Gryon ancinla Kozlov & Lê (Hymenoptera: Scelionidae): host association, expanded distribution, redescription and a new synonymy

Abstract Background Gryon Haliday (Platygastroidea: Scelionidae) is a cosmopolitan genus of egg-parasitoid wasps primarily associated with Heteroptera. New information Gryon ancinla Kozlov & Lê is reported for the first time outside of Vietnam, in China and Cambodia, and as an egg parasitoid of the pestiferous leaf-footed bug, Acanthocoris scaber (L.). Gryon ancinla is redescribed based on recently collected specimens and compared to closely related species of Gryon in the region. Gryon clavaerus Kozlov & Lê is treated as a junior synonym and some characters found in the charon species group are discussed.


Introduction
Acanthocoris scaber (L.) (Hemiptera: Coreidae) is a pest that feeds on many economically important vegetables including chili, potato, tomato, and eggplant in China (Du et al. 2014), where it has become a serious problem for chili production in recent years. Adults and nymphs usually aggregate on leaves, stems and fruits (Fig. 1). They feed by sucking plant fluids that contain sugars and other nutrients, resulting in production losses. A recent survey conducted in China to assess parasitism of A. scaber eggs yielded numerous specimens of Gryon Haliday (Scelionidae), a cosmopolitan genus of parasitoid wasps whose species primarily parasitize the eggs of Heteroptera (Johnson et al. 1996, Masner 1983).
The taxonomy of Gryon in southeast Asia is in need of thorough revision, as in much of the world. The conspicuous characters of the charon species group, to which G. ancinla belongs, enabled us to consider a limited number of species names when identifying the specimens. We considered the possibility that some species may have ranges that extend from Asia into Africa. Indeed, G. ancinla is similar to some African Gryon species but the molecular data at hand indicate that it is distinct from the African species in the current analysis. We assign the name G. ancinla to these specimens because we are confident Nymphs and adults of Acanthocoris scaber feeding on a chili plant in Xiangtoushan, China. that their morphology matches that of the holotype specimen and they are from the same geographic region. There remain a number of Asian species in the charon species group for which we have not yet examined primary types, and examination of more material is required to fully assess the morphological variation and geographic range of G. ancinla. This study should thus be considered as one of many steps forward in the advancement of the taxonomy of Gryon and the reader should be aware that future work will undoubtedly result in more nomeclatural adjustments.
The charon species group was proposed by Mineo 1983 for species with the frontal depression surrounded by a robust carina (Figs 2a,b,4). Mineo also used other characters to define this group, many of which are located on the posterior head. While characters of the posterior head can be useful, their observation may require removal of the head from the mesosoma and it is not permissible to do with many primary type specimens. Mineo 1990 erected another species group, the letus group, listing only characters that this cluster shares with the charon group and did not provide characters to separate these groups. We consider this to be an unnecessary separation and, for simplicity, consider species with a frontal depression surrounded by a robust carina to be part of the charon group. A list of these species, while not exhaustive, is provided in Table 3 to facilitate meaningful comparisons between species. Phylogenetic analysis of Gryon is currently underway and we prefer not to further define any species groups without the context of a larger study. We also emphasize the use of characters that can be seen without dissection so that recently collected material can be matched to primary types. In Gryon there are many such characters to be exploited and during this study we found two that we have not previously encountered in taxonomic literature on the genus. G. kenyotum Mineo Mineo 1982  Specimens used in the molecular analysis Table 3.
Species of Gryon in which the frontal depression is surrounded by carinae.

Cybertaxonomy
The description of G. ancinla was generated using vSysLab (vsyslab.osu.edu), an online, matrix-based tool for generating species descriptions.
a: habitus, lateral view b: head and mesosoma, lateral view All sequences generated from this study are deposited in GenBank (MT604053-MT604066) and all residual DNAs are archived at EBCL, FSCA, and SYSU (Table 2). Voucher specimens which have been reexamined following the molecular analysis were deposited in public collections ( Table 2). All barcode sequences were translated into amino acids to check for stop codons. The sequences obtained were compared with sequences present in GenBank using the Basic Local Alignment Search Tool (http:// www.ncbi.nlm.nih.gov/BLASTn) and aligned with the barcode sequences of Gryon species previously available. BOLD (Ratnasingham and Hebert 2007) was similarly datamined for Gryon CO1 barcodes and evaluated for barcode identification success. Similar sequences not identified to at least genus-level were not included in neighbor-joining analyses or K2P a b Figure 6.
a: holotype female (IEBR 0176), habitus, lateral view b: female (FSCA 00094680), head and mesosoma, anterolateral view distance calculations. Tree search was performed in MEGA7 using the K2P model with a 95% per site data coverage cutoff (Kumar et al. 2016, Kimura 1980. Branch support, calculated with 10,000 boostrap replicates, and distance calculations used the same parameters. a b Figure 7.
Comparison of length of T1 and T3.
a: Gryon ancinla (FSCA 00090590), female, dorsolateral view b: Gryon drunoris (FSCA 00094680), female, dorsal view CO1 barcodes were translated into amino acids and aligned using the default settings of ClustalW (Thompson et al. 1994) as implemented in MEGA7 (Kumar et al. 2016). The amino acid alignment was then back-translated into DNA sequences for neighbor-joining and distance analyses. The resulting alignment contains 682 bp positions and 98 terminal taxa. A Psix species was chosen as an outgroup for neighbor-joining analysis based on the phylogenetic topologies recovered by Taekul   They are also notably different in size (1.67 and 1.27 mm, respectively) and exhibit differences in the sculpture of the frons between the frontal depression and the inner orbit of the compound eye. Specimen FSCA 00094670, which is the larger specimen, has a ridge extending from the orbital carina to the margin of the frontal depression (Fig. 10a). Interestingly, the location of this ridge along the inner orbit corresponds to the transition point between the setose and glabrous portions of the orbital furrow. This ridge can clearly be seen in the holotype specimen of G. ancinla (Fig. 2b). The smaller of the two specimens, FSCA 00094672, has the frons evenly rugose between the orbital furrow and the frontal depression, without a transverse ridge (Fig. 10b).

Diagnosis
Females of G. ancinla have a 6-merous clava (Fig. 4), which is found in other species of the charon group, including the African species G. charon and G. paracharontis. The holotype of G. sponus, from Vietnam, is missing its antennae (Fig. 5), but the illustration of the female antenna in Lê 1996 suggests that it has 6 clavomeres. Each of these species can also be separated from G. ancinla by the shape of the mesoscutellum. In G. ancinla, the posterior margin of the mesoscutellar disc is directly above the posterior margin of the scutellar rim. In G. charon, G. paracharontis, and G. sponus, the mesoscutellar disc extends posteriorly well beyond the scutellar rim (Fig.  5). Gryon drunoris, which is sympatric with G. ancinla, has a mesoscutellum that is evenly convex and the clava is 5-merous (Fig. 6). Gryon ancinla and G. drunoris may also be separated by the relative lengths of the metasomal tergites: T1 is distinctly longer than T3 in G. ancinla (Fig. 7a) and they are roughly equal in G. drunoris (Fig.  7b). In the females of the charon species group that we have examined so far, the clava tends to be distinctly darker than the preceding antennomeres. This makes the clava easily distinguishable from the funicle in most cases, but we caution that using color to differentiate the clava from the funicle may not be reliable in all species or specimens and unambiguous determination of the number of clavomeres requires examination of the papillary sensilla.

Notes
Gryon clavaerus (Fig. 3) was described in the same publication as G. ancinla. Our examination of the type specimens finds no differences that justify keeping them as separate species and we thus treat G. clavaerus as a junior synonym.

Humeral pit
The mesoscutal humeral pit (Fig. 8) is located at the junction of the mesoscutal humeral sulcus and the mesoscutal suprahumeral sulcus. This pit is found in all species of the charon group that we have examined. It is also present in species that are not part of the charon group as it is currently defined, and these species have varying forms of carinae surrounding the frontal depression. The mesoscutal humeral pit thus may be useful for determining affinities between the charon group and other lineages within Gryon.

Setation of the orbital furrow
Setation of the orbital furrow can separate some species in the charon group and perhaps other species groups. Gryon ancinla has setation only in the dorsal part of the orbital furrow (Figs 2b, 4), whereas some species, including the African G. letus, have setation throughout the orbital furrow (Fig. 9). K2P neighbor-joining tree demonstrating the clustering of Gryon CO1 barcodes. The larger cluster in blue highlights the Gryon charon species group. Clusters in red and magenta highlight the Gryon ancinla haplogroups. Bootstraps values of 80 and above are indicated.

Molecular analysis
The neighbor-joining analysis revealed relatively large sequence divergences between clusters of Gryon CO1 barcodes (Fig. 11). Interpretation of sequence divergence in Gryon is currently hampered by the lack of species-level identifications that are necessary to define intra-and interspecific variation. We included 14 new CO1 barcodes from members of the Gryon charon species group and our neighbor-joining analysis recovered a cluster of these species with an additional unidentified Gryon from South Africa (BIN: BOLD:ADO2077; SAFRA3055-18, SAFRA4239-18) (Ratnasingham and Hebert 2013). We found two haplogroups of G. ancinla, indicated in Fig. 11 by the red and magenta branches. Specimens from each lineage were collected in a single Malaise trap sample in Guangzhou (FSCA 00094670-00094673), demonstrating that the haplogroups are sympatric. These Gryon ancinla haplogroups differ by K2P distances ranging from 9.6-10.4% (Table 1) and they are each other's nearest neighbor. BLASTn searches yielded poor matches to the G. ancinla barcodes (86-87% identity to other hymenopteran barcodes). In BOLD, G. ancinla from haplogroup 1 were a 97% match to an unidentified specimen from Bangladesh (GMBCB2151-15) and the available image of this specimen is consistent with our concept of G. ancinla. This suggests that G. ancinla has a wide geographical distribution in southeast Asia.

Discussion
Gryon contains widespread species and geographically broad analysis is needed to identify synonyms. This study characterizes a species found in southeast Asia to facilitate comparison with similar species of Gryon in the region and to associate ecological data with a taxonomic name.