Mitochondrial Genomes of Distant Fish Hybrids Reveal Maternal Inheritance Patterns and Phylogenetic Relationships

Abstract

As distant hybridization has profound implications for evolutionary biology, aquaculture, and biodiversity conservation, this study aims to elucidate patterns of maternal inheritance, genetic divergence, and phylogenetic relationships by synthesizing mitochondrial genome (mitogenome) data from 74 distant hybrid fish species. These hybrids span diverse taxa, including 48 freshwater and 26 marine species, with a focus on Cyprinidae (n = 35) and Epinephelus (n = 14), representing the most frequently hybridized groups in freshwater and marine systems, respectively. Mitogenome lengths were highly conserved (15,973 to 17,114 bp); however, the genetic distances between hybrids and maternal species varied from 0.001 to 0.17, with 19 hybrids (25.7%) showing distances >0.02. Variable sites in these hybrids were randomly distributed but enriched in hypervariable regions, such as the D-loop and NADH dehydrogenase subunits 1, 3 and 6 (ND2, ND3, and ND6) genes, likely reflecting maternal inheritance (reported in Cyprinus carpio × Carassius auratus). Moreover, these genes were under purifying selection pressure, revealing their conserved nature. Phylogenetic reconstruction using complete mitogenomes revealed three distinct clades in hybrids: (1) Acipenseriformes, (2) a freshwater cluster dominated by Cypriniformes and Siluriformes, and (3) a marine cluster comprising Centrarchiformes, Pleuronectiformes, Scombriformes, Cichliformes, Anabantiformes, Tetraodontiformes, Perciformes, and Salmoniformes. The prevalence of Cyprinidae hybrids underscores their importance in aquaculture for hybridization, where traits such as rapid growth and disease resistance are enhanced. In contrast, marine hybrids are valued for their market value and adaptability. While mitogenome data robustly support maternal inheritance in most cases, exceptions suggest complex mechanisms, such as doubly uniparental inheritance (DUI), in distantly related crosses. Moreover, AT-skew of genes in hybrids revealed a paternal leakage of traits in mitogenomes. This study also highlights ecological risks, such as genetic swamping in native populations, emphasizing the need for responsible hybridization practices. These findings advance our understanding of the role of hybridization in fish evolution and aquaculture, providing a genomic framework and policy recommendations for optimizing breeding programs, hybrid introduction, and mitigating conservation challenges.

1. Introduction

Distant hybridization is the cross-breeding of individuals from different species or genera. It has been widely observed in fish due to their high reproductive potential, diverse habitats, and overlapping ecological niches. This process can produce hybrid offspring with novel genetic combinations that may exhibit unique traits, such as enhanced growth rates, disease resistance, or environmental adaptability [1]. However, distant hybridization can also lead to reduced fitness, sterility, or the breakdown of species boundaries, making it a double-edged sword in both natural ecosystems and human-managed systems like aquaculture [2,3]. Fish distant hybridization, a common breeding technique, involves different species, and the traits resulting from this process have been documented [4,5].

A wide range of taxa, including freshwater, marine, and brackish species, have been documented for distant hybridization among fish. In freshwater ecosystems, hybridization is particularly common due to the dynamic nature of these habitats, which often experience fluctuations in water levels, temperature, and connectivity [6,7]. These environmental changes can bring previously isolated species into contact and facilitate hybridization. For example, in the Cyprinidae family, hybridization between carp and goldfish has been extensively studied, revealing insights into the genetic and ecological consequences of this crossing [6,8,9].

The study of distant hybridization in fish has greatly advanced through the use of molecular tools, particularly mitochondrial DNA (mtDNA) and nuclear DNA (nDNA) markers. Mitochondrial genomes are useful for tracing maternal lineages and identifying hybridization events because of their maternal inheritance and high mutation rates [10,11,12]. In contrast, nDNA provides a more comprehensive view of genetic admixture and introgression, allowing researchers to assess the extent of hybridization and its impact on genetic diversity [13,14,15].

One of the most significant outcomes of distant hybridization is hybrid speciation, in which the offspring evolve into a new species. This process, known as homoploid hybrid speciation, has been documented in several fish groups [16], including the cichlids of Crater Lake Xiloá [17] and the Great Lakes of East Africa [18]. In these lakes, hundreds of cichlids are emerging through hybridization, acting as biodiversity hotspots. Their adaptive radiation provides a compelling example of how hybridization can drive evolutionary innovation and diversification [19]. In addition to its significance in evolution, distant hybridization has practical applications in aquaculture. Hybrid fish are traditionally bred to combine the desirable traits from parent stocks, such as faster growth rates, better disease resistance, and increased tolerance to environmental stressors [20,21,22]. For example, hybrid tilapia (Oreochromis niloticus × Oreochromis aureus) are widely cultivated in aquaculture due to their rapid growth and resilience [23,24]. Similarly, hybrid sturgeon (Huso huso × Acipenser ruthenus) are bred for caviar production, combining the large size of Beluga sturgeon with the early maturation of Sterlet sturgeon [25,26].

Despite its potential benefits, distant hybridization poses challenges, particularly in conservation. Hybridization between native and non-native species can lead to genetic swamping, compromising the genetic integrity of native populations and increasing conservation threats [27,28]. This is a significant concern in freshwater ecosystems, where introduced species often hybridize with natives, leading to the decline or extinction of native populations [29,30]. For example, the introduction of rainbow trout (Oncorhynchus mykiss) has resulted in crossing with native cutthroat trout (Oncorhynchus clarkii) in North America, threatening the survival of pure cutthroat populations [31]. Similarly, hybridization of S. alburnoides with the invasive bleak (Alburnus alburnus) in Portuguese river basins not only wastes reproductive effort and impairs the genetic integrity of the endemic species but also disrupts the reproductive dynamics of the fish [27]. The study of distant hybridization in fish also provides insights into the mechanisms underlying reproductive isolation and speciation. Hybrid zones, where two species interbreed, serve as natural laboratories for studying the genetic and ecological factors that maintain species boundaries [32,33]. For example, research on hybrid zones between Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) has revealed the roles of environmental selection and genetic incompatibilities in limiting gene flow [34].

Fish distant hybridization is a complex, multifaceted phenomenon with significant implications for evolutionary biology, aquaculture, and conservation. The integration of molecular tools, ecological studies, and genomic analyses has greatly enhanced our understanding of hybridization and its effects. To date, many distant hybridization fish have been reported with mitochondrial genome (mitogenome) data but not summarized. In this study, we listed these fish species and analyzed their maternal genetic distances and phylogenetic relationships.

2. Materials and Methods

The mitogenome sequences of 74 distant hybridization fish were retrieved from GenBank (

Table 1. The list of distant hybridized fish.

SL	Scientific Names of Hybrid Fish (Names of Hybrid Fish in the Same Genera Appear as Abbreviations)	Freshwater Fish
1	Acanthopagrus schlegelii × Pagrus major	N
2	Acipenser dabryanus × A. schrenckii	Y
3	Acipenser gueldenstaedtii × A. baerii	N
4	Acipenser schrenckii × A. baerii	Y
5	Acipenser schrenckii × Huso dauricus	Y
6	Carassius auratus × Cyprinus carpio	Y
7	Channa argus × C. maculata	Y
8	Channa maculata × C. argus	Y
9	Cirrhinus mrigala × Labeo rohita	Y
10	Cromileptes altivelis × Epinephelus lanceolatus	N
11	Cromileptes altivelis × Epinephelus tukula	N
12	Ctenopharyngodon idella × Megalobrama amblycephala	Y
13	Ctenopharyngodon idellus × Elopichthys bambusa	Y
14	Culter alburnus × Ancherythroculter nigrocauda	Y
15	Culter alburnus × Megalobrama terminalis	Y
16	Cyprinus carpio haematopterus × Megalobrama amblycephala	Y
17	Cyprinus carpio wuyuanensis × Carassius auratus	Y
18	Cyprinus carpio × Megalobrama amblycephala	Y
19	Epinephelus akaara × E. lanceolatus	N
20	Epinephelus akaara × E. tukula	N
21	Epinephelus awoara × Epinephelus tukula	N
22	Epinephelus coioides × E. akaara	N
23	Epinephelus coioides × E. lanceolatus	N
24	Epinephelus fuscoguttatus × E. lanceolatus	N
25	Epinephelus fuscoguttatus × E. polyphekadion	N
26	Epinephelus fuscoguttatus × E. tukula	N
27	Epinephelus moara × E. lanceolatus	N
28	Epinephelus moara × E. tukula	N
29	Epinephelus moara × Hyporthodus septemfasciatus	N
30	Huso dauricus × Acipenser schrenckii	Y
31	Hypophthalmichthys nobilis × H. molitrix	Y
32	Hyporthodus septemfasciatus × Epinephelus lanceolatus	N
33	Kareius bicoloratus × Platichthys stellatus	N
34	Leiocassis longirostris × Tachysurus fulvidraco	Y
35	Megalobrama amblycephala × (M. amblycephala × M. terminalis)	Y
36	Megalobrama amblycephala × (M. amblycephala × Parabramis pekinensis)	Y
37	Megalobrama amblycephala × Ancherythroculter nigrocauda	Y
38	Megalobrama amblycephala × Elopichthys bambusa	Y
39	Megalobrama amblycephala × M. hoffmanni	Y
40	Megalobrama amblycephala × M. pellegrini	Y
41	Megalobrama amblycephala × M. skolkovii	Y
42	(Megalobrama amblycephala × Parabramis pekinensis) × (M. amblycephala × P. pekinensis)	Y
43	(Megalobrama amblycephala × Parabramis pekinensis) × M. amblycephala	Y
44	Megalobrama amblycephala × Xenocypris davidi	Y
45	Megalobrama skolokovii × M. amblycephala	Y
46	Megalobrama terminalis × Culter alburnus	Y
47	Megalobrama terminalis × M. amblycephala	Y
48	Misgurnus anguillicaudatus × M. bipartitus	Y
49	Misgurnus anguillicaudatus × Paramisgurnus dabryanus	Y
50	Oncorhynchus mykiss × Salmo salar	N
51	Oreochromis niloticus × O. aureus	N
52	Oxygymnocypris stewartii × Schizopygopsis younghusbandi	Y
53	Paraneetroplus synspilus × Amphilophus citrinellus	Y
54	Platichthys stellatus × (Kareius bicoloratus × P. stellatus)	N
55	Platichthys stellatus × Verasper variegatus	N
56	Pungtungia herzi × Pseudopungtungia nigra	Y
57	Pungtungia herzi × Pseudorasbora parva	Y
58	Salvelinus fontinalis × S. malma	Y
59	Scaphirhynchus albus × S. platorynchus	Y
60	Schizothorax oconnori × S. waltoni	Y
61	Scomberomorus munroi × S. semifasciatus	N
62	Siniperca chuatsi × S. knerii	Y
63	Siniperca chuatsi × S. scherzeri	Y
64	Siniperca knerii × S. chuatsi	Y
65	Siniperca scherzeri × S. chuatsi	Y
66	Squaliobarbus curriculus × Ctenopharyngodon idella	Y
67	Tachysurus fulvidraco × Leiocassis longirostris	Y
68	Tachysurus fulvidraco × T. vachelli	Y
69	Tachysurus ussuriensis × T. fulvidraco	Y
70	Takifugu fasciatus × T. flavidus	N
71	Takifugu flavidus × T. rubripes	N
72	Takifugu obscurus × T. rubripes	N
73	Takifugu rubripes × T. flavidus	N
74	Xenocypris davidi × Megalobrama amblycephala	Y

). The lengths of these genomes were calculated using BioEdit Version 7.7.1 [35]. A, T, G, C, and GC composition (in percentage) were calculated for each genome using https://punnettsquare.org/gccontent/ (accessed on 25 May 2025) and tabulated in Supplementary Table S1. Genetic distance was calculated with MEGA 11 after alignment in BioEdit Version 7.7.1 [35,36]. The comparison of variable sites between hybrids and their maternal fish was checked manually, and the percentage of variable sites (%) was computed. Genes with a greater number of variable sites include 12s rRNA, 16s rRNA, and protein-coding genes (PCGs), specifically ATP synthase subunits 6 and 8 (ATP6 and ATP8), cytochrome c oxidase subunits 1, 2 and 3 (COX1, COX2, and COX3), cytochrome b (CYTB), and NADH dehydrogenase subunits 1, 2, 3, 4, 4L, 5, and 6 (ND1, ND4, ND4L, ND5, and ND6). These genes are presented in a heatmap. Subsequently, using R statistical software (version 4.4.2) and packages including ggplot2, FactoMineR, factoextra, readxl, ggrepel, and boot, we performed Principal Component Analysis (PCA) to better understand the associations among the nucleotide characteristics of the hybrid and maternal mitogenomes, genetic distances, and variable sites in specific genomic regions, producing a PCA biplot. The PCA was conducted with 1000 bootstrap iterations. To examine selection pressure in the above PCGs, we computed the Ka/Ka ratio using the KaKs calculator under the JC model with the MS method, where the genetic codon was used for vertebrate mitochondria [37]. To prepare sequence files for the KaKs analysis, we aligned homologous genes from maternal and hybrid species using the online server at EMBL-EBI, https://www.ebi.ac.uk/jdispatcher/msa/clustalo, with ClustalW output format and default parameters (accessed on 13 July 2025). Prior to alignment, the stop codon from each coding sequence (CDS) was eliminated. Using the same CDS files, we conducted AT skew analysis with R statistical software (version 4.4.2) and packages, including Biostrings, for PCGs in both hybrids and maternal genomes. Density curve plots were created using ggplot2, readxl, and dplyr for (1) all species (hybrids and maternal genomes) and (2) individual plots for hybrids and maternal mitogenomes with genetic distances > 0.02. Lastly, we conducted a phylogenetic analysis using https://www.genome.jp/tools-bin/ete (accessed on 21 May 2025), where three Polypterus fish were used as an outgroup, which utilized ETE3 [38]. The aligner was set to ClustalW default, the alignment cleaner was set to trimal_gappyout, the tree builder was set to phyml_default_bootstrap, and the model tester was set to pmodeltest_full_slow [39,40]. The best-fit model according to BIC was TN+I. Tree branches were tested using SH-like aLRT with 1000 replicates. The fish classification was verified using FishBase [41].

3. Results

3.1. Distant Hybridization in Fish and Their Genetic Distance

Hybrid fish species represent a variety of crossbreeding efforts aimed at enhancing aquaculture productivity. These hybrids often exhibit hybrid vigor, including improved growth rates, disease resistance, and adaptability to different environmental conditions, with a focus on their growth, survival, and economic value [42]. In this study, the analyzed mitogenomes of distant hybridized fish reached 74 types, including 48 freshwater species, 38 of which belonged to the same genera.

The mitogenome length of distant hybridized fish and their maternal species is conserved, ranging from 15,973 to 17,114 bp (Figure 1 and Table S1). This conservation of length aligns with previous studies on fish mitogenomes, which indicate that genome size is generally stable across species owing to its compact and efficient structure [43,44].

The genetic distance between each hybrid and its maternal fish varied from 0.001 to 0.17, with 19 of the 74 hybrid fish showing a genetic distance greater than 0.02 (Figure 2, Table S1). Similar findings have been reported in other hybridization studies, in which genetic distances were used to assess the divergence and evolutionary relationships between hybrid offspring and their parental species [45]. Furthermore, the observed range in this study was consistent with that of previous studies [10,14].

The variable sites in the fish genome with a genetic distance greater than 0.02 were randomly distributed throughout the mitochondrial genomes, indicating maternal inheritance and genetic drift, particularly in the ND1, ND4, ND6, and D-loop regions (Table S2). Higher numbers of variable sites were observed in Cirrhinus mrigala × Labeo rohita, Cyprinus carpio × Megalobrama amblycephala, Epinephelus akaara × E. lanceolatus, Epinephelus moara × Hyporthodus septemfasciatus, Salvelinus fontinalis × S. malma, and Scomberomorus munroi × S. semifasciatus (Figure 3). In Acipenser schrenckii × Huso dauricus, variable sites were concentrated in the ND5-ND6 region. For saltwater fish, COX1, COX2, COX3, and CYTB had the most variable sites. Notably, Takifugu flavidus × T. rubripes displayed the greatest number of variable sites for these genes.

3.2. Association Between Genetic Distance and Genomic Composition and Their Variable Sites

We then performed PCA to determine the relationship between the genome length and nucleotide composition of the mitogenome groups (hybrid and maternal), their genetic distance, and variable sites (Figure 4). PC1 and PC2 together accounted for 53.4% of the variation, indicating a moderate level of structure and capturing a significant portion of the total variability in the data (Tables S3 and S4). The most significant contributions to fish species differentiation in PC1 came from mG, mA, and the variable sites in ATP6, ND2, ND3, and COX1. Certain fish groups can be distinguished primarily by their GC-rich or AT-rich compositions. The distribution may be influenced by hA and mGC variables, further differentiating between the hybrid and maternal groups. The mGC and hGC occupied distinct positions on the plot, implying that the GC content profiles of hybrid and maternal fish differed. The fish species were also divided along the PC1 axis based on the position of hT and mT compared to hA and mA. Since COX1 and ND2 are farther from the origin, they likely play a major role in the variance observed between hybrid and maternal fish. The divergence of fish samples along the axes suggests that certain genomic regions are responsible for differentiating the two groups. The mitochondrial genes COX1, ATP6, ND2, and ND4 can significantly separate hybrid and maternal groups and are often useful for species identification. The similar directions of ND2, ND4, and ATP6 indicate that these variables likely co-occur across fish samples. Additionally, a negative correlation was observed between genetic distance and mGC, suggesting that maternal GC content decreases with increasing genetic distance, and vice versa. Moreover, saltwater fish (Epinephelus akaara × E. lanceolatus, Epinephelus moara × Hyporthodus septemfasciatus, Oreochromis niloticus × O. aureus, Scomberomorus munroi × S. semifasciatus, and Takifugu flavidus × T. rubripes) were firmly separated from freshwater fish. The saltwater species and D-loop likely share comparable genetic attributes or environmental conditions, indicating that these species and the D-loop sequence may differ greatly from others. Megalābroma ambivecphala is considered a vital species in hybridization studies due to its presence in many hybrids. Certain genera (Megalābroma, Takifugu, and Epinephelus) may be considered evolutionary variable in hybridization because of their high frequency of occurrence. Additionally, the PCA biplot in hybridization studies demonstrated trait associations and genetic compatibility.

3.3. Purifying Selection Pressure and AT-Skew in Highly Variable PCGs

Later, our investigation revealed that the above 13 PCGs containing highly variable sites have Ka/Ks values less than 1, which indicated that they are under purifying selection and are highly conserved (Table S5). Furthermore, the density plots for AT-skew values revealed that these 13 PCGs have evolutionary forces present in their nucleotide composition (Figure 5A–C and Table S6). The overall AT-skew pattern across genomes showed distinct peaks, indicating variability in the directional bias of nucleotide composition between genes (Figure 5A). In maternal genomes, each of the PCGs showed distinct distribution that suggested evolutionary adaptation (Figure 5B), whereas hybrids displayed intermediate patterns or blended skewness compared to the maternal genomes, suggesting genomic integration or compensation between parental mitochondrial genomes (Figure 5C). Certain genes (like ATP8, COX1, and COX2) likely reflect evolutionary conservation, and genes (such as ND6 and ND4L) with a negative AT-skew indicate mutational hotspots or regions with evolutionary constraints, particularly in the context of hybridization.

3.4. The Phylogenetic Relationship of Distantly Hybridized Fish

The phylogenetic relationships of distantly hybridized fish based on complete mitogenomes formed three clades: one clade included Acipenseriformes, another stable freshwater clade included Cypriniformes and Siluriformes, and the third saltwater clade included Pleuronectiformes, Scombriformes, Perciformes, Cichliformes, Tetraodontiformes, and Salmoniformes (Figure 6).

In the freshwater group, most hybrids belonged to Cypriniformes, which included 35 species, while the other four species belonged to Siluriformes. In the saltwater group, 27 species were randomly distributed across six orders, with Epinephelus of Perciformes having the highest number of hybrids (n = 14). Notably, hybrids from Centrarchiformes and Anabantiformes were included in the saltwater group, despite the fact that these fish are commonly farmed in freshwater. Until recently, comparative phylogenetic information on distantly hybridized fish has been rare, except for this study.

4. Discussion

Distant hybridization in fish has been extensively studied for its applications in aquaculture, species conservation, and genetic research [1,4]. Crosses between common carp (Cyprinus carpio) and crucian carp (Carassius auratus) often produce viable hybrids with enhanced growth rates [46,47]. Hybridization between Nile tilapia (Oreochromis niloticus) and blue tilapia (Oreochromis aureus) results in hybrids with superior growth performance and salinity tolerance [48,49]. Additionally, hybridization can introduce genetic diversity into farmed fish populations, thereby reducing inbreeding depression [50,51,52]. Despite its benefits, distant hybridization faces challenges such as low survival rates, sterility, and ethical concerns [1,53]. Hybridization between more distantly related species, such as sturgeons (Acipenser spp.) and paddlefish (Polyodon spathula), may produce sterile offspring due to chromosomal mismatches [54]. Although distant hybridization in fish has significant potential for genetic improvement and aquaculture development, precautions must be taken to mitigate ecological risks. Continued research on hybridization mechanisms, coupled with advanced biotechnologies, will enhance the efficiency and sustainability of fish breeding programs. In this study, we analyzed 74 hybrid fish based on their mitogenomes to identify genetic variations, maternal inheritance, and phylogenetic relationships.

The mitogenome of fish is a circular double-stranded DNA molecule that typically ranges from 15 to 20 kilobases (kb) in length [44]. It encodes essential genes for oxidative phosphorylation, ribosomal RNAs (rRNAs), transfer RNAs (tRNAs), and a non-coding control region (D-loop) involved in replication and transcription [55]. The size and structure of fish mitogenomes are highly conserved; however, variations exist due to gene duplications, deletions, or expansions in non-coding regions [56]. The mitogenomes of the hybrid fish in this study ranged from 15,973 to 17,114 bp, and no variations due to gene duplications, deletions, or expansions in non-coding regions were observed.

mtDNA is widely used to estimate genetic distances among fish populations and species due to its maternal inheritance, high mutation rate, and lack of recombination [12]. Genetic distance, typically measured as nucleotide divergence (e.g., p-distance, Kimura 2-parameter), provides insights into evolutionary relationships, speciation events, and population structures [57]. In fish mitogenomes, genetic distance is commonly calculated using protein-coding genes (e.g., COX1 and CYTB), ribosomal RNAs (rRNAs), and the control region (D-loop) [58,59]. The COX1 gene, a standard barcoding marker, typically shows 2–3% divergence between congeneric fish species [60]. While maternal inheritance of mtDNA is predominant in fish, paternal leakage has been documented in interspecific hybrid fish. For example, crosses between Cyprinus carpio and Carassius auratus produced offspring that exhibit paternal mtDNA inheritance [61], and a possible case was reported in the mangrove killifish (Kryptolebias marmoratus), a self-fertilizing hermaphrodite [62]. In this study, the genetic distances between the hybrids and their maternal fish were calculated using complete mitogenomes. Seventy-four percent of hybrid fish had less than 0.02 genetic distance, supporting the maternal inheritance of mitogenomes, with hybrids exhibiting the same genetic distance as their maternal fish. However, a significant proportion of hybrids (25.7%) showed more than 0.02 genetic distance, particularly in Cyprinus carpio haematopterus × Megalobrama amblycephala, Cyprinus carpio × Megalobrama amblycephala, Epinephelus akaara × E. lanceolatus, Epinephelus moara × Hyporthodus septemfasciatus, Misgurnus anguillicaudatus × M. bipartitus, and Scomberomorus munroi × S. semifasciatus, which exhibited over 10% variable sites in their maternal fish, possibly due to Doubly Uniparental Inheritance (DUI) or distantly related species [63]. Distant mtDNA lineages have also been reported in Scombridae fish, including S. munroi and S. semifasciatus, which exhibit paternal inheritance due to crossing [64]. While the process of passing sex-linked traits, such as DUI, is predominantly found in mollusks, recent findings in other aquatic species highlight the significance of mitochondrial genomics for evolutionary insights [65,66,67]. This phenomenon enhances our understanding of mitochondrial inheritance and its potential applications in fish breeding, including disease tolerance, energy metabolism, and species-specific adaptation. Additionally, it can facilitate selective fish breeding and produce new lines with desirable characteristics [68].

It is clear that purifying selection is a crucial force shaping the evolution of fish mitogenomes by removing deleterious mutations and maintaining functionality. The study observed this phenomenon in the PCGs of hybrid species with a genetic distance greater than 0.02 [69]. These genes, essential for energy metabolism, are under strong selective pressures [70]. Fish species, such as Cirrhinus reba, have demonstrated purifying selection in mitogenomes through codon usage bias and negative selection, which eliminates deleterious mutations [71]. Additionally, environmental stressors like climate change and pollution have intensified this evolutionary mechanism and the purifying selection of fish populations [72,73]. The AT-skew also reveals biases in nucleotide composition that influence genome stability, as noted in our study, with hybrids showing genomic integration or compensation compared to maternal mitogenomes [69]. We found that genes such as ATP8, COX1, and COX2 exhibited conserved genomes, while a negative AT-skew was observed in genes like ND6, resulting in mutational hotspots subject to purifying selection. Studies indicate that both mutational biases and selective pressures significantly shape the mitogenomic landscape of fish, providing valuable insights into their adaptive potential [74,75,76].

Hypervariable regions in mitogenomes with more than 0.02 genetic distance were randomly dispersed throughout the genome, with hotspots located in the D-loop, ND2, ND3, ND5, and ND6 regions, consistent with other studies [77,78,79]. These hotspots can serve as DNA markers to distinguish distant hybrid fish from their maternal counterparts. Furthermore, PCA revealed similar hotspots and other genomic features that differentiate hybrids from their maternal mitochondrial genomes, as well as regions specific to freshwater and saltwater fish.

Artificial breeding and aquaculture of fish have become essential for meeting global food demands, improving production efficiency, and enhancing disease resistance. Several fish families dominate aquaculture due to their adaptability, growth performance, and market value, including Cyprinidae [80], Salmonidae [81,82], Cichlidae [83,84], Percidae [85], Sparidae [86,87], and Serranidae [88]. In the phylogenetic analysis of this study, most hybrid fish belonged to families that are dominant in artificial breeding and aquaculture. Hybrids from Cyprinidae formed a monophyly, being the most studied among 35 species, largely due to China’s pivotal role in advancing genetic breeding, hybridization, and culture technologies for key species such as common carp (Cyprinus carpio), grass carp (Ctenopharyngodon idella), silver carp (Hypophthalmichthys molitrix), and bighead carp (Hypophthalmichthys nobilis). The 14 types of hybrid fish from Epinephelus of Perciformes are the most studied in saltwater aquaculture; Epinephelus species are highly sought after in Asian markets, where live fish trade commands prices 2–3 times higher than frozen products.

Interestingly, Centrarchiformes hybrids used Siniperca chuatsi as their maternal fish, whereas Anabantiformes hybrids used Channa argus or Channa maculata. These two groups formed separate monoclades and were included in the saltwater group, with the salinity adaptations of Siniperca chuatsi and Channa argus being reported [89,90]. Hybrid fish from Acipenseriformes formed another primitive monophyletic group. Acipenseriformes, which include sturgeons (Acipenseridae) and paddlefish (Polyodontidae), are among the most valuable fish species in global aquaculture due to their high economic and ecological significance. Their farming has expanded significantly in recent decades, primarily driven by the demand for caviar and meat, as well as conservation efforts [91]. Hybridization or the crossing of different species within the Acipenseriformes order has become a key strategy in aquaculture for enhancing production traits, improving adaptability, and supporting conservation efforts [92,93]. Hybrid sturgeons often exhibit heterosis (hybrid vigor), leading to faster growth rates, better disease resistance, and higher caviar yields than purebred species [94].

Our findings yield several policy recommendations for fish management and broodstock practices. For instance, a genetic distance threshold (e.g., 0.02) should be established for stocking approval in water bodies. We recommend mandatory mitogenome sequencing of broodstocks and offspring to confirm maternal inheritance and detect anomalies, as well as conducting an environmental risk assessment before introducing hybrids, with a careful evaluation of the ecological tolerance traits of species inhabiting mixed environments.

5. Conclusions

Taken together, our study provides a detailed analysis of fish distant hybridization utilizing mitogenomic data, highlighting maternal inheritance patterns, genetic divergence, and evolutionary insights among hybrid fish. The conserved genomic features in hybrids demonstrate the stability of mtDNA structure, while varying degrees of maternal lineage are evident from their genetic distances. In particular, a genetic distance > 0.02 was observed in 25.7% of hybrids, signifying the impact of genetic drift or rare paternal leakage events. The PCGs in hybrid mitogenomes for major metabolism are usually under purifying selection to maintain genomic stability. Additionally, the phylogenetic tree uncovered three clades, confirming major taxonomic and ecological divisions: Acipenseriformes, freshwater Cypriniformes, Siluriformes, and marine hybrids. Among these, Cyprinidae hybrids are prevalent in aquaculture, underscoring their prominence. In contrast, marine hybrids, such as Epinephelus spp., show potential for improving traits like growth vigor and disease resistance. Strong support exists for the maternal inheritance of mitogenomes, though exceptions suggest complex mechanisms, such as DUI, in distantly related crosses. These findings advance our understanding of fish hybridization dynamics and provide a genomic foundation for enhancing breeding programs and mitigating biodiversity risks.