Laurel Wilt in Natural and Agricultural Ecosystems

27 Pages·2017·3.1 MB·English

Article
Laurel Wilt in Natural and Agricultural Ecosystems:
Understanding the Drivers and Scales of
Complex Pathosystems
RandyC.Ploetz1,*,PaulE.Kendra2,RobinAlanChoudhury3,JeffreyA.Rollins3,
AlinaCampbell2,KarenGarrett3,MarcHughes4andTylerDreaden5
1 TropicalResearch&EducationCenter,UniversityofFlorida,18905SW280thStreet,Homestead,
FL33031-3314,USA
2 USDA-ARSSubtropicalHorticultureResearchStation,Miami,FL33158-1857,USA;
[email protected](P.E.K.);[email protected](A.C.)
3 PlantPathologyDepartment,UniversityofFlorida,Gainesville,FL32611,USA;
ra.choudhury@ufl.edu(R.A.C.);rollinsj@ufl.edu(J.A.R.);karengarrett@ufl.edu(K.G.)
4 CollegeofTropicalAgricultureandHumanResources,UniversityofHawaiiatManoa,
875KomohanaStreet,Hilo,HI96720,USA;[email protected]
5 ForestHealthResearchandEducationCenter,SouthernResearchStation,USDA-ForestService,Lexington,
KY40546,USA;[email protected]
* Correspondence:kelly12@ufl.edu;Tel.:[phone]
AcademicEditors:MatteoGarbelottoandPaoloGonthier
Received:22December2016;Accepted:13February2017;Published:18February2017
Abstract: LaurelwiltkillsmembersoftheLauraceaeplantfamilyinthesoutheasternUnitedStates.
ItiscausedbyRaffaelealauricolaT.C.Harr.,FraedrichandAghayeva,anutritionalfungalsymbiont
of an invasive Asian ambrosia beetle, Xyleborus glabratus Eichhoff, which was detected in Port
Wentworth, Georgia, in 2002. The beetle is the primary vector of R. lauricola in forests along the
southeastern coastal plain of the United States, but other ambrosia beetle species that obtained
thepathogenaftertheinitialintroductionmayplayaroleintheavocado(PerseaamericanaMiller)
pathosystem. Susceptibletaxaarenaïve(new-encounter)hoststhatoriginatedoutsideAsia. Inthe
southeastern United States, over 300 million trees of redbay (P. borbonia (L.) Spreng.) have been
lost,andotherNorthAmericanendemics,non-Asianornamentalsandavocado—animportantcrop
thatoriginatedinMesoAmerica—arealsoaffected. However,therearenoreportsoflaurelwilton
thesignificantnumberoflauraceousendemicsthatoccurintheAsianhomelandofR.lauricolaand
X. glabratus; coevolved resistance to the disease in the region has been hypothesized. The rapid
spreadoflaurelwiltintheUnitedStatesisduetoanefficientvector,X.glabratus,andthemovement
ofwoodinfestedwiththeinsectandpathogen. Thesefactors,theabsenceoffullyresistantgenotypes,
andthepaucityofeffectivecontrolmeasuresseverelyconstrainthedisease’smanagementinforest
ecosystemsandavocadoproductionareas.
Keywords: laurelwilt;Lauraceae;redbay;avocado;Raffaelealauricola;Xyleborusglabratus;ambrosia
beetles;coevolution
1. Introduction
Newdiseasesaredevelopingatanalarmingrateontheworld’strees. Diverseforestcommunities,
pulpandtimberplantations,andagriculturalproductionareimpactedintropical,temperateandboreal
environments[1–4]. Hostjumps,pathogenhybridizationandclimatechangehavebeenassociated
withsomeofthenewdiseases[5–8],butotheroutbreakshaveresultedfromtheinvasionofnaïve
Forests2017,8,48;doi:10.3390/f8020048 www.mdpi.com/journal/forests
Forests2017,8,48 2of27
ecosyFsotreesmts 2s01b7y, 8,e 4x8 oticpathogens[9–12]. Theoccasionallycatastrophicresponsesofnew-e2 nofc 2o6 unter
hoststoalienpathogensarepoorlyunderstood,unpredictableandwarrantfurtherstudy[1].
of new‐encounter hosts to alien pathogens are poorly understood, unpredictable and warrant further 
In 2004, stands of redbay, Persea borbonia (L.) Spreng., began to die on Hilton Head Island,
study [1].  
SouthCarIno l2i0n0a4,, Ustannitdesd ofS rteadtebsay[,1 P3e]r(sFeai gbourrbeon1iaA ()L..)A Stptrhenegt.i, mbeeg,atnh teo pdrieo bonle Hmiltwona sHneaodt Iwsliadneds, pSroeuathd , and
abiotiCcafraoclitnoar,s U,snuitcehd aSstadterso [u1g3]h (tFoigrusrael t1-Aw).a Atetr tdhea mtimage,e t,hwe eprreobinlevmo kweads ansotp woisdseibsplerecaadu, saensd faobriolotiscs esof
thisnfaatcitvoerst, rseuec.hH aosw dervouegr,hat sotrh seaplt‐rwobalteerm dsapmraegaed, awenreew indvioskeeads ea,sl apuorseslibwlei lcta,uwseass froerc ologsnseizs eodf [t1h3is] .
native tree. However, as the problem spread a new disease, laurel wilt, was recognized [13].  

FigurFeig1u.r(eA 1). D(Ae)v Daesvtaatsitoantiocna ucasuesdedb ybyl alauurreellw wiilltt iinn aa ffoorrmmere rstsatnadn odf orefdrbeadyb, aPye,rsPeear bsoerabbonoriab,o inni Ga,eionrgGiae orgia
in200in9 2(p00h9o (tpoh:oRto.:P Rlo. Peltoze).tz()B. ()BA) dAudlutltf efemmaallee ooff XXyylleebboorruus sglgalbarbartautsu (sPa(PDaILD pILhoptoh o3t3o46323, 4Ju6s2t,inJu BsatrintleBtta, rtlett,
http://www.padil.gov.au/). (C) Galleries of X. glabratus in a redbay tree that was killed by laurel wilt 
http://www.padil.gov.au/).(C)GalleriesofX.glabratusinaredbaytreethatwaskilledbylaurelwilt
(photo: P. Kendra). (D) Internal staining of the sapwood of a redbay tree that is affected by laurel wilt 
(photo: P.Kendra). (D)Internalstainingofthesapwoodofaredbaytreethatisaffectedbylaurel
(photo: R. Ploetz). Note inconspicuous entrance to a gallery of X. glabratus denoted with an arrow. 
wilt (photo: R. Ploetz). Note inconspicuous entrance to a gallery of X. glabratus denoted with an
During the early development of this disease, evidence for X. glabratus is limited. (E) Laurel wilt‐
arrow. Duringtheearlydevelopmentofthisdisease,evidenceforX.glabratusislimited. (E)Laurel
induced death of redbay trees in Florida (photo: R. Ploetz). Laurel wilt‐affected trees of native Persea 
wilt-induceddeathofredbaytreesinFlorida(photo: R.Ploetz). Laurelwilt-affectedtreesofnative
spp. in the southeastern United States defoliate, as in A), only several months after dying. 
Perseaspp.inthesoutheasternUnitedStatesdefoliate,asinA),onlyseveralmonthsafterdying.
Laurel wilt affects trees and shrubs in the Lauraceae (Laurales, Magnoliid complex). It is caused 
by a fungal symbiont, Raffaelea lauricola T.C. Harr., Fraedrich and Aghayeva (Ophiostomatales), of an 
LaurelwiltaffectstreesandshrubsintheLauraceae(Laurales,Magnoliidcomplex). Itiscaused
invasive Asian ambrosia beetle, Xyleborus glabratus Eichhoff (Curculionidae: Scolytinae) [13] (Figures 
by a fungal symbiont, Raffaelea lauricola T.C. Harr., Fraedrich and Aghayeva (Ophiostomatales),
1B and 1C). Starting in its putative epicenter of Port Wentworth, Georgia, United States (i.e., where 
ofaninvasiveAsianambrosiabeetle, XyleborusglabratusEichhoff(Curculionidae: Scolytinae)[13]
X. glabratus was first detected in 2002) [14,15], the disease had spread by August 2016 as far west as 
(Figure1B,C).StartinginitsputativeepicenterofPortWentworth,Georgia,UnitedStates(i.e.,where
95° W, as far east as 78.5° W, as far north as 35° N and as far south as 25.5° N [16]. 
X.glabratuswasfirstdetectedin2002)[14,15],thediseasehadspreadbyAugust2016asfarwestas

95◦ W,asfareastas78.5◦ W,asfarnorthas35◦ Nandasfarsouthas25.5◦ N[16].
Forests2017,8,48 3of27
An estimated 300 million redbay trees have been killed by laurel wilt [17], and damage
to the ecosystems that are associated with this important species has been documented or
predicted [18–23]. Other common species in the southeastern United States, such as sassafras
(Sassafrasalbidum(Nutt.) Nees),silkbay(P.humilisNash),andswampbay(P.palustris(Raf.) Sarg.),
aredecimatedbythedisease,asarethreatened(pondspice,Litseaaestivalis(L.)Fernald)andendangered
endemics (pondberry, Lindera melissifolia (Walter) Blume) [24–26]. Commercial production of a
crop from MesoAmerica, avocado (P. americana Miller), was affected in Florida beginning in 2012,
and that outbreak poses an increasingly serious threat to commercial production there and in
other, currently unaffected areas [27,28]. Other lauraceous species from Europe and the United
States are also susceptible, including bay laurel (Laurus nobilis L.) and, after artificial inoculation,
California laurel (Umbellularia californica (Hook. andArn.) Nutt.), gulf licaria (Licariatriandra (Sw.)
Kosterm.), lancewood (Nectandra coriacea (Sw.) Griseb.), Northern spicebush (Linderabenzoin(L.)
Blume) and Viñátigo (P. indica (L.) Spreng.) [13,26,29–32]. An Asian endemic, camphortree
(Cinnamomumcamphora(L.)J.Presl.),generallytoleratesthedisease[33].
2. Origins
Rabagliaetal.[15]indicatedthatX.glabratuswasnativetoAsiaandhadbeenrecordedfrom
Bangladesh,India,Japan,MyanmarandTaiwan. Recently,HulcrandLou[34]reportedtheinsectin
mainlandChina. TheyconfirmedthatX.glabratuspreferredlauraceoushosts(PhoebezhennanS.Lee
andF.N.Wei,MachilusnanmuNees,C.camphoraandPhoebeneurantha(Hemsl.) GambleinChina),and
thattaxainotherfamilieswererarelycolonized(onlytwoof40collections).
Harringtonetal.[35]recoveredR.lauricolafromspecimensofX.glabratusfromJapanandTaiwan.
SincetherearenoreportsoflaurelwiltintheUnitedStatespriorto2004,X.glabratusprobablycarried
thepathogenwhenitwasfirstdetectedinPortWentworth[14,15]. Thefocalandtemporalspread
of laurel wilt from that area (see [16]) and the genetically uniform populations of R. lauricola and
X.glabratusthatarefoundthroughoutthesoutheasternUnitedStatessuggestthatasinglefounding
event,inorbefore2002,mayberesponsibleforthelaurelwiltepidemic[17,36,37].
Non-AsiansusceptsintheLauraceaeareallnaïve(new-encounter)hosts[1,26]. Despitethewide
geographicrangeofX.glabratusandthelargenumberofspeciesintheLauraceaethatareendemic
toAsia(seebelow), thereisonlyonereportoflaurelwiltinAsia, andthatwasontheintroduced
non-Asianhost,avocado[38]. Fraedrichetal.[33]indicatedthat“therearenoreportsthatindicate
R.lauricolacausesaplantdiseaseinAsia”,andHulcrandLou[34]doubtedthat(sic)“X.glabratus
displaystree-killingbehaviorinitsnativerange.”IflaurelwiltoccursinAsiaonAsianmembersofthe
Lauraceae,itmustbeinconspicuous.
3. Coevolution
Theterm“coevolution”wascoinedtodescribebutterflyxplantinteractions[39]. However,the
ideathatreciprocalevolutionoccurredbetweensympatricspecieswasdiscussedbyDarwin[40]and
describedinaplant-pathologicalcontextinthe1950s. Indescribingresultsfromhisclassicresearchon
flaxrust,Flor[41]suggestedthat“... obligateparasites,suchastherustfungi,musthaveevolvedin
associationwiththeirhosts”andthat“... duringtheirparallelevolution,hostandparasitedeveloped
complementarygenicsystems”. Gene-for-geneandgeneticallyquantitative/multi-genesystemshave
nowbeenidentifiedinmanyotherpathosystems,andthespecificadaptationofpathogenstohosttaxa,
suchasthosedescribedasformaespeciales,isgenerallyacceptedas“theoutcomeofcoevolution”[42].
Theserelationshipscanbeconceivedofasarmsracesinwhichincreaseddiseaseresistancedevelops
inahostinresponsetoincreasedvirulenceinapathogen[43,44].
Althoughtheseinteractionscanbedifficulttodocument[43,45,46],coevolutionappearstobe
animportantfactorinthedevelopmentofmanypathosystems[47]. Severalcriteriacanbeusedto
identifypossiblecoevolvedpathosystems[41–43,45,46,48–52];theyinclude:
Forests2017,8,48 4of27
1. Alimited,oftenspecifichostrangeforthepathogen;
2. Anoriginalgeographicdistributionofthepathogenthatoverlapswiththatofthehost;
3. Theoccurrenceofsignificantdiseaseresistanceinthehost’sprimarycenteroforigin;
4. Regionaloverlapofresistanceandpathogenicityfactorsandphenotypesintherespectivehost
andpathogenpopulations(i.e.,geographicevidenceforreciprocalselection);
5. Gene-for-gene relationships (although quantitative, non-gene-for-gene interactions can also
co-evolve);and
6. Tandemspeciation(alsoknownasparallelcladogenesis).
Duetotherigorouscriteriathatareneededtoconfirmtheserelationships,itisnotsurprising
thatrelativelyfewunequivocalexamplesofhostxpathogencoevolutionexist[51,52]. Mostproposed
coevolvedpathosystemspossesssome,butnotall,oftheaboveattributes[10,42,53,54]. Forexample,
although the first three of the above criteria are met for R. lauricola in Asia, there are no data for
criteria 4–6. Future work may provide additional support for the idea that R. lauricola coevolved
with endemic laurels in Asia. In the meantime, studies of Asian species in the Lauraceae, such as
camphortree,couldprovidevaluableinsightintohowtoleranthostsrespondtothispathogen,and
whichattributesshouldbesoughtduringthedevelopmentoflaurelwilt-resistantgenotypes.
4. AmbrosialSymbioses
Xyleborineambrosiabeetles(Curculionidae:Scolytinae:Xyleborini)exhibitahaplodiploid,sibling
matingsystem,whichisalsoknownasarrhenotoky[55–57]. Femalesarediploidandestablishcolonies
afterdispersiontouncolonizedportionsofthesameornewhosttrees. Fertilizedfemalescanestablish
new colonies of females and haploid males, whereas nonfertilized females lay haploid eggs that
becomemales. Malesareflightlessandrarelyleavetheirnatalgalleries[58],wheretheymatewith
theirmotherandsisters.
Ambrosiabeetlescarrytheirfungalsymbiontsinspecializedstructurescalledmycangia[59–61].
In the Xyleborini, paired pre-oral (also known as mandibular) mycangia are small invaginations
at mandible bases, mesothoracic mycangia are single, large invaginations between the meso- and
metanotum,andelytralmycangiaaresmallcavitiesatelytrabases[61,62]. Inthespeciesthathave
beenexamined,onlyonetypeofmycangiumispresent.However,mycangiaareabsentinsomespecies
thatplunderfungalgardensofotherspecies[63].
When adult females disperse to new trees, they bore brood galleries into host tree xylem, in
whichtheycultivategardensofthefungalsymbionts. Thedevelopingcolonyfeedsonthesefungi
(notwood),andasthecolonymaturesnewfemalesareeventuallyproduced. Theythenperpetuate
the species by dispersing and establishing new colonies. Although some ambrosia beetle species
(e.g., XylosandruscompactusEichhoff)canattackandcolonize(establishbroodcoloniesin)healthy
trees[64,65],mostreproduceonlyinstressedordeadtrees.
Duringtheearlystagesofthelaurelwiltepidemic,Fraedrichetal.[13,33]examinedX.glabratus
xredbayinteractions. Theyindicatedthatinitialattacksbytheinsectinhealthytreeswereaborted
and that reproduction by the insect was not observed in such trees. Nonetheless, aborted attacks
were sufficient to infect trees with R. lauricola. Only after laurel wilt began to develop in infected
treeswasbrooddevelopmentbyX.glabratusobserved[13,33]. Ifthissequenceistypical,X.glabratus
may resemble other ambrosia beetle species in that it preferentially colonizes and reproduces in
compromisedordeadtrees.
Mostambrosiabeetlesaregeneralistswithwidehostranges. Thus,X.glabratusisunusualasit
displaysastrongpreferencefortreesintheLauraceae. Sesquiterpenes,ratherthanethanol,thestress
metabolitetowhichambrosiabeetlesareusuallyattracted,appeartobeasignificantcomponentof
theattractionsignatureofthesetrees[66]. Thisdifferencehasbeencitedwhenindicatingthatthe
X.glabratusvectorrelationshipisexceptional,butthereappearstobenoevidencethattreesidentified
withthesesignaturesthensupportcolonizationandbrooddevelopmentbeforediseasedevelops.Thus,
Forests2017,8,48 5of27
indicationsthattheinteractionofX.glabratuswithitshosttreesisatypical[61,67,68]shouldbeclarified.
Clearly, better information on, and a distinction between, the early (attack phase) and mid-stage
interactionsofX.glabratuswithitshosttrees(colonizationandbrooddevelopment)areneeded.
In discussing the appearance of laurel wilt in the United States and its absence in the Asian
homelandofX.glabratus,HulcrandDunn[68]proposedthat“thesuddenemergenceofpathogenicity”
wasdueto“anewevolutionaryphenomenon.”Amoreparsimoniousexplanationfortheemergence
oflaurelwiltisthatcoevolutionwithR.lauricolaeliminatedsusceptiblespeciesinAsia,butnotin
theUnitedStates. TheideathatAmericanstrainsofR.lauricolabecamepathogenicaftertheirmove
from Asia is not supported by data from studies that detected no differences in pathogenicity to
avocadoandswampbaybetweenisolatesofR.lauricolafromAsiaandtheUnitedStates[36,38]. Hulcr
andDunn[68]alsosuggestedthatan“olfactorymismatch”mayberesponsiblefortheidentification
byX.glabratusofnondiseasedhosttreesassuitableforcolonization. SinceX.glabratuspropagates
in compromised host trees, as do other ambrosia beetle species, the attractive sesquiterpenes may
simplyenableX.glabratustoidentifytreesinwhichitsambrosialsymbiont,R.lauricola,willestablish,
regardlessofthehosttree’shealthstatus. Insummary,pathologicaldifferenceshavenotbeenevident
betweenR.lauricolainAsiaandtheUnitedStates,andX.glabratusandotherambrosiabeetlesappear
tohavesimilarreproductivepreferencesforcompromisedordeadtrees. However,nootherambrosial
symbiont is known to be a systemic, lethal pathogen. Even when ambrosial symbionts kill trees
(e.g., R.quercus-mongolicae K.H. Kim, Y.J. Choi and H.D. Shin, R. quercivora Kubono and Shin. Ito,
andFusariumeuwallaceaeS.Freeman,Z.Mendel,T.AokietO’Donnell),mortalityisduetolocalized
(nonsystemic)necrosisandmultipleattacksbytheassociatedambrosiabeetlevectors[1].
MoreworkisneededtounderstandtheimpactofR.lauricolaonnaïveAmericanhosts. Host
tree colonization by X. glabratus is incompletely understood, and discerning the role that other
ambrosiabeetlespeciesmayplayinthelaurelwiltepidemichasonlybegun. Severalotherspecies
areassociatedwithredbay,avocadoandotherlauraceoushosts,buttheyhavebeenconsideredas
vectorsofR.lauricolaonlyrecently. ThegeneralabsenceofX.glabratusinavocadoorchardsthatare
affectedbylaurelwilt,thepathogen’spresenceinotherspeciesofambrosiabeetlethatarerecovered
fromavocadoandotherhosttrees,andtheexperimentaldemonstrationofpathogentransmissionand
subsequentlaurelwiltdevelopmentinredbayandavocadosuggestthatspeciesotherthanX.glabratus
couldplayrolesintheepidemiologyofthisdisease[1,69,70].
5. VectorsofRaffaelealauricola
Although ambrosia beetles have an obligate association with nutritional fungi [59,61,71,72],
these can be promiscuous relationships wherein a given beetle species carries more than a single
symbiont,andthesamefungusspeciesispresentinmorethanonespeciesofbeetle. Themovementof
symbiontsamongstambrosiabeetlespecieshadbeenrecognizedpreviously[69,71,73,74]. However,
themagnitudeandspeedwithwhichthishasoccurredforR.lauricolaisunprecedented[70]. Sinceits
introductionintotheUnitedStatesinorbefore2002,R.lauricolahasbeenhorizontallytransferredfrom
X.glabratustonineadditionalambrosiabeetlespecies[1,69,70,75,76].
Reared(fromlaurelwilt-affectedhosttrees)ortrappedindividuals(intheproximityoflaurel
wilt-affectedtrees)of14speciesofambrosiabeetle(Ambrosiodmus,Euwallacea,Premnobius,Xyleborus,
Xyleborinus and Xylosandrus spp.) were assayed for R. lauricola by Ploetz et al. [70]. During
10 experiments, the pathogen was recovered from 34% (246 of 726) of the individuals that were
associatedwithPerseaspp. thatarenativetothesoutheasternUnitedStates,butonly6%(58of931)of
thosethatwereassociatedwithavocado.Raffaelealauricolawasrecoveredfrom10oftheambrosiabeetle
speciesthatwereassayed,includingX.glabratus,butwasmostprevalentinXyleboruscongeners[70].
PreviousreportshadsuggestedthatRaffaeleaspp.weretheprimarysymbiontsofXyleborusspp.[74,75].
Ingeneral,mycangiaofX.glabratuscontained10–1000timesmorecolonyformingunits(CFUs)of
R. lauricola than the other assayed species. From native Persea spp. and avocado, R. lauricola was
recoveredfromarespective91%and60%ofthelivespecimensofX.glabratusthatwereassayed[70].
Forests2017,8,48 6of27
AlthoughlittleisknownaboutsymbiosesthatareestablishedbetweenR.lauricolaanddifferent
ambrosia beetle species, some of these insects may be involved in the ongoing epidemic in the
southeasternUnitedStates[28,70]. Inno-choiceexperiments,Carrilloetal.[69]reportedthatsixand
twospeciesotherthanX.glabratustransmittedR.lauricolatopottedredbayandavocadotrees,and
thatlaurelwiltdevelopedinsixandoneoftheseinteractions,respectively.
Kostovcik et al. [74] indicated that different types of mycangia may “support functionally
and taxonomically distinct” symbioses. In summarizing a study of microbial communities in
mycangiaofXyleborusaffinisEichhoff,XyleborusferrugineusFabriciusandXylosandruscrassiusculus
Motschulsky,theyconcludedthatthemandibular(pre-oral)mycangiumfoundinXyleborusenabledthe
establishmentofabroaderarrayofsymbiontsthanthelargerandmoreexposedmesonotalmycangium
possessedbyX.crassiusculus(andXyleborinussaxeseniiRatzeburg).
Hulcr et al. [77] studied how different ambrosia beetle species responded to R. lauricola.
In olfactometer assays, X. glabratus was significantly attracted to R. lauricola in 54 of 84 assays
(p=0.004). Incontrast,adultfemalesofX.saxeseniiandX.crassiusculusweresignificantlyrepelled
bythefungus[77],whichcorrespondswithitsuncommonrecoveryfromthesespecies(1%–4%of
allassayedindividualsin[70]). Interestingly,anotherbeetlethatcarriedR.lauricolamorefrequently
thanX.saxeseniiandX.crassiusculus,X.ferrugineus(11%–57%oftheindividualsin[70]),hadanet
nonresponsetothefungus;i.e.,wasrepelledaboutasoften(156assays)asitwasattracted(132assays)
(p=0.16)[77].
Althoughbeetleattractionto,oravoidanceof,R.lauricolamayimpactwhetheritisafactorinthe
epidemiologyofthisdisease[28,70],itisunclearwhetherspeciesthatarerepelledbyR.lauricolain
olfactometerscouldstillbeoccasionalvectorsofthepathogenandwhetherattractedorneutralspecies
woulddisseminatethepathogenmorefrequently. Clearly,theinvolvementofotherambrosiabeetle
speciesinthelaurelwiltepidemicrequiresmorestudy. Nonetheless,wearebeginningtounderstand
the vector portion of this puzzle. By virtue of their affinity for Raffaelea spp. [70,74,75], species of
XyleborusmightbeexpectedtofostersymbioseswithR.lauricolamorereadilythanspeciesinother
genera. Incontrast,specieswithmesonotalmycangia,suchasX.crassiusculusandX.saxesenii,should
beabletocarrygreaterquantitiesofthepathogen,basedonthelargersizeofthisorgancomparedto
pre-oralmycangia. EventhoughtheyareinfestedwithR.lauricolainfrequentlyandarenotattractedto
thefungusinolfactometers,highlyinfestedindividualswithmesonotalmycangiacouldbevectorsof
R.lauricola.
6. VectorChemicalEcologyandHostLocation
Host-basedattractants(kairomones)havebeenstudiedmostextensivelyfortheprimaryvectorof
R.lauricola,dispersingfemalesofX.glabratus. Thisbeetleisnotattractedtoethanol[78],thestandard
lureforambrosiabeetledetection[79]. Thestrongestfemaleattractantsidentifiedtodateareterpenoid
kairomones, specifically volatile sesquiterpenes emitted from host wood [80–82]. In comparative
studies with nine lauraceous species, emissions of four sesquiterpenes (α-copaene, α-cubebene,
α-humulene,andcalamenene)werepositivelycorrelatedwithin-flightattractionofX.glabratus,and
electroantennographyhasconfirmedolfactorychemoreceptionofthesecompounds[66]. Asuccession
of field lures has been developed using essential oils naturally high in sesquiterpenes, including
manukaoil(derivedfromLeptospermumscopariumJ.R.Forst. andG.Forst.,Myrtaceae;[80]),phoebe
oil (Phoebe porosa (Nees and Martius) Barroso, Lauraceae; [80]), and cubeb oil (Piper cubeba L. f.,
Piperaceae; [83–86]). Laboratory bioassays with fractionated cubeb oil identified the negative
enantiomerofα-copaeneasaprimaryattractant,sufficienttoevokepositivechemotaxis. Currently,
the most effective lure for X. glabratus is an essential oil product enriched to 50% (−)-α-copaene
content[87,88].
Exitingthenataltreetolocateandcolonizenewresourcesiscriticalforthereproductivesuccess
ofambrosiabeetles,butthisbriefdispersaleventpotentiallyexposesthemtopredationandadverse
environmentalconditions. Tominimizerisksassociatedwithdispersion, itwouldbeadaptivefor
Forests2017,8,48 7of27
femalestohaveefficienthost-seekingbehaviors,guidedbyreliablecues. Kendraetal.[66]proposed
that host location and acceptance is a multistep process directed by a series of cues presented in
sequentialorder.
Initiation of dispersal flight in scolytine beetles is determined by light intensity, temperature,
relativehumidityandotherenvironmentalcues([89]andreferencestherein).WithX.glabratus,females
engageinhost-seekingflightduringthelateafternoonandearlyevening,severalhoursearlierthan
otherspeciesofXyleborusinFlorida[90]. Whileinflight,femalesorientinitiallytowardlong-range
olfactorycues;α-copaeneappearstobetheprimarykairomone,butotherterpenoidslikelycontribute
togenerateanattractive“signaturebouquet”oftheLauraceae[66,82,87]. Inadditiontosesquiterpenes,
severalmonoterpeneshavebeenreportedaskairomones,includingeucalyptol(1,8cineole)[91]anda
blendofredbayleafvolatiles[92]. ThereisnoevidencethatX.glabratusutilizessexoraggregation
pheromones[78].
Asfemalesapproachthefocalsourceofhostkairomones,visualcuesassistindirectingflight
towardindividualtrees.MayfieldandBrownie[93]demonstratedthatstemsilhouettediameteraidsin
hostlocation,butonlywhenpresentedinanappropriatechemicalcontext(i.e.,presentedconcurrently
withhostodors). Theseexperimentaldatasupportfieldobservationsthattheoldest,largest-diameter
trees are typically the first to be attacked by X. glabratus. In addition, there is a higher density of
beetleentranceholesonthetrunkandlargediameterbranchesofredbayandswampbay,comparedto
smallerdiameterbranches[10,58,84]. Inmatureavocadotrees,thehighestemissionsofα-copaeneand
α-cubebenearedetectedonthetrunkandlargerbranches[94]. Thischemicalgradientmayfurther
assistwithlocationofoptimalsitesforlandingandinitiationofagallery. Eventhoughlargehostsare
preferred,oncethoseresourcesaredepleted,smallerdiametertreesareutilized,therebyenablinglow
populationsofX.glabratustopersistformanyyears[95].
The short-range cues that prompt a shift from host-seeking to host acceptance and boring
behaviors have received scant study. Knowledge in this area could facilitate the development of
effectiverepellentsorboringdeterrents. Onceafemalemakescontactwithapotentialhost,shelikely
integratesavarietyofstimuli,includingolfactory,gustatory,contactchemosensory,tactileandvisual
cues, all of which must reinforce the message that an appropriate host has been found before a
reproductiveeffortisinitiated. Inshort-rangelaboratorybioassays,X.glabratusisattractedtovolatiles
emittedfromitsfungalsymbiont[77],andinfieldteststhesevolatilesincreasedbeetlecaptureswhen
combinedwithhost-basedlures[96]. Theabilitytodetectfood-basedodorsmaybeadaptiveinthe
host-choice process, confirming that the tree under evaluation is capable of supporting growth of
required(nutritional)fungalresources.
7. PathogenAttributes
Thefungalsymbiontsofambrosiabeetlestypicallycolonizeonlytheliningofthenatalgallery.
Inrarecases,thesefungicauseseriousdamageinhosttrees,forexample,R.quercus-mongolicaeand
R.quercivoraonQuercusspp.,andF.euwallaceaeondiversehostspecies[1,97,98]. However,inthese
andothercasesinwhichsignificantdamageoccurs,thefungalsymbiontdoesnotmovesystemically
inhostxylemandcausesonlylocalizeddamage;inthesesituations,treemortalityisassociatedwith
massattackbythebeetlevectors. Raffaelealauricolaisauniquesymbiont,inthatitsystemicallyinfects
treexylemandasingleinfectionofsusceptibletreescanbefatal.
ThepopulationofR.lauricolaintheUnitedStatesisgeneticallyuniform,apparentlyresulting
fromasinglefoundingevent(perhapstheimportationofX.glabratustoPortWentworth,Georgia).
LittletonogeneticvariationwasdetectedinisolatesofthepathogenacrossthesoutheasternUnited
StateswhenAmplifiedFragmentLengthPolymorphisms(AFLPs)andmicrosatellitemarkerswere
used[17]. Withthesamemicrosatellitemarkersandsequencesofthelargesubunitofribosomal(LSU)
DNA,Wuestetal.[37]alsoconcludedthatthepopulationofR.lauricolaintheUnitedStateswasquite
uniform,incontrasttofargreaterdiversitythattheydetectedinthepathogeninJapanandTaiwan.
WhenisolatesfromAsiaweretestedonavocadoandswampbay,theywereaspathogenicasisolates
Forests2017,8,48 8of27
fromtheUnitedStates[36,38]. Thus,althoughgreatergeneticdiversityinR.lauricolahasbeenfound
inAsia,differencesinpathogenicityhavenotbeenapparentwhenisolatesfromthetworegionshave
beencompared.
Theroleofsymbiontpathogenicityintheecologyofambrosiabeetlesisnotwellunderstood[1,99].
Evidentbycomparisonwithotherintroductionsisthattheaggressivenessbywhichthelaurelwilt
pathogensystemicallycolonizesitsNorthAmericantreehostsandcauseslethalvascularwiltisan
extremescenario[1]. Therangeoffungal-specificdamageobservedamongambrosiasymbiontsin
theirnativehabitatsvariesfromasymptomatic(nonpathogenic)tolocalizeddamageandhostrecovery
(mildlyvirulent).
Fromtheperspectiveoffungalpathogenicity,twohypothesesmayhelptounderstandtheextreme
aggressivenessofR.lauricolaonNorthAmericanLauraceae. ThefirstoftheseisthatR.lauricolais
an“accidentalpathogen”. AdaptationsthatenableR.lauricolatocolonizenatalgalleriesmay,when
encountered beyond the coevolved range of hosts, result in a massive host defense response that
ultimatelyleadstoxylemdysfunctionandhostdeath. Thesecondhypothesisisthatpathogenicityisa
selectiveforceinAsiathatiskeptinbalancebycoevolutionaryprocesses;susceptiblehosts,suchas
thenaïvelaurelsinthesoutheasternUnitedStates,areeliminated,whereasthosethattolerateinfection
persistandeventuallyreplacetheirsusceptiblerelatives.
Tests of these hypotheses are underway by comparing genomes of R. lauricola and its close
relatives[100].TheseanalysesareaidedbyrecentlypublishedphylogeniesofRaffaelea[101,102],which
haveidentifiedtheclosest,extantrelativesofR.lauricola;notably,allothermembersofthephylogenetic
cladeinwhichR.lauricolaresides(Raffaeleasensustricto)arenotplantpathogens. Ofparticularinterest
aregenome-widecomparisonsbetweenR.lauricolaanditscloserelativeR.aguacateD.R.Simmons,
DreadenandPloetz. Raffaeleaaguacatehasonlybeenrecoveredfromavocado[102],butitisneither
systemic nor pathogenic on avocado or redbay [103]. Evidence for the accidental versus adapted
pathogenhypothesesisbeingsoughtincomparativeanalysesbetweenthesetwogenomesandamong
genomesofrelatedspecies. Virulencegenecontent,includingthoseforputativeeffectors,elicitors
andtoxins,willbeassessed,aswilltheexpansionofthesepathogenicity-associatedgenefamilies,and
evidencefordiversifyingselectionamongcandidatevirulencefactors.
Unravelingthesefundamentalquestionsmayprovidesignificantinsightandpredictivevalue
for what might be expected in the future. If all ambrosial symbionts have the potential to be
plant pathogens, is it a matter of time before another beetle–fungus–tree combination facilitates
the emergence of another laurel wilt-like pathogen (“symbiont roulette”)? If, on the other hand,
pathogenicityisaderivedcharacter,whataretheattributesrequiredforweaklypathogenicsymbionts
tobeaggressivepathogenswhenintroducedoutsidetheirnativerange?
8. HostsofLaurelWilt
The Lauraceae is a large family that includes over 50 genera and 2500 to 3000 species [104].
The family represents some of the earliest angiosperms and has a fossil record dating back to the
Mid-Cretaceous [105]. It is well represented on both sides of the Pacific Basin, and the so-called
“amphi-Pacific tropical disjunction” of the Persea and Cinnamomum groups in the family has been
examined to understand the presence, and the origins and relatedness, of family members in the
EasternandWesternHemispheres[101,106,107].
Althoughtherearenotableexceptions,suchasnaturalizedpopulationsofcamphortreeinthe
southeasternUSAandthegloballycultivatedcrop,avocado[108,109],mostspeciesinthefamilyhave
restricted distributions. In the American tropics, lauraceous taxa comprise significant portions of
lowlandforestsandmontaneenvironments[110,111]. Consideringtheirecologicalimportance,laurel
wiltcouldhaveanevengreaterimpactasitspreadstonewareasintheWesternHemisphere.
Of economic concern is the potential impact of the disease in currently unaffected
avocado-productionareas[28]. Avocadoisasubtropical/tropicaltree, andasignificantfruitcrop.
In 2014, 5 million metric tonnes (MMT) of fruit were harvested worldwide, and Mexico was the
Forests2017,8,48 9of27
leading producer (1.5 MMT) and exporter (in 2013, 0.6 MMT valued at ca. US$1.1 billion) [112].
Depending on the cultivars that are grown, which vary considerably in cold tolerance, the crop
is grown commercially from United States Department of Agriculture Hardiness Zones 10 to 11,
withmoderateurbanproductionofsomecultivarsoccurringintoZone9. Florida’savocadoindustry
is the state’s second-largest fruit industry after citrus, and 85% of the producers have orchards of
lessthan13hectares. IntheUnitedStates,morethan6600growersonmostlysmallfarms(lessthan
45hectares)produceavocado;annualproductionworthmore$1.6billionisatrisk[112].
LaurelwilthasaffectedavocadoinFloridaforatleastadecade, butsofaronlyca. 3%ofthe
avocadotreesincommercialproductionhavebeenkilled[113]. However,theshorttimelaurelwilt
has been in South Florida (where most commercial production occurs), the great susceptibility of
avocado to this disease, and experience with other lethal, invasive diseases [11,114] suggests that
laurelwiltwill,inthefuture,causeincreasingdamageonavocadoinFlorida. Laurelwiltimperils
avocadoproductioninotherstates(e.g.,CaliforniaandHawaii),USAprotectorates(PuertoRico)and
neighboringcountriessuchasMexico[28].
PerseaamericanaisdividedintoMexican(M),Guatemalan(G)andWestIndian(WI)(alsoknown
asLowlandorAntillean)botanicalraces,respectivelyvars. drymifolia(Schltdl. andCham.) S.F.Blake,
guatemalensis(L.O.Williams)Scora,andamericanaMill.[115]. Hybridizationoccursfreelyamongthe
racesandisassociatedwitharangeofresponsestocalcareoussoil,salinity,highandlowtemperatures,
andotherstresses[109,116]. Differentresponsestolaurelwilthavealsobeennotedamongtheraces
andracialcombinationsofthespecies;ingeneral,greatersusceptibilityoccursinWIcultivars[117,118].
Althoughthemostseriousoutbreaksoflaurelwilthaveoccurredonlauraceousnativesfrom
thesoutheasternUnitedStates, afewnativesfromtheregion, gulflicariaandlancewood, display
moderatetoleranceafterartificialinoculation[32,119]. Theydevelopvascularsymptoms,butdonot
die,whichissimilartotheresponseofcamphortree[33]. Notably,raretolerancetolaurelwilthasbeen
selectedinredbay,andthereishopethattheseselectionscouldbeusedtore-establishthistreeinthe
southeasterncoastalplain[36,120].
MoredataareneededonthesusceptibilityofAmericanandnon-Americanspeciesinthefamily,
andhowsusceptibleandresistanthostsrespondtoinfectionbyR.lauricola. Todate,wehaveonly
rudimentary understandings of these processes [33,121–123]. Although natural selection against
susceptibilityhasprobablyoccurredincamphortreeandotherAsianLauraceae[1,33],hostattributes
thatareassociatedwithlaurelwilttolerancehavebeenstudiedonlyrecently[121].
9. HostResponsestoInfectionbyRaffaelealauricola
Vascularwiltdiseasestypicallyexhibitwilting,sapwooddiscoloration(Figure1D)andvascular
dysfunctionassociatedwithphysicalandhistologicalchangesinthehost[124]. Theproductionof
gels[125]andtyloses[126]inthexylemaretwoofthemostcommonattributesofaffectedwoody
plants[127–131].
Tylosesareformedinxylemlumenainresponsetopathogeninfection,embolism,aging,and
injury[131]. Theyareoutgrowthsfromadjacentparenchymacells,andeventhoughtheycanprevent
desiccation, damage, and infection of adjacent cells [131], they also reduce hydraulic conductivity
(waterconductance)[132]. Gelsinducedbypathogensgenerallyarisefromhostperforationplates,
endwalls,andpitmembranesoftheprimarywallandmiddlelamella[131,133,134]. Breakdownof
thesecellularcomponentsbythepathogenresultsintheaccumulationofgels[135].
Symptoms of laurel wilt on avocado, redbay, and swampbay include rapid wilting of foliage
and vascular discoloration (Figures 1D,E and 2). Xylem blockage associated with tylose and gel
formation appears to be at least partially responsible for the wilting symptoms [122,123]. Xylem
function (the ability to conduct water) in avocado is impaired as soon as 3 days after inoculation,
beforethedevelopmentofexternalorinternalsymptomsofthediseaseareapparent[123](Figure3).
Treemortalityisassociatedwithvascularfunctionalitiesoflessthan10%.
Forests2017,8,48 10of27
Artificial inoculations with as few as 100 conidia of R. lauricola can kill avocado and swamp
bay [136]. After laurel wilt-susceptible (avocado and swamp bay) and tolerant (camphortree)
hosts were inoculated with a green fluorescent protein (GFP)-labelled strain of R. lauricola,
Campbelletal. [120] reported that the pathogen was scarcely visible in microscopic cross sections,
evenindeadordyingplants. Althoughtheyobservedthatamaximumof0.9%ofthexylemlumena
ofavocadowerecolonizedbytheGFP-labelledstrain30daysafterinoculation(dai)(Figure4),about
40% of the lumena of avocado were occluded by tyloses 21 dai in another study [123] (Figure 5).
Mobilityofthepathogenoritsmetabolitesinthexylemseemstoberelatedtosusceptibility.
Symptomdevelopmenthasbeenassociatedwithreducedwatertransportinother,similartree
diseases. Wych elm affected by Dutch elm disease (caused by Ophiostoma novo-ulmi Brasier) [137],
Forests 2017, 8, 48  10 of 26 
bitternuthickoryaffectedbyhickorydecline(CeratocystissmalleyiJ.A.JohnsonandT.C.Harr.[138],
QuQerucuerscussp psp.p.a faffefcetcetdedb byy JaJappaanneessee ooaakk wwiilltt ((RR.. qquueerrcciivvoorraa) )[[113399],] ,aanndd NNotohtohliotlhitohcaorcpaurps u(sfo(rfmoremrlye rly
LithLoitchaorcpaursp)uds)e ndesniflsoifrloursu(sH (Hoookok..a annddA Arrnn..)) MMaannooss,, CCaannnnoonn aanndd SS.H.H.O.Oh hafaffefcetcetde dbyb syusduddedne onako adkeadteha th
(Ph(yPthoypthotphhotrhaorraa mraomrourmumW Wererrersese teta al.l).)[ 1[14400]] aallll eexxhhiibbiitteedd rreedduucceedd xxyylelmem fufnucntciotino wnhwichhi cwhaws, aisn, tiunrntu, rn,
corcreolrarteeladtewdi wthitshy smypmtpotmomd edveevleolpopmmenent.t.F Foorr eexxaammppllee,, PPaarrkk eett aal.l .[1[1338]8 d]edteetcetcetde dana innvinevrseer sreelaretiloantisohnipsh ip
betbweetweneexny xleymlemsa spapfl oflwowa nadnds aspapwwooooddi ninffeeccttiioonn bbyy CC.. ssmmaalllleeyyii. .TThheeyy pprorpoopsoesde dthtaht attyltoysleo sfeorfmoramtioanti on
indiuncdeudcebdy biyn fiencfteicotnionw wasarse rsepspoonnsisbiblelef foorrr reedduucceedd wwaatteerr ttrraannssppoorrt tinin afaffefcetcetde dtretreese; sC;.C sm.samlleayllie cyaiucsaeuds ed
mumltiupllteipclae nckaenrksearns danredd urecdeudcxeydl exmylefumn cftuinonct,iorens, urletisnugltiinngc rionw cnrowwinlt wanildt daencdl indeecolifnbei totefr nbuitttehrnicukto ry.
hickory. 

FigFuirgeu2re. F2.o Flioalriasry smymptpotmomsso fofl alauurerellw wiilltto onn aavvooccaaddoo iinncclluuddee ((AA) )wwilitlitning ganadn dana oniloyi lgyregyr edyisdcioslcooralotiroanti on
thatthraatp ridaplyidplyro pgrroesgsreesssteosn teoc rnoescirso;s(iBs;) i(nBi)t iianlitsiyaml spytmomptodmev edleovpemloepnmteinnto innl yoanlpyo art ipoonrtoiofna torfe ea’ strceaen’so py,
invcoalnvoinpgy,a inlivmolivteindgn au lmimbieterdo nfuvmasbceurl aorf vtraascceusl;aar ntrdac(eCs); ainnvdo (lCv)e imnveonltvoemfgerneta otfe rgrpeoartetiro pnosrotifotnhse ocf athneo py
follcoawnoepdyb fyoldloewfoeldia btiyo ndewfoitlihaitnioan mwoitnhtinh ao rmtwonotho fors ytmwop toofm syomnpsetotm(f aornmseot r(efarra mpiodrley rtahpaindloy ctchuarns in
occurs in redbay and other native Persea spp.) (photos: R. Ploetz). 
redbayandothernativePerseaspp.)(photos:R.Ploetz).