Fungal Genetics and Biology 56 (2013) 147–157

Contents lists available at SciVerse ScienceDirect

Fungal Genetics and Biology

journal homepage: www.elsevier .com/locate /yfgbi

An inordinate fondness for Fusarium: Phylogenetic diversity of fusariacultivated by ambrosia beetles in the genus Euwallacea on avocado andother plant hosts

1087-1845/$ - see front matter � 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.fgb.2013.04.004

⇑ Corresponding author. Address: Department of Plant Pathology, Physiology, and Weed Science, Virginia Tech University, Blacksburg, VA24061, USA. Fax: +1 540 2E-mail address: mkasson@vt.edu (M.T. Kasson).

Matthew T. Kasson a,⇑, Kerry O’Donnell b, Alejandro P. Rooney c, Stacy Sink b, Randy C. Ploetz d,Jill N. Ploetz d, Joshua L. Konkol d, Daniel Carrillo d, Stanley Freeman e, Zvi Mendel e, Jason A. Smith f,Adam W. Black f, Jiri Hulcr f, Craig Bateman f, Kristyna Stefkova g, Paul R. Campbell h,Andrew D.W. Geering h, Elizabeth K. Dann h, Akif Eskalen i, Keerthi Mohotti j, Dylan P.G. Short k,Takayuki Aoki l, Kristi A. Fenstermacher a, Donald D. Davis a, David M. Geiser a

a Department of Plant Pathology & Environmental Microbiology, The Pennsylvania State University, University Park, PA 16802, USAb Bacterial Foodborne Pathogens and Mycology Research Unit, National Center for Agricultural Utilization Research, US Department of Agriculture, Agricultural Research Service,1815 North University Street, Peoria, IL 61604, USAc Crop Bioprotection Research Unit, National Center for Agricultural Utilization Research, US Department of Agriculture, Agricultural Research Service, 1815 North University Street,Peoria, IL 61604, USAd University of Florida, IFAS Tropical Research & Education Center, Homestead, FL 33031, USAe Institute of Plant Protection, ARO, The Volcani Center, Bet Dagan 50250, Israelf School of Forest Resources and Conservation, University of Florida, Gainesville, FL 32611, USAg University of South Bohemia, 370 05 Ceske Budejovice, Czech Republich Queensland Department of Agriculture, Fisheries and Forestry & The Queensland Alliance for Agriculture and Food Innovation, The University of Queensland, Ecosciences Precinct,Dutton Park, Qld 4102, Australiai Department of Plant Pathology and Microbiology, University of California, Riverside, CA 92521, USAj Entomology and Nematology Division, Tea Research Institute of Sri Lanka, Talawakele 22100, Sri Lankak Department of Plant Pathology, University of California, Davis, 1636 E Alisal Street, Salinas, CA 93905, USAl National Institute of Agrobiological Sciences, Genetic Resources Center, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan

a r t i c l e i n f o a b s t r a c t

Article history:Received 9 February 2013Accepted 11 April 2013Available online 19 April 2013

Keywords:Ambrosia fungiDivergence datingEvolutionMutualismMycangiaSymbiosis

Ambrosia beetle fungiculture represents one of the most ecologically and evolutionarily successfulsymbioses, as evidenced by the 11 independent origins and 3500 species of ambrosia beetles. Herewe document the evolution of a clade within Fusarium associated with ambrosia beetles in the genusEuwallacea (Coleoptera: Scolytinae). Ambrosia Fusarium Clade (AFC) symbionts are unusual in thatsome are plant pathogens that cause significant damage in naïve natural and cultivated ecosystems,and currently threaten avocado production in the United States, Israel and Australia. Most AFC fusariaproduce unusual clavate macroconidia that serve as a putative food source for their insect mutualists.AFC symbionts were abundant in the heads of four Euwallacea spp., which suggests that they aretransported within and from the natal gallery in mandibular mycangia. In a four-locus phylogeneticanalysis, the AFC was resolved in a strongly supported monophyletic group within the previouslydescribed Clade 3 of the Fusarium solani species complex (FSSC). Divergence-time estimates placethe origin of the AFC in the early Miocene �21.2 Mya, which coincides with the hypothesized adap-tive radiation of the Xyleborini. Two strongly supported clades within the AFC (Clades A and B) wereidentified that include nine species lineages associated with ambrosia beetles, eight with Euwallaceaspp. and one reportedly with Xyleborus ferrugineus, and two lineages with no known beetle associa-tion. More derived lineages within the AFC showed fixation of the clavate (club-shaped) macroconidialtrait, while basal lineages showed a mix of clavate and more typical fusiform macroconidia. AFC lin-eages consisted mostly of genetically identical individuals associated with specific insect hosts indefined geographic locations, with at least three interspecific hybridization events inferred based ondiscordant placement in individual gene genealogies and detection of recombinant loci. Overall, thesedata are consistent with a strong evolutionary trend toward obligate symbiosis coupled with second-ary contact and interspecific hybridization.

� 2013 Elsevier Inc. All rights reserved.

31 7477.

148 M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157

1. Introduction

With the exception of humans, the only animals known to en-gage in ecologically significant agriculture are higher attine leaf-cutting ants, termites, some bark beetles and ambrosia beetles(Mueller et al., 2005). Ambrosia beetles are wood-inhabiting, obli-gate mutualistic insects that construct galleries within which theycultivate fungi for food (Hulcr et al., 2007). One of the most ecolog-ically successful groups of ambrosia beetles are Xyleborini (Coleop-tera: Scolytinae). The spectacular adaptive radiation of their 1200species over the past 20 million years is attributed to their obligatemutualism with xylem-inhabiting ambrosia fungi, their extremeinbreeding resulting from obligate sib-mating and haplodiploidy,and a sex-ratio highly skewed towards females (Farrell et al.,2001; Jordal et al., 2000). Almost all known ambrosia beetles pos-sess specialized pockets called mycangia within which they trans-port fungi from their natal galleries to newly colonized trees (Hulcrand Cognato, 2010).

Beetles infest dying or dead trees, but they may damage or in-cite dieback and/or mortality in apparently healthy trees followinginoculation of their symbiotic fungi (Hulcr and Dunn, 2011; Kuhn-holz et al., 2001). For example, the invasive Asian ambrosia beetleXyleborus glabratus, whose nutritional symbiont Raffaelea lauricolacauses a lethal wilt in American Lauraceae, has decimated redbay(Persea borbonia) populations in the coastal regions of the south-eastern United States (Fraedrich et al., 2008). Recently, severe die-back and mortality of avocado and other tree hosts has beenobserved in Israel and in California caused by another Asianambrosia beetle, which was originally reported as the tea shot-holeborer Euwallacea fornicatus, and its novel Fusarium symbiont (Esk-alen et al., 2012; Mendel et al., 2012). The latter two studies estab-lished that the Fusarium species alone is capable of causing diseasewhen inoculated on avocado in sufficient density. Although mor-phologically similar beetles have been recovered from avocado inAustralia and Florida, different symbionts and less damage havebeen observed than in the above situations (Ploetz et al.,unpublished). In contrast, E. fornicatus is known to damage tea inplantations in India (Gadd and Loos, 1947) and Sri Lanka (Danthan-arayana, 1968), and the muga silkworm host Persea bombycina inIndia (Kumar et al., 2011), but their Fusarium symbionts have notbeen tested as a factor in damaging the plant host.

Symbionts of ambrosia beetles are mostly ascomycetous fungi inthe orders Ophiostomatales and Microascales (Alamouti et al.,2009); however, several ambrosia beetle associations with Fusarium(Hypocreales) have been documented (Brayford, 1987; Gadd andLoos, 1947; Norris and Baker, 1967). The first known Fusarium sym-biont of ambrosia beetles was originally misclassified as Monacros-porium ambrosium because it produces clavate rather than iconicfusiform macroconidia (Gadd and Loos, 1947). Unaware of the origi-nal description, Brayford (1987) described this species four decadeslater as Fusarium bugnicourtii based on collections from E. fornicatusgalleries in tea in India and rubber tree and cacao in Malaysia. F. bug-nicourtii was later recognized as the same fungus as M. ambrosium,and synonymized under the new combination Fusarium ambrosium(Nirenberg, 1990), which remains typified based on Gadd and Loos’collections from the shot-hole borer on tea. Multilocus phylogeneticanalyses later showed that despite its morphological distinctive-ness, F. ambrosium was strongly supported as a member of the Fusar-ium solani species complex (FSSC; O’Donnell et al., 2008). The FSSCincludes three clades encompassing over 60 species recognizableby genealogical concordance phylogenetic species recognition(GCPSR; Taylor et al., 2000). In the alphanumeric system for identi-fying species and multilocus sequence types in the FSSC, F. ambro-sium was designated phylogenetic species FSSC 19; however, onlytwo isolates from shot-hole borer galleries in tea from India were

included in this analysis (haplotypes FSSC 19-a and 19-b). Two spe-cies with no known connection to ambrosia beetles are closely re-lated to F. ambrosium, namely Fusarium pseudensiforme NRRL46517 and F. cf. ensiforme NRRL 22354 (Nalim et al., 2011). F. pseu-densiforme was isolated from perithecia on a recently killed tree inSri Lanka. In contrast to the clavate conidia produced by F. ambro-sium and several other taxa within the AFC, both species producesymmetrical macroconidia typical of Fusarium and specifically theFSSC (Nalim et al., 2011).

The present study was initiated in response to the growingthreat that invasive fungus-farming Euwallacea ambrosia beetlespose to forest ecosystems (Hulcr and Dunn, 2011), urban land-scapes (Eskalen et al., 2013), and avocado production worldwide(Eskalen et al., 2012; Mendel et al., 2012). The primary researchobjectives were to (i) elucidate incidence and prevalence of ambro-sia fusaria and other fungi in several invasive Euwallacea specieswithin the United States, (ii) develop a phylogenetic frameworkfor assessing the evolutionary origins of the ambrosia fusaria andtheir phylogenetic diversity based on analyses of a four-gene dataset, (iii) formulate hypotheses of reproductive mode from DNA se-quence analyses of MAT1-1-1 and MAT1-2-1 amplicons, and (iv)determine whether ambrosia fusaria evolution coincided withthe adaptive radiation of Euwallacea. The work was facilitated bythe discovery of infestations of the ambrosia beetle Euwallacea val-idus within dying stands of the invasive Ailanthus altissima (tree-of-heaven) killed by Verticillium wilt in Pennsylvania (Schall and Da-vis, 2009).

2. Materials and methods

2.1. Sampling and isolation

Forty of the 44 fusaria analyzed were isolated from beetles,their galleries, or from trees showing extensive borer damageand dieback (Supplemental Table S1). Several Euwallacea speciespresent in the United States were targeted: E. validus (Pennsylva-nia), E. interjectus (Gainesville, Florida), and at least two Euwall-acea spp. (southern Florida and southern California) thatresemble E. fornicatus but appear to comprise cryptic phyloge-netic species. Beetles were surface disinfested for 15 s in 70%ethanol and then washed three times in sterile deionized water.Whole beetles or heads were macerated using sterile Tenbroekgrinders (Pyrex), or pellet pestles (Fisher), diluted 1:10 and1:100, and spread evenly over half-strength Potato Dextrose Agar(PDA: Difco) amended with 100 ppm streptomycin sulfate (Sig-ma–Aldrich). Plant tissue was surface disinfested for 15 s in70% ethanol, 2 min in 10% household bleach, and washed threetimes in sterile deionized water and plated either in wholepieces or macerated as described above on half-strength PDA.In addition, 11 ambrosia fusaria or close relatives within theFSSC were obtained from culture collections (SupplementalTable S1). All isolates are available from the U.S. Departmentof Agriculture, Agricultural Research Service Culture Collection(NRRL) at the National Center for Agricultural Utilization Re-search (NCAUR) in Peoria, Illinois (http://nrrl.ncaur.usda.gov).

2.2. DNA sequence acquisition and analysis

Portions of the following four genes were PCR amplified andsequenced based on their proven utility for resolving phylogeneti-cally distinct members of the F. solani species complex (O’Donnellet al., 2008): a portion of the nuclear ribosomal RNA gene repeat(rDNA) comprising the internal transcribed spacer region (ITS) plusdomains D1 and D2 of the nuclear ribosomal large subunit (LSU,

M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157 149

1004 bp total alignment); translation elongation factor 1a (EF-1a,687 bp alignment); DNA-directed RNA polymerase II subunit 1(RPB1, 1588 bp alignment) and DNA-directed RNA polymerasesubnit 2 (RPB2, 1635 bp alignment). PCR primers and conditions,sequencing and phylogenetic analyses were performed as re-ported in O’Donnell et al. (2013). Mating-type determinationwas performed as reported in Short et al. (2013), except MAT1-1 and MAT1-2 PCR reactions were performed separately; multi-plex PCR combining primer pairs targeting both idiomorphs pro-duced inconsistent results. In addition, several analysis methodsin RDP4 ver. 4.16 (Martin et al., 2010; http://darwin.uvigo.es/rdp/rdp.html) were implemented to detect possible recombina-tion breakpoints in sequences of putative interspecific hybridsand identify their parents from analyses of a concatenated 14-taxon data set that included one sequence of each of the 12 spe-cies together with putative hybrid isolates F. ambrosium NRRL46583 and 62605. The concatenated 40-taxon four-locus NEXUSfile and the most parsimonious tree were deposited in TreeBASE(Accession number S13974, tree number Tr61840) and thenucleotide sequence data was deposited in GenBank (AccessionNumbers KC691528–KC691674). Divergence times were esti-mated as described in O’Donnell et al. (2013), with Sphaerostil-bella aureonitens sequences added as an additional calibrationpoint.

2.3. Conidial morphology

Isolates were cultured on synthetic low-nutrient agar (SNA:0.2 g glucose, 0.2 g sucrose, 1 g KH2PO4, 1 g KNO3, 0.5 g MgSO4�7H2O, 0.5 g KCl, 20 g Difco agar per liter; Nirenberg, 1990) at 28 �Cin the dark for 7–10 days to investigate conidial morphology. Con-idia were suspended in 0.25% Tween 60 and examined with a ZeissAxio Imager A1 upright microscope fitted with Nomarski differen-tial interference contrast optics, utilizing a Plan-Neofluar 40�/0.75Ph2 objective, and an AxioVision ver. 4.6 LE Canon module for im-age acquisition.

2.4. Characterization of Euwallacea spp. and the fungi they vector

To assess the presence of mandibular mycangia (Nakashima,1982), six females and two males of E. validus collected in Penn-sylvania (for collection sites see Supplemental Fig. S1) wereembedded in paraffin as described in Fraedrich et al. (2008). Arotary microtome was used to take 10 lm transverse sectionsthrough each of the eight heads. Sections were double stainedwith Harris-hematoxylin and eosin–phloxine and then examinedand photographed using an Olympus BX51 compound micro-scope equipped with a Jenoptik ProgRes CFscan CCD Researchcamera.

All beetle identifications were performed using morphologicalcriteria (see http://www.ambrosiasymbiosis.org/northamericanxyleborini/). Fungal isolations and quantifications were performedin separate laboratories, using comparable protocols. For E. validus,fungi were isolated from 84 live female adults, 16 male adults, fourimmature females, 10 pupae, and 10 larvae. Fungi from separatedheads, thoraces and abdomens were isolated from 36 adult femaleE. validus and only from heads for the remaining 48. Heads from 16Euwallacea sp. from galleries in symptomatic avocado in Florida, 4Euwallacea sp. from galleries in symptomatic avocado in California,and 13 E. interjectus from galleries in water-stressed box elder inFlorida were also selected for fungal isolation and quantification.Colony forming units (CFUs) representing unique morphotypeswere counted from platings on PDA performed as described inSection 2.1.

3. Results

3.1. Phylogenetic diversity of fusaria associated with Euwallacea

All three methods of phylogenetic inference resolved all ambro-sia fusaria in a strongly supported monophyletic group, designatedthe Ambrosia Fusarium Clade (AFC), within which nine of the 11lineages were isolated exclusively from ambrosia beetles, their gal-leries, or heavily infested plant material. In the absence of formalnames, a numeric system is used to distinguish the eight unde-scribed lineages associated with ambrosia beetles, with F. ambro-sium as AF-1 and unnamed genealogically exclusive lineagesdesignated AF-2 through AF-9. The phylogeny resolved a basal splitbetween an early diverging Clade A (Fig. 1F), in which two of fourtaxa were associated with xyleborines (i.e., AF-8 ex Euwallacea sp.in southern Florida and AF-9 ex Xyleborus ferrugineus in Costa Rica),and Clade B in which all seven lineages were recovered from vari-ous Euwallacea species or their galleries, or from a rubber tree withextensive shot-hole borer damage and dieback symptoms inMalaysia (i.e., AF-5). Only two of the 11 species-level lineageswithin the AFC have been described formally: F. pseudensiformefrom a dead tree in Sri Lanka within Clade A (Nalim et al., 2011)and F. ambrosium (AF-1) from the tea shot-hole borer E. fornicatusand its galleries from Sri Lanka and India within Clade B. Thus,eight of the nine fusaria associated with ambrosia beetles or theirgalleries appear to represent novel phylogenetically distinctlineages.

MP and ML bootstrap analyses of the individual partitions(Fig. 1A–E, Supplemental Table S2) revealed RPB1 (Fig. 1B) andthe RPB2 5–7 region (Fig. 1C) possessed the highest proportion ofparsimony informative characters (both at 9.7% PIC/bp) and sup-ported the largest number of nodes (13 and 10–11 at P70% MP/ML BS, respectively). By contrast, the ITS+LSU rDNA partition wasthe least informative (Fig. 1E), with 2.3% PIC/bp and five nodesreceiving P70% MP/ML BS. MP and ML bootstrapping and Bayesiananalysis of the combined four-locus data set provided support forthe largest number of nodes: 15 with P70% MP-BS, 17 withP70% ML-BS, and 19 with P95% B-PP (Fig. 1F).

3.2. Interspecific hybridization within the AFC

Three isolates of F. ambrosium possessed RPB2 or EF-1a allelesthat suggested hybrid ancestry (Figs. 1 and 2). NRRL 62605 showedevidence for tripartite ancestry. It possessed an RPB2 allele that ap-peared to be a recombinant of sequences from an AF-4-like parentand an unknown Clade B species (Fig. 1C and D and 2A, Supple-mental Table S3), whereas its EF-1a, RPB1 and rDNA alleles wereeither identical to or phylogenetically nested within F. ambrosium.NRRL 22345 and 46583 possessed EF-1a alleles that were nearlyidentical to all known isolates of AF-4, but alleles at all other lociclearly resolved with F. ambrosium (Figs. 1A and 2B, SupplementalTable S3). Because of the hybrid ancestry of these alleles, F. ambro-sium and AF-4 do not resolve as genealogically exclusive in themultilocus phylogeny unless the hybrid alleles are removed. Sev-eral methods implemented in RDP4 ver. 4.16 (Martin et al., 2010)were used to identify recombination breakpoints in the NRRL62605 RPB2 and NRRL 22345/46583 EF-1a alleles in a concatenatedalignment (50 > RPB1-EF-1a-RPB2-rDNA) that also included one iso-late of the 12 species (Fig. 2). The GENECONV, BootScan and 3Seqmethods detected putative recombination breakpoints withinEF-1a of NRRL 22345 and 46583. NRRL 20438 F. ambrosium wasidentified as the major parent, but the minor parent could not beidentified. Seven of the methods implemented in RDP4 identifiedNRRL 62605 as a recombinant and NRRL 20438 F. ambrosium asthe major parent. However, recombination breakpoints were

A B

C D

Fig. 1. (A–E) One most-parsimonious tree (MPT) inferred from each of the four genes sequenced. Eight fusaria [AF-1 through AF-8] associated with Euwallacea spp. arehighlighted in gray. Darker gray highlight with black border is used to identify six isolates of Fusarium ambrosium. Bold font indicates three putative hybrid isolates of thelatter (i.e., NRRL 22345, 46583 and 62605); note their discordant placement in the EF-1a (A), RPB2 5–7 (C) and RPB2 7–11 (D) genealogies. The latter two regions wereanalyzed separately to highlight discordant placement of triparental hybrid NRRL 62605 F. ambrosium. Numbers above nodes represent maximum parsimony and maximumlikelihood bootstrap values P70% (MP-BS/ML-BS) based on 1000 pseudoreplicates of the data. Phylograms were rooted on sequences of F. neocosmosporiellum (formerlyknown as Neocosmospora vasinfecta). (F) One of two MPTs inferred from combined four-gene data set. Bold internodes identify two strongly supported early diverging sisterlineages designated Clades A and B. MP-BS/ML-BS bootstrap values P70% are indicated above nodes; Bayesian posterior probabilities P0.98 (B-PP) are indicated belownodes. Note the phylogeny was nearly fully resolved by the Bayesian analysis. CI, consistency index; RI, retention index.

150 M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157

Fig. 1. (continued)

A

B

Fig. 2. Putative interspecific hybrids of Fusarium ambrosium. Alleles are color-coded as follows: yellow = F. ambrosium, pink = Fusarium sp. only represented by introgressedRPB2 allele in NRRL 62605 F. ambrosium, blue = Fusarium sp. AF-4-like parent. Only variable nucleotide positions within the alignments are shown. (A) Evidence ofinterspecific hybridization and intragenic recombination within RPB2 of triparental hybrid NRRL 62605 F. ambrosium. An arrow identifies the location of the recombinationbreakpoint between nucleotide positions 536–593 within the 5–7 region. The red triangle identifies a polymorphism at nucleotide position 900 within the 7cf and 7crpriming site. (B) Evidence suggesting the EF-1a allele in NRRL 22345 F. ambrosium may have been acquired from an AF-4-like parent.

M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157 151

detected within the RPB2 region by only three of the methodsimplemented in RPB4 (i.e., GENECONV, Bootscan and 3Seq). In eachof these analyses, the minor parents of the RPB2 5–7 and 7–11regions were identified, respectively, as an unknown species andAF-4.

After the three putative hybrid F. ambrosium alleles were ex-cluded from the analyses, MP and ML bootstrapping of the individ-ual genes showed strong genealogical concordance. Monophyly ofthe eight fusaria putatively cultivated by Euwallacea spp. wasstrongly supported by one to all four individual partitions and/orthe combined data set (Supplemental Table S3). Within the earlydiverging Clade A, AF-8 from Euwallacea sp. on avocado cv. Nesbitt

in southern Florida was resolved as a basal sister to a moderatelysupported subclade (MP-BS <70%, ML-BS 91%, B-PP 1) that con-tained AF-9 from the ambrosia beetle X. ferrugineus in Costa Ricatwo sister taxa (i.e., F. pseudensiforme and F. cf. ensiforme) notknown to be associated with ambrosia beetles. F. pseudensiformewas the only species within the AFC isolated from perithecia.

Support for the branching order of the three earliest diverginglineages within Clade B was significantly supported only in Bayes-ian analysis of the combined data set (Fig. 1F). Although the earli-est diverging lineage within Clade B was supported by a B-PP of0.98, a sister-group relationship between AF-6 from Euwallaceasp. and their galleries in avocado cv. Nesbitt in southern Florida

Fig. 3. Maximum clade credibility (MCC) tree showing divergence-time estimates for the Ambrosia Fusarium Clade (AFC) generated using an uncorrelated lognormal relaxedmolecular clock with a birth–death model as the tree prior. Numbered circles identify 12 nodes for which divergence times are reported. Calibration taxa were pruned fromthe chronogram in order to focus attention on ambrosia fusaria (see Supplemental Fig. S3 for MCC tree with calibration taxa). We speculate that production of clavatemacroconidia by seven of eight fusaria (=AF-1-AF-5 and AF-7-AF-8) associated with Euwallacea spp. may represent an adaptation for the symbiosis. Geological time scale is inmillions of years before present (Mya, Walker et al., 2009). HPD = highest probability density interval, Paleo = Paleocene, Pl = Pliocene, P = Pleistocene. Scale bar, 25 lm.

152 M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157

and AF-7 from a Euwallacea sp. and its gallery in avocado inQueensland, Australia received no statistical support. Triparentalhybrid F. ambrosium strain NRRL 62605 nests within this earlydiverging lineage in analyses of the combined data set; however,this placement is discordant with the species phylogeny inferredfrom a combined three-locus data set, excluding the RPB2 partition,which strongly placed it in F. ambrosium (100% MP-BS, 100% ML-BS; result not shown). Similarly, putative hybrids NRRL 22345and 46583 were strongly supported as part of a genealogicallyexclusive F. ambrosium (100% MP-BS, 100% ML-BS; result notshown) when the EF-1a partition and NRRL 62605 were excludedfrom the analyses. The three terminal-most nodes along the back-

bone of the phylogeny and the genealogical exclusivity of the sevenEuwallacea-associated fusaria within Clade B (Fig. 1F) receivedmodest to strong clade support in analyses of the four-locus dataset (Supplemental Table S3).

To address the possibility that interspecific hybridization be-tween Fusarium ambrosium and two other species within the AFCwas mediated by a sexual process, all 44 AFC isolates werescreened for mating type to obtain a preliminary estimate ofwhether they might be self-sterile (i.e., possess only one MAT1 idi-omorph) or self-fertile (i.e., possess both MAT1 idiomorphs). Re-sults of a uniplex PCR assay targeting a portion of the MAT1-1-1and MAT1-2-1 genes (Supplemental Table S1) indicated both were

M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157 153

present in all of the species within Clade A except for NRRL 46517F. pseudensiforme, where only MAT1-1-1 was detected. This resultwas confirmed via DNA sequencing of the MAT1-1-1 amplicons (re-sult not shown). Of the seven species within Clade B, the presenceof potential homothallics was detected where both idiomorphswere amplified and confirmed in one of nine isolates of AF-2 in Cal-ifornia/Israel, two of six isolates of AF-3 from E. interjectus inGainesville, Florida, one of five isolates of AF-4 from E. validus inPennsylvania, and the two isolates of AF-6 tested from Euwallaceasp. in southern Florida, (Supplemental Table S1). Three isolates of F.ambrosium AF-1 from India, two isolates of AF-5 from Malaysia,and one isolate of AF-2 from California also produced PCR resultsindicating the presence of both idiomorphs, but only one of theamplicons in each of these strains was successfully confirmed byDNA sequencing. Of the 19 remaining strains within Clade B,amplicons could not be obtained for one isolate (SupplementalTable S1); the 18 remaining strains were typed as mating typeMAT1-1 (one putative hybrid isolate of F. ambrosium and allremaining isolates of AF-2 and AF-5) or MAT1-2 (all remaining iso-lates of AF-1, AF-3 and AF-4).

3.3. Evolution of clavate macroconidia and divergence-time estimateswithin the AFC

Seven of the nine fusaria associated with ambrosia beetles pro-duce clavate (club-shaped) macroconidia (Fig. 3). The only twoexceptions were AF-6 from Euwallacea sp. in southern Floridaand AF-9 from X. ferrugineus in Costa Rica, which only producedfusiform macroconidia in culture. AF-8 is also exceptional becauseit is the only Euwallacea-associated species that produced both cla-vate and fusiform macroconidia from colonies derived from singlemicroconidia (Fig. 3). The transition to exclusive production of cla-vate macroconidia is equivocal due to the unresolved branching or-der of AF-6 and AF-7 (Fig. 2F). However, because the five mostderived species within Clade B only produced clavate macroco-nidia, this character appears to be fixed in the most recent commonancestor of AF-5 from Malaysia. Topological constraints conductedin PAUP� (Swofford, 2003) forcing the monophyly of fusaria associ-ated with Euwallacea, Xyleborini or lineages that produce clavatemacroconidia were nine to 34 steps longer and significantly worsethan the unconstrained MPTs when subjected to the Kishino–Hase-gawa likelihood test (P < 0.05, Supplemental Table S4).

Bayesian-derived divergence time estimates obtained usingBEAST ver. 1.7.1 (Drummond and Rambaut, 2007) suggest the

Fig. 4. Female Euwallacea validus beetles possess mycangia containing novelFusarium and Raffaelea symbionts. Transverse section through paraffin-embeddedhead of female Euwallacea validus showing paired mandibular mycangia (arrows)double-stained with Harris-hematoxylin and eosin–phloxine.

AFC diverged in the early Miocene 21.2 Mya [95% HPD interval:13.3, 31.6] (Fig. 3, node 1) and the split between Clades A and B oc-curred during the middle Miocene 12.5 Mya [95% HPD interval:7.8, 19.1] (Fig. 3, node 2). Subsequent evolutionary diversificationof both clades was dated to the late Miocene, with Clade A[10.0 Mya, 95% HPD interval: 5.8, 15.5] estimated to be roughlytwice as old as Clade B [6 Mya, 95% HPD interval: 3.5, 9.6]. Theshort internodes subtending the seven ambrosia fusaria withinClade B (Figs. 1 and 3) suggests that they radiated rapidly overthe past 6 million years.

3.4. Female Euwallacea spp. harbor novel fungal symbionts withincephalic mycangia

Serial transverse sections taken through the heads of six paraf-fin-embedded adult female E. validus beetles recovered from gal-leries in A. altissima confirmed the presence of one set of pairedmandibular mycangia (sensu Six, 2003) containing yeast-like cellsin all sampled individuals (Fig. 4) similar to those in other ambro-sia beetles (for example, the redbay ambrosia beetle, X. glabratus;see Fig. 7 in Fraedrich et al., 2008). By comparison, sectionsthrough the heads of two males indicated that cephalic mycangiawere absent (results not shown). Colony counts of the mycobiotaassociated with E. validus life stages (i.e., larvae, pupae, adultmales, and immature females) and adult female body regions(i.e., head, thorax, and abdomen), as well as the heads of femaleE. interjectus (Gainesville, Florida), Euwallacea sp. (southern Flor-ida), and Euwallacea sp. (southern California) showed that AFCfusaria were consistently and abundantly found in adult femaleheads (Supplemental Table S5) Adult female E. validus heads(N = 84) harbored two phenotypically different colony morpho-types with nearly equal proportions: Morphotype 1 (473 CFUs/head) and Morphotype 2 (504 CFUs/head) (Supplemental Fig. S2).Morphotypes recovered from E. validus heads were identifiedthrough blastn searches of GenBank with EF-1a and LSU sequencedata. Morphotype 1 was identified as a Fusarium sp. based on 99%identity of its EF-1a sequence with GenBank accession DQ247552Fusarium sp. NRRL 22231 from rubber tree (Hevea brasiliensis) inSabah, Malaysia exhibiting dieback symptoms, a host reportedfor E. fornicatus (Danthanarayana, 1968). Morphotype 2 was iden-tified as a putatively novel species of Raffaelea, an ophiostomatoidanamorph and a well-known genus of ambrosia symbionts, basedon 95% identity of its LSU sequence with GenBank accessionEU177467 Raffaelea sp., associated with the ambrosia beetle Platy-pus externedentatus from South Africa. Importantly, plating of indi-vidual body regions for 36 adult female E. validus revealed that theFusarium sp. AF-4 and Raffaelea sp. were abundant in heads but notobserved in thoraces or abdomens (Supplemental Fig. S2). Despitenearly equal proportions of the two species in E. validus, Raffaeleasp. was not recovered from 13 (15%) of 84 adult female beetleheads sampled, whereas Fusarium sp. was not recovered from only2 (2%) of 84 beetle heads. Fusarium sp. and a third colony morpho-type (Morphotype 3) accounted for 80% (646) and 20% (170),respectively, of the CFUs from E. validus larvae (SupplementalFig. S2). Morphotype 3 was identified as Graphium sp., a microasca-lean synnematous fungus, based on 99% identity of its ITS sequencewith Graphium penicillioides and Graphium sp. (GenBank accessionsFJ914670.1 and FJ914689.1, respectively). Graphium sp. was recov-ered from the thoraces (�X = 98 CFUs, N = 5) and abdomen(�X = 81 CFUs, N = 6) of adult female E. validus and adult male heads(�X = 21 CFUs, N = 4). Raffaelea sp. was never recovered from larvae.By contrast, adult male E. validus heads (N = 16) and pupae (N = 10)contained only �30 CFUs that were mostly Fusarium sp. and Graph-ium sp., respectively (Supplemental Fig. S2).

Mean CFUs from the heads of adult female Euwallacea spp. pro-duced abundant Fusarium colonies as follows: Euwallacea sp. from

154 M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157

southern Florida: 2683 CFUs (N = 16); E. interjectus from Gaines-ville, Florida: 1475 CFUs (N = 13); E. validus from Pennsylvania:473 CFUs (N = 84); and Euwallacea sp. from southern California:67 CFUs (N = 4) (Supplemental Table S5). The proportion of Fusar-ium in total fungal CFUs was consistent across all insect hosts(44–63%) (Supplemental Table S5). Similar to E. validus, fusariarecovered from the heads of Euwallacea sp. from Florida and south-ern California were absent from the abdomens and thoraces of allindividuals examined, with the exception of one Euwallacea sp. col-lected in Florida, from which 2 CFUs were recovered. Heads of theEuwallacea sp. from California had nearly equal proportions ofFusarium and a synnematous fungus similar to Morphotype 3(Graphium sp.) from E. validus larvae and pupae (SupplementalTable S5). Raffaelea sp. was found only in E. validus.

4. Discussion

4.1. Phylogenetic diversity of fusaria associated with Euwallacea

A key finding of the present study is the discovery that at leastnine morphologically cryptic fusaria are associated with ambrosiabeetles, including four with Ewallacea spp. vectors that threatenavocado production in Israel, Australia, and the United States.Fusarium fungiculture by Euwallacea was initially discovered andcharacterized in the shot-hole borer of tea, E. (Xyleborus) fornicatus,in Sri Lanka (Gadd and Loos, 1947) and recently on avocado exhib-iting wilt and dieback symptoms in Israel (Mendel et al., 2012) andCalifornia (Eskalen et al., 2012). GCPSR-based analyses of multilo-cus DNA sequence data indicate the symbiont on avocado in Israeland California, designated AF-2, is phylogenetically distinct from F.ambrosium AF-1 (Eskalen et al., 2012; Mendel et al., 2012).Although the beetle was identified phenotypically as E. fornicatusin the latter two studies, DNA analyses suggest it may be a novelspecies (R. Stouthamer et al., http://www.avocadosource.com/).Morphological crypsis within the Xyleborini, which makes it diffi-cult to distinguish between closely related species, has been attrib-uted to their rapid adaptive radiation over the past 20 My and lackof selection on secondary sexual characteristics (Farrell et al., 2001;Jordal et al., 2000). Fusarium symbionts within Clade B, which arealso morphologically cryptic, appear to have radiated rapidly overthe past 6 My, judging from the short internodes and weak supportfor the branching order for the early diverging lineages (Figs. 1Fand 3).

Results of the present study indicate that, in addition to AF-2 inIsrael and California, at least three other ambrosia fusaria are cul-tivated by Euwallacea spp. in avocado, and these include AF-6 andAF-8 in southern Florida, and AF-7 in Queensland, Australia. Of thethree remaining known ambrosia fusaria, AF-3 was recovered fromE. interjectus colonies in water-stressed box elder (Acer negundo) inGainesville, Florida, AF-4 was cultivated by E. validus in Verticilliumwilt-stressed tree-of-heaven (A. altissima) in south-central Penn-sylvania (Schall and Davis, 2009), and AF-5 was isolated fromborer-damaged rubber trees (H. brasiliensis) exhibiting extensivedieback in Sabah, Malaysia (northern Borneo; Brayford, 1987).Phenotypically similar fusaria were isolated from symptomaticTheobroma cacao in Sabah, but these isolates are no longer viable.Additionally, ‘F. solani’ has been associated with sudden death ofT. cacao in Papua New Guinea (Prior, 1986). Although Danthanara-yana (1968) lists H. brasiliensis and T. cacao as hosts of the teashot-hole borer (E. fornicatus), molecular phylogenetic analyses offusaria and beetle vectors associated with these plant hosts areneeded to determine their identity, given the widespread morpho-logical crypsis within Fusarium and Euwallacea.

The mutualistic relationship suggests the AFC lineages arereproductively isolated species. This is based on the overall pattern

of genealogical exclusivity observed, and known biological forcesmaintaining predominantly vertical transmission, such as the obli-gate feeding requirement of Euwallacea sp. in Israel for its AF-2symbiont (Freeman et al., 2012). However, almost all of the intra-specific genetic variation observed in this study is better explainedby inter-lineage hybridization events rather than intraspecific alle-lic diversity. An alternative hypothesis is that the strong symbioticrelationship between the fungi and their xyleborine mutualists hasdriven rapid molecular evolution without biological reinforcementof reproductive isolation, reflected in the evidence for hybridiza-tion observed in this study. Phylogenomic data are expected tonot only help better define species-level lineages, but also eluci-date the forces that drive evolution of the invasive beetles andtheir potentially pathogenic symbionts.

4.2. Interspecific hybridization within the AFC

The non-monophyly of F. ambrosium within the EF-1a and RPB2phylogenies appears to reflect interspecific hybridization ratherthan incomplete lineage sorting or ancient paralogy followed bygene loss (Schardl and Craven, 2003). Evidence in support ofhybridization includes (i) the EF-1a allele in NRRL 22345 and46583 F. ambrosium is closely related to one found in AF-4, (ii)the recombinant RPB2 allele in NRRL 62605 F. ambrosium suggeststhis isolate has a tripartite ancestry, given that the 5–7 and 7–11regions appear to have been acquired, respectively, from an un-known parent within Clade A of the AFC and an AF-4-like parent,and (iii) F. ambrosium, AF-4 and the unknown donor of the RPB25–7 allele do not appear to be sisters. We hypothesize that AF-4,or a close relative, and a second novel Fusarium sp., from whichthe divergent RPB2 5–7 region was acquired, introgressed into F.ambrosium on a common host in India and/or Sri Lanka, or perhapsin secondary contact after introduction elsewhere. Given the tripa-rental origin of F. ambrosium NRRL 62605, it is possible that thebeetle from which it was isolated may cultivate multiple phyloge-netically distinct fusaria on the same plant host in Asia. An alterna-tive though not mutually exclusive explanation is that severalspecies of Euwallacea cultivate different fusaria on the same host.In addition to phylogenetic evidence of genetic exchange betweenspecies, several substitution methods (i.e., MAXCHI, GENECONV,3SEQ) and the BOOTSCAN phylogenetic method implemented inRPB4 (Martin et al., 2010) detected recombination blocks withinEF-1a and RPB2 in the concatenated alignment and identified theparental and recombinant sequences. These findings are notewor-thy because the overall pair wise sequence divergence was well be-low the 5% minimum threshold required to obtain statistical powerusing these methods (Posada and Crandall, 2001).

The evidence suggests that both MAT1-1 and MAT1-2 mating-types are present in F. ambrosium and AF-4, either in homothallicor heterothallic forms, possibly providing a sexual basis for the inter-specific hybridization. However, their highly clonal populationstructure, as in the asexual Neotyphodium grass endophytes (Schardland Craven, 2003), indicates that it most likely involved somatic fu-sion of vegetative hyphae followed by a parasexual cycle. Results ofthe mating type assay (Supplemental Table S1) raised a number ofquestions that require further study, including: (i) Do any of theambrosia fusaria reproduce sexually? And if so, (ii) do some species(i.e., AF-2, AF-3, AF-4 and AF-6) contain heterothallic and homothal-lic isolates, as suggested by the detection of both idiomorphs in thesame individuals (Supplemental Table S1)? Exhaustive attempts toinduce sexual reproduction in culture have not been performed,but perithecia have not been observed either in putative MAT1-1/MAT1-2 cultures or in pairings of MAT1-1 and MAT1-2 isolates. Atthe present time, the heterothallic F. pseudensiforme NRRL 46517 isthe only species within the ambrosia clade that is known to repro-duce sexually (Nalim et al., 2011).

M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157 155

If some or all of the fusaria cultivated by Euwallacea spp. onlyreproduce clonally, as is generally assumed for ambrosia fungi(Mueller et al., 2005), then interspecific hybridization, as elegantlydemonstrated in the asexual mutualistic Neotyphodium endo-phytes of grasses (Moon et al., 2004; Tsai et al., 1994), might pro-vide a means by which they counteract Muller’s ratchet(Felsenstein, 1974). The enigmatic results of the mating type assayemphasize the need for integrative studies to resolve the questionsraised above. Comparative phylogenomic studies are also neededto fully assess how much and what regions of the F. ambrosiumgenome have been impacted by introgression; such efforts willbe greatly aided by the complete genome sequence of F. solani f.sp. pisi (Coleman et al., 2009).

4.3. Evolution of clavate macroconidia and divergence-time estimateswithin the AFC

Results of the current study date the evolutionary origin of theAFC to the early Miocene �21.2 Mya (Fig. 3), which coincides withthe adaptive radiation of the Xyleborini (Jordal et al., 2000). Ourworking hypothesis is that the Euwallacea–Fusarium mutualismevolved early in the origins of the monophyletic AFC, as did the tran-sition to production of clavate macroconidia, which we speculatemay represent an adaptation for the symbiosis. Gongylidia and nod-ules appear to represent analogous adaptations in agaricalean fungifarmed, respectively, by higher attines (Schultz and Brady, 2008)and termites (Aanen et al., 2002). The inferences that four of theearly diverging lineages within Clade A of the AFC do not produceclavate conidia, and two of these are not known to be associated withxyleborines, suggest the mutualism may have been asymmetricalinitially (i.e., obligate for beetles, facultative for Fusarium) as in thebasal lineages of attines (Schultz and Brady, 2008). Several lines ofevidence, however, indicate the symbiosis within Clade B evolvedinto a mutually obligate association (i.e., symmetrical), given thatthese fusaria have only been found on or in close association withEuwallacea and the clavate conidial phenotype appears to have be-come fixed in the most recent common ancestor of AF-5 from Malay-sia (Fig. 3, node 7, 4.9 Mya). These findings highlight threeinteresting parallels between the Euwallacea–Fusarium and higherattine–Leucocoprinus symbioses. In both agricultural systems, (i)nutritious, highly adapted fungal cells (i.e., clavate macroconidiaand gongylidia) are produced exclusively within the terminal clades,(ii) beetles and ants cultivate specific symbionts that are mostlytransmitted vertically, and (iii) the mutualisms appear to be obligate(Aanen et al., 2002). An alternative hypothesis is that the mutuallyobligate association evolved in the most recent common ancestorof the AFC (Fig. 3, node 2, 12.5 Mya). If this scenario is correct, thenF. pseudensiforme NRRL 46517 and F. cf. ensiforme NRRL 22354 areprobably cultivated cryptically by ambrosia beetles because, onceobligate mutualisms have been established in insects, reversal to afree-living state appears to be highly unlikely (Aanen et al., 2002;Farrell et al., 2001). Based on the apparent success of the Euwalla-cea–Fusarium symbiosis, it is reasonable to assume that comparativephylogenomic analyses will reveal, as in the attine leaf-cutters(Schiøtt et al., 2010) and FSSC 6 gut symbiont of the Asian longhor-ned beetle (Scully et al., 2012), that they possess the metaboliccapacity to extract nutrients from living and dead wood and degradelignocellulose. AF-9 isolate NRRL 22643, for example, was able toutilize lignin as its sole carbon source, and to break down the aro-matic acids that result from lignin degradation (Norris, 1980).

4.4. Female Euwallacea spp. harbor novel fungal symbionts withincephalic mycangia

The present study confirms an earlier report by Nakashima(1982) that female E. validus possess one pair of mandibular

mycangia within which they transport fungal symbionts in andfrom the natal gallery. The flightless xyleborine males, by contrast,lack mycangia and apparently do not disperse from the natal treehost. The exclusivity of AF-6 and AF-8 from the heads of Euwallaceasp. in Florida and AF-2 from the heads of Euwallacea sp. in Califor-nia provides evidence that this type of mycangium may be fixedwithin this genus of haplodiploid ambrosia beetles (Jordal et al.,2000). However, this must be confirmed, especially given the re-cent confirmation of Euwallacea polyphyly (Cognato et al., 2011;Jordal, 2002). The present study adds to our growing knowledgeof ambrosia beetle associates by establishing that E. validus andother Euwallacea beetles are vectors of novel species of Fusariumfrom their natal galleries to their respective plant hosts.

Other fungi were abundant in adult Euwallacea spp. An unde-scribed Raffaelea species was found in both immature and matureE. validus females, but never in adult males, larvae or pupae (Sup-plemental Fig. S2). Raffaelea spp. are common symbionts of ambro-sia beetles, including the redbay laurel ambrosia beetle, where itssymbiont R. lauricola causes a lethal wilt disease of its plant host(Fraedrich et al., 2008). However, neither the Raffaelea sp. associ-ated with E. validus and tree-of-heaven nor its Fusarium associateAF-4 cause disease when inoculated into healthy tree-of-heavenplants (Kasson, unpubl.). The recovery of AF-4 in all life stages ofE. validus except pupae is consistent with a role as a primary nutri-tional symbiont, as is the absence of Raffaelea sp. from 15% of adultE. validus females sampled. Whether E. validus, like other obligatemutualistic ambrosia beetles, is dependent on one or both of thesesymbionts to complete its life cycle remains to be determined. Fur-thermore, external factors, including plant host and environmentalconditions, may determine the relative roles of different fungalsymbionts in a given beetle species, and have effects on their hostand geographic distribution (Six, 2012).

The present study suggests that shared host trees not only serveas reservoirs of potential pests and pathogens in and around areaswhere avocado is cultivated but also as a potential source of hy-brids. Given the close phylogenetic relationship of these symbi-onts, the likelihood of an olfactory mismatch between insectsand their respective symbionts may foster co-habitation (Hulcrand Dunn, 2011). The rapid expansion of E. validus’ range fromits point of introduction in New York to areas further south toGeorgia and Alabama (Miller and Rabaglia, 2009) towards key avo-cado-producing areas may pose similar risks for an ‘olfactory mis-match’ in avocado and likewise for other Euwallacea spp. inAilanthus.

Our working hypothesis is that vertical transmission of theFusarium symbionts has driven obligate mutualism and diversifica-tion within this system. However, our data suggest that occasionalsecondary contact events, particularly with human-mediatedspread of the beetles (Haack, 2006), may have facilitated interspe-cific hybridization events and introgression. While most ambrosiafungi from temperate regions are nested within the Ophiostoma-tales and Microascales (Alamouti et al., 2009), preliminary datasuggest Fusarium may be one of the dominant symbionts in ambro-sia galleries in tropical Malaysia and Papua New Guinea (Hulcr andCognato, 2010). The present results suggest that additional Fusar-ium symbionts remain undetected. Given the significant threat thatthese invasive pests pose to forest ecosystems (Rabaglia et al.,2008), urban landscapes (Eskalen et al., 2013), and agronomicallyimportant crops such as avocado, tea and cacao, it is critical thatthe identity of these novel pathogens and their insect vectors isestablished. Here we show that the association with Fusariummay have allowed some Euwallacea to exploit trees in non-nativeecosystems. We also show that only carefully chosen molecularmarkers allow for distinguishing ubiquitous yet innocuous Fusar-ium species, which are commonly transported phoretically, fromdestructive associates of aggressive woodborers. The phylogenetic

156 M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157

framework developed herein will be used in future studies to testhypotheses of transmission mode of these agronomically impor-tant symbionts.

Acknowledgments

The authors thank the Hofshi Foundation for organizing theInvasive Ambrosia Beetle Conference in August 2012, whichgreatly facilitated this collaboration. M.T.K. and D.D.D. thank theUSDA Forest Service Forest Health Technology Enterprise Team(FHTET), Morgantown, WV, for funding, and Suzanne Slack andJean Juba for their assistance in the lab. K.S. was funded from theCzech Ministry of Education, Youth and Sports Project no. LH12098 KONTAKT II CR – USA. C.B. and J.H. were funded from aCooperative Agreement with the USDA Forest Service, ForestHealth Section, and by the Norwegian Research Council. D.P.G.S.was supported by AFRI grant no. 2010-65110-20488 from theUSDA NIFA. We are pleased to thank Nathane Orwig for running se-quences in the NCAUR DNA Core Facility. A.G., P.C. and E.D. thankHorticulture Australia Limited and Avocados Australia Limited forfunding. The mention of firm names or trade products does not im-ply that they are endorsed or recommended by the US Departmentof Agriculture over other firms or similar products not mentioned.The USDA is an equal opportunity provider and employer.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.fgb.2013.04.004.

References

Aanen, D.K., Eggleton, P., Rouland-Lefèvre, C., Guldberg-Frøslev, T., Rosendahl, S.,Boomsma, J.J., 2002. The evolution of fungus-growing termites and theirmutualistic fungal symbionts. Proc. Natl. Acad. Sci. USA 99, 14887–14892.

Alamouti, S.M., Tsui, C.K.M., Breuil, C., 2009. Multigene phylogeny of filamentousambrosia fungi associated with ambrosia and bark beetles. Mycol. Res. 113,822–835.

Brayford, D., 1987. Fusarium bugincourtii sp. nov., and its relationship to F. tumidumand F. tumidum var. coeruleum. Trans. Br. Mycol. Soc. 89, 347–351.

Cognato, A.I., Hulcr, J., Dole, S.A., Jordal, B.H., 2011. Phylogeny of haplo-diploid,fungus-growing ambrosia beetles (Curculionidae: Scolytinae: Xyleborini)inferred from molecular and morphological data. Zool. Scripta 40, 174–186.

Coleman, J.J., Rounsley, S.D., Rodriguez-Carres, M., Kuo, A., Wasmann, C.C.,Grimwood, J., Schmutz, J., Taga, M., White, G.J., Zhou, S., Schwartz, D.C.,Freitag, M., Ma, L.J., Danchin, E.G., Henrissat, B., Coutinho, P.M., Nelson, D.R.,Straney, D., Napoli, C.A., Barker, B.M., Gribskov, M., Rep, M., Kroken, S., Molnár,I., Rensing, C., Kennell, J.C., Zamora, J., Farman, M.L., Selker, E.U., Salamov, A.,Shapiro, H., Pangilinan, J., Lindquist, E., Lamers, C., Grigoriev, I.V., Geiser, D.M.,Covert, S.F., Temporini, E., VanEtten, H.D., 2009. The genome of Nectriahaematococca: contribution of supernumerary chromosomes to geneexpansion. PLoS Genet 5, e1000618.

Danthanarayana, W., 1968. The distribution and host-range of the shot-hole borer(Xyleborus fornicatus Eich.) of tea. Tea Quart. 39, 61–69.

Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis bysampling trees. BioMed Cen. Evol. Biol. 7, 214. http://dx.doi.org/10.1186/1471-2148-7-214.

Eskalen, A., Gonzalez, A., Wang, D.H., Twizeyimana, M., Mayorquin, J.S., 2012. Firstreport of a Fusarium sp. and its vector tea shot hole borer (Euwallacea fornicatus)causing Fusarium dieback on avocado in California. Plant Dis 96, 1070.

Eskalen, A., Stouthamer, R., Lynch, S.C., Rugman-Jones, P.F., Twizeyimana, M.,Gonzalez, A., Thibault, T., 2013. Host range of Fusarium diebak and its ambrosiabeetle (Coleoptera: Scolytinae) vector in southern California. Plant Dis. http://dx.doi.org/10.1094/PDIS-11-12-1026-RE.

Farrell, B.D., Sequeira, A.S., O’Meara, B.C., Normark, B.B., Chung, J.H., Jordal, B.H.,2001. The evolution of agriculture in beetles (Curculionidae: Scolytinae andPlatypodinae). Evolution 55, 2011–2027.

Felsenstein, J., 1974. The evolutionary advantage of recombination. Genetics 78,737–756.

Freeman, S., Protasov, A., Sharon, M., Mohotti, K., Okon-Levy, N., Maymon, M.,Mendel, Z., 2012. Obligate feed requirement of Fusarium sp. nov., an avocadowilting agent, by the ambrosia beetle Euwallacea aff. fornicata. Symbiosis 58,245–251.

Fraedrich, S.W., Harrington, T.C., Rabaglia, R.J., Ulyshen, M.D., Mayfield, A.E., Hanula,J.L., Eickwort, J.M., Miller, D.R., 2008. A fungal symbiont of the redbay ambrosia

beetle causes a lethal wilt in redbay and other Lauraceae in the southeasternUnited States. Plant Dis. 92, 215–224.

Gadd, C.H., Loos, C.A., 1947. The ambrosia fungus of Xyleborus fornicatus Eich. Trans.Br. Mycol. Soc. 30, 13–18 plus plate I.

Haack, R.A., 2006. Exotic bark- and wood-boring Coleoptera in the United States:recent establishments and interceptions. Can. J. Forest Res. 36, 269–288.

Hulcr, J., Cognato, A.I., 2010. Repeated evolution of crop theft in fungus-farmingambrosia beetles. Evolution 64, 3205–3212.

Hulcr, J., Dunn, R.R., 2011. The sudden emergence of pathogenicity in insect-fungussymbioses threatens naive forest ecosystems. Proc. Royal Soc. B. http://dx.doi.org/10.1098/rspb.2011.1130.

Hulcr, J., Mogia, M., Isua, B., Novotny, V., 2007. Host specificity of ambrosia and barkbeetles (Col., Curculionidae: Scolytinae and Platypodinae) in a New Guinea rainforest. Ecol. Entomol. 32, 762–772.

Jordal, B.H., 2002. Elongation factor 1 a resolves the monophyly of the haplodiploidambrosia beetles Xyleborini (Coleoptera: Curculionidae). Insect Mol. Biol. 11,453–465.

Jordal, B.H., Normark, B.B., Farrell, B.D., 2000. Evolutionary radiation of aninbreeding haplodiploid beetle lineage (Curculionidae, Scolytinae). Biol. J.Linn. Soc. 71, 483–499.

Kumar, R., Rajkhowa, G., Sankar, M., Rajan, R.K., 2011. A new host plant for theshoot-hole borer, Euwallacea fornicatus (Eichoff) (Coleoptera: Scolytidae) fromIndia. Acta Entomol. Sinicia 54, 734–738.

Kuhnholz, S., Borden, J.H., Uzunovic, A., 2001. Secondary ambrosia beetles inapparently healthy trees; adaptations, potential causes and suggested research.Inter. Pest Manage. Rev. 6, 209–219.

Martin, D.P., Lemey, P., Lott, M., Moulton, V., Posada, D., Lefeuvre, P., 2010. RDP3: aflexible and fast computer program for analyzing recombination. Bioinformatics26, 2462–2463.

Mendel, Z., Protasov, A., Sharon, M., Zveibil, A., Yehuda, S.B., O’Donnell, K., Rabaglia,R., Wysoki, M., Freeman, S., 2012. An Asian ambrosia beetle Euwallaceafornicatus and its novel symbiotic fungus Fusarium sp. pose a serious threat tothe Israeli avocado industry. Phytoparasitica 40, 235–238.

Miller, D.R., Rabaglia, R.J., 2009. Ethanol and (–)-a-pinene: attractant kairomonesfor some large wood-boring beetles in southeastern USA. J. Chem. Ecol. 32, 779–794.

Moon, C.D., Craven, K.D., Leuchtmann, A., Clements, L., Schardl, C.L., 2004.Prevalence of interspecific hybrids amongst asexual fungal endophytes ofgrasses. Mol. Ecol. 13, 1455–1467.

Mueller, U.G., Gerardo, N.M., Aanen, D.K., Six, D.L., Schultz, T.R., 2005. The evolutionof agriculture in insects. Annu. Rev. Ecol. Evol. Syst. 36, 563–595.

Nakashima, T., 1982. Function and location of mycetangia in ambrosia beetles. In:Akai, H., King, R.C., Morohoshi, S. (Eds.), The Ultrastructure and Functioning ofInsect Cells. Society for Insect Cells, Sapporo, pp. 87–90.

Nalim, F.A., Samuels, G.J., Wijesundera, R.L., Geiser, D.M., 2011. New species fromthe Fusarium solani species complex derived from perithecia and soil in the OldWorld tropics. Mycologia 103, 1302–1330.

Nirenberg, H.I., 1990. Recent advances in Fusarium taxonomy. Stud. Mycol. 32, 91–101.

Norris, D.M., 1980. Degradation of C-labeled lignins and C-labeled aromatic acids byFusarium solani. Appl. Environ. Microbiol. 40, 376–380.

Norris, D.M., Baker, J.K., 1967. Symbiosis: effects of a mutualistic fungus upon thegrowth and reproduction of Xyleborus ferrugineus. Science 156, 1120–1122.

O’Donnell, K., Rooney, A.P., Proctor, R.H., Brown, D.W., McCormick, S.P., Ward, T.J.,Frandsen, R.J.N., Lysøe, E., Rehner, S.A., Aoki, T., Robert, V.A.R.G., Crous, P.W.,Groenewald, J.Z., Kang, S., Geiser, D.M., 2013. RPB1 and RPB2 phylogenysupports an early Cretaceous origin and a strongly supported clade comprisingall agriculturally and medically important fusaria. Fungal Genet. Biol. 52, 20–31.

O’Donnell, K., Sutton, D.A., Fothergill, A., McCarthy, D., Rinaldi, M.G., Brandt, M.E.,Zhang, N., Geiser, D.M., 2008. Molecular phylogenetic diversity, multilocushaplotype nomenclature, and in vitro antifungal resistance within the Fusariumsolani species complex. J. Clin. Microbiol. 46, 2477–2490.

Posada, D., Crandall, K.A., 2001. Evaluation of methods for detecting recombinationfrom DNA sequences: computer simulations. Proc. Natl. Acad. Sci. USA 98,13757–13762.

Prior, C., 1986. Sudden death of cocoa in Papua New Guinea associated withPhytophthora palmivora cankers invaded by bark beetles. Ann. Appl. Biol. 109,535–543.

Rabaglia, R., Duerr, D., Acciavatti, R., Ragenovich, I., 2008. Early Detection and RapidResponse for Non-Native Bark and Ambrosia Beetles. USDA Forest Service,Forest Health Protection, pp. 1–12.

Schall, M.J., Davis, D.D., 2009. Ailanthus altissima wilt and mortality: etiology. PlantDis. 93, 747–751.

Schardl, C.L., Craven, K.D., 2003. Interspecific hybridization in plant-associated fungiand oomycetes: a review. Mol. Ecol. 12, 2861–2873.

Schiøtt, M., Rogowska-Wrzesinska, A., Roepstorff, P., Boomsma, J.J., 2010. Leaf-cuttingant fungi produce cell wall degrading pectinase complexes reminiscent ofphytopathogenic fungi. BMC Biol. 8. http://dx.doi.org/10.1186/1741-7007-8-156.

Schultz, T.R., Brady, S.G., 2008. Major evolutionary transitions in ant agriculture.Proc. Natl. Acad. Sci. USA 105, 5435–5440.

Scully, E.D., Hoover, K., Carlson, J., Tien, M., Geib, S.M., 2012. Proteomicanalysis of Fusarium solani isolated from the Asian longhorned beetle,Anoplophora glabripennis. PLoS One 7, e32990. http://dx.doi.org/10.1371/journal.pone.0032990.

Six, D.L., 2003. Bark beetle-fungus symbioses. In: Bourtzis, K., Miller, T.A. (Eds.),Insect Symbiosis. CRC Press, New York, pp. 97–114.

M.T. Kasson et al. / Fungal Genetics and Biology 56 (2013) 147–157 157

Six, D.L., 2012. Ecological and evolutionary determinants of bark beetle – fungussymbioses. Insects 3, 339–366.

Short, D.P.G., O’Donnell, K., Thrane, U., Nielsen, K.F., Zhang, N., Juba, J.H., Geiser,D.M., 2013. Phylogenetic relationships among members of the Fusarium solanispecies complex in human infections and the descriptions of F. keratoplasticumsp. nov. and F. petroliphilum stat. nov. Fungal Genet. Biol. 53, 59–70.

Swofford, D.L., 2003. PAUP�. Phylogenetic analysis using parsimony (�and othermethods). Version 4.0b10. Sinauer Associates, Sunderland, Massachusetts.

Taylor, J.W., Jacobson, D.J., Kroken, S., Kasuga, T., Geiser, D.M., Hibbett, D.S., Fisher,M.C., 2000. Phylogenetic species recognition and species concepts in fungi.Fungal Genet. Biol. 31, 21–32.

Tsai, H.-F., Liu, J.-S., Staben, C., Christensen, M.J., Latch, G.C.M., Siegel, M.R., Schardl,C.L., 1994. Evolutionary diversification of fungal endophtes of tall fescue grassby hybridization with Epichloë species. Proc. Natl. Acad. Sci. USA 91, 2542–2546.

Walker, J.D., Geissman, J.W., compilers, 2009. Geologic time scale. Geol. Soc. Am..http://dx.doi.org/10.1130/2009.