A total of 17 species representing 14 genera/subgenera of the tribe Meconematini—Decma (Idiodecma), Decma (Decma), Kuzicus, Macroteratura (Macroteratura), Macroteratura (Stenoteratura), Megaconema, Nigrimacula, Similameconema Dou & Shi, 2018, Tamdaora, Teratura, Xiphidiopsis (Xiphidiopsis), Xiphidiopsis (Euxiphidiopsis), Xizicus, and Phlugiolopsis Zeuner, 1940—were collected from Hainan, Henan, Shaanxi, Sichuan, and Xizang provinces of China (Table S1). The species names follow the Orthoptera Species File (OSF) (Cigliano et al. 2026). All samples were preserved in 100% ethanol and stored at –20 °C in the Xi’an Botanical Garden of Shaanxi Province (Institute of Botany of Shaanxi Province). Total genomic DNA was extracted from the hind femur muscles of each individual using the DNeasy Blood and Tissue Kit (QIAGEN 69504), following the manufacturer’s protocol, and subsequently stored at –20 °C.

Total genomic DNA extracted from each of the 17 Meconematini species (Table S1) was used to construct shotgun sequencing libraries using the NEBNext® Ultra™ II DNA Library Prep Kit, with the Illumina TruSeq single-index adapters supplied in the kit. The libraries were sequenced on an Illumina HiSeq 2500 platform, generating 150 bp paired-end reads. Prior to assembly, raw reads were subjected to quality control and filtering using Trimmomatic v0.39 (Bolger et al. 2014) with the following parameters: ILLUMINACLIP:TruSeq3‑PE‑2.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:50. The resulting clean data used for assembly exceeded 2 Gb per species. Based on estimates of mitochondrial genome size, the final sequencing depth achieved was over 100× for each species. The filtered reads were assembled by mapping to the mitochondrial genome of Xizicus fascipes (JQ326212) as a reference using MIRA 4.0.2 (Chevreux et al. 2004) and MITObim v1.8 (Hahn et al. 2013) (Main parameter: -start 1 -end 100 -sample testpool -ref Xizicus_fascipes -mt -readpool input.fastq --quick JQ326212-ref.fasta --clean --NFS_warn_only >log). The complete mitochondrial genomes were initially annotated automatically using MITOS2 (Bernt et al. 2013). The resulting annotations were subsequently transferred to Geneious Prime (Kearse et al. 2012) for manual curation and boundary refinement by comparison with complete mitogenomes of related species. Transfer RNA genes predicted by MITOS2 were further confirmed in Geneious Prime using the same comparative approach (http://mitos.bioinf.uni-leipzig.de/index.py).

Nucleotide base compositions were determined using Geneious Prime (Kearse et al. 2012). Composition skew was assessed with the formulas: AT-skew = [A–T]/[A+T] and GC-skew = [G–C]/[G+C] (Perna and Kocher 1995). Relative synonymous codon usage (RSCU) values for protein-coding genes (PCGs) in the newly sequenced species were computed using MEGA v11.0 software (Tamura et al. 2021). Substitution saturation tests were performed with DAMBE7, and cumulative skew plots were generated for datasets of 13 PCGs and 2 ribosomal RNAs (PCG123R) (Xia 2018). Heterogeneity in nucleotide variation among sequences was evaluated for different datasets using AliGROOVE v1.05 (Kück et al. 2014).

We compared the Ka/Ks values for each protein-coding gene (PCG) across all species. The proportions of synonymous (Syn) and non-synonymous (Nsyn) substitutions for each PCG were calculated using DnaSP v6.12.03 (Librado et al. 2009). One-way ANOVA was conducted using SAS software to assess differences in Ka/Ks values among PCGs across all Meconematini species, with the aim of detecting signals of selection pressure.

We retrieved the 23 published mitogenome sequences of Meconematini (22 species) and two outgroup sequences (Stenopelmatus fuscus from Stenopelmatidae and Diestrammena asynamora from Rhaphidophoridae) from GenBank. These were combined with the 17 mitogenome sequences generated in this study, resulting in a total of 42 sequences for phylogenetic reconstruction (Table S1). Phylogenetic relationships were inferred from partitioned datasets of 13 PCGs and two ribosomal RNAs (rRNAs) using Bayesian inference (BI) and maximum likelihood (ML) approaches. Prior to phylogenetic inference, all PCGs and rRNAs were individually aligned using ClustalW implemented in MEGA v11.0 (Tamura et al. 2021). For the 13 PCGs, stop codons were removed and the sequences were translated into amino acids to verify alignment accuracy. Then, we concatenated the aligned genes into a single data matrix using SequenceMatrix v.1.8 (Vaidya et al. 2011). The best-fit partitioning scheme and optimal nucleotide substitution model for the concatenated data matrix were selected in PartitionFinder v2.1.1 (Lanfear et al. 2012) using the “greedy” search algorithm (heuristic search) and “unlinked” branch lengths. The maximum likelihood (ML) and Bayesian inference (BI) criteria were used to construct the phylogenetic tree in RAxML (Stamatakis et al. 2005), IQ-TREE v1.6.9 (Nguyen et al. 2015) and MrBayes v3.2 (Ronquist et al. 2012). Both ML and BI analyses employed the optimal partitioning scheme and models selected above; ML nodal support was evaluated with 1,000 bootstrap replicates. For BI, four Markov chain Monte Carlo (MCMC) chains were run for 100,000,000 generations, sampling every 5,000 generations. After discarding the first 25% of generations as burn-in, posterior probabilities (PPs) were calculated in a consensus tree. The resulting phylogenetic trees were visualized and edited using the FigTree v1.4.4 (http://tree.bio.ed.ac.uk).

As detected in previous studies (Pang et al. 2024), the new sequenced mitogenomes show circular structures and contain the typical conserved set of 37 genes, including the 13 protein-coding genes (PCGs), large and small rRNAs (rrnL and rrnS), 22 transfer RNAs (tRNAs), and a large non-coding region referred to as the A+T-rich region or control region (CR) (Table S2). Gene order and arrangement are identical to the published Meconematini mitogenomes. The mitogenomes of Meconematini exhibit a distinct AT bias (66.8% – 74.9%) and are notably deficient in G (9.9% – 11.7%). The overall GC-skew ranges from –0.3233 to –0.2047, and the AT-skew varies between 0.0069 and 0.0707 (Table S3). Similarly, the PCGs show a pronounced AT bias (66.6%–74.2%), with AT-skew and GC-skew values ranging from 0.0126 to 0.0904 and –0.3231 to –0.1818, respectively (Table S3).

For the 17 newly sequenced species, almost all PCGs have the typical initiation codon of ATN (Table S4). However, ATP6 in Xizicus (Eoxizicus) sinuatus (Liu & Zhang, 2000) initiated from a non-standard initiation codon of GTG. With respect to termination codons, about half of the PCGs have a typical termination codon TAA in all species (Table S4). The COI, ND4, and ND5 genes in all species, the COIII gene in most species, and the COII and ATP6 genes in a minority of species use the incomplete codon T as their termination codon. The PCGs exhibit a significant codon usage bias (Fig. 2). The four most preferred codons are UUA (Leu1), UUU (Phe), AUU (Ile), and AUA (Met), with average relative synonymous codon usage (RSCU) values of 3.551, 1.558, 1.712, and 1.664, respectively. With the exception of Nigrimacula xizangensis (Jiao & Shi, 2013), which shows the lowest RSCU value (0.05) for CUG (Leu), the codon AGG (Arg) exhibits the lowest usage across all other species. Furthermore, the codon AGG (Arg) is not used in the following species: Macroteratura (Macroteratura) megafurcula (Tinkham, 1944), Macroteratura sp., Kuzicus sp., Decma (Decma) tristis Gorochov & Kang, 2005, Decma (Idiodecma) birmanica (Bey-Bienko, 1971), Xiphidiopsis (Xiphidiopsis) bituberculata Ebner, 1939, and Xiphidiopsis (Xiphidiopsis) minorincisus Han, Chang & Shi, 2015.

Base substitution saturation analysis showed that the Iss value for dataset PCG123R was lower than the critical value (Iss.c), indicating that the dataset did not exhibit substitution saturation and is appropriate for phylogenetic analysis (Table S6). No obvious heterogeneous outliers were detected in AliGROOVE (Fig. S1).

The evolutionary rates evaluated by Ka/Ks values of the 13 PCGs for all 37 Meconematini mitogenomes were much less than 1, indicating that strong evolutionary constraints and purifying selection have dominated the evolution of Meconematini mitogenomes (Figs 3, S2). Among them, the ATP8 gene had the highest Ka/Ks ratio, while the COI gene had the lowest, indicating that the COI gene experienced more substantial evolutionary selection pressure. The order of the mean Ka/Ks ratios for the 13 PCGs in Meconematini was ATP8 > ND6 > ND2 > ND5 > ND4 > ND4L > ND1 > ND3 > ATP6 > COII > CYTB > COIII > COI (Fig. 3; Table S5). In pairwise comparisons of PCGs, the Ka/Ks values showed significant differences (F = 947.52, P < 0.001) among genes, except for comparisons between ATP6, ND1, and ND3 and between CYTB and COIII.

The phylogenetic trees inferred from the datasets of 13 PCGs and 2 rRNAs using the maximum likelihood (from RAxML and IQ-TREE) and Bayesian inference (from MrBayes) methods showed almost identical tree topologies, except for a minor difference in the position of the species Xiphidiopsis (Xiphidiopsis) appendiculata Tinkham, 1944 (Figs 4, S3). However, the bootstrap values of multiple branch nodes do not reach the trusted threshold range, especially in the maximum likelihood-based phylogenetic trees.

The phylogenetic reconstructions show that the monophyly of the genera Microconema, Eoxizicus, and Decma was robustly supported (BS ≥ 99 or BPP = 1), while species of the genera Phlugiolopsis, Xizicus, and Xiphidiopsis are not recovered as monophyletic (Figs 4, S3). The monophyletic Eoxizicus is clearly separated from the rest of the Xizicus species and is sister to a clade comprising of Alloxiphidiopsis Liu & Zhang, 2007 + Grigoriora Gorochov, 1993 in all phylogenetic trees. The monophyly of Phlugiolopsis is weakly supported, mainly due to the inclusion of the species Paraphlugiolopsis jiangi Bian & Shi, 2014, which was established based on the apices of posttibiae bearing two pairs of spines (Bian et al. 2014). Species of the genera Xizicus and Xiphidiopsis are intermingled. The species Xiphidiopsis (Euxiphidiopsis) gurneyi Tinkham, 1944, Xiphidiopsis (Xiphidiopsis) autumnalis Gorochov, 1998, Xizicus (Haploxizicus) maculatus (Xia & Liu, 1993), Xizicus (Xizicus) fascipes (Bey-Bienko, 1955) and X. (X.) appendiculata cluster together, forming a highly supported clade (BS = 100, BPP = 1). The remaining two species, X. (X.) bituberculata and X. (X.) minorincisus, form a distinct clade directly related to the clade of Paraphlugiolopsis Bian & Shi, 2014 + Phlugiolopsis + Chandozhinskia Gorochov, 1993 + Xizicus (excluding the subgenus Eoxizicus) + Xiphidiopsis.

The mitogenomes of Meconematini have a typical gene composition of 37 genes and variable sizes of AT-rich control regions—a common feature documented extensively in metazoan mitochondrial genomes (Clary and Wolstenhome 1985; Boore 1999). Like the vast majority of ensiferans, the gene arrangement retains the ancestral Orthoptera condition, as previously reported in related groups (Song et al. 2015; Dan et al. 2022). Regarding start codon usage, previous research has shown that mitochondrial protein-coding genes of Orthoptera exhibit not only the typical standard start codons ATN (ATA, ATT, ATC, ATG), but also unconventional ones such as GTG, TTG, AGT, TTA, CTG, and CCT (Zhao et al. 2018; Zhang et al. 2023). In our study, we identified for the first time in the subfamily Meconematinae the exceptionally rare start codon GTG in the ATP6 gene, expanding the known diversity of initiation codons in this group. Our analysis also revealed high consistency in termination codons across all examined Meconematini species: all 37 species utilize the incomplete codon T as the termination signal for COI, ND4, and ND5. According to the classical model proposed by Ojala et al. (1981), these incomplete codons may form complete termination signals through post-transcriptional polyadenylation, thereby facilitating transcription termination.

In animal mitogenomes, all 13 protein-coding genes (PCGs) are involved in aerobic metabolism, and positive selection is often associated with adaptation to new environments. In the 37 mitogenomes examined, the Ka/Ks ratios for all 13 PCGs were below 1, indicating that purifying selection has dominated the evolution of Meconematini mitogenomes. Among these genes, ATP8 exhibited the highest Ka and Ka/Ks values, whereas COI showed the lowest, suggesting that COI has experienced stronger evolutionary constraints (Fig. 3). These findings are consistent with previous studies on Orthoptera (Chang et al. 2020) and Ensifera (Dan et al. 2022). However, the ranking of Ka/Ks values among the 13 PCGs differs from their results. Meanwhile, a study on TettigoniidaePCGs by Zhao et al. (2025) reported that COIII exhibited the strongest signals of purifying selection, indicating evolutionary conservation, and positively selected sites were detected in ND1 and COI, which may reflect lineage-specific adaptations or ecological niche shifts that are not pervasive across all Ensifera.

Our study represents the first formal phylogenetic analysis of Meconematini. The tree topologies recovered by the two phylogenetic methods were consistent at the generic level, aligning with the recent findings of Pang et al. (2024). Although only 22 genera and 37 species were involved in our phylogenetic analysis, the results still provide significant insights into the morphological classification of Meconematini. In the phylogenetic tree, genera within the proximal clade (Paraphlugiolopsis, Phlugiolopsis, Chandozhinskia, Xiphidiopsis, Xizicus, Microconema, Nigrimacula, Eoxizicus, Alloxiphidiopsis, Tamdaora) are characterized by males with membranous genitalia (hagloid type in Gorochov 1993, 1998), while genera distributed in the basal half (Acosmetura Liu, 2000, Pseudocosmetura Liu, Zhou & Bi, 2010, Similameconema, Meconema Serville, 1831, Pseudokuzicus Gorochov, 1993, Shoveliteratura Shi, Bian & Chang, 2011, Decma, Kuzicus, Teratura, Macroteratura Gorochov, 1993 and Megaconema) have sclerotized male genitalia (grylloid type with partially or completely sclerotized dorsal lobe; tettigonioid type with a pair of sclerotized titillators; Gorochov 1993, 1998) (Fig. 5). These findings indicate a potential evolutionary trend in the male genitalia of Meconematini, shifting from sclerotized to membranous structures. The degree and configuration of genital sclerotization may provide phylogenetically informative characters for delimiting genus-group or subtribe taxa within the tribe.

In addition, our phylogenetic reconstructions provide robust support for the current taxonomic framework of the genera Kuzicus, Macroteratura, Teratura, and Megaconema. The genus Teratura was established by Redtenbacher (1891), and Gorochov (1993) subdivided Teratura into three subgenera, Teratura, Macroteratura, and Stenoteratura, principally according to the characteristics of the male tenth abdominal tergite, epiproct, and genitalia. Subsequent taxonomic studies (Wang 2015; Chen et al. 2020; Jin et al. 2020) elevated Teratura (Macroteratura) to full generic status, while transferring Teratura (Stenoteratura) to Macroteratura. This reclassification aligns with two critical synapomorphies: the male 10^th tergite with a pair of long lobes and hind tibia bearing three pairs of apical spurs. These diagnostic traits exhibit phylogenetic coherence with Macroteratura and Kuzicus Gorochov, 1993, forming a distinct clade that contrasts markedly with Teratura and Megaconema. The latter genera are characterized by the male 10^th tergite without long lobes and hind tibia bearing two pairs of apical spurs. Our molecular phylogenies demonstrate strong corroboration with these morphological discontinuities. Bayesian and maximum likelihood analyses resolve Macroteratura + Kuzicus and Megaconema + Teratura as reciprocally monophyletic groups with strong nodal support (BPP = 1.0, BS = 100), validating the current classification schema based on evolutionary distinctiveness.

Overall, our phylogenetic reconstructions corroborate the synapomorphic value of genital morphology in generic delineation as originally proposed by Gorochov (1993, 1998), thereby validating the taxonomic utility of these characters in Meconematini classification. However, critical issues emerge regarding current generic circumscriptions: paraphyletic tendencies observed in Xiphidiopsis sensu lato; ambiguous morphological boundaries between genera, such as Xizicus and Euxiphidiopsis; and incongruence in character-state distributions among molecular clades. Nevertheless, we will not propose any changes to generic classification herein without sufficiently thorough taxon sampling. As a species-rich and diverse group, Meconematini comprises > 780 described species, so an integrative taxonomic framework incorporating dense molecular sampling and multivariate morphometric analyses will be essential for future taxonomic studies. Future investigations should prioritize sequencing type species and examining topotypic specimens to establish robust phylogenetic hypotheses before implementing taxonomic changes.

The genera Xiphidiopsis and Xizicus are the two most diverse groups in Meconematini, and are also the most controversial in terms of species classification and subgeneric division (Jin et al. 2020; Gorochov 2022). Our phylogenetic reconstructions reject the monophyly of both Xiphidiopsis and Xizicus genera, consistent with previous phylogenetic studies (Mao et al. 2020; Pang et al. 2024).

In the genus Xiphidiopsis, whether Euxiphidiopsis should be treated as a subgenus or an independent genus has long been a matter of taxonomic debate. In our study, two species traditionally regarded by Chinese researchers as members of the genus Euxiphidiopsis—X. (E.) gurneyi and Xiphidiopsis (Euxiphidiopsis) autumnalis—formed a well-supported clade in the phylogenetic tree. This clade subsequently grouped with X. (H.) maculatus and X. (X.) fascipes, while exhibiting a distant relationship with other Xiphidiopsis species (Fig. 4). Similar results were also obtained in the previous phylogenetic analyses based on mitogenome data (Mao et al. 2020; Pang et al. 2024) or COI + ITS1-5.8S rDNA-ITS2 fragments (Han 2016). Moreover, COI-5P barcode analyses have further corroborated this relationship, recovering a monophyletic cluster comprising both Xiphidiopsis autumnalis and Xiphidiopsis gurneyi (Zhou et al. 2019). Taken together, these results support the taxonomic treatment proposed by Liu and Zhang (2000) and Bai et al. (2014), advocating the recognition of Euxiphidiopsis as a distinct genus. However, whether the disputed species transferred by Gorochov (2022) to other genera should instead be placed in Euxiphidiopsis remains to be determined and will require additional evidence—both morphological and molecular. In particular, the exploration of new taxonomic characters will be essential for refining morphological diagnoses and resolving the placement of these species.

Regarding the genus Xizicus, although Eoxizicus remains classified as a subgenus of Xizicus in recent taxonomic treatments (Di et al. 2015; Gorochov 2022; Su et al. 2023; Cigliano et al. 2026), our phylogenetic analyses provide substantial evidence supporting its elevation to generic rank as proposed by several researchers (Liu and Zhang 2000; Wang et al. 2015; Jin et al. 2020). The three examined Eoxizicus species formed a strongly supported monophyletic branch (BS = 100; BPP = 1) at the base of the membranous genitalia clade and were sister to the branch comprising the genera Grigoriora and Alloxiphidiopsis, distinctly divergent from other species of Xizicus and Xiphidiopsis. These phylogenetic results, revealing distinct evolutionary trajectories and consistent morphological differentiation, strongly corroborate the taxonomic proposal to recognize Eoxizicus as a distinct genus rather than maintain its current subgeneric status under Xizicus.