Contents lists available at ScienceDirect

Journal of Asia-Pacific Entomology

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

Molecular phylogeny of the paper wasp subgenus Polistes (Polistella)Ashmead, 1904 (Hymenoptera: Vespidae: Polistinae) from Vietnam

Lien T.P. Nguyena,b,⁎, Anh D. Nguyena,c, Trang T.P. Nguyena, Adrien Perrardd,e,James M. Carpenterf

a Institute of Ecology and Biological Resources, Vietnam Academy of Science and Technology, 18, Hoangquocviet Rd., Caugiay Distr., Hanoi, Viet NambGraduate School of Science and Technology, Vietnam Academy of Science and Technology, 18, Hoangquocviet Rd., Caugiay Distr., Hanoi, Viet Namc Duy Tan University, 254, Nguyen Van Linh, Da Nang City, Viet Namd Sorbonne Universités, UPMC Univ Paris 06, Institute of Ecology and Environmental Sciences of Paris (iEES Paris), 4 Place Jussieu, 75252 Paris, FranceeUniversité Paris Diderot, iEES-Paris UMR 7598, Sorbonne Paris Cité, F-75005 Paris, FrancefDivision of Invertebrate Zoology, American Museum of Natural History, New York, NY 10024, USA

A R T I C L E I N F O

Keywords:VespidaePolistesPaper waspPhylogenyMitochondrial genesVietnam

A B S T R A C T

This study provides the first molecular phylogeny of the social wasp subgenus Polistella (Hymenoptera: Vespidae:Polistes) from Vietnam. Fragments of the mitochondrial COI and 16S rRNA genes were used to reconstruct thephylogenetic trees among 38 Polistes species plus two out-group species (Vespa soror du Buysson and Ropalidiafasciata (Fabricius)). Our results support the existence of several species-groups, including two that are con-gruent with the previous stigma and Stenopolistes groups defined on the basis of morphology. Moreover, werecovered a clade including the stigma group and the two species P. humilis and P. variabilis that was sister to allother species of Polistella. However, the results also challenged the definition of other groups of Polistella basedon morphological data, as well as the definition of two species: P. brunus and P. affinis. This first study calls forfurther analyses including morphological characters to clarify the taxonomy and the classification of the group.

Introduction

Social vespids are of central importance to studies of behavioralevolution because among them, the polistines (paper wasps) are con-sidered to mark a clear, finely divided transition between primitivelyand highly eusocial behavior (Pardi, 1996). The vast majority of be-havioral investigations of the Polistinae have focused on the cosmo-politan genus Polistes Latreille. Their relatively small colonies (usuallywith fewer than 100 individuals) and uncovered nests often made onhuman constructions allow us to make detailed behavioral observa-tions. Therefore, Polistes has been the genus of particular interest forboth experimental and theoretical investigations of social behavior(Turillazzi and West-Eberhard, 1996; Jandt et al., 2013). However,previous phylogenetic analyses of Polistes have included only a fewspecies of Polistella (Pickett and Wenzel, 2004; Pickett et al., 2006;Arévalo et al., 2004; Pickett and Carpenter, 2010; Santos et al., 2015).

With more than 200 species, Polistes is subdivided into four mono-phyletic subgenera, with their relationships expressed as(Gyrostoma+(Polistella+(Polistes s. str. + Aphanilopterus)))(Carpenter, 1996a). Gyrostoma, Polistella and Polistes s.str. are primarily

of the Old World (including Australia), and Aphanilopterus is exclusiveto the New World. Of the four subgenera, Polistella, presently with 90species and 34 subspecies distributed from Africa to Australia(Carpenter, 1996b, updated), has been less studied taxonomically,phylogenetically and biologically than other social wasps.

The phylogenetic relationships among Polistes species (including230 species and 88 subspecies) have been studied by several authors.The first large-scale analysis within Polistes was Carpenter (1996a), whostudied the morphology of 144 species and an additional 43 subspecies,thus sampling most of the diversity within the genus. The number ofPolistella species used in that study was 36 species and 21 subspecies,the largest in number of Polistella species up to present (only a fewPolistella species were used in the studies of Pickett and Wenzel, 2004,and Pickett et al., 2006). Carpenter (1996a) synonymized Stenopolistesunder Polistella, and divided Polistella into 4 species groups, namelyPolistes adustus, P. stigma, P. sagittarius and Stenopolistes, and thosespecies groups were used until now.

The first extensive molecular treatment was by Arévalo et al.(2004), which included 33 species of Polistes (with only three species inPolistella), along with 33 species of other tribes of Polistinae. The

https://doi.org/10.1016/j.aspen.2018.04.006Received 1 February 2018; Received in revised form 5 April 2018; Accepted 10 April 2018

⁎ Corresponding author at: Institute of Ecology and Biological Resources, Vietnam Academy of Science and Technology, 18, Hoangquocviet Rd., Caugiay Distr., Hanoi, Viet Nam.E-mail addresses: phuonglientit@gmail.com, ntplien@iebr.ac.vn (L.T.P. Nguyen).

Journal of Asia-Pacific Entomology 21 (2018) 638–644

Available online 12 April 20181226-8615/ © 2018 Korean Society of Applied Entomology. Published by Elsevier B.V. All rights reserved.

T

molecular data consisted of mitochondrial CO1 sequences, three nu-clear DNA microsatellite flanking sequences, and the three repeat mo-tifs for the microsatellites represented by gap-coding. These data werecombined with the adult morphological characters of Carpenter(1996a), with characters used by Carpenter et al. (2000) in their studyof the genus Polybia, with larval characters taken from Kojima (1998)supplemented by other literature sources, and with characters of nestarchitecture from Wenzel (1993). In the combined analysis, each of theOld World subgenera was monophyletic including Polistella. And withinthe New World clade, both Aphanilopterus s.str. and Epicnemius (syno-nymized with Aphanilopterus) were monophyletic (however, thesegroups included ten and two species, respectively).

Pickett and Wenzel (2004) subsequently analyzed 40 species ofPolistes (33 of which were Aphanilopterus spp.) using slightly modifiedmorphological data and a different, non-overlapping fragment of CO1.As with the preceding studies, the New World species of Polistes formeda monophyletic group. The arrangement among the Old World sub-genera was different from both of the previous studies, but was basedon just five species. In contrast to the study by Arévalo et al. (2004), theresults of Pickett and Wenzel (2004) agreed with Carpenter's (1996a)conclusions that neither Aphanilopterus s. str. nor Epicnemius aremonophyletic.

Because there are both taxon sampling and evidentiary differencesbetween Arévalo et al. (2004), Pickett and Wenzel (2004), Pickett et al.(2006) conducted a new analysis that combined the two fragments ofCO1 for as many taxa from the two studies as possible. Pickett et al.(2006) treated 48 taxa, including all Polistes species from Arévalo et al.(2004) and all taxa, both species and subspecies from Pickett andWenzel (2004). As the result, a phylogenetic tree based on combinedmorphological and molecular data was proposed. The resolution amongthe New World subgenera was better than any previous analysis. Alltested subgenera were monophyletic. For the Old World subgenera,Gyrostoma was monophyletic and sister to all other Polistes. No previousphylogenetic studies showed this relationship.

Pickett and Carpenter (2010) presented a new phylogenetic analysisof Vespidae based on what was by far the largest taxon sample to in-clude molecular data (using four genes CO1, 12S, 16S and 28S), and thelargest phenotypic character dataset ever compiled, including 45 spe-cies of Polistes (with only four species in Polistella). The results fromPickett and Carpenter (2010) showed that only the subgenus Aphani-lopterus is paraphyletic, rendered so by Gyrostoma, and three of thesesubgenera are monophyletic: Gyrostoma, Polistella, and Polistes s.str. InPolistella stigma formed a clade with a species from the sagittarius group(P. sagittarius).

Santos et al. (2015) presented the an analysis of the phylogeny ofthe genus Polistes using morphological and behavioral characters, aswell as molecular data from six genes (COI, 12S, 16S, 28S, H3 and EF1-∂). A total of 58 species of Polisteswere included in the analyses, with anemphasis on New World species (only 8 species were in Polistella). Inthe analysis of morphology data alone, Polistella was monophyletic, andstigma formed a clade with species from the sagittarius group (P. sa-gittarius) as sister to the other groups.

No phylogenetic analysis of Polistella species has been done untilnow. There are some reasons why the species-level phylogeny ofPolistella has been neglected. First, despite behavioral salience, mor-phological variation within Polistella is slight. The other reason is mostcollections do not contain even a majority of the species, making itdifficult for researchers to gain necessary experience with the group. Asthe morphology is so uniform in Polistella, it makes sense to seek othersources of data, such as molecular data. But DNA-grade specimens ofPolistella are even scarcer than pinned specimens, and so obtainingmolecular data is difficult.

Polistella had originally been subdivided into species groups (Dasand Gupta, 1989), but Carpenter (1996a) mentioned that Das & Gupta'sgroups do not include all of the described species assignable to thesubgenus. Das and Gupta treated only species occurring in India (15

species), dividing them into the adustus group, stigma group, and ma-culipennis group. No morphological features characterized their adustusgroup; rather, it was based on color, namely a black metasoma. Thestigma group included two species, diagnosed principally by having theforewing marginal cell with a pigmented spot. The propodeal concavitywas said to be shallower and the striae finer than in the third group, themaculipennis group, however these characters do not actually allowseparation of the maculipennis group. Three species were included in themaculipennis group but one of these, maculipennis itself, has theforewing spotted apically - as stated by Das and Gupta (1989). Petersen(1987), cited by Das & Gupta, even treated maculipennis as a subspeciesof stigma.

Das & Gupta's groups as constituted were thus unsuitable, and hadbeen modified by Carpenter (1996a), who used 36 Polistella species inhis study, as follows: The adustus group comprises those species havinga tuberculate male sternum VII. The stigma group includes the specieswith an apically spotted forewing, while the sagittarius group includesall remaining species (sagittarius and strigosus were the other two spe-cies included by Das & Gupta in their maculipennis group). But as he(Carpenter, 1996a) pointed out all the species groups are as yet poorlydefined. Thus far, the phylogenetic relationships of Polistella have notbeen evaluated in a study focused on this group.

This research, for the first time, provides a phylogenetic analysisamong species of the subgenus Polistella based on molecular data, whichcan then allow us to revise the poorly established species group systemof Polistella based on monophyly, and define the phylogenetic re-lationships among Polistella species.

Materials and methods

Taxon sampling

A total of 38 species of Polistes were included in the analyses, inaddition to two outgroup species of Vespa and Ropalidia. Thirty-eightingroup species representing four former species groups defined byCarpenter (1996a) (Polistes adustus, P. stigma, P. sagittarius and Steno-polistes groups) and one species group proposed by Nguyen et al. (2011)(group of basally strongly swollen second metasomal sternum) wereincluded. Three specimens from Vietnam that could not be morpholo-gically attributed to known species were included in the analysis (sp.1,sp.2 and sp.3).

DNA extraction and amplification

Total genomic DNA was extracted from leg tissues or body musclesusing the DNAeasy Blood & Tissue Kit (Qiagen™).

Fragments of two mitochondrial genes, Cytochrome c subunit I(COI) and 16S rRNA, were amplified using polymerase chain reaction(PCR). The universal primer set (HCO-2190 and LCO-1498) (Folmeret al., 1994) was used to amplify a 680 bp fragment of COI while aprimer set (16S-R1 (5′-TTA CGC TGT TAT CCC TAA-3′) and 16S-R2 (5′-GTA CCT TGT GTA TCA GGG TT-3′)) (Saito et al., 2007) was used for a900 bp fragment of the 16S rRNA.

PCR conditions for amplification of the 16S rRNA gene was: an in-itial denaturation at 95 °C for 2min followed by 36 cycles of 95 °C for20 s, 45 °C for 40 s, and 72 °C for 1min, and a final extension at 72o for5min. The annealing temperature was changed to 42 °C for 45 s whenamplifying the COI gene.

PCR products were checked for potentially successful amplificationof a fragment of each gene using electrophoresis in 1% Agarose-TBE 1X.Successfully amplified PCR products were purified using QIA quick PCRPurification Kit (Qiagen Inc.), and sequenced at Solgen, Inc. (Korea) onan Applied Biosystems automatic sequencer (ABI3130 XL).

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

639

Alignment and phylogenetic analysis

Each successful sequence was manually checked using BioEditver.7.1 (Hall, 1999) and confirmed with BLAST searches (Altschulet al., 1997). All confirmed sequences were aligned with MUSCLE(Edgar, 2004). Ambiguous nucleotide sites and gaps were removedusing MEGA ver. 6.0 (Tamura et al., 2013).

ModelFinder was implemented in IQTREE ver. 1.5.5 for Windows tofind the most appropriate maximum likelihood substitution model forCOI and 16S rRNA (Kalyaanamoorthy et al., 2017). Models with thelowest Bayesian Information Criterion (BIC) scores were considered fordescribing the best substitution pattern for each gene. Codon positionsincluded were: 1st+ 2nd+3rd+Noncoding. The selected model wasthe General Time Reversible (GTR)+G+ I for COI and 16S rRNAgenes. Maximum likelihood (ML) bootstrap analysis was conductedusing IQTREE ver. 1.5.5 for Windows (Nguyen et al., 2015) with 1000replicates. The two datasets CO1 and 16S rRNA were analyzed sepa-rately and combined, using only specimens for which both sequenceswere available. A Bayesian inference (BI) tree was created usingMrBayes ver. 3.1.2 (Huelsenbeck and Ronquist, 2001) with 10 milliongenerations, heating parameter of 0.06, and sampling every 1000generations. Data were also analyzed under maximum parsimony (MP)with the program TNT (Goloboff et al., 2008). The analysis employedall four new technology search methods, using the default settings,except: the ratchet upweighting probability was 8% and there were 200iterations; tree-drifting used 50 cycles; tree-fusing used five rounds; andminimum length was set to be hit 15 times. The bootstrap resamplingused 1000 iterations, with 10 random addition sequences per replicate.

All nucleotide sequences were deposited at GenBank. Collectionlocalities, specimen vouchers, and GenBank accession numbers aresummarized in Table 1.

Results

DNA variation and distances

The aligned dataset of the gene COI consists of 583 bp, and hasnucleotide frequencies of 33.7, 41.0, 10.7 and 14.7 for A, T, G and C,respectively. The GC content accounts for 25.4% of total nucleotides.The COI dataset contains 194 (33.3%) parsimony informative and 249(42.7%) variable sites. The uncorrected p-distance between the taxaranges from 0.000 to 0.271; overall p-distance is 0.134.

The aligned dataset of the gene 16S rRNA contains 888 bp, hasnucleotide frequencies of 44.8, 42.9, 4.0 and 8.3 for A, T, G and C,respectively. The GC content accounts for 12.3% of total nucleotides.The 16S rRNA dataset contains 316 (35.6%) parsimony informative and394 (44.4%) variable sites. The uncorrected p-distance between thetaxa ranges from 0.000 to 0.212; overall p-distance is 0.123.

The aligned dataset of the combination COI and16S rRNA genescontains 1130 bp, has nucleotide frequencies of 37.9, 41.4, 8.2 and 12.4for A, T, G and C, respectively. The GC content accounts for 20.6% oftotal nucleotides. The dataset contains 268 (23.7%) parsimony in-formative and 343 (30.3%) variable sites. The uncorrected p-distancebetween the taxa ranges from 0.001 to 0.073; overall p-distance is0.046.

Data for the aligned dataset of the gene COI, 16S rRNA and thecombined COI and 16S rRNA are shown in Table 2:

Phylogenetic analysis

The phylogenetic trees were reconstructed for the single gene COIand 16S rRNA using the three methods of ML, BI and MP. For ML andMP analyses, we consider clades with bootstrap values below 65% to beweakly supported, between 65 and 85% to be moderately supported,and more than 85% to be strongly supported. For BI analysis, cladeswill be weakly supported if BI posterior probability is less than

0.65 bpp; be moderately supported if BI is between 0.65 and 0.85 bpp;and be well supported if BI is more than 0.85 bpp.

The COI gene

The phylogenetic trees inferred from the 583 bp fragment of the COIbased on different methods present corresponding topologies: no sup-ported node is contradicted by another supported node with a differentmethod. The resulting topology shows a structure among Polistes spe-cies (Fig. 1). Most nodes are supported using either the ML or BImethods, while many of the deeper nodes are not supported based onMP. There is a clade of all 38 Polistes species. A group of three species,(Polistes metricus+(P. chinensis+ P. riparius)), forms a sister clade ofthe other remaining Polistes species, strongly supported in the BI(0.97bpp) and moderately supported in the ML (84%) analyses but notsupported by MP. The second group consists of five species ((P.gigas+ P. tenebricosus)+ (P. rothneyi+(P. jokahamae+ P. olivaceus)),and is separated from other species with moderate BI values (0.87bpp),and weakly supported in ML and MP (less than 65%). The monophyly ofthe 30 specimens of twenty-five species of the subgenus Polistella issupported in ML, MP and BI analyses albeit with low values (less than65% and less than 0.65 bpp, respectively). This clade is not entirelyresolved, but it contains 8 species-groups found in all analyses. Group I(P. communalis + P. brunus+ P. stigma tamula+ P. stigma bernardii)includes species in the Polistes stigma group (Nguyen et al., 2017). In MLand BI analyses, this group forms a sister clade of Group II (P. hu-milis+ P. variabilis) (as in Santos et al., 2015). Group III (P. kha-sianus+(Polistes sp.1+Polistes sp.2+P. delhiensis + P. nigritarsis))includes species of the so-called “Stenopolistes” (Nguyen and Carpenter,2016, and pers. observation). Group IV (Polistes chuyangsin+(P. af-finis+ P. nipponensis)), Group V (P. lepcha+(P. dawnae+(P. snel-leni+(P. santoshae+ P. reliciniclypeus)))) and Group VI (P. mandar-inus+ Polistes sp.3) include species with a basally strongly swollensecond metasomal sternum (Nguyen et al., 2011; Nguyen and Kojima,2014)). Finally, species from the sagittarius group are split between twogroups: Group VII (P. formosanus+ P. japonicus) and Group VIII (P.strigosus+ P. sagittarius).

The 16S rRNA gene

Similarly, the phylogenetic trees inferred from the 905 bp fragmentof the 16S rRNA (Fig. 2) show corresponding topologies with the threedifferent phylogenetic methods. The monophyly of Polistella species isbetter supported with moderate ML value (84%), strong BI value(0.98bpp) but low MP value (less than 65%). The species-groups arefound again with all three methods. The sister clade of the 20 Polistellaspecies is (P. chinensis+ P. riparius), a clade strongly supported withML, MP and BI values (99% for ML, 91% for MP and 1.00bpp for BI).Sampling differences aside, the species groups of Polistella found withthe COI gene are also supported by the 16S rRNA sequences. However,the relationships among those groups are still poorly resolved. The twomain differences are that the groups I and II appear as sister-clades withstronger support (ML: 100%, MP: 89%, BI: 1.00 bp) and that the re-solution within Group V is lower and less congruent between methods,with the exception of P. lepcha being consistently at the base of theclade.

The combination of COI and 16S rRNA genes

Simultaneous analysis of the combination of a 583 bp fragment ofthe COI gene and a 905 bp fragment of the 16S rRNA was made (Fig. 3).It is noted that not all data of the 16S rRNA of all species was collected,only 25 species with 16S rRNA data from a total of 40 species with COIgene data were available. In the combined analysis, as in the phylo-genetic tree inferring from the 905 bp fragment of the 16S rRNA(Fig. 2), the group of three species, P. gigas, P. rothneyi and P. jokahamae

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

640

is a sister clade of the other remaining species, with the separation isstrongly supported with ML, MP and BI values (100% for ML and MP,and 1.00bpp for BI), another group (P. chinensis and P. riparius) isconsidered to be a sister clade of all remaining 20 Polistella species, withvery strong ML, MP and BI values (99%, 91% and 1.00bpp, respec-tively). Unlike the trees derived from COI and like the trees derivedfrom 16S rRNA, the combined analysis shows the monophyly of Polis-tella with moderate ML value (84%) and strong BI values (0.97bpp),and the relationships among species of the Polistella are mostly re-solved. It contains 6 groups: Group I (P. communalis and P. brunus) in-cludes species in Polistes stigma group (Nguyen et al., 2017), this groupforms a sister clade of Group II (P. humilis and P. variabilis); Group III(Polistes sp.1, P. delhiensis, P. sp.2, P. nigritarsis, and P. khasianus) in-cludes species of the so-called “Stenopolistes” (Nguyen and Carpenter,2016, and pers. observation), Group IV (P. affinis, P. nipponensis, and P.chuyangsin), Group V (P. dawnae, P. santohae, P. reliciniclypeus, P.

snelleni, and P. lepcha) includes species in the group of basally stronglyswollen second metasomal sternum (Nguyen et al., 2011; Nguyen andKojima, 2014), Group VII (P. formosanus and P. japonicus) and GroupVIII with only one species (P. strigosus) include species in the sagittariusgroup (Carpenter, 1996a).

Discussion

Our results based on two genes highlight the existence of well-supported species-groups within Polistella. Those species groups areonly partially congruent with the pre-existing species-groups developedby Das and Gupta (1989), Carpenter (1996a), Nguyen et al. (2011), andNguyen and Carpenter (2016). The stigma and the Stenopolistes groupsare supported by our molecular data (Group I and Group III). On theother hand, the adustus group, represented by P. dawnae, P. lepcha andP. santoshae in our sampling, is embedded in the group with a basallyswollen second sternum defined by Nguyen et al. (2011). In addition,the sagittarius group is divided between the groups II, VII and VIII(Carpenter, 1996a). Group II (P. humilis + P. variabilis) appears closelyrelated to the stigma group based on the 16S and combined analyses.This clade of group I and II was already supported by morphologicaldata, but not by molecular data, in a previous analysis including asmaller Polistella sample (Santos et al., 2015). Based on both these re-sults, the sagittarius group as previously defined appears polyphyletic.The clade of the group IV and V combined based on the 16S analysisincluding all species with a basally swollen second metasomal sternum,not divided in to two groups based on the COI analysis. In the light of

Table 1Specimen vouchers and GenBank accession numbers.

No. Samples Locality Voucher COI 16S

1 Polistes strigosus Bequaert Cai Kinh, Huu Lung, Lang Son, 22°39′42.9″N, 106°15′36″E IEBR-VesVn-1 MG733318 MG7393842 Polistes communalis Nguyen, Vu & Carpenter Cai Kinh, Huu Lung, Lang Son, 22°39′42.9″N, 106°15′36″E IEBR-VesVn-2 MG733319 MG7393853 Polistes communalis Nguyen, Vu & Carpenter Gia Phu, Phu Yen, Son La, 21°13′07.9″N, 104°32′39.7″E IEBR-VesVn-3 MG733320 MG7393864 Polistes communalis Nguyen, Vu & Carpenter Na Hang NR, Tuyen Quang, 22°21′07″N, 105°25′34.7″E IEBR-VesVn-4 MG733321 MG7393875 Polistes communalis Nguyen, Vu & Carpenter Cai Kinh, Huu Lung, Lang Son, 22°39′42.9″N, 106°15′36″E IEBR-VesVn-5 MG733317 –6 Polistes brunus Nguyen & Carpenter Easo NR, Eaka, Dak Lak, 12°55′24.3″N, 108°33′0.4″E IEBR-VesVn-6 MG733322 MG7393887 Polistes stigma tamulus (Fabricius) Cat Ba NP, Cat Hai, Hai Phong, 20°43′81″N, 107°4′74″E IEBR-VesVn-8 MG733323 –8 Polistes sp.1 Ta Dung NR, Dak Som, Dak Glong, Dak Nong, 11°39′42.9″N, 106°15′36″E IEBR-VesVn-10 MG733324 MG7393899 Polistes delhiensis Das & Goupta Kim Hi NP, Lang San, Na Ri, Bac Kan, 22°14″N, 106°05′E IEBR-VesVn-13 MG733325 MG73939010 Polistes affinis Gusenleitner Tam Dao NP, Vinh Phuc IEBR-VesVn-15 MG733326 –11 Polistes dawnae Dover & Rao Muong Lum, Yen Chau, Son La, 21°1′34″N, 104°28′57″E IEBR-VesVn-16 MG733327 MG73939112 Polistes santoshae Das & Goupta Bi Doup – Nui Ba NP, Lam Dong IEBR-VesVn-17 MG733328 MG73939213 Polistes lepcha Cameron Luong Lum, Yen Chau, Son La, 21°2′7″N, 104°26′51″E IEBR-VesVn-18 MG733329 MG73939314 Polistes mandarinus de Saussure Phia Oac NR, Nguyen Binh, Cao Bang IEBR-VesVn-19 MG733330 –15 Polistes sp.3 Phia Oac, Nguyen Binh, Cao Bang IEBR-VesVn-20 MG733331 –16 Polistes reliciniclypeus Nguyen, Kojima & Saito Dong Van, Ha Giang, 23°7′16.6″N, 105°5′39.5″E IEBR-VesVn-21 MG733332 MG73939417 Polistes khasianus Cameron Hoang Lien NP, Sa Pa, Lao Cai IEBR-VesVn-22 MG733333 MG73939518 Polistes sp.2 Kon Chu Rang NR, Son Lang, KBang, Gia Lai IEBR-VesVn-26 MG733334 MG73939619 Polistes nigritarsis Cameron Kon Chu Rang NR, Son Lang, KBang, Gia Lai IEBR-VesVn-27 MG733335 MG73939720 Polistes affinis Gusenleitner Kon Chu Rang NR, Son Lang, KBang, Gia Lai IEBR-VesVn-30 MG733336 MG73939821 Polistes chuyangsin Nguyen &Nguyen Chu Yang Sin NP, Krong Kmar, Krong Bong, Dak Lak IEBR-VesVn-31 MG733337 MG73939922 Polistes santoshae Das & Goupta Bi Doup – Nui Ba NP, Lam Dong IEBR-VesVn-32 MG733338 MG73940023 Polistes gigas (Kirby) Cat Ba NP, Cat Hai, Hai Phong, 20°47′58″N, 106°59′93″E IEBR-VesVn-34 MG733339 MG73940124 Polistes chinensis Fabricius GenBank AB284740 AB28453725 Polistes formosanus Sonan GenBank AB284759 AB28455726 Polistes humilis (Fabricius) GenBank AB284746 AB28454527 Polistes japonicus de Saussure GenBank AB284750 AB28454828 Polistes jokahamae Radoszkowski GenBank AB284738 AB28453929 Polistes metricus Say GenBank AB284742 –30 Polistes nipponensis Pérez GenBank AB284744 AB28454331 Polistes olivaceus (DeGeer) GenBank AB570104 –32 Polistes riparius Yamane & Yamane GenBank AB284741 AB28453833 Polistes rothneyi Cameron GenBank AB284739 AB28454034 Polistes sagittarius de Saussure GenBank GU596922 –35 Polistes snelleni de Saussure GenBank AB284743 AB28454236 Polistes stigma Bernardii GenBank GU596890 –37 Polistes tenebricosus Lepeletier GenBank GU596891 –38 Polistes variabilis (Fabricius) GenBank AB284745 AB28454439 Ropalidia fasciata (Fabricius) GenBank AB969808 –40 Vespa soror du Buysson GenBank KF933086 –

Table 2Average base frequencies and p-distance range for the fragments of COI and 16SrRNA.

Dataset Average base frequencies p-Distance range

A C G T

COI 33.7 14.7 10.7 41.0 0.000–0.27116S rRNA 44.7 8.2 4.0 43.1 0.000–0.212COI-16S rRNA 37.9 12.4 8.2 41.4 0.001–0.073

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

641

those new results, the definition of the adustus group, the sagittariusgroup and the group with a basally swollen second metasomal sternumwould require revision.

Besides the existence of those well-supported species groups, therelationships among the clades remain poorly resolved in the presentanalyses. Previous studies provided several hypotheses as to the re-lationships within Polistella (Carpenter, 1996a; Arévalo et al., 2004;Pickett et al., 2006; Pickett and Carpenter, 2010; Santos et al., 2015).Most of those studies included only a couple of species of Polistella, notenough to discuss the relationships between some of the species groups.One recurrent result was the presence of the stigma group as sister of theother groups, based on both morphological and molecular data. This

group was also found forming a clade with species from the sagittariusgroup (with P. sagittarius, in Pickett and Carpenter (2010) and Santoset al. (2015), molecular data and total evidence analysis, and with P.humilis and P. variabilis in Santos et al. (2015), morphological data).Such a structure was found in several of our analyses, but with rela-tively low support, except for high support for a clade of P. humilis andP. variabilis. In addition, the relationships between the groups III to VIIIwere unstable depending on the data and the methods. The monophylyof Polistella was even weakly supported to absent based on the COIsequences. These results suggest that additional data could be requiredto better resolve our picture of the evolution of Polistella.

Finally, our data enable us to relate Vietnamese species to the

Fig. 1. The phylogenetic relationships among Polistes species inferred from a 583 bp fragment of the COI gene based on Maximum Likelihood, Bayesian Inference andMaximum Parsimony analyses. #: bootstrap or BI values less than 65 or 0.65.

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

642

existing diversity. We found that two unnamed species (Polistella sp.1and Polistella sp.2) are part of the Stenopolistes group, and that anotheris related to species from the group of basally swollen second meta-somal sternum (Polistella sp.3 with P. mandarinus). The results from theCOI analysis also highlighted two potential problems in the definition

of species: Polistes stigma and P. affinis appear paraphyletic because of P.brunus and P. nipponensis respectively. Indeed, P. brunus and P. stigmatamula form a supported clade, challenging the definition of those taxa.Those results being only supported by the COI analysis, they may callfor a revision of those four species. These revisions would require more

Fig. 2. The phylogenetic relationships among Polistes species inferred from a 905 bp fragment of the 16S rRNA based on Maximum Likelihood, Bayesian Inferenceand Maximum Parsimony analyses. #: bootstrap or BI values less than 65 or 0.65.

Fig. 3. The phylogenetic relationships among Polistes species inferred from combination of a 583 bp fragment of the COI gene and a 905 bp fragment of the 16S rRNAbased on Maximum Likelihood, Bayesian Inference and Maximum Parsimony analyses. #: bootstrap or BI values less than 65 or 0.65.

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

643

data to properly define the taxa based on molecular and morphologicalevidence.

Altogether, this study brings new light to the evolution of this im-portant Polistes group by clearly defining several species groups, byconfirming the existence of a stigma group and a Stenopolistes group andby challenging the current definition of the other existing groups andsome of the species. In addition, the results suggest that the stigmagroup, together with a clade of P. humilis and P. variabilis, are sister toall the other Polistella. However, this analysis calls for the addition ofnew data, potentially with the inclusion of morphological evidence.

Acknowledgements

The present study was supported by the grant from the VietnamNational Foundation for Science and Technology Development(NAFOSTED: no. 106-NN.05-2014.40).

We have no conflicts of interest to disclose.

References

Altschul, S.F., Madden, T.L., Schäffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J.,1997. Gapped BLAST and PSI-BLAST: a new generation of protein database searchprograms. Nucleic Acids Res. 25, 3389–3402.

Arévalo, E., Zhu, Y., Carpenter, J.M., Strassmann, J.E., 2004. The phylogeny of the socialwasp subfamily Polistinae: envidence from microsatellite flanking sequences mi-tochondrial COI sequence, and morphological characters. BioMed Central Evol. Biol.4, 8.

Carpenter, J.M., 1996a. Phylogeny and biogeography of Polistes. In: Turillazzi, S., West-Eberhard, M.J. (Eds.), Natural History and Evolution of Paper-Wasps. OxfordUniversity Press, pp. 18–57.

Carpenter, J.M., 1996b. Distributional checklist of species of the genus Polistes(Hymenoptera: Vespidae; Polistinae, Polistini). Am. Mus. Novit. 3188, 1–39.

Carpenter, J.M., Kojima, J., Wenzel, J.W., 2000. Polybia, paraphyly, and polistine phy-logeny. Am. Mus. Novit. 3298 (24 pp.).

Das, B.P., Gupta, V.K., 1989. The social wasps of India and the adjacent countries (hy-menoptera: Vespidae). Oriental Insects Monogr. 11, 176–189.

Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and highthroughput. Nucleic Acids Res. 32 (5), 1792–1797. http://dx.doi.org/10.1093/nar/gkh340.

Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for am-plification of mitochondrial cytochrome c oxidase subunit I from diverse metazoaninvertebrates. Mol. Mar. Biol. Biotechnol. 3, 294–299.

Goloboff, P.A., Farris, J.S., Nixon, K.C., 2008. TNT, a free program for phylogeneticanalysis. Cladistics 24, 1–13.

Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and ana-lysis program for windows 95/98/NT. Nucleic Acids Symp. Ser. 41, 95–98.

Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of phylogenetic

trees. Bioinformatics 17, 754–755.Jandt, J.M., Tibbetts, E.A., Toth, A.L., 2013. Polistes paper wasps: a model genus for the

study of social dominance hierarchies. Insect. Soc. 61, 11–27.Kalyaanamoorthy, S., Minh, B.Q., Wong, T.K.F., von Haeseler, A., Jermiin, L.S., 2017.

ModelFinder: fast model selection for accurate phylogenetic estimates. Nat. Methods14, 587–589. http://dx.doi.org/10.1038/nmeth.4285.

Kojima, J., 1998. Larvae of social wasps (Insecta: Hymenoptera; Vespidae). Nat. HistoryBull. Ibaraki Univ. 2, 7–277.

Nguyen, L.T.P., Carpenter, J.M., 2016. Review of the Polistes (Polistella) “Stenopolistes”species-group (Hymenoptera: Vespidae: Polistinae) from Vietnam, with description ofa new species and key to species. Zootaxa 4088 (4), 583–593.

Nguyen, L.T.P., Kojima, J., 2014. Distribution and nests of paper wasps of Polistes(Polistella) in northeastern Vietnam, with description of a new species (Hymenoptera,Vespidae, Polistinae). ZooKeys 368, 45–63. http://dx.doi.org/10.3897/zookeys.368.6426.

Nguyen, L.T.P., Kojima, J., Saito, F., 2011. Polistes (Polistella) wasps (Hymenoptera:Vespidae: Polistinae) from mountainous areas of northern Vietnam, with descriptionof five news pecies. Zootaxa 3060, 1–30.

Nguyen, L.-T., Schmidt, H.A., von Haeseler, A., Minh, B.Q., 2015. IQ-TREE: a fast andeffective stochastic algorithm for estimating maximum likelihood phylogenies. Mol.Biol. Evol. 32, 268–274. http://dx.doi.org/10.1093/molbev/msu300.

Nguyen, L.T.P., Vu, T.T., John, L.X.Q., Carpenter, J.M., 2017. Taxonomic notes on thePolistes stigma group (Hymenoptera, Vespidae: Polistinae) from continental SoutheastAsia, with descriptions of three new species and a key to species. Raffles Bull. Zool.65, 269–279.

Pardi, L., 1996. Polistes: analysis of a society. In: Turillazzi, West-Eberhard, M.J. (Eds.),Natural History and Evolotion of Paper-Wasps. Oxford University Press, Oxford, NewYork, Tokyo, pp. 1–17 (xiv+400pp).

Petersen, B., 1987. Subspecies of the indo-Australian Polistes stigma (Fabricius)(Hymenoptera: Vespidae). Entomol. Scand. 18, 227–259.

Pickett, K.M., Carpenter, J.M., 2010. Simultaneous Analysis and the Origin of Eusocialityin the Vespidae (Insecta: Hymenoptera). Arthr. Syst. Phyl. 68 (1), 3–33.

Pickett, K.M., Wenzel, J.W., 2004. Phylogenetic analysis of the New World Polistes(Hymenoptera: Vespidae: Polistinae) using morphology and molecules. J. KansasEntomol. Soc. 77, 742–760.

Pickett, K.M., Carpenter, J.M., Wheeler, W.C., 2006. Systematics of Polistes(Hymenoptera: Vespidae), with a phylogenetic consideration of Hamilton's haplodi-ploidy hypothesis. Ann. Zool. Fenn. 43, 390–406.

Saito, F., Kojima, J., Nguyen, L.T.P., Kanuka, M., 2007. Polistes formosanus Sonan, 1927(Hymenoptera: Vespidae), a good species supported by both morphological andmolecular phylogenetic analyses, and a key social wasp in understanding the his-torical biogeography of the Nansei islands. Zool. Sci. 24 (9), 927–939. http://dx.doi.org/10.2108/zsj.24.927.

Santos, B.F., Payne, A., Pickett, K.M., Carpenter, J.M., 2015. Phylogeny and historicalbiogeography of the paper wasp genus Polistes (Hymenoptera: Vespidae): implica-tions for the overwintering hypothesis of social evolution. Cladistics 31, 535–549.

Tamura, K., Stecher, G., Peterson, D., Filipski, A., Kumar, S., 2013. MEGA6: molecularevolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30, 2725–2729.

Turillazzi, S., West-Eberhard, M.J., 1996. Natural History and Evolution of Paper-Wasps.Oxford University Press.

Wenzel, J.W., 1993. Application of the biogeneric law to hehavioral ontogeny: a testusing nest architecture in paper wasps. J. Evol. Biol. 6, 229–247.

L.T.P. Nguyen et al. Journal of Asia-Pacific Entomology 21 (2018) 638–644

644