Revision of the West Palaearctic Polistes Latreille, with the descriptions of two species – an integrative approach using morphology and DNA barcodes (Hymenoptera, Vespidae)

Abstract The genus Polistes is revised for the West Palaearctic region based on morphology and DNA barcodes. The revision includes all known West Palaearctic species, raising the number of species in Europe to 14 and to 17 for the West Palaearctic realm. DNA barcodes were recovered from 15 species, 14 of which belong to the subgenus Polistes, and one, P. wattii, to the subgenus Gyrostoma. An integrative taxonomic approach combining morphology and molecular data (DNA barcoding) was employed to resolve longstanding taxonomic problems in this group. Two species, P. austroccidentalis van Achterberg & Neumeyer, sp. n. (= P. semenowi auctt.) from W and SW Europe and P. maroccanus Schmid-Egger, sp. n. from Morocco are described as new. Polistes bucharensis Erichson, 1849, and P. foederatus Kohl, 1898, were restored from synonymy. The following new synonyms are proposed: P. sulcifer Zimmermann, 1930, and Pseudopolistes sulcifer var. similator Zirngiebl, 1955, under P. semenowi Morawitz, 1889, syn. n.; Polistes iranus Guiglia, 1976, Polistes gallica var. ornata Weyrauch, 1938 and Polistes gallicus muchei Gusenleitner, 1976, under P. bucharensis Erichson, 1849, syn. n.; Polistes omissus var. ordubadensis Zirngiebl, 1955, and P. hellenicus Arens, 2011, under Polistes mongolicus du Buysson, 1911, syn. n. An illustrated key includes all species and additionally three species from the subgenera Aphanilopterus Meunier, 1888 and Gyrostoma Kirby, 1828 (including a Nearctic species recently introduced to Spain and two species occurring in Egypt, the Arabian Peninsula, and SW Asia). A phylogenetic analysis using Bayesian inference provides insights into phylogenetic relationships within the genus Polistes.


Introduction
The paper wasp genus Polistes Latreille, 1802, is an important model group for behavioural and evolutionary studies (Tibbetts 2007, Hughes et al. 1993, Jandt et al. 2014. It includes many eusocial species that exhibit various forms of social organization. Moreover, the comparatively small colony size of Polistes species and their exposed nests facilitate both field observations and experiments (e.g., Cervo et al. 2008). Currently, over 220 species are recognized worldwide (Arens 2011, Buck et al. 2012, Nugroho et al. 2012, 11 of which occur in Europe (Neumeyer et al. 2014, Castro et al. 2013. Three of them, P. atrimandibularis, P. austroccidentalis sp. n. (= P. semenowi auctt.) and P. semenowi (= P. sulcifer Zimmermann, 1930), are social parasites (Cervo 2006, and references therein) and were formerly placed in a separate genus (or subgenus) Sulcopolistes Blüthgen, 1938(Blüthgen 1961, Guiglia 1972 until Carpenter (1990) synonymized Sulcopolistes with Polistes. Subsequently, a phylogenetic analysis showed that the three socially parasitic species form the sister clade to a clade consisting of P. dominula (Christ, 1791) and P. nimpha (Christ, 1791) and that this clade is nested within the European Polistes species (Choudhary et al. 1994). Blüthgen (1943) proposed the subgeneric name Leptopolistes for several nonparasitic European species, including the type species P. associus (Kohl, 1898) (Table 2). Males of these taxa share immediately narrowing temples (genae) in dorsal view (Blüthgen 1943, Guiglia 1972, giving the male head a characteristic shape. Currently, all European Polistes species are assigned to the subgenus Polistes (Carpenter 1996), although the species formerly included in Leptopolistes were still considered to be closely related (Carpenter 1997). In the present study, we are using "P. gallicus species group" to refer to the former subgenus Leptopolistes, and "P. dominula species group" for the remaining species, albeit with some changes from the traditional view (see Table 2).
The taxonomy of the P. gallicus species group has been notoriously difficult (see, for example Guiglia 1972, Arens 2011, Neumeyer et al. 2014. In the present study, an attempt was made to clarify the taxonomic status of all Polistes species of the West Palaearctic region, including a list of all available names of the genus, and to provide a key to species. Also, new names are proposed for two social parasitic species from NW Africa and SW Europe, one of them (P. maroccanus Schmid-Egger, sp. n.) new to science. Two other species (P. bucharensis Erichson, 1849, andP. foederatus Kohl, 1898) were restored from synonymy, thus raising the number of valid species in Europe to 14, and to 17 for the West Palaearctic region. Additionally, three species from other subgenera (an American species recently introduced into Spain and two southern species occurring in Egypt, Arabian Peninsula, and SW Asia) are included in the key.
The present study employs the concept of integrative taxonomy (Schlick-Steiner et al. 2010) and compares results from morphological examinations with results from DNA barcoding (see Schmidt et al. 2015 for further details). The morphological data are supplemented with published data (Neumeyer et al. 2014(Neumeyer et al. , 2015 and unpublished sources, the latter kindly provided from Aleksandar Ćetković (Belgrade) to Kees van Achterberg.

Sampling
Specimens for DNA barcoding are primarily deposited in the collections of the Zoologische Staatssammlung München and the private collection of CSE. 264 specimens representing all West Palaearctic species of the subgenus Polistes and the subgenus Gyrostoma were processed, the latter being represented by the single species P. wattii Cameron, 1900. For DNA extraction, a single leg was removed from each specimen (for further details see Schmidt et al. 2015).
For the present study, a large number of specimens from several collections were morphologically examined. The taxonomic treatment of species is primarily based on the combined analysis of morphological and molecular data, and only those specimens that were analysed both, morphologically and genetically, are listed in the Suppl. material 1.

DNA extraction and PCR
DNA extraction, PCR amplification, and sequencing were conducted at the Canadian Centre for DNA Barcoding (CCDB) using standardised high-throughput protocols (Ivanova et al. 2006, de Waard et al. 2008, available online under www.ccdb. ca/resources. The 658bp target region, starting from the 5' end of the mitochondrial cytochrome oxidase c (COI) gene, includes the DNA barcode region of the animal kingdom (Hebert et al. 2003). The DNA extracts are stored at the CCDB with aliquots being deposited at the "DNA Storage" facility at the ZSM (see www.zsm.mwn. de/einrichtungen/dna-storage/?lang=en). Specimens that were successfully sequenced are listed under Supporting Information, with sequence lengths and the number of unresolved bases. Detailed specimen and sequence data are accessible in BOLD as a single citable dataset (dx.doi.org/10.5883/DS-WPPOLIST). The sequences are also available on GenBank (for accession numbers see Suppl. material 1).

Molecular analyses
Sequence divergence statistics were calculated using the Kimura two-parameter model of sequence evolution (Kimura 1980). Sequences shorter than 500bp were excluded from the distance calculations and the phylogenetic analysis. The BIN is assigned by the BOLD system and represents a globally unique identifier for a cluster of sequences that has shown to correspond closely to a biological species (Ratnasingham & Hebert 2013), including Hymenoptera (Schmidt et al. 2015(Schmidt et al. , 2016 and other insects (Hausmann et al. 2011, Raupach et al. 2014, Hendrich et al. 2015, Mutanen et al. 2016. Genetic distances and summary indices were calculated using analytical tools in BOLD and are given as mean and maximum pairwise distances for intraspecific variation, and as minimum pairwise distances for interspecific variation (see Table 1).

Phylogenetic analysis
The COI sequence data were submitted to a phylogenetic analysis using MrBayes version 3.2 (Ronquist and Huelsenbeck 2003) after alignment using the BOLD Aligner. Duplicates of identical sequences were removed from the dataset so that each haplotype was represented by a single sequence. The analysis employed separate, unlinked substitution models for codon positions of the COI sequence fragment. Two independent runs with one cold and three heated chains were run for five million generations after which the average standard deviation of split frequencies reached values lower than 0.01. The resulting tree was rooted using two basal outgroup taxa (Peters et al. 2017), including one representative of the Eumeninae (Eumenes pedunculatus) and one species of Masarinae (Celonites abbreviatus (Villers)). Transverse yellow band of pronotum narrower medio-laterally than dorso-laterally near junction with oblique yellow stripes of pronotum (a), rarely narrowed dorso-laterally and slightly widened below it and then mesoscutum usually without yellow markings (d); scapus more or less dark brown or blackish basoventrally (b), apical half of hypopygium often entirely black or dark brown (c), sometimes with small yellow spot, rarely apical quarter of hypopygium yellow; mesoscutum usually without paired yellow spots ( -Transverse yellow band of pronotum as wide medio-laterally as dorsolaterally or wider (aa; rarely narrowed medio-laterally and slightly widened below it) and mesoscutum usually with two yellow spots (dd); scapus entirely yellow baso-ventrally (bb), rarely slightly infuscate; colour of apical half of hypopygium variable, often entirely yellow or apically brownish (cc), rarely entirely black; mesoscutum with paired yellow spots usually medium-sized to large (dd), rarely minute or absent; apical half of antenna evenly orange or yellow dorsally ( -Dorsal length of apical antennal segment 1.5-2.1 times as long as its maximal width (aa), 1.0-1.2 times as long as length of fifth antennal segment (bb); clypeus not or slightly depressed medially (cc) and lateral ridges absent dorsally or slightly developed (dd); face flat medially, without longitudinal depression (ee); antenna coloured as in ff or fff; short carina between anten-nal sockets obtuse dorsally and more or less infuscate (gg); width of clypeus 1. -Mesosternum at least with some yellow spots posteriorly (aa), and often entirely yellow, but in some males of P. dominula from southern Greece entirely black; black baso-dorsal stripe on third antennal segment at least 0.5 times length of segment (bb); clypeus laterally nearly flat and ridge between antennal socket and clypeus indistinct ( Most of clypeus with long bristles and distinct punctures (a); apical half of antenna uniformly orange-yellow dorsally and ventrally (b); apical antennal segment comparatively stout and dorsally about 1.5 times as long as wide basally (c); setae of pronotum medio-dorsally and of mesoscutum about half as long as width of posterior ocellus (d); width of clypeus 1.1 times its length (e) and medio-ventrally more flattened (f); [fourth and fifth antennal segments strongly oblique in lateral view]; Europe (for specimens from Crete see also P. bucharensis), NW Africa, W and C Turkey, Azerbaijan, probably more eastern in temperate Asia .... P. dominula (Christ, 1791) -Dorsal 0.7 of clypeus without bristles and punctures (aa); apical half of antenna usually darker dorsally than ventrally (bb); apical antennal segment more slender, dorsally about twice as long as wide basally (cc); setae of pronotum mediodorsally and of mesoscutum at least 0.8 times as long as width of posterior ocellus (dd) or longer; width of clypeus about equal to its length (ee) and medio-ventrally less flattened ( Blüthgen, 1938Blüthgen, (1937 Remarks. Polistes albellus was confused with P. bischoffi before Neumeyer et al. (2014) clarified the status of the latter by designating a neotype. He re-described the former under the new name P. helveticus. Later, Neumeyer et al. (2015) synonymised P. helveticus with P. albellus, a species described from Mongolia. Genetic results confirmed the conspecificity of the European populations with the Central Asian specimens (for detailed descriptions and discussion see Neumeyer et al. 2014Neumeyer et al. , 2015.
The species typically occurs on humid meadows or along lake shores or in fens with a large reed zone, but it also colonizes dry habitats. In contrast to P. biglumis, P. albellus does not occur at higher altitudes of the Alps. Both species may sometimes occur sympatrically in lowland habitats.
the females of the two species is problematic and they are most easily separated by the lack of an epicnemial ridge in combination with an even transition in sculpture from the coarser sculptured mesopleuron to the finer sculptured epicnemium in P. albellus. Polistes biglumis is characterized by having an epicnemial ridge, with a sudden change from the rather coarse sculpture of mesopleuron to the finer sculpture of the epicnemium. In addition, the mesoscutal setae are shorter in P. albellus than in P. biglumis.
The male is unique within the gallicus-group by the combination of narrow temples (genae) in dorsal view and the antenna, with nearly black or dark brown dorsal surface. Females from Central Asia are darker than European individuals (clypeus may be all black with small basal pale spot), and yellow markings are largely replaced by white or ivory.
Diagnosis. The recognition of P. associus females may be problematic because of their similarity to P. nimpha, in particular specimens from SW Asia. Females can be separated by colour differences only (see key to species), although in western Asia P. nimpha often exhibits high levels of colour variation. The male is unique by the combination of narrow temples (genae) in dorsal view and a markedly depressed clypeus with distinct lateral ridges. The dorsal length of the apical antennal segment is about 3.0 times its maximum width, and longer than in similar species.
Distribution. Southern Europe and Turkey, northwards to Switzerland, southwards to Israel, eastwards to Azerbaijan. Guiglia (1972) also mentions India (Jammu and Kashmir) and China, but these records may refer to the similar species P. chinensis (Fabricius, 1793).
Genetic results. We regard P. associus as a member of the P. dominula group instead of the P. gallicus group due to the results of genetic data (see discussion below for details). Specimens from Croatia and northern Italy exhibited no intraspecific variation (Table 1).    Guiglia (1972), new combination.

Polistes atrimandibularis Zimmermann
Diagnosis. The social parasitic species can be recognized by the shape of the mandibular impression and by the colour pattern of the clypeus. Females of the species group show differences in the depth of the medial impression of the mandible and the size and shape of the upper ridge. The weakest medial impression occurs in P. maroccanus, whose upper ridge is only weakly developed or even lacking in the paratype. It is followed by P. atrimandibularis with a shallow impression, and flat, but visible upper ridges. The medial impression is deep in P. austroccidentalis, with large but not modified upper ridges, whereas in P. semenowi the impression is very deep with large and narrow upper ridges. Additionally, the lower ridge is modified with a triangular margin in dorsal view. The size of the black clypeal spot is variable but as a general rule it is largest in P. austroccidentalis (only upper third of clypeus yellow), medium-sized in P. semenowi (lower third black only), small and isolated in P. atrimandibularis, and even smaller (isolated) or completely lacking in P. maroccanus.
The shape of the mandibular impression in males generally follows that of females, but is in general less developed. In addition, the male of P. atrimandibularis has the mandible and parts of clypeus black (almost entirely so in some specimens, except lateral margins), whereas mandible and clypeus are yellow in the remaining species. The male of P. maroccanus is unknown.
Distribution. Southern C and S Europe, northwards to S Germany, Turkey, Iran, Armenia (Guiglia, 1972). Records from NW Africa probably refer to P. maroccanus sp. n. A male from the MFNB from "Ägypten [Egypt], Ehrenberg [leg.]" is probably mislabelled because Egypt is far outside the known range of the species.
Genetic data. Not enough specimens were sequenced to detect any genetic variation. Remarks. Konstantin Samartsev (ZISP) kindly provided photos of the female lectotype of P. semenowi Morawitz, 1889, from Copet-dag. Their examination showed unambiguously that the specimen is conspecific with P. sulcifer (Zimmermann, 1930). This requires the species P. semenowi of authors to be described as a new species: P. austroccidentalis van Achterberg & Neumeyer, sp. n.

Polistes austroccidentalis van
Diagnosis. Large and relatively bright species with robust mandible and wide, yellow malar space (Figs 29,31,35,36), entirely yellow flagellum, black mesosternum and change in sculpture between mesepisternum and epicnemium rather abrupt (epicnemial ridge distinct) in both sexes. Outer face of mandible with distinct depression between a wide dorsal ridge and a much narrower ventral one; dorsal ridge of mandible convex and distinctly elevated above middle of mandible. Female clypeus mainly black, basally yellow (Fig. 29); mesoscutum with pair of yellow spots (Figs 26, 33); hind coxa black; yellow area along inner eye margin usually connected with yellow bar above antennal sockets (Fig. 29); hypopygium with yellow tip; basal half of mandible gradually curved in dorsal view and convex; clypeus abruptly depressed ventrally. Male with mandible mainly yellow, except for its more or less darkened margins (Figs 44,45); mandibular depression rather short and occupying less than half of outer face of mandible, and dorsal ridge wide; clypeus medially entirely yellow, face and frons yellow (Fig. 42); temple (or gena) in dorsal view convex (Fig. 43). See also comments of P. semenowi for recognising the species.
Description. FEMALE. Holotype, body length 15.8 mm; fore wing length 11.6 mm. For colour pattern, see figures.  Head. Mandible very stout and 1.5 times as long as wide (Figs 35, 36) and with a large depression on its outer face; basal half of mandible gradually curved in dorsal view and convex; dorsal ridge of mandible wide, smoothly convex without sharp edges and distinctly elevated above depression; ventral lobe of clypeus acute and step-like lowered; fine pubescence of clypeus conspicuous and comparatively long ventrally (Figs 29, 31); malar space 1.8 times POL; ocelli in equilateral triangle (Fig. 30).
Variation. Body length 13.0-17.1 mm; fore wing length 10.1-12.7 mm. Mandible either black (Europe) or partially yellowish (NW Africa), if with a yellowish or brownish area then that area always smaller than yellow area on malar space. Vertex often with pair of tiny yellow spots behind lateral ocelli. Third antennal segment often with tiny basal dark brown spot, about 5% of specimens have the yellow posterior stripes of the pronotum connected to the transverse yellow band. Tergite II with paired anterolateral yellow spots absent (NW Africa) or present (Europe). Hypopygium usually with yellow tip, but sometimes apical half or nearly entirely yellow, or entirely blackish and only dark brown apically.
MALE. Body length 11.6-17.1 mm; fore wing length 9.6-11.7 mm. For colour pattern, see figures. Similar to female, differs as follows: Head. Mandible except for darkened margins yellow (Figs 44, 45). Clypeus yellow, and laterally distinctly convex, sub-antennal depressions continued onto clypeus (Fig. 44). Malar space yellow and 1.4 times as long as POL. Temples in dorsal view convex (Fig. 43). Face and anterior half of frons yellow, at most a tiny black spot on interantennal prominence and frequently also a narrow vertical black dash originating from upper margin of each torulus. Vertex often with pair of small yellow dots behind lateral ocelli, remainder of head black. Antenna brownish yellow, but scapus and pedicellus dorsally as black as a spot on third antennal segment (Figs 37, 42). Apical antennal segment 2.2 times as long as wide (Fig. 41).
Mesosoma. Mesoscutum with paired medium-sized yellow spots, sometimes reduced to tiny dots or absent (Figs 37, 39). Dorsal yellow spot of mesopleuron large but rarely minute, often only apex of femora yellowish.
Metasoma. Pair of spots of tergite I either connected to terminal band or well separated (Fig. 37). Sternites II-VI with continuous yellow terminal bands (Fig. 40), sometimes briefly interrupted on sternite VI. Hypopygium black with brown margin (Fig. 40).
Distribution. Specimens from Algeria, Andorra, France, Italy, Morocco, Portugal, Spain, and Switzerland have been examined, indicating that the species is confined to NW Africa and SW Europe with an extension to Central Europe, and is replaced by Polistes semenowi Morawitz further east. In Switzerland, P. austroccidentalis occurs only in the SW part (Valais and one record in the Jura Mountains), whereas P. semenowi occupies mainly the SE part (Ticino and southern Grison valleys, except for two records from the canton of Valais).
Biology. According to Cervo (2006), P. austroccidentalis is an obligate social parasite, normally of P. dominula, but occasionally also of P. nimpha. Corresponding to its ubiquitous, euryoecious main host (P. dominula), P. austroccidentalis can be found in a wide variety of open and semi-open habitats, but up to now it apparently avoids the northern part of its host's range. The altitudinal records (n= 20) for P. austroccidentalis range from near sea level (Carpentras, France) to 2600 m in Spain (Sierra Nevada, Andalusia) and 2150 m in Morocco. The seasonal records (n = 67) range from March (Vaiamonte, Portugal) to 21 September in Switzerland (Ausserberg, VS) or 28. September in Morocco (males only), but most individuals were observed from May to August, at least in Switzerland (CSCF in litt.). There, the earliest record for males is 12 July (Erschmatt, VS), the latest for females 29 August (Martigny, VS).
Genetic data. Specimens from south-central Europe and Morocco only showed a small genetic distance.
Etymology. The name is a combination of the Latin adjectives "australis" (southern) and "occidentalis" (western), because of its southwestern distribution in Europe.

Remarks. See Neumeyer et al. (2014) for detailed comments about taxonomy.
Diagnosis. The female shares a dark brown to black upper side of antennal segments with P. albellus (see under P. albellus for recognition). The other species of the P. gallicus group have lighter antennal segments (except northern P. foederatus). The male has broad, convex temples in dorsal view and the clypeal disk without impressions or lateral ridges, similar to that of P. dominula and P. bucharensis.
Distribution. Europe including Norway and Sweden south of 65° N to Turkey, Central Asia. Guiglia (1972) also mentioned N Africa, Iran and Mongolia, but these records require confirmation. Polistes biglumis occurs up to 2400 m in the European Alps. The presence in Greece is confirmed by a male from Mt. Olympus, eastern slope, 2200-2500 m a.s.l., 20.ix.1989, leg. T. Osten, coll. CSE), removing doubts expressed by Arens (2011).
Genetic data. Polistes biglumis consists of two closely placed subclusters which share the same BIN (Suppl. material 2: NJ tree, and Fig. 58). Specimens from the Aosta valley in NW Italy form a geographic subcluster that is separated from a second subcluster consisting of specimens from NW Italy and SE Germany.

Polistes bischoffi Weyrauch Figs 6, 7
Polistes bischoffi Weyrauch, 1937 (Weyrauch 1939) he included a dark form of P. bischoffi (actually P. albellus) that he interpreted as geographic variations (followed by Schmid-Egger and Treiber 1989, and others). The true P. bischoffi has a southern European distribution, whereas P. albellus ranges from Central Europe to the eastern Palaearctic region. Diagnosis. This species belongs to the group of species with a short female malar space within the P. gallicus species group. The female can further be recognized by a medially interrupted yellow band on sternite IV, a large black spot or band on the clypeus, and by the dorsally black hind coxa. The sternal band IV is always continuous in P. gallicus and P. mongolicus, and the clypeus is entirely yellow in most P. mongolicus specimens or has a smaller, more excentric spot in most P. gallicus specimens. The epicnemial ridge is reduced or absent in P. bischoffi and P. albellus versus distinct or reduced in P. gallicus and P. mongolicus. The character is therefore of limited diagnostic value. The flagellum is in both sexes dorsally light orange to somewhat dark orange, whereas it is always light orange in related species, except northern P. foederatus. The male can be confused with P. gallicus (see the key to species for differences).
Distribution. S Europe and Turkey, northwards to Austria (Neusiedl am See) and Switzerland (northwards to Zürich). See Neumeyer et al. (2015) for the detailed distribution.
Genetic results. Specimens from Spain, S France and Corsica, Switzerland, and Croatia were barcoded. The species forms a separate cluster, with a small gap between SW (France, Spain) and southern central European populations.  Gusenleitner 1976), or as a synonym of P. dominula (Carpenter 1996). In our opinion P. bucharensis is clearly distinguishable from P. dominula by colour pattern, by shape of the male clypeus and by the genetic data. Specimens from Mongolia and China (= Polistes gallicus pseudopacificus) probably belong to P. bucharensis and not to P. dominula, as indicated by Giordani Soika (1970).

Polistes bucharensis
The species was first described from Uzbekistan and later as Polistes gallica var. ornata and as Polistes iranus from Iran. It was not possible to examine types of these taxa, but their descriptions agree well with examined specimens. Also, we could examine a large series of specimens from the type locality of P. bucharensis from "Bukhara" in MFNB, which also agree with description of P. bucharensis. Therefore, the valid name for this taxon is Polistes bucharensis Erichson, 1849 sp. restit.; and Polistes iranus syn. n. is a new synonym of P. bucharensis.
Another problematic taxon in this group is P. gallicus muchei. It was described by Gusenleitner (1976) from northern Caucasus and eastern Turkey as a subspecies of P. gallicus (now P. dominula) based on reduced pale marks in the male and whitish yellow or ivory instead of yellow ground colour. Six male paratypes and a non-type male from eastern Turkey (Kars, identified by J. Gusenleitner as P. gallicus muchei, in coll. CSE) were examined by CSE. They agree with typical P. bucharensis by having the mesosternum black and the clypeal ridges present. The male from Kars differs by an all-black mesonotum and by a black medial spot on clypeus. Consequently, we treat "muchei" (based on males) as a dark and whitish colour form of P. bucharensis and not as a form of P. dominula (= former P. gallicus). It is probably restricted to mountainous regions of NE Turkey and the Caucasus. This assignment is based on colour and morphology alone and requires verification through sequence data.
According to the description by Gusenleitner (1976), the female paratypes of P. muchei differ markedly from typical P. bucharensis by a reduction of the pale body colour. The mesoscutum is black and the clypeus has a transverse band. We cannot rule out that the type series of P. muchei includes P. nimpha specimens. Two nontype females from Kagisman (eastern Turkey, collected by CSE together with the above-mentioned male) belong to two species: one to P. bucharensis (typical form) and other to P. nimpha (with extreme extended pale colour pattern: clypeus and genae all yellow, however, with the mesoscutum and the hypopygium (sternite VI) all black.
Based on the material that was available for us it appears that two taxa of this lineage occur in eastern Turkey and the Caucasus: P. bucharensis and P. nimpha. Males can be recognized easily by morphological characters (see key to species), whereas the identification of females is hampered by an extraordinary colour variation (extreme pale forms occur together with extreme dark forms). They can be distinguished by characters given in the key (mainly by colour of the hypopygium (sternite VI): yellow in P. bucharensis, mainly black or reddish in P. nimpha). In addition, the sculpture of the lower half of the mesopleuron is coarser in P. bucharensis than in P. nimpha (where it is finer, with overlap in a few specimens). Results of DNA barcoding of females is needed to confirm this hypothesis. Another problem is a white coloured form of P. nimpha in Iraq ("P. nimpha irakensis", see discussion under P. nimpha).
The MFNB houses a large series of females from West Pamir, collected in 1928 by Reinig. This taxon is darker than typical P. bucharensis from Bukhara or Turkey (yellow band on gena of female medially largely interrupted, mesoscutum black, tergites with narrow pale bands, and pale colour whitish instead of yellow). The specimens have the clypeus all whitish yellow, the clypeal disc is punctate with large punctures and has some bristles. They probably also represent a dark high-elevation form of P. bucharensis. The colour variation of Central Asian specimens is not understood well and requires examination of more material.
Three barcoded specimens from Crete were assigned a separate BIN (Tab. 1). Their colour pattern is intermediate between P. bucharensis and P. dominula (see description below) but genetically they are most similar to P. bucharensis from Cyprus. Therefore, we treat them as an island form of P. bucharensis. It is possible that the population from Crete represents a distinct species but further research is required to clarify the taxonomic status of the involved species.
The specimens barcoded fall into four clusters that were assigned three different BINs (Suppl. material 2: NJ tree). The species shows distinct geographic subclustering with specimens originating from Crete, Cyprus, and Azerbaijan. Specimens from Crete differ morphologically distinctly from those of Cyprus (see discussion above), whereas two of the examined specimens from Azerbaijan are similar to the specimens from Cyprus.
A single specimen from Azerbaijan has been assigned a different BIN, whereas another specimen from Azerbaijan that is morphologically similar to the remaining females of P. bucharensis agrees genetically with P. dominula from Central Europe (Suppl. material 2: NJ tree). More specimens from this region need to be examined to be able to assess the morphological variation of each potential species in this group and its status in relation to P. dominula. It is probable that several other genetically distinct taxonomic units of this P. dominula/P. bucharensis species complex occur in this region.
Diagnosis. Polistes bucharensis is similar to P. dominula, and both sexes can be distinguished by the continuous wide yellow band on the temple (gena); seen in lateral view it is more than half as wide as the temple and extends along the entire posterior margin of the eye (specimens from Crete are different, see below). In P. dominula, this band is medially interrupted and less than half as wide as the temple, rarely continuous but then it is constricted medially. In females, the yellow band above the antennae is always connected with the band along the inner eye margin. This band is isolated from the lateral bands in P. dominula and it does not reach the inner eye margin.
Females of P. bucharensis have an entirely yellow clypeus and a somewhat denser pilosity on the clypeus (the pilosity concerns the dark bristles on the clypeal surface). In females of P. dominula the colour of the clypeus is variable: entirely yellow (mainly in specimens from Central Europe) or with a black medial spot in 50-70% of specimens from southern Europe and western Turkey ( fig. 38/39).
Sternite II is predominantly black with narrow apical yellow band in P. bucharensis, the visible part of the remaining sternites is entirely yellow (except in specimens from Cyprus, which have a larger basal part of sternite III black). In P. dominula, the visible base of the sternites III-V is always black and the apical yellow band is 0.5-0.7x as wide as the total visible part of the sternite. Additionally, the species can be recognised by the sculpture of the mesopleuron and the lateral face of the propodeum that is markedly coarser in females of P. bucharensis compared to P. dominula.
Males of P. bucharensis have the mesosternum always all black, whereas the mesosternum of P. dominula males is partly or entirely yellow. The mesosternum has at least two triangular yellow spots subapically (except in specimens from southern Greece, see below). The colour pattern of the sternites is more variable than in females. For distinction from P. nimpha and from species of higher mountains in Central Asia see the key to species and the discussion above.
Variations. Females from Crete (n = 6) differ from typical P. bucharensis by a reduction of the yellow body colour. The yellow band of the temple (gena) in lateral view is medially widely interrupted and the clypeus has nearly always a transverse band or medial spot (Fig. 35), except one female with entirely yellow clypeus. The yellow band above the antennal sockets is isolated from both lateral bands. Sternites III-VI are all yellow and sternite II has only a narrow yellow apical band.
The single male that was examined genetically has the mesosternum all black and the yellow band of the temple is medially interrupted. In specimens from Iran the yellow body colour of the only examined male from Arak is replaced by an extreme whitish yellow. For recognition of specimens from Caucasus and E Turkey see discussion on P. dominula muchei.
Distribution. From Central Turkey to Central Asia, Israel, Iran and Egypt. In Europe only known from Crete and Cyprus. Specimens described from China and Mongolia (not examined) may also belong to P. bucharensis (see Giordani Soika, 1970).

Polistes dominula (Christ) Figs 10-12
Vespa dominula Christ, 1791, Naturgesch. Insect.: 229-232 + Taf. 21, fig. 1 Remarks and genetic data. Polistes dominula and the reinstated closely related species P. bucharensis show high levels of variation in their colour patterns. Previously, both taxa were treated as a single species (Gusenleitner 1976, Carpenter 1996. Our study revealed the presence of six genetic clusters for the two species, all of which were assigned different BINs by the BOLD system, viz. 1) Morocco, 2) SW to Central Europe (subsequently referred to as western cluster), 3) SE to C Europe (subsequently referred to as eastern cluster), 4) Crete, 5) Cyprus to Azerbaijan, and 6) Azerbaijan (Table 1, Suppl. material 2: NJ tree with clusters 1-6 indicated by numbers next to each cluster). In both, the neighbour-joining and the phylogenetic analysis (Fig. 58, see below), Polistes bucharensis is nested within the P. dominula cluster. Two of these clusters (2 and 3) occur in Europe. The current data indicate for cluster 2 a south-western European distribution although we could only examine one specimen from France and records from Spain still are missing. The cluster occurs over whole Germany and it is close to the cluster of Morocco (1). The other European cluster (3) includes specimens from Greece and Azerbaijan and seems to have a more south-eastern distribution. It covers whole Germany with the westernmost records from the Aosta valley in the Italian Alps. This cluster is genetically closer to P. bucharensis than to the P. dominula clusters (1) and (2) from the western area and consequently we treat specimens from Crete as P. bucharensis (see above).
The high intraspecific variation (Table 1) and presence of multiple BINs could be the result of several species being present. However, specimens from NW Africa and Europe are very similar and indistinguishable by morphology or colour pattern. Furthermore, Neumeyer at al. (2014) found no sufficient differences in the ITS1 gene between the European clusters 2 and 3. We therefore refrain from describing each cluster as a new species until further evidence is available.
Polistes bucharensis with its eastern clusters (3-6) is genetically closer to the eastern clusters of P. dominula (1, 2). However, specimens of P. bucharensis are clearly separated by colour pattern and male morphology from P. dominula, supporting the notion that they represent distinct species.
The population of Crete seems to be isolated from the remaining populations for a long period, and the common ancestor probably came from the P. bucharensis lineage.
Because of this result, we follow the concept of morphospecies here and treat P. bucharensis as a valid species, separated from P. dominula s. str. by genetics, morphology, and colour pattern.
Diagnosis. Polistes dominula is the most common Polistes species in Europe. The female is characterised by a mostly or entirely yellow sternite VI in combination with an orange apical half of the antenna. The yellow band on the temple behind the eye in lateral view is interrupted in nearly all examined specimens, except in some from southern Greece and from Tunisia. The clypeal pattern is variable. Populations from Central Europe usually have an all yellow clypeus, most specimens from southern Europe and some from southern Central Europe (e.g. southern Germany) have one or two black medial spots, or a band on the clypeus medially (Figs 38, 39).
The female of P. dominula can be confused with P. associus (see key to species) and with P. bucharensis in western Asia. Polistes bucharensis always has a completely yellow clypeus in combination with a continuous band on the temple behind the eyes. This band is interrupted in most specimens of P. dominula, and the clypeus has often (about two-thirds of examined specimens) black spots or a band in females from southern Europe. For specimens from Crete, see under P. bucharensis.
The male of P. dominula is characterised by the orange apical half of antenna in combination with the lack of any impressions or ridges on the clypeus. It can be distinguished from the similar P. bucharensis by a (partly or entire) yellow mesosternum; the latter is all black in P. bucharensis. Also, the clypeus is laterally somewhat bulging in P. bucharensis, and always without any elevations in P. dominula.
Some specimens of P. dominula (identity confirmed by barcoding) from the Peloponnese (Greece) have some colour similarities of typical P. bucharensis (females with yellow clypeus and with wide yellow band on temple, males with mesosternum entirely black) and resemble P. bucharensis. However, the yellow band on the temple is medially constricted, and the yellow band above the antennal sockets is isolated from both lateral bands. In addition, the mesopleural sculpture is finer than that of typical P. bucharensis. We therefore regard these specimens as pale form of P. dominula at the SE border of its distribution area. The specimens from Greece occur together with typically coloured P. dominula (females: clypeus with large black spot or band, band on temple interrupted; males: mesosternum partly or all yellow). A female from Tunisia (Dougga) agrees in colour pattern with the above-described pale females from Greece and likely belongs to the same whitish form of P. dominula.
Distribution. NW Africa, C and S Europe as far north as Latvia, but missing in Great Britain, Scandinavia, Crete and Cyprus. Introduced to Australia, North America (including Canada, Buck et al. 2008) and South America. The species has also been recorded from central and eastern Palaearctic regions and from India (e.g. Guiglia 1972). These records need confirmation, because they may belong to P. bucharensis.
The easternmost records that we could examine are from western Turkey and from Azerbaijan. However, it can be expected that the species occurs farther east in Russia or Central Asia. One examined female from Egypt is clearly P. bucharensis and P. dominula probably does not occur in NE Africa.
Specimens examined. Europe: Examined from most countries in Central and S Europe. Asia: Turkey (Termessos/Antalya), Azerbaijan. Africa: Morocco, Tunisia. Remarks. The species is widespread in N Italy, the Balkans, and western Asia (Turkey to Caucasus area). The identity of P. foederatus remained unclear for a long time, and in the past the species was usually treated as P. gallicus. Arens (2011) was the first to recognize two species of the P. gallicus group in Greece (P. gallicus and P. hellenicus), using the length of the malar space as a new diagnostic character. However, he interpreted specimens with long malar space as "P. gallicus" but ignored that true P. gallicus from the western Mediterranean usually have a short malar space (with some exceptions). Type examination and genetic analysis clearly show that P. gallicus sensu Arens (2011) from Greece belong to P. foederatus. Diagnosis. Polistes foederatus is unique in the P. gallicus species group by possessing the longest malar space of all species combined with a large and mainly rectangular and central black spot on the clypeus in females. In addition, the dorsal side of the flagellum is often slightly darkened. However, especially females can be confused with P. gallicus in the transition zone of both species (N Italy to Balkans), because the latter rarely has an extreme long malar space (some genetically examined specimens from the Italian Alps). The flagellum is completely reddish in P. gallicus, but always darkened dorsally in alpine P. foederatus. The yellow spots on the mesoscutum are usually lacking in smaller specimens from Croatia and Italy. The male is variable in colour pattern; the mesoscutum is black or has a pair of large yellow spots and the base of tergite II is either black or largely yellow.

Polistes foederatus
Distribution. From NE Italy to Greece and Azerbaijan. Widespread and common in mainland Greece (Arens 2011, as P. gallicus).
Genetic data. Specimens of Polistes foederatus from several countries, including Azerbaijan, where the type locality is situated, were analysed genetically. The species exhibits little genetic variation and all specimens share the same BIN. The specimens from Crete form a distinct subcluster, perhaps because of longer isolation. However, the specimens from Crete are closer to specimens from the European and Asian mainland than Cretan P. bucharensis are from their mainland populations.  Weyrauch (1939: 161).

Polistes gallicus (Linnaeus)
Remarks. The name P. gallicus (sensu lato) was used in the past for three Mediterranean species: P. foederatus, P. mongolicus, and P. gallicus. A reassessment of morphological characters in combination with DNA barcoding shows that P. gallicus (sensu stricto) is a valid species with a mainly western Mediterranean distribution (eastwards to Corfu, but probably very rare on Balkans). Polistula omissa is regarded as a synonym of P. gallicus (sensu stricto). Diagnosis. Polistes gallicus females are characterised by a short malar space (but in a few specimens from Italy as long as in P. foederatus) and by two yellow spots on the mesoscutum; these spots are absent in most females of P. mongolicus. If there is a dark patch on the clypeus, it is small (rounded or forming a transverse band) and situated on the apical half of the clypeus. The posterior pronotal band is variable, short (in most specimens from Portugal) or reaching the anterior pronotal transverse band (in most specimens from N Italy). In the transition zone to P. mongolicus (N Italy, Balkans), P. gallicus can be confused with P. mongolicus when the yellow mesonotal spots are absent (one barcoded female of P. gallicus from Italy, Lombardia, with reduced, minute yellow spots only). Polistes gallicus has the pos-terior stripes connected to the anterior transverse band of the pronotum, whereas it remains separated from the pronotal band in all examined P. mongolicus from Croatia. Males can be recognized by the combination of the short malar space and two yellow spots on the mesoscutum.
Colour variation. All examined females from NW Africa have the clypeus yellow (one female with minute black spot), the mesoscutum has a pair of large yellow spots, and the tergite VI is entirely black (apical half yellow in one specimen). Males from NW Africa were not examined.
Distribution Genetic results. Specimens from Polistes gallicus originating from several European countries between Croatia and Portugal and from Morocco were examined genetically. They exhibit some genetic variation but all specimens share the same BIN. Diagnosis. Polistes maroccanus sp. n. is close to P. atrimandibularis and can be distinguished by the characters given in Table 2.

Polistes maroccanus
Description. FEMALE. Holotype, body length 14 mm; fore wing length 11.5 mm. For colour pattern see figures.
Head. Mandible with a large depression on its outer face; both ridges of mandible narrow, medially 0.15 times as wide as mandibular diameter, remaining space shiny, with a few large punctures on upper third; malar space 1.5 times POL. Clypeus slightly wider than long (length/width ratio 0.92 in holotype, 0.86 in paratype).
Variations. Body length of paratypes similar to holotype. Left mandible of one paratype with small yellow spot, and clypeus all yellow in one paratype.

Distribution.
Only known from the High and Middle Atlas Mountains in Morocco. Previous to this study, CSE identified a female from Quirgane (High Atlas, 22.v.1995, leg. et coll. M. Hauser) as P. atrimandibularis. It is probably referable to P. maroccanus as well but the specimen was not available for re-examination.
Biology. The species is most probably a social parasite. At the type locality, it was collected together with P. dominula that is most probably the host.
Etymology. Polistes maroccanus is named after the country of origin, Morocco. Remarks. DNA barcoding of a specimen from Morocco, formerly identified as P. atrimandibularis, indicated that it belongs to a different species that is close to the previously known social parasites. A detailed morphological examination resulted in some different character states and supports the notion that the Moroccan specimens belong to a new species. The species is morphologically close to P. atrimandibularis and probably replaces it in NW Africa.  Remarks. The species is widespread in SE Europe to C Asia and China. Apart from the original description it was later described as P. omissus var. ordubadensis Zirngiebl from Caucasus and as P. hellenicus from Greece by Arens (2011). Arens (2011) was the first who recognized two different species of the P. gallicus species group in Greece and he described P. hellenicus as new species. He based his description mainly on the short malar space in contrast to P. foederatus with long malar space, and the black venter of the males (yellow in P. foederatus). Morphological comparison, genetic examination of specimens from a wide geografic range and type study confirms the conspecificity of P. hellenicus and P. ordubadensis with P. mongolicus. Our material increases the known range of the species from Croatia to Central Asia and China, and to NE Africa. The examined type specimen of P. mongolicus from China is somewhat darker than western specimens, but agrees in general aspects with our species definition. For taxonomic status of Polistes omissus kaszabi, see Neumeyer et al. (2014). Diagnosis. Within the P. gallicus group the female of P. mongolicus is characterized by a short malar space, the lack of yellow spots on the mesoscutum (present in some females from Greece and western Asia), and usually by a yellow clypeus. Some females mainly from Greece have a very small to a medium-sized transverse spot on the clypeus. See Arens (2011, as P. hellenicus) for discussion of the colour variability. Polistes foederatus has longer malar space (see key to species).

Polistes mongolicus
The recognition of P. mongolicus is not problematic in Greece and farther east, but on the Balkans females may be confused with P. gallicus (see diagnosis of the latter). Males of P. mongolicus occur in two different colour forms. Specimens from Europe usually have the mesosternum entirely black or with a pair of yellowish spots, whereas the mesosternum of males from Asia and Egypt is largely yellow. Recognition of European males is therefore unambiguous.
In N Africa P. mongolicus is restricted to Egypt, whereas P. gallicus occurs in Tunisia, Algeria and Morocco. Specimens from Libya were not examined, but it cannot be ruled out that ranges of both species overlap in this region.
Colour variations. All examined females from Egypt have a yellow clypeus, with at most a minute black medial spot; the hypopygium (sternite VI) is partly yellow or reddish; one of the females has a pair of minute yellow spots on the mesoscutum.

Remarks.
Polistes nimpha is well defined by male morphology, in particular the long apical antennal segment and distinct lateral ridges of the clypeus, and in the female by the colour pattern (European specimens only). In western Asia, the recognition of females is not always easy since the species varies markedly in colour pattern. It can be confused with P. associus (lowlands of Turkey, Israel) and with P. bucharensis (eastern Turkey, Caucasus region, Iraq). In a small geographic area in western Asia, the dark and the pale coloured form occur in close vicinity, but probably not sympatric. Especially specimens from Iraq have an extended yellow colour pattern and can be confused with P. bucharensis. They can be recognised by the colour of the hypopygium (=sternite VI), but identification of some females remains difficult. Differences in the ocellar angle (more obtuse in P. nimpha/dominula than in P. associus), as stated by Arens (2011), cannot be confirmed here. The sculpture of the lower half of the mesopleuron is somewhat coarser in P. nimpha than in P. associus, although both species overlap in this character.
Diagnosis. The most important diagnostic character of P. nimpha females is the shape of the transverse pronotal band in that it is narrow and pointed ventrally. The lateral portion of the transverse band (seen in lateral view) is wider in front of the pronotal carina than behind it. In the remaining non-parasitic species of the P. dominula species group (P. associus, P. dominula, and P. bucharensis) the portion of the yellow band behind the carina is always wider. However, some extremely xanthic females of P. nimpha from western Asia also possess a very wide pronotal band. About 70% of females from western Asia have paired yellow drop-shaped spots on the mesoscutum. These spots are usually absent in European specimens. The visible part of the hypopygium (sternite VI) is usually black or partly reddish in P. nimpha and also in P. associus, rarely with a yellowish apical spot, while the hypopygium is entirely or predominantly yellow in P. dominula and P. bucharensis.
The latter character is used here for recognition of P. nimpha and P. bucharensis in eastern Turkey. This character is helpful in distinguishing xanthic P. nimpha females (i.e., with an all-yellow clypeus and temple), which are otherwise similar to P. bucharensis. Often only a combination of a several characters will ensure a correct identification of western Asian specimens.
The separation of P. associus and P. nimpha females can also be difficult, especially in areas where both species occur sympatrically (e.g. in western Croatia). The colour pattern of P. associus is diagnostic and exhibits little variation (based on specimens identified by barcoding): Transverse pronotal band wide laterally, separated from posterior band by 2-3 times the diameter of the anterior ocellus; mesoscutum with two large drop-shaped yellow spots. Despite significant variation, western Asian P. nimpha never show this combination of characters. The hypopygium colour is variable in both species but never all red in P. nimpha as it sometimes is in P. associus (one female from Israel).
Distribution. Europe, north to S Finland, Palaearctic Asia east to Mongolia, China, and Russian Far East.
Genetic results. Only specimens from Central Europe were examined genetically, except for one specimen from Greece. The species shows significant intraspecific genetic variation. It is possible that the examination of Asian species will yield unexpected results.  Guiglia 1972).

Remarks.
The species was formerly treated as P. sulcifer by authors. See also comments under P. austroccidentalis for the nomenclature and taxonomy of this species.
Diagnosis. The species can only be recognized based on the shape of the mandible, clypeus, and the colour of male fore and mid coxae. The upper ridge of the mandible is markedly modified in the female and forms a triangle in dorsal view (weaker and more rounded in male). The recognition of males is more difficult because the upper ridge is sometimes only weakly curved and resembles that of P. austroccidentalis. The mandibles (frontal view) differ between both species in that the mandibular depression is narrower in P. austroccidentalis and with a wider upper mandibular ridge. Furthermore, the male fore coxa is almost always and the mid coxa usually marked with yellow as opposed to P. austroccidentalis where all coxae are black.
Life history. Polistes semenowi is a social parasite of P. dominula (see Cervo 2006, as P. sulcifer), in western Asia probably also of P. bucharensis.
Distribution. S and C Europe, north to Germany (Hesse, one record from 1908; Tischendorf et al. 2015 as P. sulcifer), east to Central Asia, not recorded from Spain and Portugal. One female in coll. MFNB labelled "Egypt, Ehrenberg leg"] has more yellow coloured hind coxa than usual. Its origin is doubtful as is that of an atrimandibularis specimen with the same data (see discussion under that species). No other specimens from North Africa have been examined by the authors. Males are sometimes found at higher altitudes (e.g., five males from Italy, Dolomiti, Rif. Coldai at 2150 m; 2 males from Greece, Mt. Olympus at 2200-2500 m, in coll. CSE).
Genetic data. Barcoded specimens from south-central Europe showed little genetic variation.

Other subgenera
The following species are only discussed briefly, because they were either introduced only recently or they occur near the southern border of the study area. For further information on the nomenclature of these species see Carpenter (1996). For identification of the

Results of phylogenetic analysis and discussion
The phylogenetic analysis based on Bayesian inference resulted in a split of the included Polistes species into two major clades, with a posterior probability support value of 0.84 for the P. gallicus group (including P. foederatus, P. bischoffi, P. gallicus, P. biglumis, P. albellus, and P. mongolicus), and with a branch support of 1.0 for the P. dominula species group (including P. dominula, P. bucharensis, P. austroccidentalis, P. atrimandibularis, P. semenowi, P. maroccanus, P. nimpha, and P. associus) (Fig. 58). Polistes dominula is represented by two clades with P. bucharensis in between, illustrating the need for a closer examination of the status of the disjunct dominula clades. The sister group of the dominula/bucharensis clade is composed of a well-supported (pp = 1.0) clade consisting of the four social parasite species P. atrimandibularis, P. austroccidentalis, P. maroccanus, and P. semenowi and (Fig. 58), supporting the notion of a single origin of their biology. Within the P. gallicus species group there is some support for a clade composed of P. foederatus, bischoffi, gallicus, and biglumis (Fig. 58).
These results concur with an earlier analysis of a mitochondrial gene fragment (16S ribosomal RNA) by Choudhary et al. (1994). In their analysis, the three examined social parasites form a monophyletic group that is nested within other European Polistes. Unlike their study, where the parasitic clade came out as the sister group to dominula+nimpha, our analysis yielded the dominula/bucharensis clade as the sister group, albeit with low branch support (0.63, Fig. 58). Polistes nimpha, on the other hand, came out as the sister group to the clade comprising of P. dominula/bucharensis and the social parasites. This sister group relationship is supported by a robust branch support of 1.0 (Fig. 58). Based on the results of the present analysis, the following species groups are proposed within the subgenus Polistes.
The most important change affected the two species P. associus and P. biglumis. Polistes associus was formerly treated as a member of the P. gallicus species group (= subgenus Leptopolistes sensu Guiglia 1972) because of the narrowed temples of the male -a typical Table 3. Species and species groups of Polistes based on the present study and compared to Guiglia (1972   character of this subgenus, according to former authors. However, females of P. associus are morphologically and genetically close to P. dominula/P. bucharensis and P. nimpha, and can easily be confused with the latter species, providing support for placing the species in the dominula species group. Likewise, P. biglumis was traditionally treated as a member of the P. dominula species group (= subgenus Polistes sensu Guiglia 1972), based on the broad, convex temples of the males. The female, however, is very similar to P. albellus, and the species has close relationships to the P. gallicus species group (= subgenus Leptopolistes sensu Guiglia 1972).
Despite the comparatively limited informative value of a single mitochondrial gene region like the 658bp COI barcode fragment, it still provides insights into phylogenetic relationships within the group and the phylogenetic relationships resulting from the analysis are largely in agreement with inferred by other lines of evidence. A more comprehensive analysis of Polistes including additional genetic markers, in particular of nuclear genes (see Neumeyer et. al. 2014), should be employed for evaluating and scrutinising the results of the present study.