ZHAO L ing,ZHENG Zhem in,HUANG Yuan,ZHANG L ing,YANG L iang

(1.College of Agronomy,Xinjiang Agricultural University,Urumqi 830052,China;2.Institute of Zoology,Shaanxi Normal University,Xi'an 710062,China;3.College ofLife Sciences,Yan'an University,Yan'an 716000,China)

Catantopidae is the largest family of Acridoidea with world-wide distribution.The group includesca.4 000 known species in more than 760 genera,which are mostly distributed in the tropics and subtropics.China is the species rich Catantopidae country with ca.430 in more than 100 genera and endemism is very high.Many are Oriental species,with only a few species occurring in Palearctic region.Many species of Catantopidae are pest of agriculture,husbandry and forestry.The harmful species usually belong to the best studied insects in China.Thus,a reliable and accessible classification of these species is fundamental to research in pest control,ecology,evolutionary biology and biodiversity.Meanwhile,the systematic researches on Catantopidae are also very important.To date,there is no consensus system of Catantopidae in China.Some subfamilies,genera and species are still contentious.

Longchuanaceiswas traditionally considered a genus within the subfamily Oxyinae[1].Niu[2]removed Longchuanaceisto Caryandinae by morphological characters.On the basis of his results Sinopodism a Chang and Yunnanacris Chang are both synonymized with Podism odes Ramme.He alsosupport that Yupodism a Zhanget Xia is synonymized with Anapodism aDovnar-Zapolskii[3].In his dissertation,one genus,one subgenus and forty-two species in Cantatopidae,all from China,are new to science.Species groups are proposed in genera:OxyaServille,CaryandaSt?l,LongchuanacrisZhengetFu,Podism odesRamme,Traulia St?l[2].

Systematic research on Cantatopidae by morphology alone can not resolve these contentions completely.Recent studies have demonstrated a great potential for DNA sequence analysis in species identification,phylogeny,and gene flow[4-6].Sequences of the mitochondrial genes cytochrome oxidase subunit I(CO I),cytochrome oxidase subunit II(CO II),and cytochromeb(Cytb)have been extensively applied in rapid species identification,phylogenetic reconstructions for a variety of taxa[7-10].In this study,our objectives were to:①obtain nucleotide sequences of the CO I,CO II,and Cytb mitochondrial genes from Cantatopidae species as a basis for further species identification and phylogenetic analyses,②conduct phylogenetic analyses using these nucleotide sequences(alone and in combination),③examine the species and genera status bymolecular phylogeny,and④compare molecular findings with morphological assess ments of the contentious genera and 7 new species.

1 Materials and Methods

1.1 Taxon samples,DNA extraction,and Gen-Bank sequences

All Cantatopidae samples,including one outgroup species(Atractom orpha sinensisI.BoI.)were collected at 17 sites in six provinces of China with majority from Yunnan Province,and were preserved in 100%ethanol prior to DNA extraction.Total genomic DNA was extracted from single specimens using a standard phenol-chloroform extraction with slight modifications.Extracted genomic DNA was stored at-20℃until it was used as polymerase chain reaction(PCR)template.Published sequences from 33 species,including the other two outgroup species(Filchnerella sunanensis and Haplotropis brunneriana)obtained from Gen Bank were also included in our phylogenetic analyses besides the CO I sequence of Oxya hyla intricata(Accession NO.AF385191).All samples information from this study and GenBank are not present in this paper,but available on request.

1.2 PCR and sequencing

Fragments of the mitochondrial genes CO I,Cytb,and the complete sequence ofCO IIgenewere amplified by PCR,using established primerpairs[11-12]and reaction conditions.

PCR products were purified with DNA Gel purification kit(U-Gene).Direct sequencing of PCR products was performed by the AB I PR IS MTM3100-Avant Genetic Analyzer.

1.3 Sequencealignm entsandphylogenetic inferences

Sequenceswere trimmed by removing ambiguously resolved parts of the 3′and 5′ends and edited using the Contig1software package(Cexpress.exe).DNA sequences were aligned using the program Clustal X with default values.

Uncorrectedp-distances werecalculated with MEGA 4.0.Gaps and missing data were excluded from each pairwise calculation.A maximum likelihood[13]was carried out in PAUP＊4.0b10using GTR+I+Gmodel with parameter values as estimated by Model test 3.7[15].PAUP searches consisted of TBR heuristic searches.The three data sets,CO I,CO II and Cytb,were analyzed separately and in combination of CO I and Cytb.The robustness of the trees was tested using the bootstrap method.All bootstrap values are based on 100 replicates for CO I and Cytb,200 for combination of CO I and Cytb,and 500 for CO II.

Bayesian analyses were also conducted with Mr-Bayes V3.1.2[16],using a GTR+I+G model.We used the default priors starting with random trees,and ran four Markov chains for 150 000 generations,sampled at intervals of 10 generations.To determine the point at which the Markov chains reached stationarity,the log-likelihood scores were plotted against generation time,to determine when the log-likelihood values stabilize.After burn-in samples were discarded,trees were combined in a single majority consensus topology,and the percentage of the nodes were taken as a posteriori probabilities[16].

2 Results

2.1 Nucleotide analyses

The PCR products of CO I,CO II,and Cytb are approx.708,684 and 763 bp,respectively for the Cantatopidae species examined.For the CO I gene,the multiple sequence alignment(excluding outgroup)has 648 characters,of which 364 are constant,279 variable,and 220 are parsimony-in for mative.For CO II,multiple sequence alignments(excluding the out group but including sequences obtained from Gen Bank)had 585 characters,of which 257 are constant,328 variable,and 293 are parsimony-in for mative.For Cytb,there are 711 characters excluding the outgroup but including Gen Bank sequences,of which 377 are constant,332 variable,and 283 are pars imony-in formative.

Average nucleotide composition among Cantatopidae species(excluding the outgroup but including Gen Bank sequences)is as follows:CO I[31.6%(A),34.2%(T),15.9%(G),and 18.3%(C)];CO II[37.1%(A),32.3%(T),13.7%(G),and 16.9%(C)];and Cytb[39.3%(A),34.0%(T),11.5%(G),and 15.2%(C)].

2.2 Uncorrected“p”distance analyses

The sequence divergence between Longchuanacris fuiNiu et Zheng and Longchuanacris xiaoheishanensis Niu et Zheng,based on uncorrected“p”distance,was 0.027 across the CO I fragment,and 0.021 across the Cy tb fragment.Within Caryanda,mean CO I and Cytb divergences were 0.032 and 0.030,respectively.The pairwise sequence divergence among Caryandafour species was 0.014(Caryanda pseudodentata Niuet Zheng and Caryanda bannaensis Niuet Zheng)in CO I,0.006 in Cytb,0.006(Caryanda ruiliensis Niu and Caryanda quadrata Biet Xia)in CO I,0.009 in Cytb,0.040(Caryanda pseudodentata Niuet Zheng and Caryandaquadrata BietXia)and0.041(Caryanda bannaensisand Caryanda quadrata Biet Xia)in Cytb.Mean pairwise divergence among three genera was 0.001(Pedopodism aand Podism odes)for CO I,0.002 for Cytb,and 0.003(Sinopodism aand Podism odes)for Cytb.Furthermore,the divergence values were also calculated forPodism odes dabieshanensis Niuet Zheng and Yunnanacris yunnaneus Ramme in CO I(uncorrected“p”distance=0.056);Apalacris varicornem a Walker and Apalacris em inifronta Niu et Zheng in the CO II gene(0.001).

2.3 Phylogeneticrelationship sinferredfrom CO I,CO II,and Cytb

Phylogenies derived using maximumlikelihood and Bayesian analyses were generally congruent,and four typical trees are shown in the results(Fig.1 -4).Based on the CO I gene sequence withA tractom orpha sinensisas the outgroup,the Bayesian tree(not present)showed an overall similar topology to the likelihood tree(Fig.1 ).Three major clades within the two trees are apparent,the major differences being the position of Traulia two species(Traulia szetschuanensis and Traulia m inuta)and Catantopinae three species(Catantops pinguis pinguis,Stenocatantops splendens and Xenocatantops brachycerus),which formed separate cladeswith likelihood.In the likelihood tree,one clade includes Caryandasix species and Longchuanacristwo species with strong support.The second clade is composed of two sub-clades.These two sub-clades include Yunnanacrisyunnaneus,Podism odesdabieshanensisand Pedopodism a funiushanain one sister group,and Qinlingacris choui,Anapodism a miram ae andPrum nathree species in another.The third clade consists ofEpistaurus aberrans,Apalacris varicornem a,Ecphanthacris m irabilisandApalacris em inifronta.

Fig.1 Maximum likelihood tree of the CO I gene for Chinese Catantopidae species.Numbers near nodes represent bootstrap support(100 replicates)

For the CO II gene with Filchnerella sunanensisas the outgroup,thetopology ofthe Bayesiantree(Fig.2 )was generally identical to that of the likelihood tree(not present),therefore,onlyBayesian tree(Fig.2 )is shown in the results.Both analyses for CO II support phylogeneticrelationshipsincluding:①Coptacrinae species formed a well supported monophyletic group(posterior probability 89%)with the exception of four species,which formed a separate Clade;②Cyrtacanthacrinae was found sister to Calliptaminae,and bothsister to Eyprepocnemidinae;③Catantopinae four species also for med amonophyletic group,and have a close relationship to Cyrtacanthacrinae,Calliptaminae and Eyprepocnemid-inae.These four subfamilies constituted a distinct clade;④Spathosterninae one species,Oxyinae two species and Caryandinae one species grouped together,and formed another one clade;⑤the same truth forMelanoplinae three species and Podisminae two species with highly supportwith posterior probabilities 100%.Our results support the relationshipswithin the genus level,in agreement with the traditional view ofmorphological studies.

Fig.2 Bayesian tree of the CO II gene for Chinese Catantopidae species.Numbers near nodes representBayesian posterior probabilities

For the Cytb gene withHaplotropis brunnerianaas the outgroup,the Bayesian tree(Fig.3 )showed essential congruence with the likelihood tree(available on request).There are two distinct clades in both trees.One clade includes Caryandasix species,Longchuanacris two species and Oxyinae three species with well support.W ithin this clade,Caryandasix species andLongchuanacristwo species grouped together,and paraphyletic with respect to Oxyinae three species.The other clade mainly comprised Podis minae nine species and Melanoplinae three species plus Spathosternum prasiniferumand Qinlingacris choui.In both trees,Coptacrinae was never recovered as monophyly with Trauliaalways in uncertain position.The remaining taxa plusTrauliaeither composed a separate Clade in a comb-like arrangement(Fig.3 )or split direct from the backbone of the tree in a comb-like arrangement as well.

Fig.3 Bayesian tree of the Cytb gene for Chinese Catantopidae species.Numbers near nodes representBayesian posterior probabilities

The combined nucleotide matrix of the CO I and Cytb genes generates the two trees topology(only Bayesian tree is present,Fig.4 )similar to that obtained using CytbDNA sequences.Two clades are suggested:one clade includes Caryanda six species,Longchuanacris two species and Oxyinae three species with well support,while the other is composed totally of Podisminae seven species except Qinlingacris choui,with Yupodism a rufipennisat the basal position.Both trees show that Qinlingacris chouigrouped with Prumna two species,for med one sub-clade and Pedopodism a funiushananested in Podism odes,constituted another,that Caryandasix species and Longchuanacr is two species consistently grouped together.

3 Discussion

Sinopodism a and Yunnanacris are both considered synonyms of Podism odes[2].The result of our investigation of Cy tb gene sequence further supports such a classification,and additionally suggests Pedopodism a is also a possible synonym of Podismodes with Bayesian.However,synonymizing these two genera would be premature because of insufficient taxon and character sampling.Pedopodism ahas been previously considered a synonyms of Sinopodism a[17],but according to Niu[2]Pedopodism awas still a separate valid genus with remarkable differences from Sinopodism ain forewing.Therefore,further investigationswill be needed for addressing their relationship.The relationships inferred from Cytb Bayesian tree are((((Pedopodism a funiushana+P.tsinlingensis)((Sinopodism a houshana+S.qinlingensis)+Podism odes dabieshanensis))+S.guizhouensis)+Yunnanacris yunnaneus).

Fig.4 Bayesian tree of the combination of CO I and Cytb gene for Chinese Catantopidae species.Numbers near nodes represent Bayesian posterior probabilities

Sinopodism a houshana Huang and Podism odes dabieshanensisare two very similar morphospecies,differing only in the top of subgenital plate and furcula[2].These two similar species are also identified bymolecular data here(one individual Cytb gene sequence,and the combined CO Iand Cytb gene sequences)andDNA barcodes[18],which as species identifiers has been successfully implemented in the identification of previously described species,and discovery of new and cryptic species[5,8-9,19-21].

Traditionally the Longchuanacris was considered a genus within the subfamily Oxyinae[1],but it was recently redefined by Niu[2],who removed it from the Oxyinae and placed it in the subfamily Caryandinae.This revision is well supported by phylogenies reconstructed from Cytb gene sequence.Further,Longchuanacrisalso has shown close relationship with Caryanda in the individual genes analyses(CO I and Cytb),and the combined analyses.

Caryanda was divided into two subgenera and further 14 species groups by Niu[2],who revised the genus and provided a phylogenetic analysis based on morphology.On the basis of his results,C.ruiliensis(dehongensis-species group),C.quadrata(quadrataspecies group),C.pseudodentataandC.bannaensis(dentate-species group)in our examined belong to subgenus Yuncaryanda,and C.yunnana(yunnanaspecies group)andC.neoelegans(neoelegans-species group)subordinate within subgenus Caryanda.The species groups and subgenera of Caryandadefined by Niu[2]are in general well supported by our analyses(see Fig.1 ).Both show the relationship are(((C.ruiliensis+C.quadrata)+(C.pseudodentata+C.bannaensis)+C.yunnana)+C.neoelegans).

Storozhenko[3]considered Yupod is ma ZhangetXia was a synonym of Anapodism a Dovnar-Zapolskii,and remove Yupod is ma rufipennisto the genusAnapodism a,which supported byNiu[2],but contrary to our result.From the uncorrected“p”distance in CO I and Phylogeny inferred from CO I gene sequence,we propose that Yupodismais still a separate genus within Podisminae.

Our results also reveal Ecphanthacrisis close related to Apalacris,in agreement with Lu[22].The molecular phylogeny of seven new species was consistant with recentmor phological phylogeny by Niu[2].

In summary,our analyses show that the relationships of lower hierarchical levels are well resolved,and the three mtDNA genes are less useful for resolving deeper nodes,in agreement with general findings in insects[23-24].So additional markers,especially nuclear markers which proved pivotal to solve the problem of the deeper nodes[25-27],will be required for establishing deeper phylogenetic relationships of Cantatopidae.Acknow ledgement:We extend our sincere gratitude to Yao Niu for assistance in collecting and providing specimens.This work was supported by grants from the National Natural Science Foundation of China(NO.30670279).

[1] L IH C,X IA K L.Fauna Sinica.Insecta Vol.43.Orthoptera:Acridoidea:Catantopidae[M].Beijing:Science Press,2006.

[2] N IU Y.Systematics of Cantatopidae(s.l.)(Orthoptera:Acridoidea)from China[M].Xi'an:Shaanxi Normal University Press,2007.

[3] STOROZHENKO S Y.Revision of the genus Anapodism a Dov.-Zap.(Orthoptera,Acrididae)[J].Russian Entomol J,1993,2(3/4):21-25.

[4] FLOYD R,ABEBE E,PAPERT A,et al.Molecular barcodes for soil nematode identification[J].Mol Ecol,2002,11:839-850.

[5] MONAGHAN M T,BALKE M,GREGORY T R,et al.DNA-based species delineation in tropical beetles using nuclear and mitochondrialmarkers[J].Phil Trans R Soc B,2005,360:1925-1933.

[6] JENK INS TM,DEAN R E,VERKERK R,et al.Phylogenetic analyses of two mitochon-drial genes and one nuclear intron region illuminate European subterranean termite gene flow,taxonomy,and introductive dynamics[J].Mol Phylogenet Evol,2001,20:286-293.

[7] FRATI F,S IMON C,SULL IVAN J,et al.Evolution of the mitochondrial cytochrome oxidase II gene in Collembola[J].J Mol Evol,1997,44:145-158.

[8] HEBERT P D N,PENTON E H,BURNS J M,et al.Ten species in one:DNA barcoding reveals cryptic species in the Neotropical skipper butterfly Astraptes fulgerator[J].Proc Natl Acad Sci,2004,101:14812-14817.

[9] HEBERT P D N,STOECKLE M Y,ZE MLAK T S,et al.Identification of birds through DNA barcodes[J].PLoSBiol,2004,2:1657-1663.

[10] L IU H,BECKENBACH A.Evolution of the mitochondrial cytochrome oxidase II gene among ten orders of insects[J].Mol Phylogenet Evol,1992,1(1):41-52.

[11] FOLMER O,BLACKM,HOEH W,et al.DNA primers for amplification of mitochon-drial cytochrome c oxidase subunit I from diverse metazoan invertebrates[J].MolMar Biol Biotechnol,1994,3:294-299.

[12] S IMON C,FRATI F,BECKENBACH A,et al.Evolution,weighting,and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers[J].Ann Entomol Soc Am,1994,87:651-701.

[13] FELENSTE IN J.Evolutionarytrees from DNAsequences,a maximum likelihood approach[J].J Mol Evol,1981,17:368-376.

[14] S WOFFORD D L.PAUP.Phylogenetic Analysis using Parsimony(and othermethods),Version 4.0b10[M].Sunderland:SinauerAssociates,M A,2002.

[15] POSADA D,CRANDALL K A.Modeltest:testing the model of DNA substitution[J].Bioinformatics,1998,14:817-818.

[16] HUELSENBECK JP,RONQU ISTF.MrBAYES:Bayesian inference of phylogenetic trees[J].Bioinformatics,2001,17:754-755.

[17] STOROZHENKO S Y.To the knowledge of the tribe Melanoplini(Orthoptera,Acrididae:Catantopinae)of the Eastern Palearctica[J].Articulata,1994[1993],8(2):1-22.

[18] ZHANGL.MtDNA phylogeny of Catantopidae(Orthopter:Acridoidea)from China[D].Xi'an:Shaanxi NormalUniversity Press,2008.

[19] HEBERT P D N,RATNAS INGHAM S,DEWAARD J R.Barcoding animal life:cytochrome cytochrome c oxidase subunit 1 divergences among closely related species[J].Proc R Soc B,2003,270(Suppl.1):S96-S99.

[20] HOGG ID,HEBERT PD N.Biological identification of springtails(Hexapoda:Collembola)from the Canadian Arctic,using mitochondrial DNA barcodes[J].Can J Zool,2004,82:749-754.

[21] WARD R D,ZEMLAK T S, INNES B H,et al.DNA barcodingAustralia's fish species[J].Phil Trans R Soc B,2005,360:1847-1857.

[22] LU R S.Molecularphylogenetic analysisof some species of Catantopidae(Orthoptera:Acridoidea)in China[D].Xi'an:ShaanxiNormalUniversity Press,2005.

[23] DANFORTH B N,L IN C P,FANG J.How do insect nuclear ribosomal genes compare to protein-coding genes in phylogenetic utility and DNA substitution patterns?[J].Syst Ent,2005,30:549-562.

[24] L IN C P,DANFORTH B N.How do insect nuclear and mitochondrial gene substitution patterns differ?Insights from Bayesian analyses of combined data sets[J].Mol Phylogenet Evol,2004,30:686-702.

[25] R IBERA I,HOGAN J E,VOGLER A P.Phylogeny of hydradephagan water beetles inferred from 18S rRNA sequences[J].Mol Phylogenet Evol,2002,23:43-62.

[26] BALKEM,PONS J,R IBERA I,et al.Infrequent and unidirectional colonization of hyperdiversePapuadytes diving beetles in New Caledonia and New Guinea[J].Mol Phylogenet Evol,2007,42:505-516.

[27] WANG Y F.Molecular Evolution and Phylogenetic A-nalysis of Some Species of Arcypteridae(Orthoptera:Caelifera:Acridoidea)in China[D].Xi'an:Shaanxi Nor malUniversity Press,2006.