BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. Molecular Phylogeny of the North American Enchenopa binotata (Homoptera: Membracidae) Species Complex Author(s): Chung Ping Lin and Thomas K. Wood Source: Annals of the Entomological Society of America, 95(2):162-171. 2002. Published By: Entomological Society of America DOI: http://dx.doi.org/10.1603/0013-8746(2002)095[0162:MPOTNA]2.0.CO;2 URL: http://www.bioone.org/doi/ full/10.1603/0013-8746%282002%29095%5B0162%3AMPOTNA%5D2.0.CO %3B2 BioOne (www.bioone.org ) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/ terms_of_use . Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.
Transcript
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors nonprofitpublishers academic institutions research libraries and research funders in the common goal of maximizing access tocritical research
Molecular Phylogeny of the North American Enchenopabinotata (Homoptera Membracidae) Species ComplexAuthor(s) Chung Ping Lin and Thomas K WoodSource Annals of the Entomological Society of America 95(2)162-171 2002Published By Entomological Society of AmericaDOI httpdxdoiorg1016030013-8746(2002)095[0162MPOTNA]20CO2URL httpwwwbiooneorgdoifull1016030013-8746282002290955B01623AMPOTNA5D20CO3B2
BioOne (wwwbiooneorg) is a nonprofit online aggregation of core research in thebiological ecological and environmental sciences BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies associationsmuseums institutions and presses
Your use of this PDF the BioOne Web site and all posted and associated contentindicates your acceptance of BioOnersquos Terms of Use available at wwwbiooneorgpageterms_of_use
Usage of BioOne content is strictly limited to personal educational and non-commercialuse Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder
SYSTEMATICS
Molecular Phylogeny of the North American Enchenopa binotata(Homoptera Membracidae) Species Complex
CHUNG PING LIN1 AND THOMAS K WOOD2
Ann Entomol Soc Am 95(2) 162ETH171 (2002)
ABSTRACT TheNorthAmericanEnchenopa binotata (Say) species complex is amodel of sympatricspeciation in which phytophagous insects are hypothesized to diverge through host-plant special-ization resulting from changes in host plant usage that alter life history timing A robust phylogenyis needed to evaluate the historical relevance of the prediction that sister taxa differ in criticallife-history traits Phylogenetic analysis using parsimony and likelihood criteria of 2305 nucleotides insequences from mitochondrial COI COII tRNA-Leucine and 12S genes revealed two pairs of sistertaxa Both pairs of sister taxa differ from each other in the timing of egg hatch in the spring that ismediated by differences in host-plant phenology Host plant mediated timing of egg hatch results inasynchronous life histories among sister taxa facilitating reproductive isolation Sister taxa of Enche-nopa fromCelastrus and fromViburnumdiffer in their diurnal and temporal spansduringwhichmatingoccurs Mating of Enchenopa from Liriodendron takes place after that of its sister species on CercisThese results support the hypothesis that speciation could have been initiated through a shift to a hostplant that alters life-history timing
KEY WORDS Homoptera Enchenopa speciation sympatric
CLADES OF SPECIES are interrelated through historicalgenealogical and geographical connections that in-szliguence howextant species complexes respond to evo-lutionary processes through time (Hillis 1997) Thusthe historical phylogenetic context of a species or itsrelationship within a clade is essential to interpretingthe results of comparative studies dealing with evo-lutionary processes (Brooks and McLennan 1991Brooks et al 1995) One of the fundamental underly-ing processes of evolutionary biology is speciationThis is adifTHORNcult areabecausedeTHORNnitionsof species arevaried and often reszligect underlying assumptions con-cerning themodeof speciation(Templeton1989)Forexample the historical debate over whether geo-graphical isolation is a requisite for speciation or nothas been controversial for many years (Mayr 1982)Many of the examples of purported sympatric specia-tion are difTHORNcult to evaluate from a phylogenetic per-spective because organisms such as the host races ofRhagoletis pomonella (Walsh) (Diptera Tephritidae)have not achieved species status regardless of deTHORNni-tion (Bush 1969) In others such as the Enchenopabinotata (Say) species complex each species is de-
mographically divergent and reproductively cohesivemaking cause and effect difTHORNcult to interpret (Wood1993) Species in this complex not only have well-developed host plant preferencephilopatry duringmating and oviposition but also have been experimen-tally demonstrated in choice tests to mate by speciesFemales experimentally transferred during oviposi-tion to inappropriate host plants do not successfullyproduce offspring (Wood 1980 Wood and Guttman1982 1983)Other differences such as substrate-bornemating signals morphological and color pattern dif-ferences among nymphs have also been demonstrated(Wood 1980 Pratt andWood 1992 Hunt 1994) How-ever discrete adult morphological differentiation inthis complex of species has not developed (Pratt andWood 1993) The challenge for species complexes likeE binotata is to THORNnd sufTHORNcient phylogenetically infor-mative characters to provide a robust analysis of re-lationshipswithin the clade to evaluatemechanismsofspeciation
The E binotata species complex is a model of sym-patric speciation in which phytophagous insects arehypothesized to diverge through host-plant special-ization resulting from changes in host-plant usage(Wood 1980 Wood and Guttman 1981 1982 1983Wood et al 1990 Wood and Keese 1990 Wood 1993Tilmon et al 1998 Wood et al 1999) The hypothe-sized speciationmechanism is that asynchronousmat-ing is induced by differences in plant phenologywhich interacting with philopatry during reproduc-tion allows divergence in host-plant associated per-formance traits (Wood et al 1990 Wood and Keese
Nomenclatural acts or proposed reclassiTHORNcations inferred in thisarticle are not considered published within the meaning of the 1985International Code of Zoological Nomenclature (Article 8b) Thiswork will be published elsewhere in accordance with article eight ofthe Code
1 Department of Entomology Comstock Hall Cornell UniversityIthaca NY 14853 (e-mail cl135cornelledu)
2 Department of Entomology and Applied Ecology University ofDelaware Newark DE 19717
0013-8746020162ETH0171$02000 2002 Entomological Society of America
1990Wood1993Tilmonet al 1998Woodet al 1999)The North American E binotata species complex iscomprised of nine species associated with eight plantgenera distributed among six plant families (Table 1)Because the species in this complex have not beenformally named we refer to them by their host-plantgenus or by host-plant species as in the case of the twospecies using Juglans nigra (L) and J cinerea (L)(see footnote for nomenclatural disclaimer) The Ebinotata species on Carya and Viburnum are polyph-agous in that theyoccuron several specieswithin theirrespective plant genera but the remaining species ap-pear to be monophagous (Wood 1993) The Enche-nopa species and their host plants are sympatricthroughout the eastern United States (Wood 1993)(Fig 1)There is substantial overlapof thedistributionof Enchenopa species with that of its host plant (Little1971)
InNorthAmerica the univoltineE binotata speciesbegin their life history in the spring when overwin-tered eggs hatch synchronously within a 10 d periodon their host-plant Nymphs mature to the adult stage1 mo after egg hatch and males eclose 1ETH2 d beforefemales Mating begins 1 mo after adult eclosionMales can mate several times whereas females mateonce (Wood 1993) Oviposition begins 1 mo afterthe onset of mating Females insert eggs into woodystems cover eggs with froth and deposit multiple eggmasses throughout the fall until mid-November(Wood and Patton 1971) Eggs of each of the nineEnchenopa species hatch at different times in thespring as a result of the differences in changes ofwatercontent among host plants (Wood et al 1990) Asyn-chrony of egg hatch subsequently results in asyn-chrony of time of adult eclosion mating and oviposi-tion among Enchenopa species (Wood and Keese1990) The consequence of life history coordinatedwith plant phenology is that the nine Enchenopa onphenologically different host plants have asynchro-nous life histories within the same locality across theirgeographic range For anEnchenopa species there arealso latitudinal and longitudinal gradients in life his-tory timing (Wood 1993)
The genus Enchenopa is placed within the tribeMembracini in the subfamily Membracinae on thebasis of morphology (Metcalf and Wade 1965 Deitz1975 McKamey 1998) There is debate over whetherthe tribe Membracini is monophyletic and its rela-tionship to the other tribes within the subfamily(Dietrich and McKamey 1995) Because pronotalshape is a primary character for distinguishing generaEnchenopa is not well delineated from other generasuch as Campylenchia and Enchophyllum Thus it ispossible that many species presently assigned toEnchenopa Campylenchia and Enchophyllum are mis-placed Although adults of the nine species in the Ebinotata complex differ in body length and pronotalshape the genitalia (Pratt andWood 1993) and othermorphological features do not provide diagnosticcharacters suitable for the development of a rigorousphylogenetic hypothesis Attempts have beenmade tounderstand relationshipswithin theEbinotata speciescomplex using allozymes female pronotal shape andnymphal characters (Wood 1993)
Relationships within the E binotata species com-plex were THORNrst inferred using allozyme data (Guttmanet al 1981 Wood 1993) using Campylenchia latipes(Say) as an outgroup to derive a distance matrixThere is considerable debate on whether allozymesprovide appropriate information reszligecting evolution-ary history (Swofford et al 1996) The two majorconcerns are whether or not to transform allozymedata to a genetic distance and how evolutionarily im-portant is the presenceabsence of alleles versus thefrequency of alleles (Mickevich and Johnson 1976Swofford andBerlocher 1987)Unless sample sizes arelarge taxa that are in reality polymorphic for somealleles may be scored as THORNxed if one allele is rare Thesame holds true if only a few populations within aspecies are sampled since the frequency of an allelecan vary dramatically among populations over a geo-graphic range (Swofford and Berlocher 1987) In theallozyme study of theE binotata complex sample sizeand number of localities appear to be adequate tocounter these objections but a distancematrix cannotbe analyzedbymodern character-based cladistic anal-ysis and homology assessments remain controversialThe only cladistically based analysis is a phylogenyusing sevenTHORNrst and24THORNfth-instar nymphal characters(Pratt and Wood 1992)
In the above studies (Pratt and Wood 1992 Wood1993) C latipes was used as an outgroup because itwas the only related North American genus (Metcalfand Wade 1965) where adequate fresh material wasavailable for allozymeandnymphal character analysisThe phylogenetic relationship of C latipes to theNorth American E binotata complex is unknown andits use as an outgroup subjective Regardless of thediversity of approaches the results of previous studiesare in general agreement that the Enchenopa on Ro-binia Liriodendron and Carya are basal whereasEnchenopa on Cercis Ptelea Celastrus and Viburnumare more apical (Wood 1993) Enchenopa on J nigraand J cinerea and those on Ptelea andCelastrus appearto be two sister groups but there are disagreements
Table 1 Host plants of the Enchenopa binotata species com-plex (Wood 1993)
Genus Species Family
Ptelea trifoliata (L) RutaceaeJuglans nigra (L) JuglandaceaeJuglans cinerea (L) JuglandaceaeCarya illinoensis (Wang) K Koch
ovalis (Wang) Sargcordiformis (Wang) K Kochlaciniosa (Michx) Loudovata (Mill) K Koch
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 163
Fig 1 Distribution of the North American Enchenopa species and their respective host plants Distribution of Enchenopaspecies (a) was redrawn from line drawings of Pratt et al (unpublished data) Additional collecting records of Enchenopaspecies are from Guttman and Weigt (1989) and TKW (unpublished data) Distribution of the host plant of an Enchenopaspecies is presented in (b) which has been redrawn from Little (1971)
164 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
among trees in the placement of Enchenopa on Liri-odendron
Recently Liu (1996) examined THORNve individuals ofeach of two treehopper species (Atymna querci Fitchand E binotata) using partial sequences of mitochon-drial cytochrome oxidase II (COII) gene Only onenucleotide in THORNve A querci individuals and two nu-cleotides in THORNve E binotata individuals were found todiffer (intraspeciTHORNc variations are equal or 03)among 379 nucleotides in each sequence In additionto low intraspeciTHORNc variation LiuOtildes study (1996) dem-onstrated that interspeciTHORNc sequence differences existin the partial COII gene between these two species oftreehoppers suggesting the mitochondrial COI andCOII genes could be a source of phylogenetic char-acters to resolve the nine E binotata species
Comparedwithmitochondrial proteincodinggenessuch as COI and COII the small subunit ribosomalgene (12S) which has a critical role in protein assem-bly evolves more slowly as a result of its structuralconservation (Simon et al 1994) Preliminary workshowed that partial sequences of 12S were phyloge-netically useful for tribal levels in treehoppers (Liu1996) and could be used to determine the placementof thenineEnchenopabinotata specieswithin the tribeMembracini of the subfamily Membracinae and tofacilitate the selection of closely related species orgenera for outgroups
A host-shift THORNeld experiment to directly test theassumptions of the sympatric speciation hypothesis isin progress but a robust phylogeny is necessary toevaluate the historical relevance of this mechanism tothe extant North American E binotata species com-
plex The objectives of this study are as follows (1) todetermine whether partial DNA sequences for fourmitochondrial genes provide sufTHORNcient phylogeneti-cally informative characters to infer a phylogeny (2)to determine monophyly of the North American Ebinotata species complex and (3) to determine theconcordance between phylogeny and a host shift hy-pothesis of speciation
Materials and Methods
Intraspecific Variation Before DNA sequences canbe used for phylogenetic analysis of a cryptic speciescomplex it is important to determine whether nucle-otide characters are polymorphic among different in-dividuals within a species Therefore for each of thepartial sequences of the three genes COI (397 bp)COII (357 bp) and 12S (339 bp) we sampled threeindividuals fromeach of nineNorthAmericanE bino-tata species to determine if intraspeciTHORNc variation ex-ists within each species (total 27 individuals or 81sequences)
Outgroup Taxa To choose appropriate outgrouptaxa to polarize the characters among the nine speciesofE binotata 31ETH63 treehopper taxa representing THORNvetribes in the subfamilyMembracinaeandCentrodontusatlas (subfamilyCentrodontinae)were sequenced forthe small COI (391 bp) COII (347 bp) and 12S (347bp) fragments (Lin2000) Parsimonyanalyses of thesedata showed twoCentralAmericanEnchenopa species(Enchenopa sp1 and Enchenopa sp 2 Table 2) are themost closely related taxa to the North American
Table 2 Locality data of specimens examined
Species Locality Date Collector Code
Campylenchia latipes Newark DE 30895 T Wood LCLE binotata on Carya Cecil County MD 26896 C P Lin LE24-1E binotata on Carya Cecil County MD 29696 M AdamsC P Lin E14-1E binotata on Carya Ottawa County OK 3996 M Adams E27E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-2E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-3E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-4E binotata on Cercis Newark DE 24896 C P Lin LE20-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-3E binotata on J nigra Ithaca NY 6796 K TilmonT Wood LE12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-3E binotata on Liriodendron Cecil County MD 82596 C P Lin LE22-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood LE6-1E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-3E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E26E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-1E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-2E binotata on Viburnum Newark DE 82596 C P Lin E23E binotata on Viburnum Newark DE 82596 C P Lin K7E binotata on Viburnum Newark DE 82596 C P Lin K8Enchenopa sp 1 Panama City Panama 2798 R Cocroft ENCAEnchenopa sp 2 Guanacaste Costa Rica 5796 R Cocroft LE31-1
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 165
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
SYSTEMATICS
Molecular Phylogeny of the North American Enchenopa binotata(Homoptera Membracidae) Species Complex
CHUNG PING LIN1 AND THOMAS K WOOD2
Ann Entomol Soc Am 95(2) 162ETH171 (2002)
ABSTRACT TheNorthAmericanEnchenopa binotata (Say) species complex is amodel of sympatricspeciation in which phytophagous insects are hypothesized to diverge through host-plant special-ization resulting from changes in host plant usage that alter life history timing A robust phylogenyis needed to evaluate the historical relevance of the prediction that sister taxa differ in criticallife-history traits Phylogenetic analysis using parsimony and likelihood criteria of 2305 nucleotides insequences from mitochondrial COI COII tRNA-Leucine and 12S genes revealed two pairs of sistertaxa Both pairs of sister taxa differ from each other in the timing of egg hatch in the spring that ismediated by differences in host-plant phenology Host plant mediated timing of egg hatch results inasynchronous life histories among sister taxa facilitating reproductive isolation Sister taxa of Enche-nopa fromCelastrus and fromViburnumdiffer in their diurnal and temporal spansduringwhichmatingoccurs Mating of Enchenopa from Liriodendron takes place after that of its sister species on CercisThese results support the hypothesis that speciation could have been initiated through a shift to a hostplant that alters life-history timing
KEY WORDS Homoptera Enchenopa speciation sympatric
CLADES OF SPECIES are interrelated through historicalgenealogical and geographical connections that in-szliguence howextant species complexes respond to evo-lutionary processes through time (Hillis 1997) Thusthe historical phylogenetic context of a species or itsrelationship within a clade is essential to interpretingthe results of comparative studies dealing with evo-lutionary processes (Brooks and McLennan 1991Brooks et al 1995) One of the fundamental underly-ing processes of evolutionary biology is speciationThis is adifTHORNcult areabecausedeTHORNnitionsof species arevaried and often reszligect underlying assumptions con-cerning themodeof speciation(Templeton1989)Forexample the historical debate over whether geo-graphical isolation is a requisite for speciation or nothas been controversial for many years (Mayr 1982)Many of the examples of purported sympatric specia-tion are difTHORNcult to evaluate from a phylogenetic per-spective because organisms such as the host races ofRhagoletis pomonella (Walsh) (Diptera Tephritidae)have not achieved species status regardless of deTHORNni-tion (Bush 1969) In others such as the Enchenopabinotata (Say) species complex each species is de-
mographically divergent and reproductively cohesivemaking cause and effect difTHORNcult to interpret (Wood1993) Species in this complex not only have well-developed host plant preferencephilopatry duringmating and oviposition but also have been experimen-tally demonstrated in choice tests to mate by speciesFemales experimentally transferred during oviposi-tion to inappropriate host plants do not successfullyproduce offspring (Wood 1980 Wood and Guttman1982 1983)Other differences such as substrate-bornemating signals morphological and color pattern dif-ferences among nymphs have also been demonstrated(Wood 1980 Pratt andWood 1992 Hunt 1994) How-ever discrete adult morphological differentiation inthis complex of species has not developed (Pratt andWood 1993) The challenge for species complexes likeE binotata is to THORNnd sufTHORNcient phylogenetically infor-mative characters to provide a robust analysis of re-lationshipswithin the clade to evaluatemechanismsofspeciation
The E binotata species complex is a model of sym-patric speciation in which phytophagous insects arehypothesized to diverge through host-plant special-ization resulting from changes in host-plant usage(Wood 1980 Wood and Guttman 1981 1982 1983Wood et al 1990 Wood and Keese 1990 Wood 1993Tilmon et al 1998 Wood et al 1999) The hypothe-sized speciationmechanism is that asynchronousmat-ing is induced by differences in plant phenologywhich interacting with philopatry during reproduc-tion allows divergence in host-plant associated per-formance traits (Wood et al 1990 Wood and Keese
Nomenclatural acts or proposed reclassiTHORNcations inferred in thisarticle are not considered published within the meaning of the 1985International Code of Zoological Nomenclature (Article 8b) Thiswork will be published elsewhere in accordance with article eight ofthe Code
1 Department of Entomology Comstock Hall Cornell UniversityIthaca NY 14853 (e-mail cl135cornelledu)
2 Department of Entomology and Applied Ecology University ofDelaware Newark DE 19717
0013-8746020162ETH0171$02000 2002 Entomological Society of America
1990Wood1993Tilmonet al 1998Woodet al 1999)The North American E binotata species complex iscomprised of nine species associated with eight plantgenera distributed among six plant families (Table 1)Because the species in this complex have not beenformally named we refer to them by their host-plantgenus or by host-plant species as in the case of the twospecies using Juglans nigra (L) and J cinerea (L)(see footnote for nomenclatural disclaimer) The Ebinotata species on Carya and Viburnum are polyph-agous in that theyoccuron several specieswithin theirrespective plant genera but the remaining species ap-pear to be monophagous (Wood 1993) The Enche-nopa species and their host plants are sympatricthroughout the eastern United States (Wood 1993)(Fig 1)There is substantial overlapof thedistributionof Enchenopa species with that of its host plant (Little1971)
InNorthAmerica the univoltineE binotata speciesbegin their life history in the spring when overwin-tered eggs hatch synchronously within a 10 d periodon their host-plant Nymphs mature to the adult stage1 mo after egg hatch and males eclose 1ETH2 d beforefemales Mating begins 1 mo after adult eclosionMales can mate several times whereas females mateonce (Wood 1993) Oviposition begins 1 mo afterthe onset of mating Females insert eggs into woodystems cover eggs with froth and deposit multiple eggmasses throughout the fall until mid-November(Wood and Patton 1971) Eggs of each of the nineEnchenopa species hatch at different times in thespring as a result of the differences in changes ofwatercontent among host plants (Wood et al 1990) Asyn-chrony of egg hatch subsequently results in asyn-chrony of time of adult eclosion mating and oviposi-tion among Enchenopa species (Wood and Keese1990) The consequence of life history coordinatedwith plant phenology is that the nine Enchenopa onphenologically different host plants have asynchro-nous life histories within the same locality across theirgeographic range For anEnchenopa species there arealso latitudinal and longitudinal gradients in life his-tory timing (Wood 1993)
The genus Enchenopa is placed within the tribeMembracini in the subfamily Membracinae on thebasis of morphology (Metcalf and Wade 1965 Deitz1975 McKamey 1998) There is debate over whetherthe tribe Membracini is monophyletic and its rela-tionship to the other tribes within the subfamily(Dietrich and McKamey 1995) Because pronotalshape is a primary character for distinguishing generaEnchenopa is not well delineated from other generasuch as Campylenchia and Enchophyllum Thus it ispossible that many species presently assigned toEnchenopa Campylenchia and Enchophyllum are mis-placed Although adults of the nine species in the Ebinotata complex differ in body length and pronotalshape the genitalia (Pratt andWood 1993) and othermorphological features do not provide diagnosticcharacters suitable for the development of a rigorousphylogenetic hypothesis Attempts have beenmade tounderstand relationshipswithin theEbinotata speciescomplex using allozymes female pronotal shape andnymphal characters (Wood 1993)
Relationships within the E binotata species com-plex were THORNrst inferred using allozyme data (Guttmanet al 1981 Wood 1993) using Campylenchia latipes(Say) as an outgroup to derive a distance matrixThere is considerable debate on whether allozymesprovide appropriate information reszligecting evolution-ary history (Swofford et al 1996) The two majorconcerns are whether or not to transform allozymedata to a genetic distance and how evolutionarily im-portant is the presenceabsence of alleles versus thefrequency of alleles (Mickevich and Johnson 1976Swofford andBerlocher 1987)Unless sample sizes arelarge taxa that are in reality polymorphic for somealleles may be scored as THORNxed if one allele is rare Thesame holds true if only a few populations within aspecies are sampled since the frequency of an allelecan vary dramatically among populations over a geo-graphic range (Swofford and Berlocher 1987) In theallozyme study of theE binotata complex sample sizeand number of localities appear to be adequate tocounter these objections but a distancematrix cannotbe analyzedbymodern character-based cladistic anal-ysis and homology assessments remain controversialThe only cladistically based analysis is a phylogenyusing sevenTHORNrst and24THORNfth-instar nymphal characters(Pratt and Wood 1992)
In the above studies (Pratt and Wood 1992 Wood1993) C latipes was used as an outgroup because itwas the only related North American genus (Metcalfand Wade 1965) where adequate fresh material wasavailable for allozymeandnymphal character analysisThe phylogenetic relationship of C latipes to theNorth American E binotata complex is unknown andits use as an outgroup subjective Regardless of thediversity of approaches the results of previous studiesare in general agreement that the Enchenopa on Ro-binia Liriodendron and Carya are basal whereasEnchenopa on Cercis Ptelea Celastrus and Viburnumare more apical (Wood 1993) Enchenopa on J nigraand J cinerea and those on Ptelea andCelastrus appearto be two sister groups but there are disagreements
Table 1 Host plants of the Enchenopa binotata species com-plex (Wood 1993)
Genus Species Family
Ptelea trifoliata (L) RutaceaeJuglans nigra (L) JuglandaceaeJuglans cinerea (L) JuglandaceaeCarya illinoensis (Wang) K Koch
ovalis (Wang) Sargcordiformis (Wang) K Kochlaciniosa (Michx) Loudovata (Mill) K Koch
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 163
Fig 1 Distribution of the North American Enchenopa species and their respective host plants Distribution of Enchenopaspecies (a) was redrawn from line drawings of Pratt et al (unpublished data) Additional collecting records of Enchenopaspecies are from Guttman and Weigt (1989) and TKW (unpublished data) Distribution of the host plant of an Enchenopaspecies is presented in (b) which has been redrawn from Little (1971)
164 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
among trees in the placement of Enchenopa on Liri-odendron
Recently Liu (1996) examined THORNve individuals ofeach of two treehopper species (Atymna querci Fitchand E binotata) using partial sequences of mitochon-drial cytochrome oxidase II (COII) gene Only onenucleotide in THORNve A querci individuals and two nu-cleotides in THORNve E binotata individuals were found todiffer (intraspeciTHORNc variations are equal or 03)among 379 nucleotides in each sequence In additionto low intraspeciTHORNc variation LiuOtildes study (1996) dem-onstrated that interspeciTHORNc sequence differences existin the partial COII gene between these two species oftreehoppers suggesting the mitochondrial COI andCOII genes could be a source of phylogenetic char-acters to resolve the nine E binotata species
Comparedwithmitochondrial proteincodinggenessuch as COI and COII the small subunit ribosomalgene (12S) which has a critical role in protein assem-bly evolves more slowly as a result of its structuralconservation (Simon et al 1994) Preliminary workshowed that partial sequences of 12S were phyloge-netically useful for tribal levels in treehoppers (Liu1996) and could be used to determine the placementof thenineEnchenopabinotata specieswithin the tribeMembracini of the subfamily Membracinae and tofacilitate the selection of closely related species orgenera for outgroups
A host-shift THORNeld experiment to directly test theassumptions of the sympatric speciation hypothesis isin progress but a robust phylogeny is necessary toevaluate the historical relevance of this mechanism tothe extant North American E binotata species com-
plex The objectives of this study are as follows (1) todetermine whether partial DNA sequences for fourmitochondrial genes provide sufTHORNcient phylogeneti-cally informative characters to infer a phylogeny (2)to determine monophyly of the North American Ebinotata species complex and (3) to determine theconcordance between phylogeny and a host shift hy-pothesis of speciation
Materials and Methods
Intraspecific Variation Before DNA sequences canbe used for phylogenetic analysis of a cryptic speciescomplex it is important to determine whether nucle-otide characters are polymorphic among different in-dividuals within a species Therefore for each of thepartial sequences of the three genes COI (397 bp)COII (357 bp) and 12S (339 bp) we sampled threeindividuals fromeach of nineNorthAmericanE bino-tata species to determine if intraspeciTHORNc variation ex-ists within each species (total 27 individuals or 81sequences)
Outgroup Taxa To choose appropriate outgrouptaxa to polarize the characters among the nine speciesofE binotata 31ETH63 treehopper taxa representing THORNvetribes in the subfamilyMembracinaeandCentrodontusatlas (subfamilyCentrodontinae)were sequenced forthe small COI (391 bp) COII (347 bp) and 12S (347bp) fragments (Lin2000) Parsimonyanalyses of thesedata showed twoCentralAmericanEnchenopa species(Enchenopa sp1 and Enchenopa sp 2 Table 2) are themost closely related taxa to the North American
Table 2 Locality data of specimens examined
Species Locality Date Collector Code
Campylenchia latipes Newark DE 30895 T Wood LCLE binotata on Carya Cecil County MD 26896 C P Lin LE24-1E binotata on Carya Cecil County MD 29696 M AdamsC P Lin E14-1E binotata on Carya Ottawa County OK 3996 M Adams E27E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-2E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-3E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-4E binotata on Cercis Newark DE 24896 C P Lin LE20-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-3E binotata on J nigra Ithaca NY 6796 K TilmonT Wood LE12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-3E binotata on Liriodendron Cecil County MD 82596 C P Lin LE22-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood LE6-1E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-3E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E26E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-1E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-2E binotata on Viburnum Newark DE 82596 C P Lin E23E binotata on Viburnum Newark DE 82596 C P Lin K7E binotata on Viburnum Newark DE 82596 C P Lin K8Enchenopa sp 1 Panama City Panama 2798 R Cocroft ENCAEnchenopa sp 2 Guanacaste Costa Rica 5796 R Cocroft LE31-1
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 165
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
1990Wood1993Tilmonet al 1998Woodet al 1999)The North American E binotata species complex iscomprised of nine species associated with eight plantgenera distributed among six plant families (Table 1)Because the species in this complex have not beenformally named we refer to them by their host-plantgenus or by host-plant species as in the case of the twospecies using Juglans nigra (L) and J cinerea (L)(see footnote for nomenclatural disclaimer) The Ebinotata species on Carya and Viburnum are polyph-agous in that theyoccuron several specieswithin theirrespective plant genera but the remaining species ap-pear to be monophagous (Wood 1993) The Enche-nopa species and their host plants are sympatricthroughout the eastern United States (Wood 1993)(Fig 1)There is substantial overlapof thedistributionof Enchenopa species with that of its host plant (Little1971)
InNorthAmerica the univoltineE binotata speciesbegin their life history in the spring when overwin-tered eggs hatch synchronously within a 10 d periodon their host-plant Nymphs mature to the adult stage1 mo after egg hatch and males eclose 1ETH2 d beforefemales Mating begins 1 mo after adult eclosionMales can mate several times whereas females mateonce (Wood 1993) Oviposition begins 1 mo afterthe onset of mating Females insert eggs into woodystems cover eggs with froth and deposit multiple eggmasses throughout the fall until mid-November(Wood and Patton 1971) Eggs of each of the nineEnchenopa species hatch at different times in thespring as a result of the differences in changes ofwatercontent among host plants (Wood et al 1990) Asyn-chrony of egg hatch subsequently results in asyn-chrony of time of adult eclosion mating and oviposi-tion among Enchenopa species (Wood and Keese1990) The consequence of life history coordinatedwith plant phenology is that the nine Enchenopa onphenologically different host plants have asynchro-nous life histories within the same locality across theirgeographic range For anEnchenopa species there arealso latitudinal and longitudinal gradients in life his-tory timing (Wood 1993)
The genus Enchenopa is placed within the tribeMembracini in the subfamily Membracinae on thebasis of morphology (Metcalf and Wade 1965 Deitz1975 McKamey 1998) There is debate over whetherthe tribe Membracini is monophyletic and its rela-tionship to the other tribes within the subfamily(Dietrich and McKamey 1995) Because pronotalshape is a primary character for distinguishing generaEnchenopa is not well delineated from other generasuch as Campylenchia and Enchophyllum Thus it ispossible that many species presently assigned toEnchenopa Campylenchia and Enchophyllum are mis-placed Although adults of the nine species in the Ebinotata complex differ in body length and pronotalshape the genitalia (Pratt andWood 1993) and othermorphological features do not provide diagnosticcharacters suitable for the development of a rigorousphylogenetic hypothesis Attempts have beenmade tounderstand relationshipswithin theEbinotata speciescomplex using allozymes female pronotal shape andnymphal characters (Wood 1993)
Relationships within the E binotata species com-plex were THORNrst inferred using allozyme data (Guttmanet al 1981 Wood 1993) using Campylenchia latipes(Say) as an outgroup to derive a distance matrixThere is considerable debate on whether allozymesprovide appropriate information reszligecting evolution-ary history (Swofford et al 1996) The two majorconcerns are whether or not to transform allozymedata to a genetic distance and how evolutionarily im-portant is the presenceabsence of alleles versus thefrequency of alleles (Mickevich and Johnson 1976Swofford andBerlocher 1987)Unless sample sizes arelarge taxa that are in reality polymorphic for somealleles may be scored as THORNxed if one allele is rare Thesame holds true if only a few populations within aspecies are sampled since the frequency of an allelecan vary dramatically among populations over a geo-graphic range (Swofford and Berlocher 1987) In theallozyme study of theE binotata complex sample sizeand number of localities appear to be adequate tocounter these objections but a distancematrix cannotbe analyzedbymodern character-based cladistic anal-ysis and homology assessments remain controversialThe only cladistically based analysis is a phylogenyusing sevenTHORNrst and24THORNfth-instar nymphal characters(Pratt and Wood 1992)
In the above studies (Pratt and Wood 1992 Wood1993) C latipes was used as an outgroup because itwas the only related North American genus (Metcalfand Wade 1965) where adequate fresh material wasavailable for allozymeandnymphal character analysisThe phylogenetic relationship of C latipes to theNorth American E binotata complex is unknown andits use as an outgroup subjective Regardless of thediversity of approaches the results of previous studiesare in general agreement that the Enchenopa on Ro-binia Liriodendron and Carya are basal whereasEnchenopa on Cercis Ptelea Celastrus and Viburnumare more apical (Wood 1993) Enchenopa on J nigraand J cinerea and those on Ptelea andCelastrus appearto be two sister groups but there are disagreements
Table 1 Host plants of the Enchenopa binotata species com-plex (Wood 1993)
Genus Species Family
Ptelea trifoliata (L) RutaceaeJuglans nigra (L) JuglandaceaeJuglans cinerea (L) JuglandaceaeCarya illinoensis (Wang) K Koch
ovalis (Wang) Sargcordiformis (Wang) K Kochlaciniosa (Michx) Loudovata (Mill) K Koch
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 163
Fig 1 Distribution of the North American Enchenopa species and their respective host plants Distribution of Enchenopaspecies (a) was redrawn from line drawings of Pratt et al (unpublished data) Additional collecting records of Enchenopaspecies are from Guttman and Weigt (1989) and TKW (unpublished data) Distribution of the host plant of an Enchenopaspecies is presented in (b) which has been redrawn from Little (1971)
164 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
among trees in the placement of Enchenopa on Liri-odendron
Recently Liu (1996) examined THORNve individuals ofeach of two treehopper species (Atymna querci Fitchand E binotata) using partial sequences of mitochon-drial cytochrome oxidase II (COII) gene Only onenucleotide in THORNve A querci individuals and two nu-cleotides in THORNve E binotata individuals were found todiffer (intraspeciTHORNc variations are equal or 03)among 379 nucleotides in each sequence In additionto low intraspeciTHORNc variation LiuOtildes study (1996) dem-onstrated that interspeciTHORNc sequence differences existin the partial COII gene between these two species oftreehoppers suggesting the mitochondrial COI andCOII genes could be a source of phylogenetic char-acters to resolve the nine E binotata species
Comparedwithmitochondrial proteincodinggenessuch as COI and COII the small subunit ribosomalgene (12S) which has a critical role in protein assem-bly evolves more slowly as a result of its structuralconservation (Simon et al 1994) Preliminary workshowed that partial sequences of 12S were phyloge-netically useful for tribal levels in treehoppers (Liu1996) and could be used to determine the placementof thenineEnchenopabinotata specieswithin the tribeMembracini of the subfamily Membracinae and tofacilitate the selection of closely related species orgenera for outgroups
A host-shift THORNeld experiment to directly test theassumptions of the sympatric speciation hypothesis isin progress but a robust phylogeny is necessary toevaluate the historical relevance of this mechanism tothe extant North American E binotata species com-
plex The objectives of this study are as follows (1) todetermine whether partial DNA sequences for fourmitochondrial genes provide sufTHORNcient phylogeneti-cally informative characters to infer a phylogeny (2)to determine monophyly of the North American Ebinotata species complex and (3) to determine theconcordance between phylogeny and a host shift hy-pothesis of speciation
Materials and Methods
Intraspecific Variation Before DNA sequences canbe used for phylogenetic analysis of a cryptic speciescomplex it is important to determine whether nucle-otide characters are polymorphic among different in-dividuals within a species Therefore for each of thepartial sequences of the three genes COI (397 bp)COII (357 bp) and 12S (339 bp) we sampled threeindividuals fromeach of nineNorthAmericanE bino-tata species to determine if intraspeciTHORNc variation ex-ists within each species (total 27 individuals or 81sequences)
Outgroup Taxa To choose appropriate outgrouptaxa to polarize the characters among the nine speciesofE binotata 31ETH63 treehopper taxa representing THORNvetribes in the subfamilyMembracinaeandCentrodontusatlas (subfamilyCentrodontinae)were sequenced forthe small COI (391 bp) COII (347 bp) and 12S (347bp) fragments (Lin2000) Parsimonyanalyses of thesedata showed twoCentralAmericanEnchenopa species(Enchenopa sp1 and Enchenopa sp 2 Table 2) are themost closely related taxa to the North American
Table 2 Locality data of specimens examined
Species Locality Date Collector Code
Campylenchia latipes Newark DE 30895 T Wood LCLE binotata on Carya Cecil County MD 26896 C P Lin LE24-1E binotata on Carya Cecil County MD 29696 M AdamsC P Lin E14-1E binotata on Carya Ottawa County OK 3996 M Adams E27E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-2E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-3E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-4E binotata on Cercis Newark DE 24896 C P Lin LE20-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-3E binotata on J nigra Ithaca NY 6796 K TilmonT Wood LE12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-3E binotata on Liriodendron Cecil County MD 82596 C P Lin LE22-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood LE6-1E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-3E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E26E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-1E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-2E binotata on Viburnum Newark DE 82596 C P Lin E23E binotata on Viburnum Newark DE 82596 C P Lin K7E binotata on Viburnum Newark DE 82596 C P Lin K8Enchenopa sp 1 Panama City Panama 2798 R Cocroft ENCAEnchenopa sp 2 Guanacaste Costa Rica 5796 R Cocroft LE31-1
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 165
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
Fig 1 Distribution of the North American Enchenopa species and their respective host plants Distribution of Enchenopaspecies (a) was redrawn from line drawings of Pratt et al (unpublished data) Additional collecting records of Enchenopaspecies are from Guttman and Weigt (1989) and TKW (unpublished data) Distribution of the host plant of an Enchenopaspecies is presented in (b) which has been redrawn from Little (1971)
164 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
among trees in the placement of Enchenopa on Liri-odendron
Recently Liu (1996) examined THORNve individuals ofeach of two treehopper species (Atymna querci Fitchand E binotata) using partial sequences of mitochon-drial cytochrome oxidase II (COII) gene Only onenucleotide in THORNve A querci individuals and two nu-cleotides in THORNve E binotata individuals were found todiffer (intraspeciTHORNc variations are equal or 03)among 379 nucleotides in each sequence In additionto low intraspeciTHORNc variation LiuOtildes study (1996) dem-onstrated that interspeciTHORNc sequence differences existin the partial COII gene between these two species oftreehoppers suggesting the mitochondrial COI andCOII genes could be a source of phylogenetic char-acters to resolve the nine E binotata species
Comparedwithmitochondrial proteincodinggenessuch as COI and COII the small subunit ribosomalgene (12S) which has a critical role in protein assem-bly evolves more slowly as a result of its structuralconservation (Simon et al 1994) Preliminary workshowed that partial sequences of 12S were phyloge-netically useful for tribal levels in treehoppers (Liu1996) and could be used to determine the placementof thenineEnchenopabinotata specieswithin the tribeMembracini of the subfamily Membracinae and tofacilitate the selection of closely related species orgenera for outgroups
A host-shift THORNeld experiment to directly test theassumptions of the sympatric speciation hypothesis isin progress but a robust phylogeny is necessary toevaluate the historical relevance of this mechanism tothe extant North American E binotata species com-
plex The objectives of this study are as follows (1) todetermine whether partial DNA sequences for fourmitochondrial genes provide sufTHORNcient phylogeneti-cally informative characters to infer a phylogeny (2)to determine monophyly of the North American Ebinotata species complex and (3) to determine theconcordance between phylogeny and a host shift hy-pothesis of speciation
Materials and Methods
Intraspecific Variation Before DNA sequences canbe used for phylogenetic analysis of a cryptic speciescomplex it is important to determine whether nucle-otide characters are polymorphic among different in-dividuals within a species Therefore for each of thepartial sequences of the three genes COI (397 bp)COII (357 bp) and 12S (339 bp) we sampled threeindividuals fromeach of nineNorthAmericanE bino-tata species to determine if intraspeciTHORNc variation ex-ists within each species (total 27 individuals or 81sequences)
Outgroup Taxa To choose appropriate outgrouptaxa to polarize the characters among the nine speciesofE binotata 31ETH63 treehopper taxa representing THORNvetribes in the subfamilyMembracinaeandCentrodontusatlas (subfamilyCentrodontinae)were sequenced forthe small COI (391 bp) COII (347 bp) and 12S (347bp) fragments (Lin2000) Parsimonyanalyses of thesedata showed twoCentralAmericanEnchenopa species(Enchenopa sp1 and Enchenopa sp 2 Table 2) are themost closely related taxa to the North American
Table 2 Locality data of specimens examined
Species Locality Date Collector Code
Campylenchia latipes Newark DE 30895 T Wood LCLE binotata on Carya Cecil County MD 26896 C P Lin LE24-1E binotata on Carya Cecil County MD 29696 M AdamsC P Lin E14-1E binotata on Carya Ottawa County OK 3996 M Adams E27E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-2E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-3E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-4E binotata on Cercis Newark DE 24896 C P Lin LE20-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-3E binotata on J nigra Ithaca NY 6796 K TilmonT Wood LE12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-3E binotata on Liriodendron Cecil County MD 82596 C P Lin LE22-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood LE6-1E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-3E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E26E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-1E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-2E binotata on Viburnum Newark DE 82596 C P Lin E23E binotata on Viburnum Newark DE 82596 C P Lin K7E binotata on Viburnum Newark DE 82596 C P Lin K8Enchenopa sp 1 Panama City Panama 2798 R Cocroft ENCAEnchenopa sp 2 Guanacaste Costa Rica 5796 R Cocroft LE31-1
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 165
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
among trees in the placement of Enchenopa on Liri-odendron
Recently Liu (1996) examined THORNve individuals ofeach of two treehopper species (Atymna querci Fitchand E binotata) using partial sequences of mitochon-drial cytochrome oxidase II (COII) gene Only onenucleotide in THORNve A querci individuals and two nu-cleotides in THORNve E binotata individuals were found todiffer (intraspeciTHORNc variations are equal or 03)among 379 nucleotides in each sequence In additionto low intraspeciTHORNc variation LiuOtildes study (1996) dem-onstrated that interspeciTHORNc sequence differences existin the partial COII gene between these two species oftreehoppers suggesting the mitochondrial COI andCOII genes could be a source of phylogenetic char-acters to resolve the nine E binotata species
Comparedwithmitochondrial proteincodinggenessuch as COI and COII the small subunit ribosomalgene (12S) which has a critical role in protein assem-bly evolves more slowly as a result of its structuralconservation (Simon et al 1994) Preliminary workshowed that partial sequences of 12S were phyloge-netically useful for tribal levels in treehoppers (Liu1996) and could be used to determine the placementof thenineEnchenopabinotata specieswithin the tribeMembracini of the subfamily Membracinae and tofacilitate the selection of closely related species orgenera for outgroups
A host-shift THORNeld experiment to directly test theassumptions of the sympatric speciation hypothesis isin progress but a robust phylogeny is necessary toevaluate the historical relevance of this mechanism tothe extant North American E binotata species com-
plex The objectives of this study are as follows (1) todetermine whether partial DNA sequences for fourmitochondrial genes provide sufTHORNcient phylogeneti-cally informative characters to infer a phylogeny (2)to determine monophyly of the North American Ebinotata species complex and (3) to determine theconcordance between phylogeny and a host shift hy-pothesis of speciation
Materials and Methods
Intraspecific Variation Before DNA sequences canbe used for phylogenetic analysis of a cryptic speciescomplex it is important to determine whether nucle-otide characters are polymorphic among different in-dividuals within a species Therefore for each of thepartial sequences of the three genes COI (397 bp)COII (357 bp) and 12S (339 bp) we sampled threeindividuals fromeach of nineNorthAmericanE bino-tata species to determine if intraspeciTHORNc variation ex-ists within each species (total 27 individuals or 81sequences)
Outgroup Taxa To choose appropriate outgrouptaxa to polarize the characters among the nine speciesofE binotata 31ETH63 treehopper taxa representing THORNvetribes in the subfamilyMembracinaeandCentrodontusatlas (subfamilyCentrodontinae)were sequenced forthe small COI (391 bp) COII (347 bp) and 12S (347bp) fragments (Lin2000) Parsimonyanalyses of thesedata showed twoCentralAmericanEnchenopa species(Enchenopa sp1 and Enchenopa sp 2 Table 2) are themost closely related taxa to the North American
Table 2 Locality data of specimens examined
Species Locality Date Collector Code
Campylenchia latipes Newark DE 30895 T Wood LCLE binotata on Carya Cecil County MD 26896 C P Lin LE24-1E binotata on Carya Cecil County MD 29696 M AdamsC P Lin E14-1E binotata on Carya Ottawa County OK 3996 M Adams E27E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-2E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-3E binotata on Celastrus Cecil County MD 25896 C P Lin LE15-4E binotata on Cercis Newark DE 24896 C P Lin LE20-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-1E binotata on Cercis Wilmington OH 18696 K TilmonT Wood E3-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-2E binotata on J cinerea Ithaca NY 16696 K TilmonT Wood E4-3E binotata on J nigra Ithaca NY 6796 K TilmonT Wood LE12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-2E binotata on J nigra Ithaca NY 6796 K TilmonT Wood E12-3E binotata on Liriodendron Cecil County MD 82596 C P Lin LE22-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-1E binotata on Liriodendron Cecil County MD 62996 M AdamsC P Lin E10-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood LE6-1E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-2E binotata on Ptelea Wilmington OH 61896 K TilmonT Wood E6-3E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E26E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-1E binotata on Robinia Ithaca NY 61696 K TilmonT Wood E2-2E binotata on Viburnum Newark DE 82596 C P Lin E23E binotata on Viburnum Newark DE 82596 C P Lin K7E binotata on Viburnum Newark DE 82596 C P Lin K8Enchenopa sp 1 Panama City Panama 2798 R Cocroft ENCAEnchenopa sp 2 Guanacaste Costa Rica 5796 R Cocroft LE31-1
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 165
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
Enchenopa species complex and that C latipes is anappropriate outgroup (Lin 2000)
After an outgroup was chosen COI and COII frag-ments that encompass an additional 822 bp in COI 69bp in t-RNA-Leucine and an additional 335 bp inCOIIwere sequenced for both outgroup and the ingroupEnchenopa species complex We analyzed the NorthAmerican E binotata species using a ldquototal evidencerdquoapproach by reconstructing the phylogeny with allavailable DNA sequences
Specimen Treatment Specimens were collected asadults or nymphs at various localities in North andCentral America (Table 2) Field-collected treehop-pers were immediately preserved in 95 ethanol fol-lowed by long-term storage at20CDNAextractionfollowed protocols outlined in Danforth (1999) withthe remainder of the specimen preserved as vouchersin 95 ethanol at 20C
Primers Initially the small partial mitochondrialCOI COII and small ribosomal subunit gene (12S)fragments were ampliTHORNed via polymerase chain reac-tion (PCR)with three sets of primers (see Table 3 forprimer sequences and locations) Ron-Nancy primersproduced a PCR product of 400 bp in the COI geneA combination of A-298 and B-LYS primers produceda PCR product of nearly 350 bp between the 3 half ofCOII and tRNA-LysgeneThe12Sbi and12SaiprimersproducedaPCRproductof330bp in12SgeneOnceappropriate outgroups were determined another twosets of PCR products of 1000 and 1200 bp in theCOI tRNA-leucine and COII regions were ampliTHORNedforEnchenopaandCampylenchiausing thenewprimercombinations of Ron-Calvin and Rick (Dick)-Barb
Sequencing Protocols A Perkin-Elmer thermal cy-cler (GeneAmp PCR System 2400 Foster City CA)was used for double-stranded ampliTHORNcations of theCOI COII and 12S gene The cycling proTHORNle beganwith one cycle of DNA denaturation at 94C for 2 minand followed by 35ETH45 cycles of sequence ampliTHORNca-tion (DNA denaturation at 94C for 30 s primer an-nealing at 50ETH53C for 30 s and sequence extension at72C for 1 min) The PCR products were puriTHORNed bya gel puriTHORNcation method provided by J McDonald
(University of Delaware) Sequences were obtainedfrom both sense and antisense strands using the Ap-plied Biosystems 373A DNA sequencer (Foster CityCA) The chromatograph of each sequence was THORNrstexamined using the 373 DNA Data Analysis Program(Foster City CA) to determine the quality of eachsequence and subsequently edited in SeqEd (version103 Applied Biosystems 1992) by manually compar-ing the aligned chromatograph of both sense and an-tisense strands toconTHORNrmambiguousbases Sequencesused in this study can be obtained from GENBANK(accession number AY057846-AY057857)
Sequence Alignment DNA sequences of each spe-cies were transferred to Editseq THORNles and aligned withEDITSEQ and MEGALIGN programs in Lasergene(DNASTAR Madison WI) The Clustal method inMEGALIGN was used with the pairwise alignmentparameter Ktuple set to 2 For COI and COII proteincoding genes the multiple alignment parameter gappenalty was set to 100 to minimize gap formationSequence alignment for protein-coding gene se-quences like COI and COII was straightforward be-cause codon reading frames could be determined byalignment withDrosophila yakuba (Burla) (Clary andWolstenholme 1985) Alignment of ribosomal genesequences of distantly related species may be difTHORNcultbecause of insertion and deletion events For Enche-nopa species the sequence alignments of both pro-tein-coding and ribosomal genes are relatively unam-biguous because of low sequence divergence For the12S ribosomal gene sequences weremanually alignedwith reference to the secondary structure of the thirddomain using Cicadidae as a model (Kjer 1995 Hick-son et al 1996) Each data matrix was subsequentlysaved as MEGALIGN and PAUP (NEXUS format)THORNles Combining data matrices from different se-quence fragments was done using MacClade (version305 Maddison and Maddison 1992)
Phylogenetic Analysis Maximum parsimony andlikelihoodanalysesweredonebyusingPAUP400d64(Swofford 1998) As a result of outgroup analyses twoCentralAmericanEnchenopa specieswere included inthe ingroupandC latipeswaschosenasoutgroup(Lin
The standardized primer names are in parentheses (Simon et al 1994)a Position number refer to 5 end of primer sequence in Drosophila yakuba (Clary and Wolstenholme 1985)b Designed by Harrison Laboratory at Cornell Universityc Designed by C Keeler at the University of Delawared Designed by Liu and Beckenbach (1992)e Designed by Kocher et al (1989)
166 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
2000) Because of the small number of taxa (1 out-group and 11 ingroup taxa) parsimony tree searcheswereperformedusing themoreexhaustivebranchandbound search method with equally weighted charac-ters In separate analyses gap coded characters intRNA-Leucine and 12S gene were treated as missingdata or as a new (THORNfth) state To assess the level ofbranch support bootstrap values were calculatedbasedon1000 replicationsusing thebranchandboundsearch method (Felsenstein 1985) Bremer support(Bremer 1988) was calculated using the TreeRot pro-gram (Sorenson 1999) based on 20 replicate heuristicsearches with random addition of taxa
Formaximum likelihood analyses equallyweightedtrees obtained from parsimony analysis were used toestimate the log likelihood of each tree under 20 dis-tinctmodels of sequence evolution (Huelsenbeck andCrandall 1997) The four basic models were Jukes-Cantor (JC) which has single substitution type andequal base frequency Kimura two-parameter (K2P)whichhas two substitution types (transition and trans-version) and equal base frequency Hasegawa-Kishino-Yano (HKY) which has two substitutiontypes (transition and transversion) and nonequal basefrequency and General Time Reversible (GTR)which has six substitution types and nonequal basefrequency Within each model there were four meth-ods of accounting for rate heterogeneity no rate het-erogeneity gamma distributed rates (G) proportionof invariant sites (I) gamma invariant sites (IG)and site-speciTHORNc rates (SSR) For the site-speciTHORNc ratemodel (SSR) we assigned THORNve different rate catego-ries the THORNrst second third codon positions t-RNA-Leu gene and 12S gene After likelihood scores werecalculated for each model we used the equallyweighted parsimony trees as starting trees and per-formed searches using increasingly exhaustive branchswapping methods in the following order Nearestneighbor interchange (NNI) subtree pruning and re-grafting (SPR) second round of SPR tree bisectionand reconnection (TBR) and second round of TBRAt each iteration themaximum likelihood parameterswere reestimated from the trees whichwere obtainedfrom the previous round of branch swapping
Results
Intraspecific VariationOf these 1093 bp from smallfragments of COI COII and 12S only one nucleotide(insertion or deletion of Thymine in the 12S se-quence) difference was found among three individ-uals of Enchenopa on J cinerea This nucleotide dif-ference needs to be conTHORNrmed by sampling additionalindividuals The remaining sequences of all threegenes are identical among the three individualswithineach of the other eight E binotata species With thisone possible exception the nucleotide differencesamong the nineE binotata species are THORNxed andusefulas phylogenetic characters Of these 1093 nucleotidesites examined 1018(93)areconstant 39(35)areuninformative and 36 (33) are phylogenetically(parsimony) informative characters
Nucleotide Composition and Codon BiasA total of2305 bp of aligned sequences for the four genes wasobtained for 11 Enchenopa species and the outgrouptaxon The distribution of nucleotides is as follows1219 (position 1ETH1219) in COI 65 (1220ETH1288) andfour gap coded sites (1235ETH38) in tRNA-Leucine 681(1289ETH1969) in COII 329 (1970ETH2305) and seven gapcoded sites (2091ETH92 2123ETH38 2233 2248 and 2256) in12S The overall base composition was AT biased(745 Table 4) as in other insect mitochondrial ge-nomes (60ETH80 Simon et al 1994) Codons for aminoacids were inferred by alignment with mitochondrialDNA sequences of D yakuba (Clary and Wolsten-holme 1985) The base composition of protein codinggenes varies among codon positions and between thetwo genes The AT bias is highest in the third codonof both COI (887) and COII (911) whereas thesecond codon of the COI has the least AT bias(627) Chi-square tests show no signiTHORNcant devia-tion fromhomogeneity of base frequencies across taxa(P 065)
Parsimony Analyses The E binotata species com-plex was analyzed using C latipes as an outgroup withthe expanded character of 2305 nucleotides Of 249(70 for ingroup) informative characters most arefound in COI and COII protein-coding genes (22390)with182or82(56 90 for ingroup) in the third
Table 4 Nucleotide composition of 2305 bp of COI COII t-RNA-Leucine and 12S of Enchenopa binotata species complex andoutgroup taxon
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 167
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
codon position The t-RNA-Leu and 12S genes com-bined had 26 or 10 (8 11 for ingroup) of theinformative characters
Four equally parsimonious trees of length 890 wereobtained Treating gap-coded characters either asmissing data or as a THORNfth state yielded the same resultThe strict consensus (Fig 2) of four equally parsimo-nious trees shows support (bootstrap value of 100)for the basal position of the two Enchenopa speciesfrom Central America This tree strongly suggests themonophyly of the North American E binotata speciescomplex with bootstrap value of 100 Two pairs ofsister species within the complex Enchenopa fromCelastrus and from Viburnum (99) and Enchenopafrom Cercis and from Liriodendron (88) are alsostrongly supported by this tree However the rela-tionships among all of the nine species of the NorthAmerican Enchenopa binotata complex were not com-pletely resolved
Likelihood Analyses Log likelihood scores of 20models are shown in Fig 3 Allowing for variabletransitiontransversion ratios and non-equal base fre-quencies (HKY model) greatly improved the likeli-hood scores among the four basic models (Fig 3arrow 1) Within HKY models accounting for rateheterogeneity among sites (SSR) improved the like-lihood scores compared with the other four differentmethodsof accommodating rateheterogeneity(Fig 3arrow 2) Therefore we chose HKY model with SSR
for maximum likelihood analysis because it requiredthe least assumptions while substantially improvingthe likelihood scores
One tree (Fig 4) was obtained after branch swap-ping using the HKYSSR model that had the sametree topology as themore complexGTRSSR and lesscomplex HKYIG model The topologies of thesetreeswere congruentwith that of strict consensus treederived from the four equal parsimony trees In ad-
Fig 2 Strict consensus of four equally parsimonioustrees from 2305 bp of mitochondrial COI tRNA-LeucineCOII and 12S gene (tree length 890 CI 0849 RI 0680) Numbers above branch are bootstrap scores of 1000replicates (bootstrap values 50 not shown) Numbersbelow branch are the decay index of 20 replicate heuristicsearches with random addition of taxa
Fig 3 Log likelihood scores of 20 models of sequenceevolution (JC Jukes and Cantor K2P Kimura two-parame-ter HKY Hasegawa-Kishino-Yano GTR General Time Re-versible G Gamma distribution rates I Proportion of in-variant sites SSR Site-speciTHORNc rates)
Fig 4 Maximum likelihood tree based on HKYSSRmodel (-Ln likelihood 682481164) The numbers abovebranch are bootstrap scores of 100 replicates (bootstrap val-ues 50 not shown)
168 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
dition they revealed a basal relationship ofEnchenopaspecies from J cinerea J nigra and Carya (all in theJuglandaceae) relative to the remaining six NorthAmerican Enchenopa species
Discussion
TheexpandedDNAsequences fromCOI(1219bp)COII (681bp) t-RNA-Leucine (69bp including gaps)and 12S (336 bp including gaps) provide sufTHORNcientcharacters to resolve the relationships of closely re-lated North American Enchenopa species with theexception of two internal nodes in the parsimonyanalyses (Fig 2) This lack of complete resolutionmaybe due to character conszligicts or simply a result of notenough informative characters Additional sequencesfrom other mitochondrial genes may provide moreinformative characters for complete resolution of thisspecies complex With the exception of one nucleo-tide difference the mitochondrial sequences fromthree individuals of each of nine North AmericanEnchenopa species are identical However the extentof intraspeciTHORNcvariationneeds tobeexpandedbeyondthe limited data present here to reszligect the geographicranges For future study answering the question ofwhether there is intra or interspeciTHORNc genetic struc-ture throughout the geographic range of the extantnine Enchenopa species requires more variable mito-chondrial genes and extensive geographic sampling ofeach species throughout the eastern North America
Because the topologies of maximum parsimony andstrict consensus of maximum likelihood trees are inconcordance we chose the maximum likelihood tree(Fig 4) as a working phylogenetic hypothesis for theE binotata species complex Despite the relativelyshort branch lengths of internodes in the maximumlikelihood tree applying likelihood analysis provides auseful method of investigating regions of the cla-dogram which lack parsimony resolution Choosingamongmodels of DNA substitution is among themostimportant steps when applying likelihood criterion inphylogenetic analysis Applying theHKYSSRmodelfor likelihood analyses is appropriate because there isa highly unequal base frequency (AT bias Table 4)These data also had an unequal transitiontransver-sion ratio of 12ETH28 depending on the model of se-quence evolution Accounting for rate heterogeneityis also reasonable because the rate of evolution variesnot only among three codon positions of protein cod-inggenesbutalso in transferRNAandribosomalgenes(Simon et al 1994)
The phylogenetic hypothesis derived from molec-ular characters is not in concordance with that ofnymphal morphology These two hypotheses suggestdifferent sets of sister taxa relationships The treebased on nymphal characters suggests three sets ofsister taxa Enchenopa from Cercis and ViburnumEnchenopa from Ptelea and Celastrus and Enchenopafrom J nigra and J cinerea (Pratt and Wood 1992)However themitochondrial tree suggests another twosets of sister taxaEnchenopa fromCercis andLirioden-dron and Enchenopa from Celastrus and Viburnum
The nymphal character based tree suggests thatEnchenopa from Robinia is basal to the remainingNorth American Enchenopa species whereas the mi-tochondrial based tree suggests Enchenopa from Jcinerea is basal Several technical gene or organismallevel factors may account for the discordance amongtrees derived from different sources of characters(Wendel and Doyle 1998) It is likely that the discor-dance between nymphal and mitochondrial trees isdue to difference in taxon sampling and the nature ofcoding continuous nymphal characters Althoughboth trees use the same outgroup C latipes the mi-tochondrial tree includes twomore Central AmericanEnchenopaHowever the discordance of the two treescannot be resolved until the nymphal data are rean-alyzed including the two additional Central AmericanEnchenopa species
Thehypothesis that thenineextant species ofNorthAmerican E binotata are monophyletic is supportedby this mitochondrial phylogeny because the twoEnchenopa species from Central America are basal tothe remaining E binotata species complex and thesetwo ingroup nodes are strongly supported by boot-strap values (Figs 2 and 4) This result suggests thatthe North American Enchenopa species were derivedfrom a common ancestor in Central or South Americaand subsequently speciated through host shifts inNorth America Additional taxon sampling of MexicoCentral and South American Enchenopa species is re-quired to fully test the hypothesis
Sympatric speciation is one of the more controver-sial subjects in evolutionary biology Sympatric spe-ciation could occur if biological traits (eg life historytiming philopatry) impeded gene szligow between pop-ulations in the absence of geographic isolation Exceptfor polyploidy in plants where speciation events takeplace almost instantaneously most examples of sym-patric speciation require ecological or habitat differ-ences topromote reproductive isolation Inaddition totheE binotata species complex examples of sympatricspeciation through shifts in host use or prey special-ization can be found in other insect groups likeRhago-letis (Bush 1969) and Chrysopidae (Tauber andTauber 1982) The E binotata species complex is hy-pothesized to diverge through host-plant specializa-tion resulting from changes in host-plant use Thissympatric hypothesis of speciation predicts that sistertaxa should differ in critical life-history traits like timeof egg hatch length of development to mating tem-poral span of mating which are hypothesized to haveinitiated divergence (Wood 1993) Therefore sistertaxa relationships revealed by a phylogenetic analysisand the differences in life history traits among speciestogether could be used to test the validity of thisprediction Sequences from mitochondrial DNA pro-vide THORNxed characters for reconstructing a robust phy-logenetic hypothesisWell-supported sister taxa of theEnchenopa from Celastrus and Viburnum in the mito-chondrial tree differ both in their diurnal and tempo-ral spans during which mating occurs Although thecomplete life-history data of the Enchenopa from Li-riodendron is limited the available data suggest that
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 169
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
mating of this species takes place after that of its sistertaxon theEnchenopa fromCercis (Wood andGuttman1985) Thus this mitochondrial phylogeny supportsthehypothesis that shifts tohost plants that disrupt lifehistory synchrony couldhave initiated speciation Therelatively few informative characters (70 in 2305sites) found along with little adult morphological dif-ferentiation (Pratt andWood 1993) suggest theNorthAmerican Enchenopa species complex have speciatedrecently
Acknowledgments
WethankDTallamy JMcDonald andCKeeler for theircontinuous advice and use of laboratory facilities during thecompletion of this study Our appreciation is given to BDanforth for his help on maximum likelihood analysis Wealso thank R Cocroft for providing specimens from CentralAmerica The following people provided helpful commentsof the manuscript B Danforth K Magnacca S Sipes CSimon and an anonymous reviewer Funding support fromthe National Science Foundation to TKW is greatly appre-ciated
References Cited
Bremer K 1988 The limits of amino acid sequence data inangiosperm phylogenetic reconstruction Evolution 42795ETH803
Brooks D R and D A McLennan 1991 Phylogeny ecol-ogy and behavior a research program in comparativebiology University of Chicago Press Chicago IL
Bush G L 1969 Sympatric host race formation and spe-ciation in frugivorous szligies of the genus Rhagoletis(Diptera Tephritidae) Evolution 23 237ETH251
ClaryDO andDRWolstenholme 1985 Themitochon-drial DNA molecule of Drosophila yakuba nucleotidesequence gene organization and genetic code J MolEvol 22 252ETH271
Danforth B N 1999 Phylogeny of the bee genus Lasio-glossum (Hymenoptera Halictidae) based on mitochon-drial cytochrome oxidase Syst Entomol 24 377ETH393
Deitz L L 1975 ClassiTHORNcation of the higher categories ofthe New World treehoppers (Homoptera Membraci-dae) NC Agric Exp Stn Tech Bull 225 1ETH177
Dietrich C H and S H McKamey 1995 Two new neo-tropical treehopper genera and investigation of the phy-logeny of the subfamily Membracinae (HomopteraMembracidae) Proc Entomol Soc Wash 97 1ETH16
Felsenstein J 1985 ConTHORNdence limits on phylogenies anapproach using the bootstrap Evolution 39 783ETH791
Guttman S I and L A Weigt 1989 Macrogeographicgenetic variation in theEnchenopabinotata complex (Ho-moptera Membracidae) Ann Entomol Soc Am 82156ETH65
Guttman S I T K Wood and A A Karlin 1981 Geneticdifferentiation along host plant lines in the sympatricEnchenopa binotata Say complex (Homoptera Mem-bracidae) Evolution 35 205ETH17
Hickson R E C Simon A Cooper G S Spicer J Sullivanand D Penny 1996 Conserved sequence motifs align-ment and secondary structure for the third domain ofanimal 12S rRNA Mol Biol Evol 13 150ETH169
Hunt R E 1994 Vibrational signals associated withmatingbehavior in the treehopper Enchenopa binotata Say (Ho-moptera Membracidae) J NY Entomol Soc 102 266ETH270
Huelsenbeck J P and K A Crandall 1997 Phylogenyestimation and hypothesis testing using maximum likeli-hood Annu Rev Ecol Syst 28 437ETH466
Kjer K M 1995 Use of rRNA secondary structure in phy-logenetic studies to identify homologous position an ex-ample of alignment and data presentation from the frogsMol Phylogenet Evol 4 314ETH330
Kocher T D W K Tomas A Meyer S V Edwards SPaabo F X Villablanca and A C Wilson 1989 Dy-namics of mitochondrial DNA evolution in animals am-pliTHORNcation and sequencing with conserved primers ProcNatl Acad Sci USA 86 6196ETH6200
Lin C P 2000 Molecular phylogeny of the Enchenopabinotata species complex (Homoptera Membracidae)MS thesis University of Delaware Newark
Little E 1971 USA Department of Agriculture Miscella-neous Publication No 1146 Atlas of the USA Trees vol1 Conifer-G and important hardwoods USDA MiscPubl 202
Liu D 1996 Evaluation of mitochondrial cytochrome ox-idase II and small subunit ribosomal RNA (12S) genes forphylogenetic inference in the Membracidae MS thesisUniversity of Delaware Newark
Liu H and A T Beckenbach 1992 Evolution of the mi-tochondrial cytochromeoxidase II geneamong tenordersof insects Mol Phylogenet Evol 1 41ETH52
McKamey S H 1998 Taxonomic catalogue of the Mem-bracoidea (exclusive of leafhoppers) Second supple-ment to fascicle 1ETHMembracidae of the general catalogueof the Hemiptera Mem Am Entomol Inst 60 1ETH377
Maddison W P and D R Maddison 1992 MacClade ver-sion 305 Sinauer Sunderland MA
Mayr E 1982 Processes of speciation in animals pp 1ETH19In C Barigozzi [ed] Mechanisms of speciation LissNew York
Metcalf Z P and V Wade 1965 General catalogue of theHomoptera A supplement to fascicle INtildeMembracidae ofthe general catalogue of the Hemiptera MembracoideaIn two sections NorthCarolina State University Raleigh
Mickevich M F and M S Johnson 1976 Congruence be-tween morphological and allozyme data in evolutionaryinference and character evolution Syst Zool 25 260ETH270
Pratt G and T K Wood 1992 A phylogenetic analysis ofthe Enchenopa binotata species complex (HomopteraMembracidae) using nymphal characters Syst Entomol17 351ETH357
Pratt G and T K Wood 1993 Genitalic analysis of malesand females in the Enchenopa binotata (Say) complex(Membracidae Homoptera) Proc Entomol Soc Wash95 574ETH582
Simon C F Frati A Beckenbach B Crespi H Liu and PFlook 1994 Evolution weighting and phylogeneticutility of mitochondrial gene sequences and a compila-tion of conserved polymerase chain reaction primersAnn Entomol Soc Am 87 651ETH701
Sorenson M D 1999 TreeRot version 2 Boston Univer-sity Boston MA
Swofford D L 1998 PAUP phylogenetic analysis usingparsimony (and other methods) test version 400d64Illinois Natural History Survey Champaign IL
170 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol 95 no 2
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
Swofford D L and S H Berlocher 1987 Inferring evo-lutionary trees from gene frequency data under the prin-ciple of maximum parsimony Syst Zool 36 293ETH325
Swofford D L G J Olsen P J Waddell and D M Hillis1996 Phylogenetic Inference pp 407ETH514 In D M Hil-lis C Moritz and B K Mable [eds] Molecular system-atics Sinauer Sunderland MA
Tauber C A and M J Tauber 1982 Sympatric speciationin Chrysopa further discussion Ann Entomol Soc Am75 1ETH2
Templeton A R 1989 The meaning of species and specia-tion a genetic perspective pp 3ETH27 In D Otte and JEndler [eds] Speciation and its consequences SinauerSunderland MA
Tilmon K J T K Wood and J D Pesek 1998 Geneticvariation in performance traits and the potential for hostshifts inEnchenopa treehoppers (HomopteraMembraci-dae) Ann Entomol Soc Am 91 397ETH403
Wendel J F and J J Doyle 1998 Phylogenetic incongru-ence window into genome history and molecular evo-lution pp 265ETH296 In D E Soltis P S Soltis and J JDoyle [eds] Molecular systematics of plants II DNAsequencing Kluwer Academic Nowell MA
Wood T K 1980 IntraspeciTHORNc divergence in Enchenopabinotata Say (Homoptera Membracidae) effected byhost plant adaptation Evolution 34 147ETH160
Wood T K 1993 Speciation of the Enchenopa binotatacomplex (Insecta Homoptera Membracidae) pp 299ETH317 In D R Lees and D Edwards [eds] Evolutionarypatterns and processes Academic New York
Wood T K and S I Guttman 1981 The role of host plantsin the speciation of treehoppers an example from theEnchenopa binotata complex pp 39ETH54 In R F Dennoand H Dingle [eds] Insect life history patterns habitatand geographic variation Springer New York
Wood T K and S I Guttman 1982 Ecological and be-havioural basis for reproductive isolation in the sympatricEnchenopa binotata complex (Homoptera Membraci-dae) Evolution 36 233ETH242
Wood T K and S I Guttman 1983 The Enchenopa bino-tata complex sympatric speciation Science 220 310ETH312
Wood T K and S I Guttman 1985 A newmember of theEnchenopa binotata Say complex on tulip tree (Lirioden-dron tulipifera) Proc Entomol Soc Wash 87 171ETH75
Wood T K and M Keese 1990 Host plant induced assor-tative mating in Enchenopa treehoppers Evolution 44619ETH628
Wood T K and R Patton 1971 Egg froth distribution anddeposition by Enchenopa binotata (Homoptera Mem-bracidae) Ann Entomol Soc Am 65 1190ETH1191
Wood T K K L Olmstead and K L Guttman 1990Insect phenology mediated by host-plant relations Evo-lution 44 629ETH636
Wood T K K J Tilmon A B Shantz C K Harris and JPesek 1999 The role of host-plant THORNdelity in initiatinginsect race formation Evol Ecol Res 1 317ETH332
Received for publication 6 April 2001 accepted 23 October2001
March 2002 LIN AND WOOD MOLECULAR PHYLOGENY OF Enchenopa COMPLEX 171
