The terminology used in this study follows that described by Archer (1987, 1999).

The images were captured using a Leica DFC 495 camera mounted on a Leica M205A stereozoom microscope (Leica Microsystems, Solms, Germany) and acquired by using LAS v.4.1.0 (Leica Microsystems, Switzerland). In addition, to observe the pronotal rugae of D. kuami and D. flora, scanning electron microscopy (SEM) images of specimens selected from each species were obtained. Subsequently, the head, metasoma, wings and legs were removed and only the mesosoma was retained. The separated mesosomes were immersed in 10% NaOCl to remove excess tissue. The detached mesosoma was washed with distilled water using a soft brush to remove the remaining tissues. After cleaning, the samples were dried and coated with gold ion particles for SEM observation (Hitachi SU8220 & SU8230, Tokyo, Japan). Image plates were prepared using Adobe Photoshop CS6 (Adobe Systems Inc., San Jose, CA, USA).

Genomic DNA was extracted from the hind legs of dried or 100% alcohol-preserved specimens using the DNeasy Blood and Tissue Kit (Qiagen) after incubating for 24 h at 56 °C in lysis buffer and Proteinase K, as per the manufacturer’s instructions. A total of 26 specimens of D. kuami, eight specimens of D. media, and two specimens of D. flora were sequenced. The cytochrome c oxidase I barcoding region was amplified using the primer pair LepF1 and LepR1 (Hebert et al. 2004). PCR was performed using Solg^TM 2X Taq PCR Pre-Mix (Solgent). We prepared 30 μL of reaction mixture containing 15 μL of PCR Pre-Mix, 11 μL of nuclease-free dH₂O, 2 μL DNA template, and 1 μL of 10 pmol of each primer. A typical PCR program started with 4 min of initial denaturation at 94 °C, followed by 40 cycles of 30 s at 94 °C, 45 s of annealing at 45 °C, and 45 s of elongation at 72 °C, ending with a 6-min period of final elongation at 72 °C.

Sequence reads were edited and assembled using Geneious 11 (Biomatters, Auckland, New Zealand). The optimal model (GTR+I+G) for each partition was selected using the Akaike information criterion in jModelTest (Posada 2008). The tree was constructed using the maximum-likelihood method with RAxML v.8.1.2 (Stamatakis 2014), with 1000 bootstrap replications. Successful sequences were submitted to the NCBI GenBank (accession numbers presented in Table 1).

In addition, for phylogenetic analysis, other Dolichovespula species and outgroups, excluding the D. maculata group (D. flora, D. maculata, D. media and D. kuami), included were as follows: KU874880 (Dolichovespula norwegica), KM568773 (D. alpicola), MG374965 (D. norvegicoides), KJ147234 (D. saxonica), KJ147233 (D. pacifica), KJ147230 (D. arenarina), KM567260 (D. adulterina), KJ147235 (D. sylvestris), KJ147244 (Vespa crabro) and LT977378 (Formica fusca) (Table 1). The latter two species, from Vespidae and Formicidae, respectively, were included to test the monophyly of the family and root the tree, respectively.

Thirty-five specimens were used in this study to review the taxonomic positions of the focus species, D. kuami (25) from the Korea National Arboretum (Pocheon, Republic of Korea) and D. flora (2) from Northwest University (Xi’an, China), and D. media (8).

The pronotum rugae of D. kuami were generally very dull and faint (Fig. 1A), whereas those of D. flora were relatively more distinct (Fig. 1D). We found that D. kuami has faint longitudinal rugae on the pronotal lateral face next to the pronotal pit (Fig. 1B), and the remaining posterior area has fine rugae running vertically (Fig. 1C). On the other hand, D. flora has distinct longitudinal rugae (Fig. 1E) that run downward, except in the upper pronotal area (Fig. 1F). Thus, there is a clear difference between the pronotal rugae of these two species. Additionally, the pronotum carina was sharper in D. kuami than in D. flora.

Figure 1. 

Comparison between the pronotal rugae and carinas of D. kuami and D. flora. Pronotum of D. kuami (A) and D. flora (D); rugae and carinas in the pronotal lateral part to the pronotal pit of D. kuami (B) and D. flora (E); rugae on the pronotum of D. kuami (C) and D. flora (F). Arrows in figures A, D indicate the enlarged parts in B, C, E, F. Scale bars: 0.5 mm.

Genitalia (Fig. 2A–F). The external features of the genitalia of D. kuami and D. flora are very similar (Fig. 2A, D; also refer to Tan et al. 2014 for D. flora). However, D. kuami had a triangular parameral spine (Fig. 2C, arrow), whereas that of D. flora is short and slender (Fig. 2F, arrow). In addition, D. kuami has a truncated dorsal terminal process (Fig. 2A, arrow), whereas it is somewhat edged or shortly rounded in D. flora (Fig. 2D, arrow).

Figure 2. 

Genitalia of Dolichovespula kuami (A–C) and D. flora (D–F) A genitalic capsules, in dorsal view (truncated dorsal terminal process, arrow) B aedeagus C gonostipes and triangular parameral spine (arrow) D genitalic capsules, in dorsal view (shortly rounded terminal process, arrow) E aedeagal tip F gonostipes and slender parameral spine (arrow). Scale bars: 1 mm.

Phylogenetically, D. maculata clearly clustered with other Dolichovespula spp. (Fig. 3). In the D. maculata group, D. flora was more closely related to D. maculata and D. media, whereas D. kuami clustered as a sister species to the clade that included these three species (Fig. 3). Thus, the two species D. flora and D. kuami stat. rev. were clearly separated into well-supported clusters, and the genetic distance between them was relatively high (average DNA barcode distance: 0.0996), suggesting that they represent two biological species. These results were further supported by those of the morphological examination.

Figure 3. 

A maximum-likelihood phylogenetic tree of the successfully DNA barcoded Dolichovespula specimens. The numbers above the branches indicate bootstrap proportions.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This study was supported by the Korea National Arboretum [project no. KNA1-1-20, 16-1].

Conceptualization, MB Choi and JK Kim; data curation, CJ Kim, MB Choi and JL Tan; formal analysis, CJ Kim and MB Choi; funding acquisition, CJ Kim; investigation, MB Choi, JK Kim and JL Tan; methodology, CJ Kim and JK Kim; project administration, CJ Kim and MB Choi; supervision, MB Choi and JK Kim; writing–original draft, MB Choi, JK Kim and CJ Kim; writing–review and editing, JK Kim and JL Tan. All authors have read and agreed to the published version of the manuscript.

Chang-Jun Kim https://orcid.org/0000-0002-5823-8703

Jeong Kyu Kim https://orcid.org/0000-0003-2175-5798

All of the data that support the findings of this study are available in the main text.