Unraveling the Phylogenetic Relationships and Taxonomic Status of a Puzzling Freshwater Mussel Genus Inversidens (Bivalvia, Unionidae) through Multilocus Phylogeny and Mitochondrial Phylogenomics

,


Introduction
Freshwater mussels of the family Unionidae occur in freshwater habitats worldwide except Antarctica and South America, with the most diverse regions being East Asia and North America [1][2][3].These bivalves are ecologically significant benthic organisms that play pivotal roles in ecosystem functioning and services [4][5][6][7].Additionally, they also possess substantial economic value; for instance, soft body serves as a natural food source for poultry and livestock, while shell holds high ornamental and collection value [8].Regrettably, anthropogenic activities and environmental changes have posed severe threats to the habitats of freshwater mussels, resulting in a significant decline in their biodiversity over recent decades [9][10][11].In many countries, especially in North America and Europe, the number of conservation actions for these mussels has increased substantially over the latest years [12,13].China is a biodiversity hotspot for freshwater mussels, but limited understanding of their biology and lack of precise systematic classification hinder the implementation of conservation measures for certain taxa [14].
The freshwater mussel genus Inversidens Haas, 1911 was initially described as a subgenus of Nodularia Conrad, 1853, containing two species endemic to Japan, i.e., Unio brandtii Kobelt, 1879 and Nodularia parcedentata Haas, 1911 [15].Later, Thiele [16] elevated it to the generic level and designated U. brandtii as the type species, classifying it under the subfamily Unioninae in Unionidae.Haas [17] also placed Inversidens in Unioninae based on shell morphology and described six species within the genus.The subsequent efforts of malacologists focused on refining the classification on the genus level and assessing the species validity [18][19][20][21].
The classification of the upper genus for Inversidens based on shell morphology has been subject to uncertainty [26,27].With the recent advancement of molecular systematics, researchers have attempted to employ molecular fragments (COI, 16S rRNA, and 28S rRNA) from type species Inversidens brandtii for resolving phylogenetic positions [28][29][30][31].At present, the taxonomic status of this genus under the subfamily Gonideinae in Unionidae has been recognized [22,23].However, the classification position at the tribe level has not been solved (Figure 1).Based on threegene data, Lopes-Lima et al. [30] resolved the diversity, biogeography, and evolutionary relationships of the Far East Asian Unionidae and assigned Inversidens brandtii to the tribe Gonideini based on a phylogenetic context of five tribes within Gonideinae.Later, Dai et al. [32] expanded the taxonomic sampling and found that the results do not support the classification of Inversidens to Gonideini and considered Gonideini to be polyphyletic.Recently, a phylogenetic framework for eight tribes of Gonideinae proposed by Zieritz et al. [33] by incorporating new taxa and molecular markers (five genes) has been accepted.Therefore, the systematic status of the genus Inversidens in the context of the updated phylogeny of Gonideinae necessitates reassessment.
Additionally, the species Inversidens rentianensis Wu & Wu, 2021, which was discovered and described by our research team [25], only had limited molecular data (i.e., mitochondrial COI) and shell morphological description.This species has a limited distribution range and specific ecological requirements, exhibiting a preference for inhabiting smaller rivers and streams.The vulnerability of this distinctive habitat to both anthropogenic and natural influences makes the species vulnerable to threats [34,35].The progress of modern biodiversity and conservation research heavily relies on the phylogenetic information [36].Yet the lack of phylogenetic data for many imperiled species hinders our understanding of their basic biology and the implementation of conservation efforts [37][38][39].Thus, phylogenetic information on I. rentianensis must be gathered to ensure scientific assessment of the endangered status and development of conservation and management strategies.
To determine the phylogenetic position of Inversidens, we constructed a combined gene dataset (including COI-mtDNA, ND1-mtDNA, mt 16S-rRNA, 18S-rRNA, and 28S-rRNA) based on Zieritz et al.'s [33] comprehensive sampling of eight tribes within the Gonideinae and utilized various software programs with partitioning strategies to construct phylogenies.Moreover, mitochondrial genomes are known for providing significantly more informative characters for phylogenetics and have proven highly effective in resolving both shallow and deep relationships in freshwater mussels [40][41][42][43].In order to further validate the systematic position of Inversidens obtained from the five-gene dataset, we also conducted phylogenetic analyses using mitochondrial genomes.Therefore, a new mitogenome from Inversidens rentianensis was acquired.
Herein, the purpose of this study is to (1) describe the anatomical characteristics of Inversidens rentianensis and decipher its female complete mitochondrial genome, (2) elucidate the phylogenetic position of Inversidens based on both multilocus dataset and mitogenomic dataset, and (3) infer the evolutionary history of the genus Inversidens by timecalibrated phylogeny of Unionoidea.

Sample Collection and Anatomical Feature Observation.
From 2021 to 2022, we conducted multiple field expeditions to the type locality of Mianshui River in Ganzhou City, Jiangxi Province, China, for collecting Inversidens rentianensis.However, it was not until October 2022 that we fortuitously obtained two live specimens (Figure 2).Both voucher specimens were deposited at the Museum of Zoology, Shanxi Normal University (SXNU), China (SXNU22102201-SXNU22102202).Anatomical features of the soft body were observed by visual examination and stereoscopic microscopy.
Genomic DNA from samples was extracted from dissected foot tissue using the TIANamp Marine Animals DNA Kit (Tiangen Biotech, Beijing, China) according to the manufacturer's instructions.Five gene regions (COI, ND1, 16S rRNA, 18S rRNA, and 28S rRNA) were amplified and sequenced using the same primers from Zieritz et al. [33].PCR conditions were performed according to the TaKaRa Ex manufacturer's protocol, including an initial denaturation step at 98 °C for 10 s, followed by 35 cycles of amplification consisting of a denaturation step at 94 °C for 1 min, annealing at 50 °C for 1 min, and extension at 72 °C for 1 min.The final extension was performed at 72 °C for seven minutes.Amplified PCR products were purified and sequenced by Sangon Biotech (Shanghai).
Based on the dataset provided by Zieritz et al. [33], we expanded the taxa, resulting in a five-gene dataset comprising 35 Gonideinae taxa, with Margaritifera dahurica and Margaritifera margaritifera included as outgroup.The sequences of certain gene fragments from Postolata guangxiensis, Ptychorhynchus pfisteri, Koreosolenaia sitgyensis, and Parvasolenaia rivularis are newly obtained in this study (Table 1).
To further confirm the results of the multilocus datasets and clarify the phylogenetic relationship between Inversidens and other close relative, we sequenced the mitochondrial genome of Inversidens rentianensis.Qualified samples were sent to Novogene Co., Ltd.(China) for library construction and sequencing.The complete mitogenome was concatenated by Geneious (ver.11.0, see https://www .geneious.com)[44].Protein-coding genes were confirmed by the NCBI ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder/) and BLAST search (http://blast.ncbi.nlm.http:// nih.gov/) analysis.Subsequently, the mitogenome genomic sequence was submitted to GenBank under accession number OR823224.
Sequence length, base composition, and AT content were calculated using the built-in EditSeq program in DNAstar.Strand asymmetry was determined by calculating the GC skew = G% − C% / G% + C% and AT skew = A% − T% / A% + T% .The mitochondrial genome was visualized using GenomeVx (GenomeVx (ucd.ie))[46] with manual modifications.
We combined the new genome with 92 Unionida mitogenomes from GenBank to construct the complete mitogenome dataset (Table 2).The nucleotide sequences of 12 protein-coding genes (12 PCGs, excluding ATP8 due to high sequence variation) and two rRNA genes were used for subsequent phylogenetic analyses.

Phylogenetic Analyses.
The molecular phylogenetic reconstruction is consistent with the methods used in our previous studies [47].In the datasets constructed above, protein-coding genes (PCGs) were aligned by built-in MUS-CLE [48] with default settings implemented in PhyloSuite [49].rRNA genes were aligned using MAFFT v7.2 [50] with the L-INS-i algorithm.Ambiguous alignment areas were trimmed by Gblocks (ver.0.91b; see http://molevol.cmima.csic.es/castresana/Gblocks.html) [51], the parameter ribosomal gene block with a minimum length was set to 2 base pairs (bp), and allowed gap position was selected with half; the minimum length of protein-coding gene block was set to 3 bp, and allowed gap position was also selected with half.For the five-gene dataset, the COI, ND1, 16S, 18S, and 28S sequences were aligned and trimmed to lengths of 657 bp, 882 bp, 469 bp, 642 bp, and 743 bp, respectively.The sequences of various multigene datasets were concatenated using PhyloSuite v1.2.3.The mitogenomic data consisted of concatenated twelve protein-coding genes (12 PCGs) and two ribosomal RNA genes with a total length of 12,751 bp.
All built datasets were performed with partition schemes based on genes and codons.PartitionFinder (ver.2.1.1;see http://www.robertlanfear.com/partitionfinder/)[52] was used to select Bayesian inference (BI) and BEAST analysis models for partitioning schemes.ModelFinder (ver.1.4.2;see http://www.iqtree.org/ModelFinder/)[53] was used to select the maximum likelihood (ML) analysis models in IQ-TREE.The selection for best-fit models was based on    ML analysis was performed in IQ-TREE [54] based on generated models in ModelFinder, using 1000 ultrafast bootstraps.BI analysis was used in MrBayes v2.01 [55] with generated models in PartitionFinder.Four independent Markov chain Monte Carlo (MCMC) were run simultaneously for ten million generations, and sampling was conducted every 1000 generations, with a burn-in of 25%.We terminated the process when the average standard deviation of splitting frequency falls below 0.01.The visualization and editing of phylogenetic trees were performed by using iTOL online software (http://itol.embl.de/itol.cgi)[56].

Divergence Time Estimation.
The time-calibrated phylogenetic tree was generated using BEAST v1.8.4 [57] based on the concatenated five-gene dataset.We employed the uncorrelated lognormal clock model with a priori model of birth-death speciation process for tree construction.Three reliable fossil calibrations were selected according to our previous study [58]: (1) the minimum stem age of Margaritiferidae was set to 230 Mya (exponential prior, lambda = 30), (2) the minimum age of Margaritifera dahurica and Margaritifera margaritifera for the most recent common ancestor (MRCA) was set to 34 Mya (exponential prior, lambda = 9 3), and (3) the minimum age of the most recent common ancestor (MRCA) of Lamprotula leaii and Lamprotula caveata was also set to 34 Mya (exponential prior, lambda = 9 3).
Four independent MCMC ran with 300 million generations each with sampling every 50,000 generations.Convergence and effective sample size (ESS) of parameters were checked using Tracer 1.7.1 [59].LogCombiner v1.8.4 [57] was used to combine trees from these four runs with discarding the first 25% generations.A maximum clade credibility tree was summarized in TreeAnnotator v1.8.4 [57], with a burn-in of 25%.

Soft-Body Anatomical
Features.The fresh soft-body morphology of Inversidens rentianensis is depicted in Figure 3.The color of gills is milky white, and the inner gill is larger than the outer one (Figure 3(a)).The foot is butter yellow and darker than the gills (Figure 3(a)).The papillae of incurrent aperture are weakly developed, arranged in two rows, and almost stick together (Figure 3(b)); no papillae are observed in the excurrent aperture (Figure 3(c)).The pigmentation on both the incurrent and excurrent aperture is remarkable.The anal opening, located on the dorsal margin of the posterior adductor muscle, is connected to the excurrent aperture (Figure 3(d)).Labial palps are milky white and subtriangular (Figure 3(e)).

Mitochondrial Genome Structure. The mitogenome of
Inversidens rentianensis is 15,987 bp in length (Figure 4).A +T bias has been observed in many unionid mitogenomes [60], and I. rentianensis genome is no exception: AT content of 60.91% (38.48% A, 22.42% T, 11.35% G, and 27.75% C).The AT skew and GC skew are 0.26% and -0.42%, respectively, indicating a bias towards A over T and C over G.
The 13 PCGs have a total length of 10,995 bp, which accounts for 68.77% of the entire mitochondrial genome.The base composition is A, 21.44%; T, 38.73%; G, 24.09%; and C, 15.74%, with an AT content of 60.17%.The mitochondrial genome has four start codons: ATA, ATT, ATG, and TTG.Among 13 PCGs, cox1 starts with TTG while the other genes use standard codons ATN (ATA, ATG, and ATT).The mitogenome contains 22 tRNA genes, including two trnS and two trnL.Most of tRNA genes are located on the light strand, except for the trnH and trnD (Figure 4).The length of tRNA genes varies from 52 bp (trnA) to 68 bp (trnS1 and trnN).Both rrnS and rrnL are encoded on the light strand and are separated by trnK, trnT, and trnY (Figure 4), as in all the unionoid mt genomes studied so far [60,63].

Multilocus Phylogenetic Analyses.
Based on the mtDNA gene dataset, the phylogenetic trees obtained from Bayesian inference (BI) and maximum likelihood (ML) analyses show that all eight tribes in the subfamily Gonideinae are monophyletic groups, although most are weakly supported (BS < 70%, PP < 0 9).The sisterly group between Inversidens brandtii and Inversidens rentianensis is closely related to the tribe Lamprotulini, but with low support values (BS = 60% and PP = 0 62, Figure 5(a)).The BI and ML trees based on the nuDNA dataset do not form monophyly for all eight   tribes, while taxa of Inversidens independently form distinct clades that do not belong to any tribe (Figures 5(b) and 5(c)).
Both BI and ML phylogenetic analyses consistently support a sister group relationship between Inversidens brandtii and Inversidens rentianensis, with strong support values (BS = 100%, PP = 1 0; Figure 6).Furthermore, these two species are placed in a basal clade within the tribe Gonideini, as strongly supported by maximum likelihood bootstrap (BS) values of 97% and Bayesian posterior probabilities (PP) of 1.0 (Figure 6).

Fossil Calibrations and Molecular
Dating.The tree topology obtained using the AICc partitioning scheme in BEAST does not align with our BI and ML phylogeny hypothesis at the tribe level; however, it consistently exhibits a phylogenetic relationship among taxa of Gonideini (Figures 6 and 7).
The genus Inversidens originates in the mid-Cretaceous (102.73Mya, 95% HPD = 72 22-137 03 Mya).The divergence time between I. rentianensis and I. brandtii is estimated to be 37.92 Mya (95% HPD = 20 39-60 59 Mya), which occurs during the Late Paleogene (Figure 7).The phylogenetic results support the recognized four monophyletic groups of Unionidae, i.e., (Ambleminae+(Goni-deinae+(Unioninae+Parreysiinae))) (Figure 8).Focusing on the subfamily Gonideinae, it is divided into six robust monophyletic clades, corresponding to six tribes with the following relationships: (Chamberlainiini+((Rectidentini+Contradentini)+(Lamprotulini+(Gonideini+Pseudodontini)))) by high support values (BS ≥ 99%, PP = 1 0; Figure 8).In both trees, Inversidens rentianensis is positioned as a basal clade within the tribe Gonideini with strong support values (BS = 100% and PP = 1 0; Figure 8).[33,[64][65][66][67]. Lopes-Lima et al. [30] classified Inversidens in Gonideini under the phylogenetic framework of the five tribes of Gonideinae based on the (COI+28S) dataset.However, Dai et al.'s [32] phylogenetic trees inferred from the three-gene (COI+16S+28S) dataset did not support this classification and suggested that "Inversidens was not within any tribe of Gonideinae."Recently, Zieritz et al. [33] established and recognized a taxonomic system for eight tribes within Gonideinae based on five-gene phylogeny (COI+ND1+16S+18S+28S).Under this updated phylogenetic framework, we attempt to utilize multiple molecular datasets to address the phylogenetic location of Inversidens.The nuDNA trees were unable to effectively resolve the tribal relationship (Figures 5(b) and 5(c)), suggesting that nuclear genes 18S and 28S lack reliability in analyzing relationships at the tribe and lower levels within unionids.The phylogenetic relationships of all eight tribes were consistent between the mtDNA tree and the combined five-gene tree (Figures 5(a systematics further corroborate its taxonomic status (Figure 8, BS/PP = 100%/1 0).Accurate systematics is essential for comprehending the evolutionary lineage of organisms.The previous divergence time estimation for the tribe Gonideini was 63.7 Mya, with Ptychorhynchus pfisteri as the basal clade [41].In this study, the basal clade group of Gonideini was determined to be Inversidens.By reconstructing the species differentiation time, we revised the divergence time for Gonideini to be 102.73Mya (95% HPD = 72 22-137 03 Mya) (Figure 7).Despite we speculate that Inversidens rentianensis and Inversidens brandtii, which are endemic to China and Japan, respectively, originated in the Late Paleogene, their historical biogeography remains enigmatic due to a lack of comprehensive understanding on species diversity and distribution for this group.The recent discovery of new species in Southwest China [32,68,69] underscores the imperative for a more comprehensive examination of the overall diversity and distribution of freshwater mussels in other inadequately surveyed regions across China.

Discussion
The accuracy of the phylogenetic framework is influenced by insufficient information loci and inadequate taxon selection [70,71].The phylogenetic reconstruction of eight tribes in Gonideinae that was conducted by Zieritz et al. [33] only included three taxa from the species-rich tribes Gonideini and Lamprotulini.The present study incorporated additional sampling taxa from Gonideini and Lamprotulini, which predictably revealed discrepancies in the phylogenetic relationships at the tribe level.The mitochondrial genomics in this study covered six out of the eight tribes within the subfamily Gonideinae, excluding Schepmaniini and Ctenodesmini due to unavailability of mitochondrial genome data.We found strong support for the monophyly of six tribes within Gonideinae, which is consistent with recent phylogenomic analyses utilizing anchored hybrid enrichment data [66] and mitogenomic data [32,41,63].In this study, a novel phylogenetic hypothesis for the eight tribes of Gonideinae was proposed based on the five-locus dataset; however, further verification is required due to the low nodal support.To reconstruct and validate       IUCN Red List, while the third species, I. rentianensis, remains unevaluated due to its recent discovery and description [25,72,73].I. brandtii, an endemic species in Japan, was assessed as vulnerable based on extensive research conducted by Kondo [21] on its morphology, ecology, and reproductive biology.On the contrary, the understanding of the endemic species I. pantoensis and I. rentianensis in China is obviously insufficient.The knowledge about I. pantoensis is currently limited to its morphological shell descrip-tion, which poses significant challenges in determining its validity due to the considerable shell morphological plasticity [74][75][76].It is regrettable that the taxonomic status of this particular species has not been determined yet due to difficulties in collecting specimens for molecular examination.The species I. rentianensis is restricted to a narrow geographic range and primarily inhabits creeks and small streams characterized by substrata ranging from gravelly to sandy muddy compositions.Despite multiple visits to the  13 Journal of Zoological Systematics and Evolutionary Research type locality following the publication of this new species, only two live specimens were successfully collected.Through careful examination of these invaluable specimens, our understanding of the anatomy and molecular systematics of I. rentianensis has been significantly enriched.Since no brooding individuals were collected in this study, we could not determine the exact brooding period and examine the glochidia of this species.The accuracy of phylogenies is crucial for comprehending the evolutionary histories, reproductive traits, and ecological habits of organisms.The breeding period of the closely related species I. brandtii is from May to September [18].The specimens of I. rentianensis were collected in October, and the gill exhibited characteristics indicative of postegg release.Therefore, we speculated that the breeding period of I. rentianensis likely occurs during August and September.Recently, Liu et al. [14] established a new conservation priority method for freshwater mussels, i.e., Quantitative Assessment of Species for Conservation (QASCP).According to this conservation assessment method, our preliminary evaluation ranks I. rentianensis as the second priority.
The impacts of urbanization and human activities have left river habitats containing endemic freshwater mussel species vulnerable to threats and require urgent attention and protection [38,77].Fortunately, from 2020, Chinese government has implemented comprehensive watershed protection measures (e.g., a decade-long fishing prohibition in crucial sections of the Yangtze River) and legislation (e.g., Yangtze River Protection Law) to effectively restore aquatic ecosystems, yielding remarkable outcomes [78,79].The recent discovery of new species, including Inversidens rentianensis, highlights the crucial role that small rivers and tributaries play as sanctuaries for freshwater mussels.However, the lack of comprehensive biodiversity surveys in these tributary basins hampers our understanding of aquatic biodiversity and the factors posing threats to it.Furthermore, the conservation efforts for unionid fauna are significantly hindered by a lack of knowledge regarding their distribution, population  dynamics, and accurate taxonomic information pertaining to freshwater mussels.Therefore, we advocate researchers to pay more attention to the species diversity and population distribution of freshwater mussels in small tributaries.Simultaneously, we urge the government to formulate scientific protection measures aimed at safeguarding other endemic or endangered mussel species inhabiting small rivers and streams, in addition to those residing in larger rivers.

Conclusion
This study provides anatomical features and mitogenome data for Inversidens rentianensis.

Figure 1 :
Figure 1: Summary of recent multilocus phylogenetic hypotheses at tribe level of Gonideinae and the position of Inversidens.

3. 5 .
Mitochondrial Phylogenomic Analyses.ML and BI trees based on the mitogenome dataset yield nearly identical topologies and are statistically well supported by 100% maximum likelihood bootstrap (BS) support values and Bayesian posterior probabilities (PP) in most nodes.Some discrepancies are observed only in specific branches, i.e., Anodonta nuttalliana, Anodonta anatina, and Pseudanodonta complanata (Figure S1-S2).

4. 1 .
Systematic Position of Inversidens from Multilocus and Mitogenomic Data.The classification of Gonideinae has undergone consistent revisions in recent years by the incorporation of additional taxa

Figure 4 :
Figure 4: Gene maps of the newly generated mitochondrial genome of Inversidens rentianensis.Genes positioned inside the circle are encoded on the heavy strand, and genes outside the circle are encoded on the light strand.Color codes: small and large ribosomal RNAs (red), transfer RNAs (purple), and PCGs (green).

4. 2 .
Threats and Conservation Implications.The genus Inversidens currently recognizes three species, with both I. pantoensis and I. brandtii classified as data deficient on the

Figure 5 :
Figure 5: Phylogenetic trees inferred from Bayesian inference (BI) and maximum likelihood (ML) analyses.(a) Phylogram of BI and ML tree based on the mtDNA gene dataset (COI+ND1+16S rRNA).Numbers at nodes are support values for ML (bootstrap support (BS))/ BI (posterior probability (PP)).(b) Phylogram of BI tree based on the nuDNA gene dataset (18S rRNA+28S rRNA).Numbers close to the nodes indicate for PP values.(c) Phylogram of ML tree based on nuDNA gene dataset (18S rRNA+28S rRNA).Numbers close to the nodes indicate BS values.Both Inversidens species are shown in red font.Colored clades represent eight tribes of the subfamily Gonideinae.

Figure 6 :
Figure 6: Maximum likelihood (ML) and Bayesian inference (BI) trees of the Gonideinae subfamily based on the five-gene dataset (COI +ND1+16S rRNA+18S rRNA+28S rRNA).BI tree generated by Bayesian analysis and ML tree generated by IQ-TREE analysis produce completely congruent topology.Numbers at nodes are statistical support values for ML (bootstrap support (BS))/BI (posterior probability (PP)).Color-coded clades are eight tribes within Gonideinae.

Figure 7 :
Figure 7: Fossil-calibrated phylogeny based on the five-gene dataset.Node bars represent the mean ages of 95% highest posterior density (HPD).Black numbers near nodes are mean ages (Mya).Fossil calibrations are marked by star signs.Mya: million years ago; Q: Quaternary; N: Neogene.

Figure 8 :
Figure 8: Phylogenetic trees based on mitogenome dataset inferred from Bayesian inference (BI) and maximum likelihood (ML) analyses.(a) BI tree.(b) ML tree.Numbers close to the nodes for (a) indicate posterior probabilities (PP), and those for (b) indicate bootstrap support (BS).Colored backgrounds and colored clades represent six tribes of the subfamily Gonideinae.Red font indicates the species sequence from this study.Here, outgroups are not shown, and both Unioninae and Ambleminae taxa are collapsed in the phylogenetic tree, but complete trees can be found in supplementary Figure 1 and 2.

Table 1 :
List of sequences used in multilocus phylogenetic analyses.

Table 2 :
List of sequences used in mitochondrial phylogenomic analyses.

9
Journal of Zoological Systematics and Evolutionary Research The mitochondrial phylogenomics and five-gene phylogeny clearly show that Inversidens rentianensis and Inversidens brandtii have a wellsupported sister group relationship and form a basal clade in the tribe Gonideini, which should be classified to Gonideini in the subfamily Gonideinae.Molecular dating analysis suggests that Inversidens originated in the mid-Cretaceous era (102.73Mya, 95% HPD = 72 22-137 03 Mya) and diversified during the Late Paleogene era (37.92 Mya, 95% HPD = 20 39-60 59 Mya).Additionally, preliminary assessment ranks Chinese endemic species Inversidens rentianensis as second priority based on the QASCP, which aids in managing and conserving this rare group.Taken together, this study shed light on the systematic position of Inversidens, while enriching information on the biology of Inversidens rentianensis, an endemic species to China.These findings pave the way for further comprehensive studies on the ecology, reproductive behavior patterns, and conservation biology of this fascinating group.