Molecular and morphological evidence reveals a new genus of the subfamily Heteropterinae (Lepidoptera, Hesperiidae) from China

Abstract Molecular phylogenetic analysis indicates that the genus Carterocephalus is not monophyletic. Based on combined molecular and morphological evidence, we propose a new genus, Pulchroptera Hou, Fan & Chiba, gen. nov., for Pamphilapulchra Leech, 1891. The adult, wing venation, and male genitalia of Pulchropterapulchracomb. nov., Carterocephaluspalaemon, and related genera are illustrated.


Introduction
In recent years, the molecular phylogeny of the family Hesperiidae has attracted the attention of an increasing number of researchers (Warren et al. 2008(Warren et al. , 2009Sahoo et al. 2016;Toussaint et al. 2018;Cong et al. 2019;Li et al. 2019;Liu et al. 2020). At the subfamily level, however, the phylogeny of the family Hesperiidae has yet to be established, and multiple new subfamilies  and genera Huang et al. 2016Huang et al. , 2019Cong et al. 2019;Li et al. 2019) have been proposed in recent years.
The genus Carterocephalus includes more than 20 species distributed in the Holarctic and Oriental regions. However, a cursory inspection of the male genitalia indicates that C. pulchra (Leech, 1891) is not a congener of the type species Papilio palaemon Pallas, 1771. Indeed, the findings of our morphological and molecular phylogenetic studies have revealed closer relationships with species in the genera Heteropterus and Leptalina. Accordingly, we consider that Carterocephalus pulchra should be placed in a new genus.
In the present study, we sought to assess the monophyly of the genus Carterocephalus and its relationship with other genera of Heteropterinae. On the basis of the evidence obtained, we describe a new genus.

Morphological examination
For the morphological study, we followed the methods described by Fan et al. (2010). To examine wing venation, wings were removed from the thorax and cleaned with a 1:1 mixture of bleaching liquid (Blue Moon, Guangzhou, China) and water for approximately 3 to 4 min. Photographs of the wing venation and male genitalia were taken using a Keyence VHX-5000 digital microscope (Keyence, Osaka, Japan).

Taxon sampling
We sampled specimens from all genera listed in the subfamily Heteropterinae (Warren et al. 2008(Warren et al. , 2009Cong et al. 2019;Toussaint et al. 2020), including as many species as possible. We used a total of 44 specimens of 38 species in 13 genera as ingroup taxa, along with 12 species from other subfamilies (Coeliadinae, Pyrginae, Eudaminae, Euschemoninae, Barcinae, Trapezitinae, and Hesperiinae) as outgroup taxa. Among these specimens, 31 were newly sequenced in this study, with the remaining sequences being obtained from the GenBank database along with supplementary data presented by Sahoo et al. (2016) and Toussaint et al. (2020). The respective voucher specimens and additional information are listed in Suppl. material 1: Table S1. Vouchers bearing codes beginning with the abbreviation SCAU have been deposited in the collection of South China Agricultural University (SCAU), Guangzhou, China, and the specimens (JU19), (Dalla), and (SZSMETI) are retained in the private collections of J. Uehara, H. Chiba, and S. Sáfián, respectively.

Laboratory protocols
DNA was extracted from two or three legs of dried adult specimens using a TIANamp Genomic DNA Kit (Tiangen, Guangzhou, China) following the manufacturer's instructions. We amplified a single mitochondrial gene (658 bp of COI) and three nuclear genes (1066 bp of EF-1α, 610 bp of RPS5, and 403 bp of Wingless), for a total of 2737 bp. The primers used to amplify each gene were synthesised by Sangon Biotech (Shanghai, China) and are shown in Suppl. material 2: Table S2. DNA amplification was performed in 20-µL reaction volumes containing 1 µL of template DNA, 0.8 µL of each primer (10 µM), 10 µL of 2× EasyTaq PCR superMix (+dye) (Transgen, Beijing, China), and 7.4 µL of ddH 2 O. The amplification protocol adopted is the one described by Huang et al. (2019). Sequencing of the amplicons thus obtained was performed by Sangon Biotech (Shanghai, China) and Tsingke Biological Technology (Beijing, China), and new sequences have been deposited in GenBank (Suppl. material 1: Table S1).

Phylogenetic analyses
Sequences were aligned using Clustal W (Thompson et al. 1997) and edited manually using MEGA 7.0 (Kumar et al. 2016). Gene data from Cong et al. (2019) were extracted from the genomic assembly in IDBA-UD (Peng et al. 2012). Partition-Finder v2.1.1 (Lanfear et al. 2012(Lanfear et al. , 2016Guindon et al. 2010) was used to select the optimal codon partitioning scheme under Akaike information criterion correction (AICc) (Suppl. material 3: Table S3). We inferred the phylogenetic trees using two methods, namely maximum likelihood (ML) and Bayesian inference (BI), for which we used the partition scheme produced by PartitionFinder. ML analyses were performed using IQ-TREE (Nguyen et al. 2015) as implemented in the IQ-TREE web online server (iqtree.cibiv.univie.ac.at, Trifinopoulos et al. 2016), with branch support values evaluated based on 1000 replicates for ultrafast bootstrap (UFBoot) (Minh et al. 2013) and SH-aLRT (Guindon et al. 2010). BI analyses were performed using the CIPRES Science Gateway (https://www.phylo.org/) (Miller et al. 2010) with Markov Chain Monte Carlo (MCMC) randomisation in MrBayes using XSEDE 3.2.6 (Ronquist et al. 2012). Reversible-jump MCMC was used to facilitate sampling across the entire subduction rate model. We conducted two independent MCMC runs, with four Markov chains (5 × 10 6 generations) for each analysis, of which the initial 25% of samples were discarded as burn-in. Bayesian posterior probabilities (PP) were used to evaluate branch support, and trees were visualised using FigTree v1.4.0.

Phylogenetic relationships
The topological structures of the concatenated dataset inferred by ML and BI analyses were found to be generally consistent and strongly supported at most nodes (PP ≥ 0.98, SH-aLRT ≥ 95, UFBoot ≥ 98) (Fig. 1). Moreover, the two analyses provided strong support for the monophyly of Heteropterinae (PP = 1, SH-aLRT = 99.9, UFBoot = 100), which excludes the genera Apostictopterus, Barca, Lepella, and Tsitana originally assigned to this subfamily, and is consistent with the findings of the most recent studies (Toussaint et al. 2018(Toussaint et al. , 2020Cong et al. 2019;Zhang et al. 2019). Within the subfamily Heteropterinae, four major clades were differentiated, with 14 well-supported monophyletic subclades, corresponding to the 13 currently recognised genera and the Carterocephalus pulchra clade. Certain results were consistent with those of previous studies Toussaint et al. 2020): (1) of the 13 genera, 12 genera, excluding Carterocephalus, were monophyletic; (2) Argopteron and Butleria formed a strongly supported monophyletic group (PP = 1, SH-aLRT = 99.1, UFBoot = 100) that is sister to all other genera in Heteropterinae (PP = 1, SH-aLRT = 98.3, UFBoot = 99); (3) Carterocephalus, excluding the species C. pulchra, was sister to the clade containing Metisella, Hovala, and Willema with strong support (PP = 1, SH-aLRT = 99.5, UFBoot = 100); and (4) Piruna, Dardarina, Freemaniana, Ladda and Dalla formed a strongly supported monophyletic clade (PP = 1, SH-aLRT = 98.8, UFBoot = 99). Two findings, however, are inconsistent with those reported previously. Firstly, Piruna is sister to Dardarina (PP = 0.76, SH-aLRT = 87.2, UFBoot = 94), as opposed to sister to the four genera Dardarina, Freemaniana, Ladda, and Dalla. Based on the morphology of the male genitalia (Evans, 1955), Piruna shows a relatively close similarity to Dardarina, whereas species of Dalla show extensive variation. However, previous molecular phylogenetic studies, as well as our own, sampled only some representatives of Dalla. Accordingly, the monophyly of Dalla as well as the relationships among these five genera should be subjected to further studies. Secondly, we found that Carterocephalus is not a monophyletic group, given that the 11 species analysed in the present study were recovered in two distinct clades, with C. pulchra clustering with Leptalina and Heteropterus with strong support (PP = 1, SH-aLRT = 97.7, UFBoot = 100). The other ten species, including the type species C. palaemon, were recovered as a strongly supported monophyletic clade.
Although in this study we focused on relationships among the genera of Heteropterinae, it is worth mentioning that certain intra-generic relationships, namely, those between C. abax Oberthür, 1886 and C. patra Evans, 1939, C. avanti (de Nicéville, 1886 and C. argyrostigma (Eversmann, 1851), C. longimaculatus Hou, Fan & Chiba, 2021and C. alcina Evans, 1939, C. palaemon (Pallas, 1771 and C. silvicola (Meigen, 1828) are strongly supported. As described by Toussaint et al. (2020), despite the lack of strong support (PP = 0.73, SH-aLRT = 85.3, UFBoot = 69), C. houangty and C. dieckmanni were clustered in a clade comprising C. palaemon, C. silvicola, C. longimaculatus, and C. alcina. In our previous study (Hou et al. 2021), we established that C. dieckmanni is sister to C. abax and C. patra. However, owing to an oversight, the names C. dieckmanni and C. argyrostigma were confused, which explains the discrepancy compared with the results reported herein. Accordingly, to determine relationships more comprehensively in the genus Carterocephalus, we ideally need to undertake additional and more extensive sampling.
Morphologically, although C. pulchra is similar to the type species of Carterocephalus with respect to wing shape and pattern (Fig. 2), the origin of vein R s on the hindwing is located nearly midway between the termen and the base in C. pulchra, Heteropterus, and Leptalina, whereas in other species of Carterocephalus the origin of vein R s is closer to the termen than to the base (Fig. 3). With regards to the male genitalia, the uncus in C. pulchra, Heteropterus, and Leptalina is deeply bifurcated, with arms distant from each other, whereas in the type species of Carterocephalus the uncus bifurcates with arms closely aligned (Fig. 4). These morphological similarities would accordingly appear to indicate that C. pulchra is more closely related to Heteropterus and Leptalina than to other species of Carterocephalus. Of these related genera, C. pulchra is autapomorphous with respect to its male genitalia. Notably, the gnathos is weakly sclerotized, membranous, and rounded at the tip, the valvae are asymmetrical, and the juxta is a heart-shaped ring with a narrow and long latero-central process. In summary, we propose a new genus, Pulchroptera Hou, Fan & Chiba gen. nov., for the Carterocephalus pulchra clade based on its autapomorphies and molecular evidence.

Pulchroptera Hou, Fan & Chiba, gen. nov.
http://zoobank.org/3C184FA2-423E-43F5-B77A-5D06C5639245 Figures 2-4 Type species. Pamphila pulchra Leech, 1891 Description. Forewing length 11-12 mm. Antennae approximately half the length of forewing; nudum 8 on apiculus, dark brown. Palpi on second segment long and erect, yellow with long black hairs; on third segment black, thick, short, and porrect. Wing venation (Fig. 3): forewing: length of discoidal cell almost equal to 2/3 forewing length, Sc ends at 1/2 forewing length; origin of vein R 4 before vein R 5 ;  origin of vein M 2 in middle of veins M 1 and M 3 ; veins CuA 1 , CuA 2 , and 1A+2A almost parallel to each other; origin of vein CuA 2 nearly midway between vein CuA 1 and base. Hindwing: costa longer than dorsum; length of discoidal cell almost equal to 3/5 hindwing; origin of vein Rs midway between base and termen; origin of vein M 2 slightly nearer M 1 than M 3 . Wing ground colour and wing patterns: upper side dark brown with small yellow spots in central and submarginal areas; underside light brown, forewing patterns similar to upper side, hindwing with small silvery spots in spaces Rs, M 3 , CuA 1 , and CuA 2 , and with a silvery longitudinal central streak. Mid and hind tibiae each with pair of spurs. Male genitalia: Tegument small and narrow, constricted at middle in dorsal view; uncus deeply bifurcated, V-shaped dorsally; gnathos long and wide, longer than tegument, membranous, undivided from basal 1/3; saccus long; valvae asymmetrical, bifid, distal end of left valva more sclerotized than right valva; aedeagus long, subzonal sheath shorter than suprazonal sheath, ratio of subzonal sheath to suprazonal sheath approximately 1:2, vesica with cornuti; juxta a heart-shaped ring with membranous extensions dorsally.
Remarks. The new genus superficially resembles Carterocephalus Lederer, 1852, although it is distinguishable from the latter with regards to the following characters: hindwing undersides with silver spots, a deeply bifurcated V-shaped uncus, juxta a heart-shaped ring, and valvae asymmetrical.
The new genus contains only the type species Pulchroptera pulchra (Leech, 1891) comb. nov., with the nominotypical subspecies and a further subspecies, Pulchroptera pulchra ops (Grum-Grshimaïlo, 1891) comb. nov. According to the description of Evans (1949), in Pulchroptera pulchra pulchra comb. nov. the upper side of the hindwing has a cell spot and the submarginal markings are notably more conspicuous, whereas in Pulchroptera pulchra ops comb. nov. the upper side of the hindwing lacks a cell spot and has conspicuous submarginal markings. Whether the subspecies status of the latter is valid is subject to further verification.
Etymology. The name of the genus is taken from the specific epithet of the type species 'pulchr-', meaning beautiful, and 'optera', meaning wing. The gender is feminine.