Hazard risk assessment of tropical cyclones based on joint probability theory

1. Introduction

With the rapid development of next-generation sequencing (NGS) technologies that can effectively analyze huge pools of molecular data, an increasing number of mitogenomes provide important insights into species evolution and phylogenetic relationships (Tan et al., 2018; Ruan et al., 2020, Yang et al., 2021). Generally, gene order in most vertebrate mitogenomes is considered conserved, e.g., less than 4% rearrangement ratio in fish mitogenomes (Li et al., 2013). However, extensive gene rearrangements have been observed in invertebrate mitogenomes (Wu et al., 2012; Liu et al., 2017; Jiang et al., 2018). Recent studies have shown that some of these rearrangements contain useful information for phylogeny, and many scholars have applied gene rearrangement as a new molecular marker in phylogenetic studies. For example, Akasaki et al. (2006) compared the gene rearrangement of subclass Coleoidea and proposed that the arrangements of mitochondrial genes in Oegopsida and Sepiida were derived from those of Octopoda. This conclusion is consistent with the results of their phylogenetic analysis based on mitochondrial genes. Through a comparative study of gene rearrangement and phylogenetic relationships of five species from the superfamily Tellinoidea, Yuan et al. (2012) suggested that the genus Sinonovacula should be placed in the superfamily Solenoidea instead of the superfamily Tellinoidea. Besides, Tan et al. (2018) compared the published mitogenome sequences of two infraorders (Anomura and Brachyura) and affirmed the potential value of using rearrangement information to investigate the phylogeny of Anomura.

In contrast, there are also some scholars consider that gene order is not suitable for phylogenetic reconstruction. For example, Xie et al. (2019) demonstrated that approach based on gene order alone is clearly inferior to sequence-based approaches to resolve major phylogenetic relationships. In their research, none of the relationships among major stylommatophoran groups were resolved in the gene order tree. Recently, Zhang et al. (2021b) reconstructed the phylogeny of Paguroidea based on both sequence data and gene order. The results indicated that gene order data did not seem to work well for phylogenetic analysis within families. From their gene order tree, the relationships within families are suspicious because two close relatives belonging to the same genus (Dardanus arrosor and D. aspersus) owned two different gene rearrangements. Of course, increasing the availability of mitogenomic data from different taxa will help to validate the applicability of gene order data in inferring phylogenetic relationships.

The infraorder Brachyura contains approximately 7 250 known species inhabiting marine, freshwater, and terrestrial habitats (Chen et al., 2018; Ma et al., 2019). Brachyura, as the oldest crab, originated in the Jurassic period (Schweitzer and Feldmann, 2010; Davie et al., 2015a), and a group of its members with extremely diverse morphology and ecology was finally formed after massive radiative evolution. However, the diversity has also caused remarkably challenges for species identification, and their real phylogenetic relationships remain controversial (Camargo et al., 2020; Tan et al., 2018). Grapsoidea and Ocypodoidea, two of the most abundant and economically important groups in Brachyura, are of commercial value to fisheries and aquaculture. However, the classification of Grapsoidea and Ocypodoidea has been controversial for a long time. Previous studies based on morphological features considered them to be monophyletic clades (Ng et al., 2008; Davie et al., 2015b). Recently, an increasing number of molecular studies have challenged the monophyly of these taxa (Chen et al., 2018, 2019; Lu et al., 2020). For example, molecular study of Wang et al. (2020) revealed that Ocypodoidea and Grapsoidea are divided into three clades, and similar findings were presented in Tan et al. (2018). Larger taxon samples are required to fully understand the phylogenetic relationships between Ocypodoidea and Grapsoidea in future studies.

Members of the family Camptandriidae Stimpson, 1858 are commonly found in the estuarine, mangrove mudflat, and open mudflat habitats in the Indo-West Pacific regions (Jones and Clayton, 1983). Species of this family share a distinct condition in the male first gonopod, in which the distal part is bent or twisted over the proximal base by about 180°, producing a strongly recurved structure (Naruse et al., 2015). Initially, this family was regarded as a subfamily of Ocypodidae. Subsequently, Camptandriinae was raised to the family level and a complete diagnosis of this family was carried out (Cheryl, 1997). According to WoRMS (http://www.marinespecies.org/), Camptandriidae consists of 42 species belonging to 24 genera. Most studies of this family focused on morphological features (Naderloo, 2017a; Trivedi et al., 2017). Although there are few researches on molecular level, most of them were based on partial mitochondrial gene sequences (Kitaura et al., 1998; Miura et al., 2007). To date, no complete mitogenome from Camptandriidae has ever been reported. The phylogenetic relationships among Camptandriidae and even the evolutionary status of this family have not been well resolved due to limited mitogenomic data.

Members of the family Macrophthalmidae Dana, 1851 occur throughout the Indo-West Pacific, with most of the known species living in intertidal habitat (Mendoza and Ng, 2007). The macrophthalmids are distinguished primarily by having antennules that fold transversely or obliquely, a narrow inter-antennulary septum, external maxillipeds that do not completely close the buccal cavern, and eyestalks that are usually elongate (Davie, 2002). Although this family had long been regarded as a subfamily of Ocypodidae, Kitaura et al. (2002) provided clear molecular evidence that it should be regarded as a distinct family. At present, it includes 86 species belonging to 13 genera. Similarly, the genus Euplax H. Milne Edwards, 1852, it was initially regarded as a subgenus of Macrophthalmus Desmarest, 1823. Afterward, McLay et al. (2010) updated it to a valid genus. According to WoRMS, the genus Euplax only contains two species, Euplax dagohoyi (Mendoza and Ng, 2007) and Euplax leptophthalmus (H. Milne Edwards, 1852). To date, only three complete mitogenomes of this family are available from the National Center for Biotechnology Information (NCBI) dataset, and all of them belong to the genus Macrophthalmus. The phylogenetic relationships among Macrophthalmidae have been poorly resolved.

Accordingly, in the present study, two newly sequenced mitogenomes of Ocypodoidea (C. dilatatum and Euplax sp.) were reported for the first time, one of which (C. dilatatum) is the first species in the family Camptandriidae whose complete mitogenome was sequenced. The characteristics of these two mitogenomes and the other 17 mitogenomes clustering in one branch of the phylogenetic tree were compared. Genome collinearity analysis of 19 mitogenomes showed that 18 of them shared the same gene rearrangement, while that of C. dilatatum mitogenome was consistent with the ancestral gene arrangement of Brachyura. Possible models were proposed to explain the current mitogenomic rearrangements. The phylogeny of Brachyura was reconstructed and the evolutionary status of Camptandriidae was revealed for the first time from the mitogenomic level. These results will not only enrich the mitogenomes of Ocypodoidea and mitogenomic rearrangements, but also lay a foundation for further evolutionary studies of Brachyura.

2. Materials and methods

2.1 Sampling, DNA extraction, mitogenome sequencing, and assembly

Specimens of C. dilatatum and Euplax sp. were collected from Jiangsu Province, China (34°47′48.80″N, 119°13′42.38″E) and Hainan Province, China (18°24′39.48″N, 109°58′20.60″E), respectively. Specimens were immediately preserved in 95% ethanol until DNA extraction. According to the key morphological features of crabs, these two specimens were identified with a stereo dissecting microscope (Naderloo, 2017a, 2017b). The SQ Tissue DNA Kit (OMEGA) was used to extract the total genomic DNA from muscle tissue following the manufacturer’s instructions. The genomic DNA was sent to Shanghai Origingene Biopharm Technology Co., Ltd. for library preparation and high-throughput sequencing. The libraries were constructed by using the VAHTS Universal Plus DNA Library Prep Kit, with an insert size of 150 bp. Paired-end sequencing with a read length of 150 bp was performed on an Illumina Hiseq 6000 platform. Adapters and low-quality bases were removed using cutadapt v1.16 (Martin, 2011) with the following parameters: q, 20; m, 20. Trimmed reads shorter than 50 bp were discarded. Quality control of raw and trimmed reads was performed using FastQC v0.11.5 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The filtered clean data were assembled and mapped to complete mitogenome sequence using NOVOPlasty v2.7.2 (Dierckxsens et al., 2017).

2.2 Mitogenome annotation and sequence analysis

The newly assembled mitogenomes of C. dilatatum and Euplax sp. were annotated using the software of Sequin (version 15.10, http://www.ncbi.nlm.nih.gov/Sequin/). The boundaries of protein-coding and ribosomal RNA genes were performed using NCBI-BLAST (http://blast.ncbi.nlm.nih.gov). Transfer RNA genes were manually plotted, according to the secondary structure predicted by the MITOS Web Server (Bernt et al., 2013) and tRNAscan-SE 1.21 (Lowe and Chan, 2016). The control region was determined by the locations of adjacent genes. Finally, circular mitogenome maps of C. dilatatum and Euplax sp. were drawn with the BLAST Ring Image Generator v0.95 (Alikhan et al., 2011).

The base composition and relative synonymous codon usage (RSCU) were obtained using MEGA X (Kumar et al., 2018). The strand asymmetry was calculated using the following formulas: AT-skew = (A − T)/(A + T); GC-skew = (G − C)/(G + C) (Perna and Kocher, 1995). Synteny analysis between the genomes was performed using Mauve v2.4.0 (Darling et al., 2004). To estimate the evolutionary-selection constraints on 13 PCGs, the nonsynonymous (dN) and synonymous (dS) substitution rates were calculated using Mega X. The genetic distances of 13 PCGs were also estimated using Mega X based on the Kimura 2-parameter (K2P) substitution model.

2.3 Phylogenetic analysis

The phylogeny of Brachyura was inferred based on 107 available complete mitogenomes and two newly determined ones (Table S1). The species Pagurus nigrofascia and P. gracilipes from Anomura were used as outgroups. PhyloSuite (Zhang et al., 2020a) was used to extract the nucleotide sequences of 13 PCGs for each of the above species from the GenBank files. The MAFFT program (Katoh et al., 2002) integrated into PhyloSuite was executed to align multiple sequences in normal-alignment mode, and ambiguously aligned regions were identified and moved by Gblocks (Talavera and Castresana, 2007). The alignments of individual genes were then concatenated and used to generate input files (Phylip and Nexus formats) for phylogenetic analysis. The best-fit models were selected by ModelFinder (Kalyaanamoorthy et al., 2017) based on the Bayesian Information Criterion (BIC). Phylogenetic trees were built under maximum likelihood (ML) and Bayesian inference (BI) methods. The ML analysis was carried out in IQ-TREE (Nguyen et al., 2015) using an ML+rapid bootstrap (BS) algorithm with 1000 replicates. The BI analysis was performed in MrBayes 3.2.6 (Ronquist et al., 2012) with default parameters and 3×106 Markov Chain Monte Carlo generations. The trees were sampled every 1000 generations with a burn-in of 25%. The average standard deviation of split frequencies below 0.01 was considered to reach convergence.

3. Results and discussion

3.1 General features of C. dilatatum and Euplax sp. mitogenomes

The complete mitogenomes of C. dilatatum and Euplax sp. are 15 444 bp and 16 129 bp in length, respectively (GenBank accessions MW191756 and MT176431; the order of the following data is the same as these) (Fig. 1; Tables 1 and 2). These two mitogenomes both contain a typical set of 37 genes (13 PCGs, 22 tRNAs, and two rRNAs) and a putative control region (CR). Nine PCGs and 14 tRNAs are encoded by the heavy (H-) strand, while the remaining genes are encoded by the light (L-) strand. There are 140 intergenic nucleotides dispersed in 13 locations in C. dilatatum mitogenome, and 537 intergenic nucleotides in 17 locations in Euplax sp. mitogenome; respectively. The longest one is 53 bp (between ND5 and ND4) and 169 bp (between ND4L and ND6) in these two mitogenomes (Tables 1 and 2). The base composition of C. dilatatum mitogenome is 34.4% A, 34.7% T, 11.4% C, 19.5% G, and that of Euplax sp. is 34.9% A, 34.0% T, 10.4% C, 20.7% G; the AT contents are 69.1% and 68.9%, suggesting a strong AT bias (Tables S2 and S3).

Figure  1.  Gene maps of Cleistostoma dilatatum (a) and Euplax sp. (b) mitogenomes. Genes encoded on the heavy or light strands are shown outside or inside the circular gene map, respectively.

DownLoad: Full-Size Img PowerPoint

Table  1.  Features of the mitochondrial genome of Cleistostoma dilatatum

Gene	Position		Length/bp	Amino acid	Start/Stop codon	Anticodon	Intergenic region	Strand
	From	To
COI	1 bp	1534 bp	1534	511	ATG/T	–	0	H
Leu (L2)	1535 bp	1599 bp	65	–	–	TAA	6	H
COII	1606 bp	2293 bp	688	229	ATG/T		0	H
Lys (K)	2294 bp	2363 bp	70	–	–	TTT	0	H
Asp (D)	2364 bp	2424 bp	61	–	–	GTC	1	H
ATP8	2426 bp	2584 bp	159	52	ATG/TAA	–	–4	H
ATP6	2581 bp	3252 bp	672	223	ATA/TAA	–	–1	H
COIII	3252 bp	4041 bp	790	263	ATG/T	–	0	H
Gly (G)	4042 bp	4105 bp	64	–	–	TCC	–3	H
ND3	4103 bp	4456 bp	354	117	ATT/TAA	–	4	H
Ala (A)	4461 bp	4525 bp	65	–	–	TGC	4	H
Arg (R)	4530 bp	4593 bp	64	–	–	TCG	0	H
Asn (N)	4594 bp	4662 bp	69	–	–	GTT	0	H
Ser (S1)	4663 bp	4729 bp	67	–	–	TCT	0	H
Glu (E)	4730 bp	4795 bp	66	–	–	TTC	2	H
His (H)	4798 bp	4862 bp	65	–	–	GTG	1	L
Phe (F)	4864 bp	4928 bp	65	–	–	GAA	–1	L
ND5	4928 bp	6643 bp	1716	571	ATT/TAA	–	53	L
ND4	6697 bp	8034 bp	1338	445	ATG/TAA	–	–7	L
ND4L	8028 bp	8330 bp	303	100	ATG/TAA	–	9	L
Thr (T)	8340 bp	8405 bp	66	–	–	TGT	0	H
Pro (P)	8406 bp	8470 bp	65	–	–	TGG	2	L
ND6	8473 bp	8976 bp	504	167	ATT/TAA	–	–1	H
Cyt b	8976 bp	10 110 bp	1135	378	ATG/T	–	0	H
Ser (S2)	10 111 bp	10 177 bp	67	–	–	TGA	15	H
ND1	10 193 bp	11 131 bp	939	312	ATA/TAA	–	34	L
Leu (L1)	11 166 bp	11 232 bp	67	–	–	TAG	0	L
16S	11 233 bp	12 546 bp	1314	–	–	–	0	L
Val (V)	12 547 bp	12 619 bp	73	–	–	TAC	0	L
12S	12 620 bp	13 435 bp	816	–	–	–	0	L
CR	13 436 bp	14 024 bp	589	–	–	–	0	H
Ile (I)	14 025 bp	14 090 bp	66	–	–	GAT	–3	H
Gln (Q)	14 088 bp	14 156 bp	69	–	–	TTG	8	L
Met (M)	14 165 bp	14 234 bp	70	–	–	CAT	0	H
ND2	14 235 bp	15 245 bp	1011	336	ATT/TAG	–	–2	H
Trp (W)	15 244 bp	15 315 bp	72	–	–	TCA	1	H
Cys (C)	15 317 bp	15 380 bp	64	–	–	GCA	0	L
Tyr (Y)	15 381 bp	15 444 bp	64	–	–	GTA	–1	L
Note: – represents no data.

 | Show Table

DownLoad: CSV

Table  2.  Features of the mitochondrial genome of Euplax sp.

Gene	Position		Length/bp	Amino acid	Start/Stop codon	Anticodon	Intergenic region	Strand
	From	To
COI	1 bp	1539 bp	1539	512	ATG/TAA	–	–5	H
Leu (L2)	1535 bp	1600 bp	66	–	–	TAA	8	H
COII	1609 bp	2296 bp	688	229	ATG/T	–	28	H
ATP8	2325 bp	2486 bp	162	53	ATT/TAA	–	–4	H
ATP6	2 483 bp	3154 bp	672	223	ATA/TAA	–	–1	H
COIII	3154 bp	3943 bp	790	263	ATG/T	–	0	H
Gly (G)	3 944 bp	4006 bp	63	–	–	TCC	–3	H
ND3	4004 bp	4357 bp	354	117	ATA/TAA	–	1	H
Ala (A)	4359 bp	4422 bp	64	–	–	TGC	1	H
Arg (R)	4424 bp	4487 bp	64	–	–	TCG	0	H
Asn (N)	4488 bp	4554 bp	67	–	–	GTT	0	H
Ser (S1)	4555 bp	4621 bp	67	–	–	TCT	6	H
Thr (T)	4 628 bp	4689 bp	62	–	–	TGT	16	H
Pro (P)	4 706 bp	4770 bp	65	–	–	TGG	10	L
ND1	4781 bp	5707 bp	927	308	ATA/TAG	–	33	L
Leu (L1)	5741 bp	5807 bp	67	–	–	TAG	0	L
16S	5808 bp	7170 bp	1363	–	–	–	0	L
12S	7171 bp	8048 bp	878	–	–	–	0	L
His (H)	8 049 bp	8113 bp	65	–	–	GTG	–1	L
ND5	8113 bp	9813 bp	1701	566	ATG/TAA	–	125	L
Val (V)	9939 bp	10 011 bp	73	–	–	TAG	0	L
CR	10 012 bp	10 806 bp	795	–	–	–	0	H
Gln (Q)	10 807 bp	10 875 bp	69	–	–	TTG	7	L
Cys (C)	10 883 bp	10 944 bp	62	–	–	GCA	0	L
Tyr (Y)	10 945 bp	11 010 bp	66	–	–	GTA	37	L
Lys (K)	11 048 bp	11 116 bp	69	–	–	TTT	0	H
Asp (D)	11 117 bp	11 182 bp	66	–	–	GTC	4	H
Glu (E)	11 187 bp	11 249 bp	63	–	–	TTC	–1	H
Phe (F)	11 249 bp	11 314 bp	66	–	–	GAA	7	L
ND4	11 322 bp	12 659 bp	1338	445	ATG/TAA	–	–7	L
ND4L	12 653 bp	12 955 bp	303	100	ATG/TAA	–	169	L
ND6	13 125 bp	13 649 bp	525	174	ATT/TAA	–	–20	H
Cyt b	13 630 bp	14 764 bp	1135	378	ATG/T	–	0	H
Ser (S2)	14 765 bp	14 830 bp	66	–	–	TGA	76	H
Ile (I)	14 907 bp	14 971 bp	65	–	–	GAT	2	H
Met (M)	14 974 bp	15 042 bp	69	–	–	CAT	0	H
ND2	15 043 bp	16 053 bp	1011	336	ATG/TAG	–	–2	H
Trp (W)	16 052 bp	16 121 bp	70	–	–	TCA	7	H
Note: – represents no data.

 | Show Table

DownLoad: CSV

All the 13 PCGs initiate with typical ATN codons in the two mitogenomes. The majority of PCGs terminate with TAA or TAG, while four PCGs in C. dilatatum mitogenome (COI, COII, COIII, and Cyt b) and three PCGs in Euplax sp. mitogenome (COII, COIII, and Cyt b) use a single T as a stop codon (Tables 1 and 2). Incomplete stop codons are common in metazoan mitogenomes and may be recovered via post-transcriptional polyadenylation (Ojala et al., 1981). The GC-skew values of nine PCGs (COI, COII, ATP8, ATP6, COIII, ND3, ND6, Cyt b, and ND2) are negative, indicating they are encoded by the H-strand, whereas the remaining four exhibit positive values, indicating they are encoded by the L-strand (Tables S2 and S3). The most frequently used amino acids are Leu and Ser. In comparison, the least common amino acids are Cys and Arg (Figs 2a, b). The RSCU values of each codon in these two mitogenomes are roughly identical (Figs 2c, d; Table S4). It is worth noting that the RSCU values for the codons NNU and NNA are usually greater than one, suggesting a strong AT bias in the third codon position.

Figure  2.  Amino acid composition in the mitogenome of Cleistostoma dilatatum (a) and Euplax sp. (b); relative synonymous codon usage in the mitogenome of C. dilatatum (c) and Euplax sp. (d). RSCU: relative synonymous codon usage.

DownLoad: Full-Size Img PowerPoint

Twenty-two tRNAs are scattered throughout the entire mitogenome (Tables 1 and 2). All of them can be folded into typical cloverleaf secondary structures except for S₁ in both two mitogenomes (Figs S1 and S2). The lack of DHU arm in S₁ is thought to be a common phenomenon in metazoan mitogenomes (Gong et al., 2020; Lu et al., 2020; Ruan et al., 2020). The 16S rRNA and 12S rRNA genes of C. dilatatum mitogenome are located between L₁ and V, V and CR, respectively. While Euplax sp. mitogenome shares different rRNA arrangements (L₁- 16S- 12S- H).

3.2 Comparative mitogenomic analysis clustering in one branch

To estimate the evolutionary-selection constraints on 13 PCGs in 19 mitogenomes, we perform dN/dS analysis for each PCG. It is commonly accepted that dN/dS>1, dN/dS=1, and dN/dS<1 generally indicate positive selection, neutral mutation, and purifying selection, respectively (Yang, 2006). All of the dN/dS ratios are lower than one (<1), indicating that all 13 PCGs are evolving under purifying selection. ATP8 gene exhibits a highly relaxed purifying selection with the highest dN/dS value (0.619), whereas COI gene exhibits the strongest purifying selection with the lowest dN/dS value (0.077) (Fig. 3). The lowest dN/dS value of COI gene indicates that this gene is bound by the protein-coding function and bears strong natural selection pressure, thus ensuring the normal function of its encoded protein, which means that COI gene has an important role in the survival and evolution of the above species. Besides, we conduct genetic distance analysis for 13 PCGs. COI gene possesses the least genetic distance (average 0.214), and ATP gene captures the largest value (average 0.409), representing the most conserved and variable genes, respectively (Fig. 3).

Figure  3.  Genetic distance (on average) and dN/dS substitution rates of 13 PCGs among 19 mitogenomes.

DownLoad: Full-Size Img PowerPoint

Genomic synteny analysis reveals that four large genomic homologous regions are prevalent in all 19 mitogenomes (marked A–D in Fig. 4). It is evident that the homologous regions B and C are rearranged in C. dilatatum mitogenome when choosing Eriocheir sinensis (Brachyura: Varunidae) mitogenome as the reference sequence (Fig. 4). The two homologous regions show a C-B order in C. dilatatum mitogenome, while that the remaining crabs display a B-C order (Fig. 4). Further analysis indicated that C. dilatatum mitogenome was consistent with the ancestral gene arrangement of Brachyura, while that of the remaining crabs shared exactly the same gene rearrangements.

Figure  4.  Multiple genome alignments of 19 mitogenomes. The mitogenome of Eriocheir sinensis at the top as the reference genome. All genomes are started from the COI gene. The number at the top of each genome shows nucleotide positions. Within each of the alignments, local collinear blocks are represented by blocks of the same color connected by lines.

DownLoad: Full-Size Img PowerPoint

3.3 Gene rearrangement

Gene arrangements in C. dilatatum and Euplax sp. mitogenomes are shown in Fig. 5. For C. dilatatum mitogenome, only a single H moves from the downstream of ND5 to downstream of E (Fig. 5A①) when compared with the gene order in ancestral crustaceans (the pancrutacean ground pattern) mitogenomes (Boore, 1999). In contrast, gene order in Euplax sp. mitogenome underwent large-scale gene rearrangements. At least nine gene clusters (or genes) significantly differ from the typical order, involving 12 tRNA genes (K, D, E, F, H, T, P, L₁, V, Q, C, and Y), two rRNAs (16S rRNA and 12S rRNA), one PCG (ND1), and a putative CR (Fig. 5B). Of these gene rearrangements, three tRNA gene pairs (K-D, E-F, and C-Y) and two single tRNA genes (V and Q) are moved into the ND5 and ND4 junction (Fig. 5B①②⑥⑧⑨), forming an eight-tRNA cluster (V-Q-C-Y-K-D-E-F) if CR is not considered. The CR is shifted from the typical area between 12S rRNA and I to the V and Q junction (Fig. 5B⑦). A single H gene, one tRNA gene pair (T-P), and the ND1- L₁-16S-12S gene cluster are moved to the position between S₁ and ND5 (Fig. 5B③④⑤).

Figure  5.  Gene arrangements in Cleistostoma dilatatum (A) and Euplax sp. (B) mitogenome.

DownLoad: Full-Size Img PowerPoint

Currently, four widely-accepted mechanisms have been used to account for mitogenomic rearrangements, including tandem duplication and random loss (TDRL) model (Moritz and Brown, 1987), intramitochondrial recombination model (Poulton et al., 1993), tandem duplication and non-random loss model (Lavrov et al., 2002), and double replications and random loss model (Shi et al., 2014). How did the gene orders in these two newly sequenced mitogenomes emerge? Here, we proposed that the TDRL mechanism resulted in the generation of these two mitogenomes. The hypothesized intermediate steps are as follows. Firstly, the F-ND5-H genes underwent a complete copy, forming a dimeric block, (F-ND5-H)-(F-ND5-H). Consecutive copies were then followed by a random loss of the duplicated genes, forming a novel H-F-ND5 gene order (Fig. 6B). The H-F-ND5 gene cluster is a common phenomenon in the mitogenome of ancestral and most living species of Brachyura (Lu et al., 2020; Zhang et al., 2020b), including Portunidae, Grapsidae, Ocypodidae, Leucosiidae, Eriphiidae, and the C. dilatatum mitogenome in this study. In the second rearrangement event, the gene block from K to Y underwent a complete copy, forming a dimeric block (K-D-ATP8-ATP6-COIII-G-ND3-A-R-N-S₁-E-H-F-ND5-ND4-ND4L-T-P-ND6-Cyt b-S₂-ND1-L₁-16S-V-12S-CR-I-Q-M-ND2-W-C-Y)-(K-D-ATP8-ATP6-COIII-G-ND3-A-R-N-S₁-E-H-F-ND5-ND4-ND4L-T-P-ND6-Cyt b-S₂-ND1-L₁-16S-V-12S-CR-I-Q-M-ND2-W-C-Y). Consecutive copies were then followed by a random loss of supernumerary genes, forming a new gene block, (K-D-ATP8-ATP6-COIII-G-ND3-A-R-N-S₁-E-F-ND4-ND4L-T-P-ND6-Cyt b-S₂-ND1-L₁-16S-12S-I-M-ND2-W-H-ND5-V-CR-Q-C-Y). In the following step, the newly formed gene block from K to Y underwent a second copy and likewise experienced a random loss of redundant genes. Finally, the ultimate gene arrangement in Euplax sp. mitogenome was generated (Fig. 6C), which is consistent with the ancestral gene arrangement of Varunidae and Macrophthalmidae (Wang et al., 2020). Summarily, all the rearrangement events mentioned above can be explained by TDRL model, which supposes that the rearranged gene order occurs via tandem duplications followed by random deletion of certain duplications (Moritz et al., 1987).

Figure  6.  Inferred intermediate steps between the ancestral gene arrangement of crustaceans and two newly sequenced mitogenomes. The ancestral gene arrangement of crustaceans (A); the results of one tandem duplication and random loss (TDRL) event, the ancestral gene arrangement in Brachyuran mitogenome, and the final gene arrangement in Cleistostoma dilatatum mitogenome (B); the results of two TDRL events, the ancestral gene arrangement in Varunidae and Macrophthalmidae mitogenomes, and the final gene arrangement in Euplax sp. mitogenome (C). The duplicated gene block is underlined and the lost genes are labeled with gray.

DownLoad: Full-Size Img PowerPoint

3.4 Phylogenetic analysis

The phylogenetic trees obtained using BI and ML methods resulted in identical topological structures except for supporting values. Here, only one topology (BI) with both support values was presented (Fig. 7). The results show that all Macrophalmidae species cluster together as a group, wherein Euplax sp. shows the closest relationship with Macrophthalmus darwinensi. Our phylogenetic trees firstly show the evolutionary status of Camptandriidae that it has the most closely related relationship with Macrophalmidae. These two families (Camptandriidae and Macrophalmidae) as a group then form a sister clade with Varunidae. Macrophalmidae and Varunidae sharing exactly the same mitogenomic rearrangements gather together in the phylogenetic tree, which is in consistence with most molecular results (Chen et al., 2018; Wang et al., 2020; Zhang et al., 2021a). Camptandriidae mitogenome, however, capturing the conserved gene arrangement (ancestral gene arrangement of Brachyura) forms a clade with the taxa that share the identically large-scale gene rearrangements. Similar phenomena have been reported in increasing number of crab mitogenomes (Tan et al., 2018; Li et al., 2020; Zhang et al., 2020c, 2021b). For instance, our recent work found that two closely related species belonging to the same genus (D. arrosor and D. aspersus) possessed two different gene rearrangements (Zhang et al., 2021b). More complex situations exist in Potamidae mitogenomes (Zhang et al., 2020c). Thus it echoes the viewpoint that the mitogenomic gene rearrangement is likely a continuous and dynamic process and may occur very recently even after speciation events (Zhang et al., 2021b). Of course, since here C. dilatatum is the only species of the family Camptandriidae, the phylogenetic status of Camptandriidae and the aforesaid thought-provoking hypothesis should be confirmed with more species.

Figure  7.  Phylogenetic tree of brachyuran species inferred from the nucleotide sequences of 13 PCGs based on maximum likelihood (ML) and Bayesian inference (BI) analyses. The node marked with a solid circle indicates 100 ML bootstrap support and 100% BI posterior probability. The numbers after the species name are the GenBank accession number.

DownLoad: Full-Size Img PowerPoint

Of the 30 families in our phylogenetic tree, except for Xanthidae, Gecarcinidae, and Homolidae, each family forms a monophyletic clade (Fig. 7). Regarding the non-monophyly of Xanthidae, four Xanthidae species are divided into two clades. Three of them cluster together as a clade, and the remaining one (Leptodius sanguineus) forms a sister clade with the single representative of the family Oziidae (Epixanthus frontalis), which calls attention to authoritative identification of these two species (L. sanguineus and E. frontalis). Of course, the increasing samples of Oziidae will also help to clarify the suspicious classification and relationships. For two Gecarcinidae species, one of them (Gecarcoidea natalis) forms a sister clade with Sesarmidae species, and then clusters with the remaining one (Cardisoma carnifex). As far as the non-monophyly of Homolidae, the single representative of Latreilliidae (Latreillia valida) forms a sister clade with a member of the family Homolidae (Moloha majora), which calls attention to authoritative identification of L. valida. Furthermore, it is worth noting that almost one-third of the families (11/30) include only one representative, so the non-monophyly of relevant families should be treated with caution.

Viewed from a higher taxonomic level, most superfamilies of Brachyura are found to be monophyletic, with the exception of Eriphioidea, Ocypodoidea, and Grapsoidea (Fig. 7). Although the polyphyly of the above three superfamilies is well supported in our phylogenetic tree, the interrelationships of these groups remain largely disputable. Regarding the interrelationships among Ocypodoidea and Grapsoidea, no consensus has been reached in current studies. For example, Sesarmidae (Grapsoidea) have a close relationship with Gecarcinidae (Grapsoidea), and Dotillidae (Ocypodoidea) form a sister clade with Grapsidae (Grapsoidea) in our phylogenetic tree. However, in Tan et al. (2018) , Sesarmidae (Grapsoidea) first clustered with Dotillidae (Ocypodoidea), and then formed a sister clade with Gecarcinidae (Grapsoidea). While in Wang et al. (2020) , Dotillidae (Ocypodoidea) and Xenograpsidae (Grapsoidea) formed a sister clade, and then clustered with Sesarmidae (Grapsoidea). These three families as a group then formed a clade with Gecarcinidae (Grapsoidea). Therefore, more sampling across a breadth of taxonomic groups and integration of additional molecular data need to be mined in order to substantially resolve the interrelationships of these groups.

4. Conclusions

In this study, two newly sequenced mitogenomes of Ocypodoidea, C. dilatatum and Euplax sp., were reported for the first time. TDRL model is proposed to be involved in the evolution of these two mitochondrial gene rearrangements. Comparative mitogenomic analyses of the species clustering in one branch in the tree display two types of gene arrangements. The dN/dS ratio analysis of all PCGs indicates that purifying selection plays a leading role in the evolution of mitochondrial PCGs. Phylogenetic analyses show that Camptandriidae and Macrophalmidae are the most closely related species, and the polyphyly of three superfamilies (Ocypodoidea, Eriphioidea, and Grapsoidea) is well supported. Nevertheless, large-scale taxonomic samplings are still needed to confirm the phylogenetic status of Camptandriidae and the non-monophyly of relative families due to limited representatives. Also, the authentic relationships within Brachyura will be better understood with the help of increasing samplings and data.