Mitochondrial genomes of Dactylogyrus wunderi (Monopisthocotyla: Dactylogyridae): structural features, codon usage patterns, and phylogenetic implications Authors: Yuan-Yuan Yang, Cheng Yue, Li Zhang, Shen-Zhen Xia, Yan-Jun Liu, Nian-Wen Wei, Cui-Lan Hao Published: 2026 ABSTRACT AbstractBackgroundCodon usage bias (CUB) is a common phenomenon reported among many species and genes, but its unique characteristics in the mitochondrial genome of class Monopisthocotyla remain unknown.MethodsThe complete mitochondrial genome of Dactylogyrus wunderi was sequenced and characterized, and the mitochondrial genome compositions and CUB of six Dactylogyrus species and 35 Monopisthocotyla species were analyzed using bioinformatics methods.ResultsThe mitochondrial genome of D. wunderi is a typical circular structure in length of 14,920 bp. The A + T contents of the six Dactylogyrus species (58.4% ± 5.7%) were significantly lower than that of Monopisthocotyla species (71.0% ± 5.80%, p < 0.01). Neutrality plot analysis showed slopes of 0.3136 and 0.389 in the six Dactylogyrus and the 35 Monopisthocotyla species, respectively. Furthermore, 98.3% and 77.4% of the genes in the six Dactylogyrus and the Monopisthocotyla species, respectively, had effective number of codons (ENC) higher than 35, but 23.3% and 0.5% genes of ENC ratio ranged from − 0.05 to 0.05 in the six Dactylogyrus and Monopisthocotyla species. Phylogenetic analysis revealed that, within the context of the sampled taxa, the families of Monopisthocotyla were monophyletic groups, except for Ancyrocephalidae.ConclusionsThe nucleotide composition had AT base bias in Monopisthocotyla, and natural selection was the main factor affecting CUB in the mitochondrial genomes of Monopisthocotyla species. These results provided insights into the factors affecting CUB in Monopisthocotyla species and deepened our insight of phylogeny, evolution, and codon usage of Monopisthocotyla.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12864-026-13106-x.Keywords: Dactylogyrus wunderi, Mitochondrial genome, Monopisthocotyla, Base composition, Codon usage bias, Phylogeny Background Monopisthocotyla Brabec et al. (2023) (previously belonged to class Monogenea), a class within Platyhelminthes, most members exhibit direct development and high host specificity, and parasitize freshwater fishes and marine fishes [1]. The Monopisthocotyla mainly inhabits the fins, skin or gills of fishes, and often causes diseases and death in farmed fish populations globally. When Monopisthocotyla parasitizes their host in large numbers, it can result in substantial economic losses [2, 3]. The classification and evolution in Monopisthocotyla have always been the hot research topics [4, 5]. The first recognition of monogeneans as a group went back to the Belgian scientist Pierre-Joseph Van Beneden in 1858, and the first species of monogenean was discovered and described in 1776 by Otto Friedrich Müller [6, 7]. Monogeneses were classified as a group within class Trematoda in early taxonomic systems [8]. Bychowsky (1937) first classified the Monogenea as a distinct class, separating it from the class Trematoda, and was divided it into two subclasses: Polyonchoinea and Oligonchoinea [9]. Lebedev (1988) established a new subclasse, Polystomatoinea, by isolating the monogeneans primarily parasitizing the bladder of amphibians and reptiles from the subclasse Oligonchoinea [10]. In addition, Brabec et al. (2023) revealed that Monogenea is not monophyletic by transcriptomic analyses, and suggested the two former subclasses (Monopisthocotylea and Polyopisthocotylea) shall be independent class ranks: Monopisthocotyla and Polyopisthocotyla [11]. This revised taxonomic framework has been adopted for several subsequent investigations. Vanhove (2024) indicated that ‘Monogenea’ is a non-monophyletic group [7]; Yamada P O F et al. (2025) used ‘Monopisthocotyla’ as a class in the integrative taxonomy study of genus Tereancistrum in freshwater fishes from the Caatinga Domain of Brazil [12]. The classification problem also exist contentious within in the family of the previous class Monogenea, Bychowsky and Nagibina (1978) elevated the subfamily Ancyrocephalinae to the family rank, and thereby established the family Ancyrocephalidae in order Dactylogyridea [13, 14]. Kritsky & Boeger (1989) demonstrated that the family Ancyrocephalidae contains both para- and polyphyletic components and is paraphyletic, they proposed a preferred conservative revision: rejecting Ancyrocephalidae as a valid family and returning the taxon to subfamily status (Ancyrocephalinae) within the Dactylogyridae [15]. Zhang et al. (2018) recovered a sister-group relationship between the subfamilies Dactylogyrinae and Ancyrocephalinae by 36 mitogenome sequences [16], Hao et al. (2023) revealed that Dactylogyridae is nested within a paraphyletic Ancyrocephalidae based on two mitogenomic datasets from 39 monopisthocotylean and 67 species covering key related families [17]. The results further adds support to the revision of Kritsky & Boeger [15]. The mitochondrial genome serves as a powerful tool for inferring the phylogenetic relationships among species, owing to its advantageous characteristics, such as maternal inheritance, minimal genetic recombination, and diverse molecular markers (e.g., cox1) [18]. Moreover, the inheritance of the mitogenome is independent of the nuclear genome, and has unique evolution rate. Consequently, mitogenome sequences are extensively utilized in species identification, phylogeography, and various aspects of molecular evolutionary studies [19, 20]. Nikol Kmentová et al. (2024, 2026) revealed the population genetics and spatial dynamics of Monopisthocotyla, and highlighting the importance of high-resolution molecular markers and host ecological factors [21, 22]. The codon usage bias (CUB) of mitogenome serves as a relevant indicator of molecular evolution in phylogenetic studies. CUB refers to the non-random preferential usage of synonymous codons within a gene or genome [23] and is closely associated with evolutionary forces such as mutation pressure and natural selection [24]. The analysis of CUB not only reflects molecular evolutionary patterns [24] but also elucidates evolutionary dynamics across species, providing supportive and complementary evidence for inferring phylogenetic relationships, particularly among closely related taxa [25]. The relative contribution of natural selection in shaping gene or genome evolution can be assessed by quantifying CUB using metrics such as the effective number of codons (ENC) and relative synonymous codon usage (RSCU). Consequently, CUB offers valuable insights into the genetics and molecular evolution of genes [26]. So far, more than 5,000 species of Monopisthocotyla have been described in the world, however, available mitochondrial genome data for Monopisthocotyla remain very limited. As of May 2025, the mitogenomes of 72 Monopisthocotyla species (including 177 sequences) have been public accessible in GenBank, representing less than 2% of the total Monopisthocotyla diversity. Generally, the mitogenomes of Platyhelminthes comprise 36 genes, including 12 protein-coding genes (PCGs), 2 ribosomal RNA (rRNA) genes and 22 transfer RNA (tRNA) genes. Gene atp8 is typically absent in the mitogenomes of Platyhelminthes. In Platyhelminthes, Gene cytb in Platyhelminthes reportedly exhibits a stronger CUB [27], which is primarily influenced by evolutionary factors, including natural selection as a dominant role and mutation pressure acting as a secondary factor. Additionally, nucleotide compositional features such as AT content and AT skew contribute to variations in CUB of gene cytb across different classes within the phylum Platyhelminthes [27]. Mazumder et al. (2018) indicated a preference for A/T-ending codons in gene atp6 [28], as well as Mazumder indicated that high AT content in ND genes of Platyhelminthes [29]. However, the extent of CUB varies among different ND genes and across Platyhelminthes classes, due to differential impacts of natural selection and mutation pressure. Hao et al. (2023) indicated a distinct CUB pattern has been documented in the genus Dactylogyrus, notably, the RSCUs in Dactylogyrus are generally lower than in other Monopisthocotyla species [30]. While a strong bias toward AT-ending codons is observed in the mitogenomes of about 82% of Monopisthocotyla species, this trend is not followed in four Dactylogyrus species (Dactylogyrus simplex, D. tuba, D. crucifer, and D. zandti). Furthermore, comparative analyses reveal distinct codon usage patterns in Dactylogyrus mitogenomes relative to other Monopisthocotyla. These results suggest that mitochondrial CUB may represent a genus-specific genomic characteristic within this taxonomic group. Dactylogyrus wunderi Bychowsky, 1931 parasitizes the gills of Abramis brama orientalis (infection prevalence > 55% in the Irtysh River) [31]. D. wunderi exhibiting high host specificity and geographical distribution is a model species well-suited for investigating adaptive evolution. In this study, we sequenced the mitogenome of D. wunderi and analyzed its gene content, structural, codon usage, and nucleotide composition. We compared D. wunderi mitogenome with 40 other Dactylogyrus species to determine the nucleotide composition, parity rule 2 bias, ENC and evolutionary rate in the class Monopisthocotyla. Moreover, codon usage patterns were clarified by grouping species into their natural evolutionary lineages (i.e., monophyletic families) based on phylogenetic relationships. Methods Mitochondrial genome sequencing, assembly, annotation The total DNA of Dactylogyrus wunderi was amplified using a Discover-sc single cell WGA kit (Nanjing Vazyme Biotech., China). Mitogenome sequences were obtained on Illumina NovaSeq 6000 system, and assembled and spliced on DNAstar 7.1 (Nanjing Genepioneer Biotech., China). Each gene in the mitogenome of D. wunderi was annotated on basis of genetic code 9 (Echinoderm and Flatworm Mitochondrial Code) and using MITOS2 web server (https://mitos2.bioinf.uni-leipzig.de/index.py). Then the uncertain genes in the mitogenome of D. wunderi were aligned by choosing and referring to 3 congeneric mitogenomes: D. lamellatus, D. crucifer and D. zandti [16, 30]. Data retrieval The complete mitogenome sequences of 38 Monopisthocotyla species were retrieved from the NCBI database (www.ncbi.nlm.nih.gov/) as of May 2025. The selection criteria required that the sequences of Monopisthocotyla species should be publicly available and fully annotated. and their relevant information was summarized in Table 1. A final set of 41 species was selected for analysis (3 in-house + 38 retrieved), these 41 species were divided into six Dactylogyrus and 35 other Monopisthocotyla. The study included three species of Dactylogyrus - D. wunderi Bychowsky, 1931 (GenBank: PV847579), D. crucifer Wagener, 1857 (GenBank: OR466407) and D. zandti Bychowsky, 1933 (GenBank: PP779068) sequenced by the College of Veterinary Medicine, Xinjiang Agricultural University. Complete sets of 12 PCGs were annotated in all species except for Ancyrocephalus mogurndae, Capsaloides cristatus and Cichlidogyrus halli, which had incomplete PCG annotations. Table 1GenBank accession number, species, length and reference of mitogenomesIDOrganismorderFamilylength (bp)Reference1NC_014591.1 Benedenia hoshinai CapsalideaCapsalidae13,554[32]2NC_014291.1 Benedenia seriolae CapsalideaCapsalidae13,498[1]3MN746369.1 Capsaloides cristatus CapsalideaCapsalidae13,948Unpublisheda4NC_062638.1 Capsala katsuwoni CapsalideaCapsalidae13,265Unpublisheda5NC_063093.1 Capsala martinieri CapsalideaCapsalidae13,984Unpublisheda6NC_047185.1 Capsala pricei CapsalideaCapsalidae13,851Unpublisheda7NC_082221.1 Megalobenedenia derzhavini CapsalideaCapsalidae14,235Unpublisheda8JQ038228.1 Neobenedenia melleni CapsalideaCapsalidae13,270Unpublisheda9NC_073001.1 Tristoma integrum CapsalideaCapsalidae13,968Unpublisheda10MN095192.1 Ancyrocephalus mogurndae DactylogyrideaAncyrocephalidae14,760[33]11MZ703276.1 Cichlidogyrus casuarinus DactylogyrideaAncyrocephalidae15,575[34]12MG970255.1 Cichlidogyrus halli DactylogyrideaAncyrocephalidae15,047[35]13JQ038226.1 Cichlidogyrus sclerosus DactylogyrideaAncyrocephalidae15,052Unpublisheda14OR466407 Dactylogyrus crucifer DactylogyrideaDactylogyridae14,403[30]15NC_035610.1 Dactylogyrus lamellatus DactylogyrideaDactylogyridae15,187[16]16OP058752.1 Dactylogyrus simplex DactylogyrideaDactylogyridae14,648[17]17OP058753.1 Dactylogyrus tuba DactylogyrideaDactylogyridae15,406[17]18PV847579 Dactylogyrus wunderi DactylogyrideaDactylogyridae14,920This study19PP779068 Dactylogyrus zandti DactylogyrideaDactylogyridae18,584[30]20MH700477.1 Euryhaliotrema johni DactylogyrideaAncyrocephalidae14,132[36]21NC_048529.1 Enterogyrus malmbergi DactylogyrideaAncyrocephalidae14,107[33]22NC_039617.1 Lepidotrema longipenis DactylogyrideaDiplectanidae15,433Unpublisheda23MH328204.1 Lamellodiscus spari DactylogyrideaDiplectanidae14,614Unpublisheda24NC_036305.1 Paratetraonchoides inermis DactylogyrideaTetraonchoididae14,654[37]25MW565922.1 Rhabdosynochus viridisi DactylogyrideaDiplectanidae13,863[38]26NC_056186.1 Scutogyrus longicornis DactylogyrideaAncyrocephalidae14,241[39]27NC_053548.1 Thaparocleidus asoti DactylogyrideaAncylodiscoididae16,074[40]28NC_018031.1 Tetrancistrum nebulosi DactylogyrideaAncyrocephalidae13,392[41]29NC_053547.1 Thaparocleidus varicus DactylogyrideaAncylodiscoididae14,088[42]30NC_030339.1 Aglaiogyrodactylus forficulatus GyrodactylideaOogyrodactylidae14,371[43]31NC_031337.1 Gyrodactylus brachymystacis GyrodactylideaGyrodactylidae14,767[44]32NC_010976.1 Gyrodactylus derjavinoides GyrodactylideaGyrodactylidae14,741[45]33NC_041379.1 Gyrodactylus gurleyi GyrodactylideaGyrodactylidae14,771[46]34NC_030050.1 Gyrodactylus kobayashii GyrodactylideaGyrodactylidae14,786[47]35NC_046757.1 Tetraonchus monenteron GyrodactylideaTetraonchidae14,791[48]36NC_038214.1 Gyrodactylus nyanzae GyrodactylideaGyrodactylidae14,885[35]37NC_031438.1 Gyrodactylus parvae GyrodactylideaGyrodactylidae14,702[44]38NC_008815.1 Gyrodactylus salaris GyrodactylideaGyrodactylidae14,790[49]39KP780991.1 Gyrodactylus sp. FY-2015 GyrodactylideaGyrodactylidae14,755[44]40MW464989.1 Gyrodactylus sp. FZ-2021 GyrodactylideaGyrodactylidae18,165Unpublisheda41NC_024754.1 Paragyrodactylus variegatus GyrodactylideaGyrodactylidae14,517[50]42NC_061193 Eudiplozoon nipponicum MazocraeideaDiplozoidae17,328Out group43NC_062047 Paradiplozoon yunnanesis MazocraeideaDiplozoidae14,829Out groupathe sequences listed above as “unpublished” are publicly available in the NCBI GenBank accession number, species, length and reference of mitogenomes athe sequences listed above as “unpublished” are publicly available in the NCBI Comparative analysis of mitogenome data The base composition and RSCUs of 41 mitogenome in Monopisthocotyla species were analysed on PhyloSuite v2 [51, 52]. RSCU less than 1 indicates the codon is employed at a lower frequency than anticipated, whereas an RSCU higher than 1 indicated the codon is employed more frequently than anticipated [53]. The RSCU clustering heat map was generated on Origin2022. GC12, GC3, A3, T3, C3, and ENCs were quantified on CodonW 1.4.2 [54]. The p values was evaluated using GraphPad Prism 8.0.2. The range of ENC was 20–61 [55] and low ENC (< 35) signifies strong CUB [55, 56]. Parity rule 2 (PR2) plot PR2 bias was detected using AT-bias [A3/(A3 + T3)] as y-axis and GC-bias [G3/(G3 + C3)] as x-axis [57]. In the coordinate system, each point represents the composition preference at the third codon position of a gene [58]. This visualization clearly assessed both the extent and direction of CUB in genes. Neutrality plot The neutrality plot was utilized on GraphPad Prism 8.0.2, with GC3 as x-axis and GC12 as y-axis. In this coordinate system, each dot represents an individual species. The slope of the regression line reflects the dominant evolutionary force (mutation pressure or natural selection). A slope close to 1 indicates mutation pressure similarly affects GC12 and GC3, suggesting mutation pressure is the main driver of CUB. In contrast, a slope close to 0 implies natural selection acts differently on GC12 and GC3, decoupling GC12 from GC3, which suggests natural selection is the dominant force shaping CUB [59–61]. Mutation pressure refers to the non-adaptive, stochastic bias inherent in the generation of new genetic variants, whereas natural selection acts as an adaptive, non-random filtering process that discriminates among these variants based on the phenotypic effects of mutations on organismal fitness [62, 63]. ENC-GC3s plot The ENC-GC3s plot is a common method to define if the codon use in a specific gene is impacted only by mutation or other factors [64]. ENC plot was generated using Nanjing Genepioneer Biotechnology Co Bioinformatics Cloud (http://112.86.217.82:9929/#/home). Under a model of pure mutation pressure, the actual ENC value of a gene is expected to align closely with the theoretical curve predicted solely using its GC3s content. Therefore, when points corresponding to these genes clustered around the anticipated curve, it can be inferred that mutation pressure is the principal factor shaping CUB. Conversely, points diverging from the expected curve signify a probable influence of other factors on CUB formation [55]. The ENC ratio indicates the scope of variation between the expected and observed ENCs [65], calculated as (ENCexp - ENCobs)/ENCexp. Evolutionary rate The average rate of non-synonymous substitutions (Ka) and the average rate of synonymous substitutions (Ks) were used to evaluate the differences in PCGs among closely related species [66]. Ka, Ks and the mean Ka/Ks ratio across 12 PCGs were computed across 12 PCGs on DnaSP 5.0 [67]. A Ka/Ks ratio below 1 indicates negative (or purifying) selection, a ratio equal to 1 indicates neutral evolution, and a ratio above 1 indicates positive selection. A ratio near 1 signifies minimal selection pressure [68]. Phylogenetic analysis The mitochondrial genomes of 38 Monopisthocotyla species were selected (41 − 3 excluded), with Eudiplozoon nipponicum and Paradiplozoon yunnanesis were chosen as outgroups. For all 40 species (38 ingroup + 2 outgroup), nucleotide sequences of 36 mitochondrial genes (PCGs, rRNAs, and tRNAs) were extracted from their complete mitogenomes using PhyloSuite v2 [51, 52]. Multiple sequence alignment of PCGs was performed with MAFFT 7.313 [69]. Then codon alignment of PCGs was performed using MACSE [70], and the resulting alignments were pruned on Gblocks [71] under default parameters to retain conserved blocks. The optimal partitioning schemes and substitution models were selected separately for Maximum Likelihood (ML) and Bayesian Inference (BI) analyses using ModelFinder [72] implemented in PhyloSuite under the Bayesian Information Criterion (BIC). A Bayesian phylogenetic tree was conducted using MrBayes [73], with four Markov chain Monte Carlo (MCMC) chains operating independently for 2,000,000 generations and sampling every 1, 000 generations. The first 25% of the chains were discarded as burn-in. ML trees were built using IQ-TREE [74] with 1,000 fast bootstrap replicates. The final phylogenetic trees were displayed on iTOL web server (http://itol.embl.de). Results Characterization of mitochondrial genome of D. wunderi Bychowsky, 1931 The mitochondrial genome of D. wunderi was a typical circular structure in length of 14,920 bp. There were 36 genes in the mitogenome of D. wunderi, including 12 PCGs, 22 tRNA genes, 2 rRNA genes, and 2 non-coding regions (NCRs, 477 and 1,392 bp, respectively). The 36 genes of the mitogenome were positioned on the heavy strand in D. wunderi (Fig. 1, Table 2). The mitogenome of D. wunderi was closely arranged, with 6 gene overlapping regions (1–28 bp) and 18 intergenic regions (1–18 bp). The start codon ATG was used in 9 PCGs, while GTG was used in genes nad4L, nad3, and nad5. The most used complete stop codons were TAA and TAG, and the incomplete T codon was used in cox2 and cox3 (Table 2). To-scale circular architectural maps of mitogenomes of D. wunderi Table 2Comparison of mitogenomic architecture of Dactylogyrus wunderiGenePositionSizeIntergenic regionCodonStrandFromToStartStop trnG 16666H cox3 67712646ATGTH trnH 71377462H cytb 7761,8521,0771ATGTAGH nad4l 1,8702,10323417GTGTAAH nad4 2,0763,3021,227-28ATGTAAH trnQ 3,3043,367641H trnF 3,3663,44176-2H trnM 3,4223,48362-20H atp6 3,4863,9955102ATGTAAH nad2 4,0024,8328316ATGTAGH trnV 4,8474,9076114H trnA 4,9094,972641H trnD 4,9725,03362-1H nad1 5,0345,924891ATGTAGH trnN 5,9265,992671H trnP 5,9976,063674H trnI 6,0636,12967-1H trnK 6,1316,195651H nad3 6,2146,54333018GTGTAAH trnS1 6,5446,59956H trnW 6,6016,663631H cox1 6,6718,2241,5547ATGTAAH trnT 8,2398,3046614H rrnL 8,3059,243939H trnC 9,2449,30764H rrnS 9,30810,029722H cox2 10,03010,633604ATGTH NCR-1 10,63411,110477H trnE 11,11111,17565H nad6 11,18011,6264474ATGTAGH trnY 11,62911,693652H trnL1 11,69411,76067H trnS2 11,76011,82768-1H trnL2 11,83111,896663H trnR 11,89711,96266H nad5 11,96613,5281,5633GTGTAGH NCR-2 13,52914,9201,392H Comparison of mitogenomic architecture of Dactylogyrus wunderi To investigate the base composition and its correlation with CUB in the mitogenome of D. wunderi, we analyzed the base composition of the complete mitogenome, PCGs, tRNAs and rRNAs. The A + T content was 57.5%, with the highest content of T bases (32.5%) and the lowest content of C bases (18.5%) in the mitogenome of D. wunderi. The AT skew value of D. wunderi was − 0.13, and the GC skew value was 0.129. The minimum AT skew and GC skew values were − 0.211 and 0.118 across all PCGs in D. wunderi, respectively. The PCGs were 9,912 bp in length, and the AT content was 57.3% in D. wunderi. The T content was 34.7% in PCGs, followed by G content (23.9%), A content (22.6%), and C content (18.8%). The AT content was 56.9% in rRNAs, and the AT skew value was 0.026, GC skew value was 0.19. The AT content was 57.7% in tRNAs, and the AT skew value was − 0.05, GC skew value was 0.202. The secondary structures of the 22 tRNA genes were predicted in D. wunderi, The most tRNA genes were able to fold into typical cloverleaf structures (Supplementary Fig. 1), Gene trnS1 lacked the DHU arm, and genes trnH and trnM genes lacked the TψC loop. The results indicated that the base composition of PCGs had AT bias in D. wunderi, and the base usage of PCGs preferred T rather than C (Table 2). Genome structure and composition of mitogenomes in Dactylogyrus and other Monopisthocotyla species To understand the genome structure and composition of mitogenomes in Monopisthocotyla species, we compared the size and AT content of PCGs, tRNA and rRNA from 41 Monopisthocotyla species (except D.wunderi, D. crucifer and D. zandti, the remaining 38 sequences were downloaded from GenBank). The contents of the four bases ranked as T > A > G > C among the five types of genes in six Dactylogyrus species, and the A + T content (57.27%−63.90%) was higher than the C + G content (36.08%−42.72%) in the six Dactylogyrus species (Fig. 2). Average base composition of full genome, PCGs, tRNAs, rRNAs and NCR of mitogenomes of six Dactylogyrus species The length of PCGs ranged from 9,912 to 9,960 bp in six Dactylogyrus species, the length of rRNAs ranged from 1,650 to 1,824 bp, and the length of tRNAs ranged from 1,408 to 1,429 bp (Fig. 3). The length of PCGs ranging from 7,038 to 10,017 bp in other Monopisthocotyla species, the length of rRNAs ranged from 1,647 to 1,734 bp, and the length of tRNAs ranged from 1,237 to 1,500 bp. Significant differences were observed in the lengths of PCGs, tRNAs and rRNAs between the six Dactylogyrus species and other Monopisthocotyla species. The mean AT skew and GC skew values of PCGs were − 0.15 ± 0.042 and 0.16 ± 0.082 in the six Dactylogyrus species, and that of PCGs were − 0.19 ± 0.061 and 0.22 ± 0.088 in the 35 other Monopisthocotyla species, respectively, there was significant differences between the six Dactylogyrus and 35 other Monopisthocotyla groups. In addition, there were significant differences in the A + T content between the six Dactylogyrus species (58.6% ± 6.2%) and other Monopisthocotyla species (71.0% ± 5.80%, p < 0.01). The A + T contents of PCGs, tRNAs and rRNAs were higher than 50.0% (Fig. 4). These results indicated that the PCGs, tRNAs and rRNAs of Monopisthocotyla had AT base bias, and there were significant differences in sizes between the six Dactylogyrus species and other Monopisthocotyla species. Box plots of sizes of full genome, PCGs, tRNAs, and rRNAs for six Dactylogyrus and 35 Monopisthocotyla species. n: not significant, *: p < 0.05 Box plots of the A + T content of mitogenome, PCGs, tRNAs, and rRNAs of six Dactylogyrus and 35 Monopisthocotyla species. Boxes indicate first and third quartiles, and horizontal lines indicate median. ****: p < 0.001 PR2 plot To investigate the degree of bias at the third codon position in the PCGs of Monopisthocotyla species, we calculated the G3/(G3 + C3) and A3/(A3 + T3) of 41 Monopisthocotyla species. There were 7, 1, 8 and 56 genes at the first, second, third and fourth quadrants respectively in the six Dactylogyrus (Fig. 5). There were 29, 31, 45 and 295 genes at the first, second, third and fourth quadrants in other Monopisthocotyla species, respectively (Fig. 5). Most of genes in Monopisthocotyla were located in the fourth quadrant, and the third nucleotide of mitogenome codon were T preferred to A, and G preferred to C. These results indicated that the G/T were more used than A/C at the third nucleotide of mitogenomes codons in Monopisthocotyla. PR2-bias plot of six Dactylogyrus and 35 Monopisthocotyla species Neutrality plot To comprehend the effects of natural selection and mutational pressure on codon use preference in Monopisthocotyla, we analyzed the correlation between GC3 and GC12. The GC3 content of PCGs was 23.8%−55.5%, and GC12 content of PCGs was 34.8% −44.4% in the six Dactylogyrus species, and a significant correlation between GC3 and GC12 was observed (R = 0.9603, p < 0.05) (Fig. 6). Similarly, the GC3 content of PCGs was 8.0%−45.0%, the GC12 content of PCGs was 23.2%−41.7% in the other Monopisthocotyla species, and their correlation was also significant (R = 0.9344, p < 0.05). These high correlations suggested that the nucleotide composition of Monopisthocotyla was largely shaped by mutational pressures over evolutionary time in mitogenomes. Furthermore, the slopes were 0.3136 and 0.389 in the six Dactylogyrus and 35 other Monopisthocotyla species, respectively. In mitogenome codon usage pattern, the relative contributions of natural selection and mutational pressures were approximately 69.0% and 31.0% in the six Dactylogyrus species respectively, and were 61.0% and 39.0% respectively in other Monopisthocotyla species. These results implied that codon usage bias in Monopisthocotyla species was collectively shaped by natural selection and mutation pressure, with natural selection being the dominant force. Neutrality plot of six Dactylogyrus and 35 Monopisthocotyla species ENC-GC3s plot To better understand the impact of mutation pressure on codon usage preferences in the mitogenomes of Monopisthocotyla species, we examined the connection between the ENC and the GC3s content (Fig. 7). The ENCs of the six Dactylogyrus species scoped from 33.39 to 61, and the ENCs were ‌above 35 in 97.2% of the genes. Other Monopisthocotyla species scoped from 25.21 to 61, and the ENCs were above 35 in 77.4% of the genes. The results indicated the CUB were weak in mitogenomes of Monopisthocotyla species. The 20.83% genes ENC ratio ranged from − 0.05 to 0.05 in the six Dactylogyrus species, and 0.5% gene ENC ratio varied from − 0.05 to 0.05 in other Monopisthocotyla species (Table 3), indicating that only a few genes had very small differences between their ENCexp and ENCobs values. Conversely, the CUBs of the majority of genes (with ENC ratios outside this range) were mainly determined by natural selection. ENC-GC3s plot of 12 PCGs in six Dactylogyrus and 35 Monopisthocotyla species. The standard curve ENC = 2 + GC3s + 29/[GC3s²+(1-GC3s)²] represents the expected ENC to GC3s Table 3Distribution of ENC ratiosENCratio interval6 frequency (%)35 frequency (%)< -0.0566.6798.78-0.05 ~ 0.0520.830.50> 0.0512.500.72total100100 Distribution of ENC ratios Analysis of RSCU in 41 Monopisthocotyla species To clarify the codon usage of the Monopisthocotyla, we counted the RSCU in PCGs of mitogenomes from 41 Monopisthocotyla species. The RSCU values of the codon UUA (5.29) was the highest in 41 Monopisthocotyla species, and the RSCU values of the codon CUC (0.00) was the lowest in the 41 Monopisthocotyla species (Fig. 8). Most synonymous codons ending in U or A exhibited RSCU > 1.00, such as UUA, AUU, GUU, and GCU, implying a positive usage bias toward these codons. Conversely, most codons ending in C or G, such as CUC, GCC, CGC, and CUG, exhibited RSCU < 1.00, indicating a lower usage frequency. The six Dactylogyrus species exhibited distinct codon usage patterns compared to other Monopisthocotyla species. For instance, the codon GCC had an RSCU of 1.49 in D. tuba but 1.21 in D. zandti, while its RSCU was consistently < 1.00 across all 34 Monopisthocotyla species examined. Similarly, the RSCUs of codon UGC were > 1 in D. wunderi, D. simplex, D. tuba, and D. zandti, whereas it remained < 1 in other Monopisthocotyla species. Meanwhile, certain codons were preferred in other Monopisthocotyla species, but they were avoided in Dactylogyrus. For example, the RSCU value of codons UUA and GGU were higher than 1 in 35 Monopisthocotyla species, but less than 1 in three Dactylogyrus species. Moreover, codon usage also varied in different Dactylogyrus species. The RSCUs values of codon CUG were higher than 1 in D. tuba (1.56), D. crucifer (1.03), and D. zandti (1.40), but less than 1 in D. wunderi (0.67), D. simplex (0.91), and D. lamellatus (0.23). The results revealed that Dactylogyrus exhibits a distinct and variable codon usage pattern within other Monopisthocotyla species. Hierarchical clustering and heat map of RSCU for 41 Monopisthocotyla species, with a colour change from blue to red in the heat map indicating an increase in RSCU Selection pressure To assess the impact of evolutionary selection on codon usage in Monopisthocotyla mitogenomes, we computed the average Ka/Ks ratio of 12 PCGs in 41 Monopisthocotyla species (Fig. 9). The average Ka/Ks ratio scoped from 0.15 (cox1) to 0.60 (nad 2) in 12 PCGs of six Dactylogyrus, and the selective pressure was in the order of nad 2 > nad5 > nad 4 > nad 6 > atp 6 > nad 4 L > nad 3 > cox 2 > nad 1 > cox 3 > cytb > cox1. The average Ka/Ks ratio scoped from 0.31 (cox1) to 0.89 (nad 2) in 12 PCGs of the 35 Monopisthocotyla species, and selective pressure ranked in the order of nad 2 > nad 6 > nad5 > nad 4 > nad 4 L > nad 3 > atp 6 > cox 3 > cox 2 > nad 1 > cytb > cox1. The results indicate the genes of 12 PCGs were affected by purifying selection in both six Dactylogyrus and 35 Monopisthocotyla species. The Ka/Ks value was the largest in nad 2 among the genes of 12 PCGs, and was the smallest in cox1. The Ka/Ks ratio of six Dactylogyrus species was lower than that of 35 Monopisthocotyla species. The results indicated that there were affected by greater selection pressure in the six Dactylogyrus and 35 Monopisthocotyla species. Comparisons of average Ka/Ks ratios for 12 PCGs in six Dactylogyrus and 35 Monopisthocotyla species. Data below the x-axis show the Ka/Ks values for the two groups. ***: p < 0.05 between the two groups Phylogenetic associations To investigate the relationships among Monopisthocotyla members, we constructed phylogenetic trees using the nucleotide sequences of PCGs from Monopisthocotyla through Bayesian inference (BI) and maximum likelihood (ML) assays. Both BI and ML analyses were performed to establish comparable topologies boasting elevated posterior probability values/bootstrap support (Fig. 10). The resultant BI and ML analysis trees bifurcated into two main independent clades: Gyrodactylidea (Clade I) and Capsalidea - Gyrodactylidea - Dactylogyridea (Clade II). The six Dactylogyrus species were grouped into one clade in Dactylogyridea, and the species of Ancylodiscoididae, Dactylogyridae, Diplectanidae, Capsalidae and Gyrodactylidae were all grouped into one clade. Cichlidogyrus sclerosus was sister to Scutogyrus longicornis, which both clustered with C. casuarinus. The phylogenetic tree revealed that Dactylogyrus was monophyletic and Cichlidogyrus was paraphyletic. To evaluate the family-level distribution of ENCs, we performed an analysis of variance (ANOVA) (Table 4). The coefficient of variation (CV) analysis revealed that family-level CV values ranged from 4.8% (Ancylodiscoididae) to 18.4% (Gyrodactylidae), with a weighted mean CV of 11.0%. The ANOVA showed 33.6% of the total variation was attributed to within-family differences, while 66.4% originated from between-family variation. This pattern corresponds to the phylogenetic topology, the monophyletic family Dactylogyridae showed relatively high ENCs (53.25 ± 6.08), whereas the families Diplectanidae (34.95 ± 2.39) and Ancylodiscoididae (34.24 ± 1.63) displayed lower values. Phylogenetic analysis revealed that within the context of the sampled taxa, the families of Monopisthocotyla were monophyletic groups, except for Ancyrocephalidae. The pattern of variation in CUB was correlated with phylogenetic relationships. Phylogenetic tree utilizing 12 PCGs from the mitogenomes of Monopisthocotyla on basis of BI and ML analyses Table 4Analysis of ENC variation across families of Monopisthocotyla based on phylogenetic relationshipsFamilyNumber of SpeciesENC Average value ± SD(s²)Coefficient of Variation (%)Dactylogyridae653.25 ± 6.0836.9911.4 Diplectanidae334.95 ± 2.395.736.8Ancylodiscoididae234.24 ± 1.632.654.8Gyrodactylidae1042.42 ± 7.860.8318.4Capsalidae944.81 ± 3.5912.928.0Ancyrocephalidae646.13 ± 3.310.867.2total3643.46 ± 8.0364.5417.8 Analysis of ENC variation across families of Monopisthocotyla based on phylogenetic relationships Discussion The present study represents first comprehensive characterization of the complete mitogenome of Dactylogyrus wunderi and a comparative analysis of CUB and structural features across 41 species in class Monopisthocotyla. The mitogenome of D. wunderi was 14,920 bp in length, and had a high A + T content in 12 PCGs, 22 tRNAs, and 2 rRNAs. The RSCUs of codons was high and tended to end with A/U bases in Monopisthocotyla species, and the Monopisthocotyla had strong CUB. The neutrality plot and ENC showed that natural selection primarily influenced the CUB. Phylogenetic analyses showed that, within the context of the sampled taxa, most families of Monopisthocotyla were monophyletic groups, except for family Ancyrocephalidae. The newly sequenced mitogenome of D. wunderi was 14,920 bp in length and contains the standard 36 genes (12 PCGs, 22 tRNAs, and 2 rRNAs) found in most Monopisthocotyla. The whole mitogenomes of five Dactylogyrus species have been made public, and the length of D. crucifer is the shortest (14,402 bp) and the length of D. zandti is the largest (18,579 bp) in five Dactylogyrus species. The mitogenome length of D. wunder was between D. crucifer and D. zandti [16, 17, 30]. This length is within the published range of the mitogenome lengths, and is consistent with the length characteristics of published Dactylogyrus mitogenomes. The gene order of mitogenome of D. wunderi was identical to that reported for other Dactylogyrus species (e.g., D. simplex, D. crucifer, D. zandti and D. tuba) and was largely conserved across the Monopisthocotyla, reinforcing the utility of mitogenome arrangement as a stable phylogenetic marker at higher taxonomic levels within this group [75, 76]. In addition, the compact structure, with several gene overlaps and short intergenic regions, is a hallmark of Trematode and Monopisthocotyla mitogenome, likely reflecting strong selection for compactness and efficient replication [77]. A prominent feature across all analyzed species is a pronounced A + T bias. The AT content of D. wunderi was 57.5%, which aligns with the range observed in other five Dactylogyrus species (54.0% to 70.6%) but was generally lower than that found in many other Monopisthocotyla species, such as Neobenedenia melleni (75.9%) and Gyrodactylus pseudorasborae (73.1%) [78, 79]. In addition, the Dactylogyrus species primarily distributed in Hunan province (70.6%) and Xinjiang (54.0%-58.4%) of China, there was significantly differ between the base compositions of Xinjiang and Hunan province (p < 0.01), and the AT content of Dactylogyrus from Xinjiang was lower than that of Hunan province. This universal AT-richness is a common trait in metazoan mitogenomes, particularly in invertebrates and parasites, and is primarily attributed to asymmetric mutational pressure during replication and transcription, often linked to strand-specific deamination events [77]. The high A + T content in some families may reflect more intense or prolonged mutational pressure, potentially associated with differing life-history traits such as shorter generation times or higher metabolic rates [80], though this requires further investigation. Our results corroborate prior studies on Monopisthocotyla mitogenomes, which report strand compositional bias characterized by a negative AT skew and a positive GC skew [42]. The mean AT skew and GC skew of PCGs were negative and positive in the six Dactylogyrus and 35 other Monopisthocotyla species, respectively. Strand asymmetry directly affects nucleotide usage, which in turn shapes codon usage and the amino acid composition of mitochondrial proteins [81]. Among different species, the likelihoods of synonymous codons being employed are not always the same [82]. The codon use modes of intimate species separated from intimate hosts are highly similar [30]. Our results indicated the codon preferences of Dactylogyrus were relatively similar in closely-related hosts. For example, five Dactylogyrus species (D. wunderi, D. Zandti, D. crucifer, D. tuba and D. simplex) exhibited preferred ending with C or G (e.g., AGG, CAG, CCC), and hierarchical clustering of RSCU showed D. Zandti, D. crucifer and D. tuba grouped into one clade. Moreover, the high RSCUs of codons exhibited preferred ending with A or U (e.g., UUA, UCU, CCU) in D. lamellatus, and the D. lamellatus grouped with Megalobenedenia derzhavini and Enterogyrus malmbergi into one clade. The five Dactylogyrus species parasitized subfamilies Leuciscinae and Schizothoracinae, which belong to cold-water fishes in Irtysh River and Yarkand River, while D. lamellatus parasitized the fish in subfamily Xenocypridinae, which belongs to warm-water fishes in central China. The RSCUs verified the codon preferences of Dactylogyrus were not conservative, a phenomenon was also found in genus Gyrodactylus. The Dactylogyrus was isolated from different hosts in different environments, which impacted the codon preferences. The physicochemical properties of aquatic habitats influence parasitism [83], and especially the water temperature significantly impacts the interaction between parasites and hosts [84]. In long-term isolated populations, genetic drift, combined with various local adaptations (including adaptive adjustments to host physiological characteristics and the local environment), may have fixed different genetic traits, such as the codon usage patterns [85]. Reportedly, demonstrated that environmental temperature can affect the polypurine (AG) tracts in mRNAs, and thereby exerting an indirectly affects genomic GC content levels [86]. The variations in GC content regulate codon usage, especially the GC content at the third position of codons, and directly affects the amino acid composition. These findings indicate that the environment serves as a key ecological driver in shaping genetic traits, including CUB, and it enables genetic traits to adapt rapidly to the environment. Thus, the codon preferences may be associated with host species isolated from geographical location and the evolutionary rate of Monopisthocotyla [17, 87]. Nucleotide composition affects CUB within genomes, supporting the inter connectedness of genetic components [88]. Van Hemert et al. (2016) showed that nucleotide bias is the primary cause of CUB, a phenomenon that constrains the regulatory effects of codon selection and translation [89]. The nucleotide composition of mitochondrial genes demonstrated a significant A/T bias in six Dactylogyrus and 35 Monopisthocotyla species. This finding complies with other research on the mitogenome of Monopisthocotyla, including Capsala katsuwoni and Capsala martinieri (66.5% and 63.9%) [90]. Nucleotide bias is a common feature of mitogenomes, and the A + T content of most species is high [32]. Our analysis indicated the AT content was relatively high in the nucleotide compositions of the 41 Monopisthocotyla species. This AT content varies widely among parasitic flatworms. For example, the cestode Parabreviscolex niepini has an AT content of 59.6% [91], whereas in Nippotaenia mogurndae it reaches 72.2% [92]. And the third place of the codon were mostly A/T base, the preference for U-ending codons was especially strong, as evidenced by the highest average RSCU value for UUA (Leu). This aligns with the PR2 plot results, where the majority of genes across all species fell into the fourth quadrant (G3/(G3 + C3) > 0.5, A3/(A3 + T3) < 0.5), indicating a significant preference for T over A and G over C at the third codon position. Additionally, CUB was affected by the third nucleotide, the G/T content of the third nucleotide were higher than A/C content among the 12 PCGs in 41 Monopisthocotyla species. This T/G bias at the synonymous sites is a distinctive feature not universally observed in all AT-rich genomes and may point to specific asymmetries in the mitogenome replication process in this lineage [93–95]. This pattern is a direct consequence of the underlying A + T bias and the compositional constraints at the third codon position (wobble position), where mutation pressure is often least constrained by purifying selection [96, 97]. As well known, CUB reflects an equilibrium‌ between mutation pressure and natural selection, two primary forces driving CUB [24, 98]. ENC-GC3s plot can efficiently explore the main characteristics of codon usage among genes [55]. A combination of neutral plot and ENC analysis implied CUB may be influenced by natural selection in the mitogenomes of 41 Monopisthocotyla species. The high correlation coefficient suggests that the overall nucleotide composition of these genomes is largely shaped by mutational pressure, as supported by previous studies [99]. However, the regression slopes of GC12 and GC3 were 0.3136 and 0.389 in PCGs of six Dactylogyrus and 35 Monopisthocotyla species, respectively, and were significantly lower than 1, indicating that the base compositions of the 1st, 2nd, and 3rd codon positions were greatly different in codons of PCGs, and that natural selection imposes a stronger constraint than mutation pressure on the GC content at the 1st and 2nd codon positions in codons of PCGs [100, 101]. The ENC analysis confirmed that natural selection influenced CUB in six Dactylogyrus and 35 Monopisthocotyla species. The mean ENCs were 51.59 and 41.84 in six Dactylogyrus and 35 Monopisthocotyla species, respectively. ENC > 35 indicates low CUB. Most points were not gather near the standard curve in Fig. 7 implying other factors influenced CUB. Chen’s results shown that there are a minority‌ of genes close to the anticipated ENC-plot curve in eight Nemertea species, implying that mutation pressure contributes to shaping CUB among these genes [102]. However, our results shown that a large gap in ENC values between expected and observed for most genes, this large gap is primarily attributed to natural selection, which favors optimal codons to enhance translational efficiency and accuracy, thus causing a marked deviation of codon usage bias from neutral expectations [103, 104], and the ENC ratio further confirmed that natural selection more affects CUB than mutational pressure. Importantly, the distribution of CUB parameters, particularly ENC values, correlates with this phylogenetic structure. Families that form coherent clades (e.g., Dactylogyridae, Diplectanidae) show relatively homogeneous ENC values among their members, whereas the polyphyletic pattern suggested for Ancyrocephalidae in our tree is reflected in more variable ENC values. This congruence suggests that aspects of CUB, especially those shaped by long-term, lineage-specific mutational pressures, can serve as a valuable complementary phylogenetic character. They provide insights into the evolutionary history of genomic compositional evolution that may be independent of, yet informative alongside, primary nucleotide sequence data. Mutation pressure and natural selection interplay to shape CUB by driving changes and preserving favorable codon usage patterns, respectively. For instance, natural selection, often in the form of translational selection, favors the fixation of optimal codons that enhance both translational speed and accuracy [104]. Conversely, mutational biases can systematically alter nucleotide composition, thereby changing CUB, at the same time, selection might act to counteract such mutational biases at a genome-wide level to maintain compositional stability [105]. The CUB can generate selection pressures in protein translation [106]. Furthermore, CUB can predict gene functions [107], assist in discovering novel genes, and design primers [108]. Investigating CUB across different species provides insights into their genetic architecture and evolutionary trends across different species [109]. Natural selection critically affects CUB formation across the genome in various species, including Caenorhabditis, Drosophila and Arabidopsis [110]. This study combined a neutrality plot with ENC analysis and evidently, PCGs were primarily governed by natural selection in the mitogenomes of the six Dactylogyrus and 35 Monopisthocotyla species. Conclusions For the first time, this study presents the complete mitogenome of Dactylogyrus wunderi, and compares and analyzes the codon usage patterns and phylogenetic relationships were compared and analyzed in Monopisthocotyla. This mitogenome of Dactylogyrus wunderi was 14,920 bp in length and has a high A + T content. A pervasive A + T bias driven largely by mutational pressure in Monopisthocotyla, which establishes a foundational preference for NNU/NNA codons. However, natural selection plays a dominant role in shaping the final CUB pattern of Dactylogyrus and other Monopisthocotyla species, acting to constrain extreme bias. Genus Dactylogyrus exhibits a distinctive and phylogenetically coherent CUB signature, characterized by unique codon preferences and stronger purifying selection on its PCGs. The results established a foundational framework for understanding molecular evolution in these parasites and provided a rich resource for future studies on their systematics, population genetics, and host-parasite coevolution. Supplementary Information Supplementary Material 1. Acknowledgements Not applicable. Abbreviations Codon usage bias Parity rule 2 Relative synonymous codon usage Effective number of codons Effective number of codons values Protein-coding genes Bayesian inference Maximum likelihood Authors’ contributions Conceptualization, CLH; methodology, YYY and LZ; validation, CY, LZ and SZX; formal analysis, YYY and SZX; investigation, YJL, NWW, LZ, SZX and CY; resources, CLH; data curation, YJL and NWW; writing-original draft preparation, YYY; writing-review and editing, CLH; visualization, CY; supervision, CY; project administration, CY. All authors have read and agreed to the published version of the manuscript. Funding This work was funded by the National Natural Science Foundation of China (32403085 to C-L H), Key research and development Program for Special Project of Xinjiang Uygur Autonomous Region (2023B02012 and 2025B02020-2 to C-L H). Data availability A newly sequenced mitogenomes is available from GenBank under the accession numbers: PV847579 ( Dactylogyrus wunderi ). Other datasets supporting the conclusions of this article are included within the article. Declarations Ethics approval and consent to participate In compliance with all relevant legal and ethical regulations concerning laboratory animal welfare and wildlife sampling, this study was approved and supervised by the Animal Welfare and Ethics Committee of Xinjiang Agricultural University (Approval No. 2023045). Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement A newly sequenced mitogenomes is available from GenBank under the accession numbers: PV847579 ( Dactylogyrus wunderi ). Other datasets supporting the conclusions of this article are included within the article.