We collected adult Perlodidae from different provinces of China: Filchneria zhouchangfai from Anhui, Stavsolus manchuricus from Liaoning, Neowuia wuyishana from Fujian, Perlodinella kozlovi and Perlodinella epiproctalis from Qinghai, Isoperla qinlinga and Filchneria songi from Shaanxi, Tibetisoperla wangluyui from Xizang (Huo et al. 2020, 2021, 2022a, 2022b, 2022c, 2023), Megarcys ochracea and Skwala compacta from Heilongjiang (unpublished records).

Throughout the research, we followed ethical guidelines, ensuring that all activities were authorized and no endangered or threatened species were harmed. All specimens were preserved in 100% ethanol and stored at -20°C. For DNA extraction, we took samples from the thoracic muscles and legs of adult specimens, using the column mtDNAoutKit (Axygen Biotechnology, Hangzhou, China) according to the manufacturer’s instructions. The extracted genomic DNA was then stored at -20°C until it was needed for PCR analysis.

The whole genome was amplified using LA-PCR with the following conditions: initial denaturation at 93°C for 2 minutes, 40 cycles at 92°C for 10 seconds, annealing at 54°C for 30 seconds, and extension at 68°C for 8 minutes, increased by 20 seconds for the final 20 cycles, concluding with a 10-minute extension at 68°C. Quality control was performed using a Qubit 3.0 fluorometer to quantify DNA concentration and 1% agarose gel electrophoresis to assess the integrity of the PCR products. High-quality DNA samples were then used to create a 500 bp paired-end library with the NEBNext Ultra DNA Library Prep Kit for Illumina sequencing on the Illumina NovaSeq 6000 platform (BIOZERON Co., Ltd, Shanghai, China). De novo assembly was conducted using GetOrganelle v1.6.4, referencing the other mitochondrial genomes of Plecoptera to generate contigs. The Contigs of the mitogenome were generated by de novo assembly using GetOrganelle v1.6.5 using the mitochondrial genomes of closely related species as a reference. The potential mitochondrial reads were isolated from Illumina reads using BLAST against the mitogenomes species and GetOrganelle outcome. To complete the process, the accumulated sequences were rearranged and oriented to match the reference mitochondrial genome, providing a complete genetic sequence of mitochondria (BIOZERON Co., Ltd., Shanghai, China).

We analyzed 18 previously published Plecoptera mito­chondrial genomes for phylogenetic assessment (­Table 1). Phylogenetic trees were constructed focusing on the family Perlodidae, with two species from the family Leuctridae included as the out-group. The analysis included all protein-coding genes (PCGs) and rRNA genes. Four datasets were utilized for the phylogenetic investigation: (1) sequences of the 13 PCGs (PCG matrix); (2) sequences of the 13 PCGs plus two rRNAs (PCGR matrix); (3) sequences of the 13 PCGs excluding third codon positions (PCG12 matrix); and (4) sequences of the 13 PCGs excluding third codon positions plus two rRNAs (PCG12R matrix). The optimal partitioning scheme and models were identified with ModelFinder (Kalyaanamoorthy et al. 2017) for maximum likelihood (ML) and Bayesian inference (BI) analyses. Bayesian inference was performed using MrBayes (version 3.1.2) with the following parameters: 10 million generations with four chains, sampling every 100 generations, and discarding the initial 25% as burn-in (Zhang et al. 2020; Ronquist and Huelsenbeck 2003; Zhao et al. 2020). Maximum likelihood analysis was conducted using IQTree v. 1.6.12 (Nguyen et al. 2015), employing 1,000 ultra-fast bootstrap estimates to assess the support for the branches in the resulting phylogenetic tree. The final phylogenetic tree was visualized using FigTree v1.4.2. (Available at: http://tree.bio.ed.ac.uk/software/figtree).

The phylogenetic analyses were conducted using concatenated mitochondrial genes from 20 stoneflies, and analyzed as four data matrixes (PCG, PCGR, PCG12, PCG12R). These mitogenomes included 16 species from the family Perlodidae (Table 1). Two species of Leuctridae (Rhopalopsole subnigra and Rhopalopsole bulbifera) were included as outgroup taxa (Table 1).

The phylogenetic trees generated by Bayesian inference (BI) and maximum likelihood (ML) analyses showed similar topologies, with species grouping into well-supported clades (Figs 1, 2). In both analyses, the subfamily Isoperlinae and most members of the subfamily Perlodinae were represented as distinct branches, with the exception of Pseudomegarcys, which was considered as a member of Arcynopterygini in Perlodinae. In the subfamily Perlodinae, the three genera of Arcynopterygini, the three genera of Perlodini, and the two genera of Diploperlini each formed monophyletic groups. However, it is noteworthy that Perlodes sp. nested within ­Filchneria, and was closely related with Filchneria zhouchangfai. The tribes Perlodini and Diploperlini formed sister groups, which in turn are together sister to Arcynopterygini. The final relationship within the Perlodinae subfamily is: (Perlodini + Diploperlini) + Arcynopterygini.

Within the Isoperlinae, Isoperla qinlinga is positioned at the base of the entire subfamily tree. The other three genera of Isoperlinae and Pseudomegarcys (Perlodinae) are contained within the clade of Isoperla. Additionally, Tibetisoperla from the Tibetan Plateau in China unexpectedly forms the sister group of I. bilineata from America, rather than of any Chinese (or even Asiatic) Isoperlinae taxa. The bootstrap value for this branch is relatively low (only 60), suggesting uncertainty regarding the robustness of this sister group relationship.

This study raises significant concerns regarding the accuracy of species identification related to the data for Pseudomegarcys and Perlodes. Morphologically, Pseudomegarcys Kohno, 1946 undoubtedly belongs to tribe Arcynopterygini of Perlodinae, and the etymology “Pseudo + Megarcys” suggests that the genus is most similar to its fellow Megarcys Klapálek, 1912. There is only one species, P. japonica Kohno, 1946, which occurs in Japan. However, the only mitochondrial genome data of P. japonica in NCBI (NC_038168.1) was sequenced by Wang et al. (2018) (materials without morphological figures presented), and this species has consistently appeared in Isoperlinae in previous molecular phylogenetic studies (South et al. 2021; Ding et al. 2019; Wang et al. 2022; Cao et al. 2022). Therefore, there is a high probability that the “P. japonica” in NCBI is misidentified.

The entry for Perlodes sp. in NCBI (MF197377.1) raises significant issues regarding species identification. Perlodes Banks, 1903 is the type genus of Perlodidae, primarily distributed from Europe to Northeast Asia, with only a few species found in southwest China (DeWalt et al. 2024). However, all Perlodes species in China were originally described based on the outline of the female subgenital plates, with poor descriptions and unknown critical features of their males (Wu 1938, 1962, 1973). Du (1999) and Yang and Li (2018) concluded that the taxonomic status of certain Perlodes species in China remains uncertain due to inadequate morphological data. Given that there is extensive intraspecific variation in the wing veins and subgenital plate of females (Lillehammer 1974; Huo et al. 2022a), these traits cannot be reliably used for genus-level identification. Huo et al. (2022c) further discussed the morphology of eggs from some related genera in the tribe Perlodini and concluded that the identity of all Perlodes species recorded in China is suspect .

Huo et al. (2022c) subsequently confirmed that Sino­perlodes is a synonym of Filchneria and found F. zhouchangfai in Zhejiang, Anhui, Fujian, and Guizhou in China. The first author examined this specimen and found its morphological characteristics to be consistent with those of Filchneria zhouchangfai (Fig. 3), rather than those of Perlodes. In this study, the branch of Perlodes sp. appeared between the two F. zhouchangfai from Anhui and Zhejiang in the phylogenetic tree, indicating that these three sequences correspond to the same species. Consequently, this indicates that there is currently no available mitochondrial genome data for Perlodes in the NCBI database.

In conclusion, our results generally support the monophyly of the two subfamilies of Perlodidae and the three tribes of the Perlodinae, which is consistent with the current taxonomic status of Perlodidae based on morphological taxonomy. Additionally, Perlodinae and Isoperlinae originated at the same time as they are sister taxa; the same is true for Diploperlini and Perlodini. However, the affiliation of Pseudomegarcys and the monophyly of the genera within Isoperlinae have not yet been verified.

Our phylogenetic tree presents the chronological order of differentiation within the Perlodidae: Perlodinae and Isoperlinae diverged from a common ancestor at the same time, while the split between Arcynopterygini and Perlodini + Diploperlini occurred after the split between the two subfamilies. This result is largely supported and consistent with the transcriptome-based analysis by South et al. (2021), reflecting the relative stability of the internal structure of this family. Although molecular data of many genera and species are still missing, the existing research has revealed problems in our understanding of their evolution, and offers new insights and hypotheses for their zoogeography and evolutionary outcomes.

At present, only eight genera of perlodids are distributed across the Holarctic realm (Fig. 4), with the most widely distributed being Arcynopteryx, Megarcys, ­Skwala (all Arcynopterygini) and Isoperla, while Diploperlini (Pictetiella, Kogotus) and Perlodini (Diura) have few Holarctic genera and species. It should be noted that only one species of Helopicus (Perlodini), H. infuscatus (Newman, 1838), is reported from Asia, but its generic placement is, in our opinion, doubtful. DeWalt et al. (2024) provided the original references of this species, indicating its type locality was Hong Kong, China. However, for more than a century, Chinese taxonomists have largely overlooked the existence of this species (Du 1999; Yang and Li 2018; Huo and Du 2023). At present, there are no specimen images or collecting reports to confirm the generic placement, valid status, or distribution of H. infuscatus.

In terms of longitudinal distribution, most holarctic genera of Perlodidae (except Isoperla) have been recorded in the western part of North America, with few or no occurrences in the eastern part. Fochetti and Tierno de Figueroa (2008) provided a summary of the four postulated origins of Plecoptera in North America from Hynes (1988): 1) Original distribution prior to the separation of North America / Europe / Asia; 2) Eastward spread within North America from the West to the East due to the formation of the Cordillera mountain range; 3) Origin of one genus from the South; 4) Very recent spread of some species via the Bering Strait. We tend to consider that some of the genera of Isoperlinae and Arcynopterygini could have diverged after the breakup of Laurasia but prior to the formation of the Bering land bridge. However, for the genera with true Holarctic distributions (West and East Palaearctic, Nearctic), e.g., Arcynopteryx, Diura and Isoperla, especially the circumboreal Arcynopteryx, their dispersal could be more complex and require the combination of independent evidence sources for further study. The remaining genera of this family are distributed across North America and Eurasia, which may indicate independent evolutionary events resulting from geographical isolation.

In regards to latitude, overall poor dispersal capability, their dependence on coldwater, and flow direction of the water systems (Zwick 2000) appear to be important factors in dispersal of Perlodidae from north to south across the Holarctic realm . They are rarely reported (Huo 2023; DeWalt et al. 2024) both in the southern Nearctic realm (southern United States to Central America) and the northern Oriental region (mainly southern China). The major Chinese rivers (Yangtze/Yellow/Pearl Rivers) and mountainous terrain likely hindered stonefly dispersal from high to low latitudes. Huo (2023) suggested that the boundary of Oriental-Palearctic region in China may be the differentiation center of Isoperlinae in East Asia. On the east of Tibetan Plateau and south of Qinling Mountains, reports of Perlodidae are fewer, while discoveries of new genera/species in Isoperlinae (Huo et al. 2020, 2022b; Huo and Du 2023; Cao et al. 2020) are more frequent (Fig. 5). However, the Nearctic realm is where Isoperlinae is most diverse, with most of its genera found only in North America. The results of the current study do not prove whether Isoperlinae in the Nearctic Region evolved independently after the separation of the American continent from Asia or if it is a relict group following an extinction event. Future work should aim to obtain more species sequences and conduct more detailed analysis of the phylogenetic relationships within other genera of this family.