ambrosia beetle attacks and DAT
for trees injected with 90% ethanol (r (cid:8) 0.87; P (cid:9)
0.0001).
The mean ((cid:10)SE) cumulative number of attacks
occurringontreesinjectedwith45%ethanolwerenot
signiÞcantlydifferentbetween1DAT(2.0(cid:10)0.7)and
3DAT(6.8(cid:10)2.4),buttheyweresigniÞcantlylower
comparedwith7DAT(25.8(cid:10)5.3),14DAT(44.0(cid:10)
11.4),and21DAT(50.8(cid:10)11.6)(F(cid:8)12.27;df(cid:8)4,20;
P (cid:9) 0.0001). However, no signiÞcant difference in
cumulativeattacksweredetectedamong7,14,and21
Fig.3. Relationshipbetweenconcentrationofethanol
DAT. PearsonÕs correlation coefÞcient analysis de-
injectedintoM.virginianaandcumulativeambrosiabeetle
tectedapositivecorrelationbetweencumulativeam-
attacks.DifferentlettersindicatesigniÞcantdifferences21
DAT (DuncanÕs MRT; (cid:3)(cid:8) 0.05; see Results section for brosiabeetleattacksandDATfortreesinjectedwith
statisticalanalyses). 45%ethanol(r(cid:8)0.78;P(cid:9)0.0001).
NosigniÞcantdifferenceincumulativeattacksoc-
curredfortreesinjectedwith22.5%ethanolbetween
between1DAT(0.0(cid:10)0.0)and4DAT(1.8(cid:10)1.0), 1DAT(3.4(cid:10)0.8)and3DAT(8.2(cid:10)0.9)(F(cid:8)6.65;
whichwerebothsigniÞcantlylowerthantheremain- df(cid:8)4,20;P(cid:8)0.0014)(Fig.3).AsigniÞcantdifference
ingdates,butdifferenceswerenotdetectedamong7 was also not detected between 3 DAT and 7 DAT
DAT(4.3(cid:10)1.4),10DAT(5.3(cid:10)1.3),13DAT(5.5(cid:10) (26.0(cid:10)5.9),oramong7DAT,14DAT(39.6(cid:10)12.5),
1.4),and16DAT(6.3(cid:10)1.9)(F(cid:8)11.76;df(cid:8)5,24;P(cid:9) and 21 DAT (46.8 (cid:10) 15.5) PearsonÕs correlation co-
0.0001)(Fig.2).Apositivecorrelationwasdetected efÞcient analysis detected a positive correlation be-
betweencumulativeattacksandDATfortreesbaited tween cumulative ambrosia beetle attacks and DAT
withethanol(r(cid:8)0.76;P(cid:9)0.0001).Water-injected fortreesinjectedwith22.5%ethanol(r(cid:8)0.70;P(cid:8)
treeswerenotattackedthroughoutthedurationofthe 0.0001).
experiment(Fig.2). NosigniÞcantdifferenceincumulativeattacksoc-
Concentration Response. An association between curredfortreesinjectedwith11.25%between1DAT
concentration of injected ethanol ranging from 0 to (2.6(cid:10)0.9)and3DAT(6.4(cid:10)0.9),oramong3DAT,
90% and cumulative ambrosia beetle attacks was as- 7DAT(15.0(cid:10)4.9),14DAT(18.6(cid:10)7.6),and21DAT
sessedunderÞeldconditions(Fig.3).Amean(cid:10)SEof (19.4(cid:10)8.2)(F(cid:8)3.23;df(cid:8)4,20;P(cid:8)0.034)(Fig.3).
82.8(cid:10)11.0cumulativeattackswereassociatedwith Apositivecorrelationbetweencumulativeambrosia
treesinjectedwith90%ethanolat21DAT,whichwas beetle attacks and DAT was also detected for trees
signiÞcantlyhighercomparedwith11.2(cid:10)5.5attacks injectedwith11.25%ethanol(r(cid:8)0.53;P(cid:8)0.007).
ontreesinjectedwith5.6%ethanol,19.4(cid:10)8.2attacks NosigniÞcantdifferenceincumulativeattacksoc-
ontreesinjectedwith11.25%,46.8(cid:10)15.52attackson curredfortreesinjectedwith5.6%ethanolamong1
treesinjectedwith22.5%,and50.8(cid:10)14.6attackson DAT(1.6(cid:10)0.7),3DAT(4.2(cid:10)1.2),7DAT(8.6(cid:10)3.6),
trees injected with 45% (F (cid:8) 8.40; df (cid:8) 5, 24; P (cid:9) 14DAT(10.0(cid:10)4.5),and21DAT(11.2(cid:10)5.5)(F(cid:8)
0.0001)(Fig.3).SigniÞcantdifferenceswerealsode- 1.52;df(cid:8)4,20;P(cid:8)0.234)(Fig.3).Apositivecor-
tectedintheoverallequalityofslopesassociatedwith relationbetweencumulativeambrosiabeetleattacks
cumulativeattacksandDATfortreesinjectedwitha and DAT was detected for trees injected with 5.6%
range of 0 to 90% ethanol (F (cid:8) 11.44; df (cid:8) 5; P (cid:9) ethanol(r(cid:8)0.40;P(cid:8)0.045).Noattacksoccurredon
0.0001).SubsequentspeciÞccontrastsdeterminedthe treesinjectedwithwaterthroughoutthedurationof
slopeassociatedwithcumulativeattacksandDATfor theexperiment(Fig.3).
trees injected with 90% ethanol (0.34) was signiÞ- VolatileEmissions.SPME-GC-MSwasusedtochar-
cantlylargercomparedwithtreesinjectedwith45% acterize the time-course emission of volatiles from
(0.28),22.5%(0.22),11.25%(0.12),5.6%(0.07),and ethanol-injectedandwater-injectedM.virginianaat1,
0% (0.0) (F (cid:8) 23.02; df (cid:8) 1; P (cid:9) 0.0001). PearsonÕs 2, 10, and 16 DAT (Fig. 4; Table 1). Ethanol was
correlationcoefÞcientanalysisalsodetectedapositive detectedinemissionsfromethanol-injectedtreesat
correlationbetweenconcentrationofinjectedetha- each time point, but was not associated with water-
nolandcorrespondingcumulativeattacksat21DAT injectedtrees(Table1;Fig.4).Avarietyofmonoter-
(r(cid:8)0.76;P(cid:9)0.0001). peneswerealsoassociatedwiththeethanol-injected
RepeatedmeasuresANOVAdetectedasigniÞcant andwater-injectedtrees,namely,(cid:3)-pinene,((cid:7))-cam-
within-subjecttreatment(cid:4)timeeffect(F(cid:8)7.44;df(cid:8) phene, (-)-(cid:5)-pinene, ((cid:7))-(cid:5)-pinene, myrcene, (cid:6)-cy-
20;P(cid:9)0.0001).Inparticular,signiÞcantdifferencesin mene, limonene, and eucalyptol (Fig. 4; Table 1).
mean((cid:10)SE)cumulativeattacksweredetectedamong RepeatedmeasuresANOVAdetectedsigniÞcantbe-
treesinjectedwith90%ethanolat1DAT(4.8(cid:10)1.2), tween-subjecttime(cid:4)volatileeffectsbetweenetha-
3DAT(15.2(cid:10)2.8),7DAT(46.4(cid:10)2.8),and14DAT nol-injectedandwater-injectedtreesforethanol(F(cid:8)
(73.6 (cid:10) 8.9), but differences were not detected be- 56.94;df(cid:8)1,6;P(cid:8)0.0003),(cid:3)-pinene(F(cid:8)23.64;df(cid:8)
642 ENVIRONMENTALENTOMOLOGY Vol.41,no.3
1 1 1 1 1
F,P 0,0.0001Ñ4,0.042(cid:9)0,0.0008,0.082,0.0072,0.88(cid:9)1,0.0007,0.12(cid:9)7,0.0003,0.024,0.00020,0.297,0.0026,0.17(cid:9)3,0.0009,0.36(cid:9)7,0.000
1 7485263800449911
7. 3.3.2.6.0.5.2.8.5.6.1.9.1.6.1.0.
1 1 2 1 1 1 3
b
20
1
(cid:4) 5baa 9aa7aab aa6aa5aa
2 146 23216 224450
16(cid:10)7.0Ñ(cid:10)10.(cid:10)0.0(cid:10)0.6Ñ(cid:10)39.(cid:10)0.5(cid:10)27.(cid:10)0.3(cid:10)0.4Ñ(cid:10)8.5(cid:10)0.4(cid:10)15.(cid:10)0.4(cid:10)58.(cid:10)0.8
(cid:4)310 22.900.042.36 59.640.8453.480.521.07 42.251.1028.351.2880.172.31
1
2
0
1.
a) b
amber (cid:4)310 4bbab0ab1abbb8ab1ab36ab
vequantities(ng/ch DAT 10(cid:4)(cid:10)39.26103.64Ñ(cid:10)68.1910.3(cid:10)7.750.90(cid:10)21.084.44(cid:10)1.000.35(cid:10)88.9156.1(cid:10)23.544.27(cid:10)114.2254.4(cid:10)11.111.56(cid:10)9.982.18(cid:10)3.040.65(cid:10)151.9920.5(cid:10)7.291.59(cid:10)120.7128.6(cid:10)11.771.12(cid:10)521.76170.(cid:10)14.360.48 (cid:8)RT;df3,12).
ati M
rel nÕs
(cid:10)Mean(SE) 2(cid:10)(cid:4)52.3710aÑ(cid:10)260.93a(cid:10)0.99b(cid:10)36.51aÑ(cid:10)10.99a(cid:10)4.10b(cid:10)116.82a(cid:10)1.56b(cid:10)87.31a(cid:10)0.65b(cid:10)81.71a(cid:10)1.37b(cid:10)67.41a(cid:10)1.34b(cid:10)220.53a(cid:10)0.55b erent(Dunca
M.virginiana (cid:4)56.7110 481.337.6564.06 67.7123.20266.3910.69160.733.49180.998.18144.4812.40460.7116.23 aresigniÞcantlydiff
ndWI (cid:4)410a 66aaaaaa5aa4aa11aa65aa1aa letters
oundsfromEIa 1(cid:4)(cid:10)52.96107.2Ñ(cid:10)276.48109.(cid:10)0.330.16(cid:10)25.738.89(cid:10)0.120.07(cid:10)42.359.02(cid:10)0.990.06(cid:10)89.8717.6(cid:10)2.331.29(cid:10)143.6250.0(cid:10)1.020.37(cid:10)225.87114.(cid:10)2.161.03(cid:10)238.63128.(cid:10)5.011.44(cid:10)175.6580.2(cid:10)3.461.54 wedbydifferent
omp follo
c )
olatile EIWIEIWIEIWIEIWIEIWIEIWIEIWIEIWIEIWI orWI
ofv (EI 4.
(asMr(cid:6)ien4-o.cs)dFnpyvismoi2(grfn)-.ger)4.dion-n.(cid:5)mteioa-,Rp:en((it8aenh1)pe)aarlntneeimeo1sth,el,o-na2i(nnnt,5eajo)1netl0ic,ev(,(t,e(cid:7)ea2andc)n)hdd-((cid:3)r(cid:5)E-1o(-pI6m9p)i)niDanaeetenAonundgeTcer,wa.,a(lPmay3(tep)6seat)(rook-(cid:7)flimnnv()ujosy-eemclrcaecattbemeiTelndeparesbh(e,lWemce(nosi7Iers1))--, 1.Time-courseemission Compound Ethanol (cid:3)-pinene (cid:7)()-camphene (cid:5)-pinene(-)- (cid:5)(cid:7)-pinene()- Myrcene (cid:6)-cymene Limonene Eucalyptol withinarowandtreatmentetected(Ñ).numberscorrespondtoFig.
Table beak 1 2 3 4 5 6 7 8 9 MeansaNotdbPeak
P
June2012 RANGERETAL.:AMBROSIABEETLERESPONSESTOVOLATILEEMISSIONS 643
Table 2. t-test statistics comparing time-course emission of
individualvolatilecompoundsfromEIvsWIM.virginiana(see
Table1)
DAT
Peaka Compound
1 2 10 16
1 Ethanol t 11.26 6.64 5.15 2.52
P (cid:9)0.0001 0.0006 0.002 0.04
2 (cid:3)-pinene t 3.77 2.80 8.43 3.96
P 0.009 0.031 0.0002 0.008
3 ((cid:7))-camphene t 3.64 3.42 7.41 9.80
P 0.011 0.014 0.0003 (cid:9)0.0001
4 (-)-(cid:5)-pinene t 3.53 4.03 0.96 1.81
P 0.012 0.007 0.37 0.12
5 ((cid:7))-(cid:5)-pinene t 7.23 3.47 2.63 3.01
P 0.0004 0.013 0.039 0.024
6 Myrcene t 3.60 2.55 3.16 7.70
P 0.011 0.043 0.02 0.0003
7 (cid:6)-cymene t 2.80 3.04 10.36 7.33
P 0.031 0.023 (cid:9)0.0001 0.0003
8 Limonene t 2.62 2.76 5.83 2.43
P 0.04 0.033 0.001 0.052 Fig.5. SPME-GC-EADresponseofX.germanustovola-
9 Eucalyptol t 2.81 2.63 3.66 3.68
tilescollectedfromethanol-injectedM.virginianaat2DAT.
P 0.031 0.039 0.012 0.01
FigureincludeschromatogramassociatedwiththeFIDand
correspondingresponsebytheEAD.A0.5spuffofethanol
Unpairedt-testcomparedrelativequantitiesofindividualcom-
poundsbetweenEIandWItreesateachtimepoint(df(cid:8)6). wasprovidedattheendoftheruntoconÞrmviabilityofthe
aPeaknumberscorrespondtoFig.4. antennalpreparation,asindicatedbytheverticalbarinthe
digital(DIG)trace.Peaknumberscorrespondto:(1)etha-
nol,(2)(cid:3)-pinene,(3)((cid:7))-camphene,(4)(-)-(cid:5)-pinene,(5)
1,6;P(cid:8)0.003),((cid:7))-camphene(F(cid:8)19.97;df(cid:8)1,6; ((cid:7))-(cid:5)-pinene,(6)myrcene,(7)(cid:6)-cymene,(8)limonene,
P (cid:8) 0.004), (-)-(cid:5)-pinene (F (cid:8) 7.15; df (cid:8) 1, 6; P (cid:8) and(9)eucalyptol.
0.037),((cid:7))-(cid:5)-pinene(F(cid:8)98.28;df(cid:8)1,6;P(cid:9)0.0001),
myrcene(F(cid:8)22.54;df(cid:8)1,6;P(cid:8)0.0032),(cid:6)-cymene 87.31ngofmyrcenepersamplingchamberwereas-
(F(cid:8)26.01;df(cid:8)1,6;P(cid:8)0.0022),limonene(F(cid:8)28.03; sociatedwithethanol-injectedtreesat1and2DAT,
df(cid:8)1,6;P(cid:8)0.002),andeucalyptol(F(cid:8)10.30;df(cid:8) whichweresigniÞcantlyhighercomparedwith9.98(cid:10)
1,6;P(cid:8)0.018).Emissionsofallofthemonoterpenes 2.18and1.07(cid:10)0.46ngat10and16DAT,respectively
weresigniÞcantlyhigherfromethanol-injectedtrees (F(cid:8)5.03;df(cid:8)3,12;P(cid:8)0.02)(Table1).Anegative
comparedwiththewater-injectedtreesateachtime correlationwasdetectedbetweenmyrceneemission
point,exceptfor(-)-(cid:5)-pineneat10and16DAT(Ta- andDAT(r(cid:8)(cid:6)0.75;P(cid:8)0.001).SigniÞcantcorrela-
bles1and2). tions between the emission of individual monoter-
For ethanol-injected trees, repeated measures penesandDATwerenotdetectedfortheremaining
ANOVArevealedasigniÞcantwithin-subjecttime(cid:4) compounds(P(cid:2)0.05).
volatileeffectassociatedwithethanol(F(cid:8)18.77;df(cid:8) For water-injected trees, repeated measures
3;P(cid:8)0.0003)(Fig.4;Table1).Amean(cid:10)SEof2.96(cid:4) ANOVArevealedasigniÞcantwithin-subjecttime(cid:4)
105 (cid:10) 7.2 (cid:4) 104 and 6.71 (cid:4) 105 (cid:10) 2.37 (cid:4) 105 ng of volatile effect for (cid:3)-pinene (F (cid:8) 69.26; df (cid:8) 3; P (cid:9)
ethanolpersamplingchamberwereassociatedwith 0.0001),camphene(F(cid:8)5.56;df(cid:8)3;P(cid:8)0.02),(-)-
ethanol-injected trees at 1 and 2 DAT, which were (cid:5)-pinene(F(cid:8)6.50;df(cid:8)3;P(cid:8)0.013),((cid:7))-(cid:5)-pinene
signiÞcantlyhighercomparedwith9.26(cid:4)103(cid:10)3.64(cid:4) (F(cid:8)5.15;df(cid:8)3;P(cid:8)0.024),myrcene(F(cid:8)4.87;df(cid:8)
103and1.02(cid:4)103(cid:10)7.02(cid:4)102ngat10and16DAT, 3;P(cid:8)0.03),(cid:6)-cymene(F(cid:8)8.21;df(cid:8)3;P(cid:8)0.006),
respectively(F(cid:8)17.10;df(cid:8)3,12;P(cid:8)0.0001)(Table limonene(F(cid:8)4.4;df(cid:8)3;P(cid:8)0.04),andeucalyptol
1).PearsonÕscorrelationcoefÞcientanalysisdetected (F(cid:8)5.51;df(cid:8)3;P(cid:8)0.02),whichresultedinsignif-
anegativecorrelationbetweenethanolemissionand icantdifferencesdetectedamongvarioustimepoints
DAT(r(cid:8)(cid:6)0.62;P(cid:8)0.011). from 1 to 16 DAT (Table 1). However, signiÞcant
A signiÞcant within-subject time (cid:4) volatile effect correlationsbetweentheemissionofindividualmono-
for ethanol-injected trees was also associated with terpenesandDATforwater-injectedtreeswerenot
(cid:3)-pinene(F(cid:8)4.45;df(cid:8)3;P(cid:8)0.035)andmyrcene detected(P(cid:2)0.05).
(F(cid:8)5.57;df(cid:8)3;P(cid:8)0.019).Amean(cid:10)SEof276.48(cid:10) Electrophysiological Responses. The antennal re-
109.66 and 481.33 (cid:10) 260.93 ng of (cid:3)-pinene per sam- sponseofX.germanustoemissionsfromethanol-in-
plingchamberwereassociatedwithethanol-injected jectedM.virginianaat2DATwassuccessfullychar-
treesat1and2DAT,whichweresigniÞcantlyhigher acterizedusingSPME-GC-EAD(Fig.5).Aconsistent
comparedwith68.19(cid:10)10.34and22.90(cid:10)10.15ngat antennalresponseoccurredtoethanol,withamean
10and16DAT,respectively(F(cid:8)3.74;df(cid:8)3,12;P(cid:8) ((cid:10)SE) depolarization of 0.95 (cid:10) 0.18 mV. Antennal
0.042) (Table 1). PearsonÕs correlation coefÞcient responseswerenotobservedtotheaforementioned
analysis detected a negative correlation between monoterpenes.
(cid:3)-pineneemissionandDAT(r(cid:8)(cid:6)0.66;P(cid:8)0.006). Lure Formulation. An ethanol plus monoterpene
Similarly,amean(cid:10)SEof143.62(cid:10)50.04and160.73(cid:10) mixturewaspreparedtomimicratiosassociatedwith
644 ENVIRONMENTALENTOMOLOGY Vol.41,no.3
Table3. Ambrosiabeetlecapturesassociatedwithtrapsbaitedwithethanolalonecomparedwithsyntheticmixtureofethanolplus
variousmonoterpenesformulatedtomimicethanol-injectedM.virginiana
Mean((cid:10)SE)totaltrapcounts
Species F P
Ethanol(cid:7)monoterpenesa Ethanol Blank
X.germanus 95.8(cid:10)19.9a 94.4(cid:10)15.9a 0.0(cid:10)0.0b 58.97 (cid:9)0.0001
E.validus 5.1(cid:10)2.8a 4.0(cid:10)1.4a 0.0(cid:10)0.0b 5.64 0.011
A.sayi 2.0(cid:10)0.5a 3.0(cid:10)1.1a 0.0(cid:10)0.0b 7.51 0.004
X.saxeseni 1.5(cid:10)0.5a 0.6(cid:10)0.4b 0.0(cid:10)0.0b 5.93 0.009
M.mali 0.3(cid:10)0.2a 0.1(cid:10)0.1a 0.0(cid:10)0.0a 1.11 0.35
H.eruditus 0.3(cid:10)0.3a 1.8(cid:10)1.2a 0.0(cid:10)0.0a 2.18 0.14
X.crassiusculus 0.1(cid:10)0.1a 0.3(cid:10)0.2a 0.0(cid:10)0.0a 1.11 0.35
aSeeMaterialsandMethodsforconcentrationsofindividualcomponents.
DifferentletterswithinarowindicatesigniÞcantdifferences(ANOVA;DuncanÕsMRT;(cid:3)(cid:8)0.05;df(cid:8)2,21).
ethanol-injected M. virginiana at 2 DAT (Fig. 4). A centration response to ethanol-baited traps (Kli-
signiÞcant difference in total trap captures was not metzek et al. 1986, Ranger et al. 2011a). This
detected between the ethanol plus monoterpene phenomenon may help to explain why traps baited
mixture and ethanol alone for X. germanus, X. cras- with bolts from ethanol-injected M. virginiana were
siusculus, Monarthrum mali (Fitch), Anisandrus sayi moreattractivetoX.germanusthantheethanollure
(Hopkins), Euwallacea validus (Eichhoff), and Hy- usedaspartofourstudy,whichhadareleaserateof
pothenemuseruditusWestwood(Table3).However,a 65mg/dat30(cid:5)C.Additionalstudiesarewarrantedto
mean((cid:10)SE)of1.5(cid:10)0.5Xyleborinussaxesenii(Rat- quantifythetime-coursereleaseratesofethanolfrom
zeburg) were collected from traps baited with the injectedbolts,butpreliminaryGC-MSanalysesindi-
ethanol plus monoterpene mixture, which differed catedthatmoreethanolwasemittedfromtheboltsat
signiÞcantly from the 0.6 (cid:10) 0.4 specimens collected 1 DAT compared with the lure (C. M. Ranger, per-
fromtrapsbaitedwithethanolalone(F(cid:8)5.93;df(cid:8) sonalobservation).Higheramountsofethanolemis-
2,21;P(cid:8)0.009)(Table3). sionmayalsoexplainwhybasalinjectionsofethanol
X.germanusrepresented91.2%oftotalcapturesfor inducedmoreattacksthanirrigatingwithacompara-
trapsbaitedwithethanolplusthemonoterpenemix- blevolumeofethanolorbaitingwithanethanollure.
ture,followedbyE.validus(4.9%),A.sayi(1.9%),X. Conceivably, more ethanol would have been deliv-
saxesenii(1.4%),M.mali(0.3%),H.eruditus(0.3%), ered into the vascular system through pressurized
and X. crassiusculus (0.1%). Similarly, X. germanus injection compared with uptake by the roots after
represented 90.6% of total captures for traps baited irrigation.Ourcurrentstudyandpreviousstudieshave
withethanolalone,followedbyE.validus(3.8%),A. demonstratedthatethanolcanbeemittedfromstem
sayi(2.9%),H.eruditus(1.7%),X.saxesenii(0.6%),M. and leaf tissue (Ranger et al. 2010). However, the
mali(0.1%),andX.crassiusculus(0.3%). amountofethanolemittedfromanethanol-injected
or-irrigatedM.virginianatreewasnotcalculatedas
partofourcurrentstudyforcomparisontotheethanol
Discussion
lure.Perhapsethanolbeingemittedoveralargersur-
Collectively,theseresultssupporttheroleofeth- face area also contributed to the ethanol-injected
anolasaprimaryvolatilecuethatattractsandinduces treesinducingmoreattacksthanthosebaitedwithan
attacksbyX.germanus.Trapsbaitedwithboltsfrom ethanollure.Itshouldbenotedthatattacksoneth-
M.virginianatreesinjectedwithethanolweremore anol-injected trees were vertically distributed along
attractivetoX.germanusthananethanollureunder thewoodystemtissueofM.virginiana,butattackson
thetestedconditions.Inaddition,basalinjectionsof ethanol-baited trees almost exclusively occurred in
ethanol into M. virginiana induced more ambrosia theregionwheretheethanollurerestedagainstthe
beetleattacksthanirrigatingwithacomparablevol- barktissue(C.M.Ranger,personalobservation).
ume of ethanol or baiting with an ethanol lure. A ThepositiveconcentrationresponsebyX.germanus
positive correlation was also demonstrated between toethanolshouldbeconsideredtooptimizeandmake
theconcentrationofinjectedethanolandcorrespond- odor-basedtraptreesasattractiveaspossible.How-
ing ambrosia beetle attacks. Such information will ever,toohighofanethanolreleaseratecouldhavea
assistwithpursuinganodor-basedtraptreestrategy repellenteffectonambrosiabeetleorientation(Salom
formonitoringandpotentiallymasstrappingambrosia and McLean 1990). For instance, Montgomery and
beetles.Strategicdeploymentofodor-basedtraptrees Wargo(1983)founda2g/dethanolreleaseratewas
in ornamental nurseries and potentially integrating moreattractivetoScolytinaethanhigherreleaserates.
themwithrepellentsaspartofaÔpush-pullÕstrategy Additionalconcentrationshigherthanthosetestedas
mayalsobeeffectiveforambrosiabeetlemanagement partourcurrentstudyarerequiredtodeterminethe
purposes. upperconcentrationthresholdofethanolforattract-
Inadditiontothepositiveconcentrationresponse ing and inducing attacks by X. germanus and other
exhibitedtoethanol-injectedtreesaspartofourcur- selected ambrosia beetles. Understanding the dura-
rentstudy,X.germanusalsoexhibitedapositivecon- tion such trap trees will remain attractive is also of
June2012 RANGERETAL.:AMBROSIABEETLERESPONSESTOVOLATILEEMISSIONS 645
importance.Time-courseanalysisbySPME-GC-MSof 2009,Gandhietal.2010,Rangeretal.2011a).Itshould
emissionsfromethanol-injectedM.virginianadeter- also be considered that the various monoterpenes
mined ethanol emission was negatively correlated wereemittedfromwater-injectedM.virginiana,but
withtime,suchthatethanolsteadilydeclinedoverthe attacksdidnotoccuronthesetrees.BecausemoreX.
courseof16DAT.Temperature,lightlevels,andad- saxensiwereattractedtotrapsbaitedwiththeethanol
ditionalabioticfactorscanaffecttherateofvolatile plus monoterpene mixture compared with ethanol
emissions(Nielsenetal.1995,StaudtandBertin1998). alone, additional Þeld-trapping studies over a larger
Notably, 90% of cumulative ambrosia beetle attacks geographicdistributionarewarranted.X.saxesenihas
occurred on ethanol-injected M. virginiana by 10 been recovered from naturally infested ornamental
DAT.Similarly,aspartoftheconcentrationresponse Þeldstock(Redingetal.2010)andethanol-injected
experiment, 90% of cumulative ambrosia beetle at- trees(Rangeretal.2010).
tacksoccurredontreesinjectedwith90%ethanolby X. germanus is the most abundant species of am-
14 DAT. In each case, attacks presumably began to brosia beetle emerging from infested nursery stock
subside as ethanol emission decreased. Ethanol was (Reding et al. 2010) and ethanol-injected trees
notemittedfromwater-injectedtrees,whichwerenot (Rangeretal.2010)inOhio.Ithastraditionallybeen
attacked by ambrosia beetles during these experi- considered a pest of physiologically stressed trees
mentsandothers(Rangeretal.2010). (Hoffman1941,Solomon1995),presumablybecause
Inadditiontoethanol,atotalofeightacyclic(i.e., oftheassociationofethanolwithsuchhosts(Byers
myrcene) and cyclic (i.e., (cid:3)-pinene, (cid:5)-pinene, cam-
1995,Kelsey2001,Rangeretal.2010).Notably,attacks
phene, (cid:6)-cymene, limonene, and eucalyptol) mono-
havenotoccurredonneighboringwater-injectedor
terpenes were emitted in relatively trace amounts
uninjectedtreesaspartofourcurrentstudyandpre-
from ethanol-injected M. virginiana. Monoterpenes
viousones(Rangeretal.2010).AttacksbyX.germanus
belongtotheC classofisoprenoidsandprovidethe
10 havebeensuspectedtooccuronÔapparentlyhealthyÕ
characteristicaromasofessentialoils,ßoralscents,and
trees (Weber and McPherson 1984, Gre´goire 2001),
certain defensive resins (Mahmoud and Croteau
but the possibility of such trees being Ôinapparently
2002).Theserelativelycommonterpenoidsareasso-
stressedÕatthetimeofattackcannotberuledout.In
ciatedwiththebarkofM.virginiana(Shaetal.2004),
somecases,physiologicallystressedtreescanappear
particularly tissue that has been mechanically dam-
visuallyhealthy,butstillemitstress-relatedvolatiles
aged(Azumaetal.1997).Consideringthattheafore-
thatsignifytheirsuitabilitytocolonizationbyX.ger-
mentionedmonoterpenesweregenerallyemittedin
manus(C.M.Ranger,personalobservation).Forin-
higheramountsfromthebarkofethanol-injectedver-
stance,blackwalnut(JuglansnigraL.)treesthatwere
suswater-injectedtrees,ethanol-injectionmayhave
colonized by X. germanus exhibited slower growth
inßuenced monoterpene biosynthesis or increased
rates in the season before attack than uncolonized
theirvolatilizationfromthebarktissue.
trees,yettheywerereferredtoasapparentlyhealthy
The various monoterpenes emitted from ethanol-
(WeberandMcPherson1984).Ratherthanrandomly
injectedM.virginianadidnotappeartoactadditively
attackingapparentlyhealthytrees,resultsfromthese
orsynergisticallywithethanolinattractingorinduc-
experimentsandothers(Rangeretal.2010)indicate
ing attacks by X. germanus. For instance, antennal
thatX.germanusspeciÞcallytargetshostsbasedonthe
responsesbyX.germanusviaSPME-GC-EADtovol-
emissionofethanol.Indeed,WeberandMcPherson
atile emissions from ethanol-injected M. virginiana
(1984)concludedthatX.germanuscoulddifferentiate
onlyoccurredtoethanolandnotthevariousmono-
betweenevenslightdifferencesintreevigor.
terpenes.Toourknowledge,theseresultsrepresent
Becauseoftheimplicationsofethanolbeingasso-
theÞrstknownSPME-GC-EADresponsesmeasured
forX.germanus.DoseÐresponseelectroantennogram ciated with physiologically stressed hosts (Kelsey
(EAG)studiesarewarrantedtodetermineifX.ger- 2001),theimportanceofmaintaininghostvigorand
manus Õ antennae do in fact recognize the various minimizing stressors should be considered the pri-
monoterpenes identiÞed in emissions from ethanol- maryfoundationofamanagementplanforX.germa-
injected M. virginiana. Such information could then nusandotherselectedambrosiabeetlesinornamental
leadtoadditionaltrappingstudiesaimedatdeveloping nurseries. For instance, water-logging and ßooding,
animprovedlurebytestingmixturescontainingan- whichareknowntoinducetheproductionandemis-
tennallyactivemonoterpenesinhigheramountsthan sionofethanol(KimmererandKozlowski1982,Liao
thosenaturallyassociatedwithethanol-injectedtrees. andLin2001),havecoincidedwithtargetedambrosia
Becauseresultsfromelectroantennographicexper- beetleattacksonsuchtreesinornamentalnurseries
iments should be compared with behavioral and/or (C. M. Ranger, personal observation). Additional
Þeldresponses,theattractivenessofasyntheticmix- stressorsthatcanoccurwithinornamentalnurseries,
tureofethanolandmonoterpenesthatmimickedeth- includingpathogens,waterdeÞcits,heat,freezing,gir-
anol-injectedM.virginianawasalsoexamined.How- dling,andpollutantscanalsoleadtotheproduction
ever,theethanolandmonoterpenemixturewasnot and emission of ethanol (Kimmerer and Kozlowski
more attractive to X. germanus than ethanol alone. 1982, Kimmerer and MacDonald 1987, Sjo¨din et al.
Previousstudiesfound(-)-(cid:3)-pinenegenerallyhada 1989, Joseph and Kelsey 1997, Kelsey and Joseph
negligible effect on increasing the attraction of X. 1998).Maintaininghostvigorisparticularlyimportant
germanustoethanol-baitedtraps(MillerandRabaglia becausepreventativeapplicationsofconventionalin-