Comparison of Mitochondrial Genome Sequences between Two Palaemon Species of the Family Palaemonidae (Decapoda: Caridea): Gene Rearrangement and Phylogenetic Implications

To further understand the origin and evolution of Palaemonidae (Decapoda: Caridea), we determined the mitochondrial genome sequence of Palaemon macrodactylus and Palaemon tenuidactylus. The entire mitochondrial genome sequences of these two Palaemon species encompassed 37 typical genes, including 13 protein-coding genes (PCGs), 2 ribosomal RNA genes (rRNAs), and 22 transfer RNA genes (tRNAs), and a control region (CR). The lengths of their mitochondrial genomes were 15,744 bp (P. macrodactylus) and 15,735 bp (P. tenuidactylus), respectively. We analyzed their genomic features and structural functions. In comparison with the ancestral Decapoda, these two newly sequenced Palaemon species exhibited a translocation event, where the gene order was trnK-trnD instead of trnD-trnK. Based on phylogenetic analysis constructed from 13 PCGs, the 12 families from Caridea can be divided into four major clades. Furthermore, it was revealed that Alpheidae and Palaemonidae formed sister groups, supporting the monophyly of various families within Caridea. These findings highlight the significant gene rearrangements within Palaemonidae and provide valuable evidence for the phylogenetic relationships within Caridea.


Introduction
The family Palaemonidae Rafinesque, 1815, belonging to the order Decapoda, infraorder Caridea, and superfamily Palaemonoidea, encompasses numerous economically valuable species and represents one of the largest taxonomic units at the family level within the true shrimp classification [1]. There are approximately 980 species in the world belonging to the family Palaemonidae, with the extant species inhabiting marine, estuarine, and freshwater environments [2]. These diverse habitats have contributed to high physiological, biochemical, morphological, and ecological diversity observed throughout the evolutionary history of the Palaemonidae family. Symbiosis is a widespread and crucial ecological process in nature, contributing significantly to biodiversity [3]. By enabling organisms to receive or exchange mutualistic services and access previously unreachable resources, symbiosis offers opportunities for expanding ecological niches and diversification [4,5]. The family Palaemonidae is one of the biological groups actively engaged in symbiosis, particularly in coral reef ecosystems, displaying a wide array of species, ecological roles, and morphological variations [6]. Some researchers argue that an evolutionary framework would facilitate a comprehensive understanding of the complexity of symbiosis in the Palaemonidae family [3]. However, the scope covered by the Palaemonidae family has been a subject of debate since its establishment. De Grave et al. [2] provided an overview of the most recent classification of the superfamily, and subsequent revisions have relied heavily on molecular techniques surpassing traditional morphological examinations. Multiple independent molecular studies have demonstrated the inclusion of Gnathophyllidae, Hymenoceridae, and Kakaducarididae nested within the Palaemonidae family [7][8][9][10]. These findings have prompted morphological reappraisals aimed at distinguishing family roles. For instance, the studies by Short et al. [11] and De Grave et al. [12] revealed shared morphological characteristics among these three families within the Palaemonidae family, thus considering them synonymous. The taxonomic status of the Palaemonidae family has also been a subject of interest in previous molecular systematic studies of the suborder Caridea. Most studies have proposed a close relationship between the Palaemonidae family and Alpheidae [13][14][15][16][17][18][19], but Li et al.'s research yielded different results [20]. Using five nuclear genes, they constructed both Bayesian inference (BI) and maximum likelihood (ML) phylogenetic trees, which indicated a closer affinity between Alpheidae and Hippolytidae, while Palaemonidae formed a separate branch. Moreover, the internal phylogenetic positions of various families within Caridea have also shown discrepancies in previous molecular systematic studies [13,14,[17][18][19].
The rapid advancement of modern molecular biology greatly propelled the research in molecular systematics. By studying molecular sequences to investigate the phylogenetic relationships between species, it was possible to effectively supplement the limitations of traditional taxonomy and address many contentious issues in the fields of classification and systematic evolution [17,21,22]. The mitochondrial genome, characterized by its simple structure, rich genetic information, ease of isolation, and maternal inheritance [23], has been widely utilized in research areas such as population genetic structure, species identification, and systematic evolution [18,24,25]. The expanded availability of complete mitogenomes has the potential to aid in unraveling the phylogeny of Palaemonidae. This can be accomplished by offering multiple loci with varying rates of evolution, thus enhancing our understanding of their evolutionary relationships. In terms of mitochondrial gene arrangement, gene rearrangements are commonly observed in the mitochondrial genomes of crustaceans [20]. Previous studies have also identified non-conservative gene arrangement patterns within the Palaemonidae family, highlighting the necessity of exploring the mitochondrial genome characteristics of this family [1,13,14]. These findings not only contribute to the field of systematic evolution, but also aid in the understanding of the molecular aspects of Palaemonidae. Despite the ecological and economic importance of Palaemonidae species, the available mitogenome data for Palaemonidae is currently quite limited. In GenBank, there are only 18 complete mitogenomes available (until July 5th, 2023, excluding unverified records) (https://www.ncbi.nlm.nih.gov/nuccore).
The Palaemon macrodactylus Rathbun, 1902 and Palaemon tenuidactylus Liu, Liang & Yan, 1990 belong to the family Palaemonidae. Both species have significant economic value, especially P. macrodactylus, which had been discussed by American scholar regarding its potential introduction methods [26], other researchers have also documented the expansion of its range [27][28][29]. Studies on P. macrodactylus mainly focused on its morphology [30], life history [31], ecological behavior [32], the influence of temperature and salinity on larval survival and development [33], and geographical distribution [34,35]. On the other hand, studies on P. tenuidactylus were relatively scarce, mainly concentrated on its morphology [30] and larval development process [36], with no reports on the complete mitochondrial genomes of these two species. Until now, only 10 species of the genus Palaemon with a complete mitogenome available in the GenBank database.
In this study, we conducted the complete mitochondrial genomes of P. macrodactylus and P. tenuidactylus. Our objectives are as follows: (1) to enhance taxonomic research methods and provide additional references for the molecular classification of Palaemonidae; (2) to analyze the characteristics and functions of the mitochondrial genomes of two Palaemon species, and gain insights into gene function through the assessment of AT skew values and relative synonymous codon usage (RSCU) of protein-coding genes (PCGs); (3) to investigate the patterns of mitochondrial gene rearrangements within Palaemonidae; (4) to elucidate the taxonomic position of the Palaemonidae family within Caridea; (5) to explore the phylogenetic relationships within Caridean shrimp.

Sampling, Identifcation and DNA Extraction
Samples of P. macrodactylus and P. tenuidactylus were collected from the coastal area of Zhoushan (122 • 50 N, 30 • 09 E), Zhejiang Province, in the East China Sea. Fresh specimens were preserved in 95% ethanol and transported to the laboratory for subsequent morphological identification by experts from the Marine Biology Museum of Zhejiang Ocean University, with reference to the sixth volume of "An Illustrated Guide to Species in China's Seas" [37]. Total genomic DNA was extracted from muscle tissue using a salt-extraction method and stored at −20 • C for sequencing [38].

Mitogenome Sequencing, Assembly, and Annotation
The complete mitochondrial genome sequences of P. macrodactylus and P. tenuidactylus were sequenced on the Illumina HiSeq X Ten platform by Origin gene Bio-pharm Technology Co. Ltd., Shanghai, China. Genomic DNA of the sample was first quality-checked, and after passing the quality control, 1 µg DNA was used to construct the library. The DNA was randomly fragmented into 300-500 bp fragments using a Covaris M220 ultrasonic disruptor, followed by end-repair, A-tailing, adapter ligation, purification, and PCR amplification to complete the library preparation, and each library produced approximately 10 Gb of raw data. The constructed library was sequenced using the sequencing by synthesis (SBS) technique with an Illumina HiSeq X Ten platform. After the trimming and quality control of the raw data using Cutadapt software [39], the preliminary assembly results were obtained using GetOrganelle (https://github.com/Kinggerm/GetOrganelle (accessed on 9 May 2023)) [40]. The best assembly results were obtained through multiple rounds of correction and iteration. The stack cluster was compared with the genomes of other Palaemon species in the GenBank and mitogenomic sequences were verified by checking the cox1 and 16S rRNA sequences using NCBI BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi (accessed on 9 May 2023)) [41]. Similar codons in other invertebrate species were compared to identify aberrant start and stop codons. Structural and functional annotation was performed using the online software MITOS (http://mitos2.bioinf.uni-leipzig.de/index.py (accessed on 10 May 2023)) [42] and manual corrections were made to obtain the final complete mitogenome. Finally, the sequenced mitogenomes were uploaded to the GenBank database at the National Center for Biotechnology Information (NCBI). The GenBank accession numbers for P. macrodactylus and P. tenuidactylus are OQ512152 and OP650931, respectively.

Sequence Analysis
The complete mitogenome was annotated using the Sequin software (version 15.10, http://www.ncbi.nlm.nih.gov/Sequin/ (accessed on 14 May 2023)). NCBI-BLAST was employed to determine the boundaries of protein-coding and ribosomal RNA genes. The correctness of transfer RNA genes and their secondary structures were verified using MITOS WebServer [42]. The base composition was analyzed using DAMBE 7 [43], while the nucleotide composition and relative synonymous codon usage (RSCU) of each proteincoding gene were calculated using MEGA-X [44]. To estimate the strand asymmetry, the formulas AT-skew = (A − T)/(A + T); GC-skew = (G − C)/(G + C) [45] were utilized. Additionally, the circular visualization of the mitogenomes of P. macrodactylus and P. tenuidactylus was performed using the CGView server (https://cgview.ca/ (accessed on 14 May 2023)) [46]. ORFfinder (https://www.ncbi.nlm.nih.gov/orffinder/ (accessed on 2 May 2023)) was used to find the ORFs (open reading frames) and determine the boundaries between genes.

Gene Order Analysis
In the mitochondrial genomes of Malacostraca, gene rearrangement is commonly observed [47]. The arrangement of mitochondrial genes serves as an important tool in systematic biogeography and phylogenetics, providing significant insights into the evolution of metazoans [48,49]. Currently, four models are primarily used to explain mitochondrial genome rearrangement: (1) duplication-random loss model, where genes are duplicated and individual copies are randomly lost or deleted; (2) tRNA gene-initiated replication errors, where replication starts at a tRNA and is retained as an incorrect replication origin, leading to gene rearrangement; (3) recombination, where gene order changes upon reconnection after double-strand breaks in the DNA; (4) replication-nonrandom loss, where gene duplication forms a dimer, and the loss of transcription promoter function in one set of the dimers leads to directional nonrandom loss or deletion of genes [50].
In addition to the two newly sequenced Palaemon species mitochondrial genomes sequenced in this study, we obtained an additional 18 complete Palaemonidae mitochondrial genomes from GenBank (Table 1) for comparative analysis. The gene arrangements of these 20 mitochondrial genomes were compared with the ancestral Decapoda [13,14] in order to investigate the gene rearrangement patterns within the Palaemonidae family. To ensure that observed differences in gene arrangement were not attributed to misannotations, any Palaemonidae mitogenomes that deviated from the ancestral pattern underwent reannotation using MITOS.

Phylogenetic Analysis
To explore the phylogenetic relationships within the Palaemonidae family, we downloaded sequences of 89 species from 12 Caridea families from GenBank (Table 1). We used the mitogenomes of Solenocera crassicornis (MF379621) and Metapenaeopsis dalei (NC_029457) from Dendrobranchiata as outgroups, and analyzed the phylogenetic relationships based on the 13 PCGs of these 93 species. We used DAMBE 7 software [43] to identify the sequence of the 13 PCGs from each downloaded sample. The nucleotide sequences for all 13 PCGs were individually aligned using the default settings of ClustalW [44] in MEGA X, and then concatenated by PhyloSuite [51]. Afterward, Gblocks v.0.91b [52] was employed with default parameter settings to remove divergent and ambiguously aligned blocks, selecting conserved regions. The substitution saturation was calculated using the GTR substitution model via DAMBE 7, and the third position of the codons was excluded from subsequent analyses due to saturation. We tested the selected DNA sequences for nucleotide models using jModelTest2.1.7 (https://code.google.com/p/jmodeltest2/ (accessed on 15 May 2023)) [53].
We employed two methods to analyze the phylogenetic relationships: the maximum likelihood (ML) method using IQ-tree 2.1.3 [54], and the Bayesian inference (BI) method using MrBayes 3.2.7a [55]. Two partitions (first and second codon positions of 13PCGs) were set in the combined data set for partitioned Bayesian analyses using MrBayes v.3.2, we used PAUP 4 [56] for format conversion, and then used a combination of PAUP 4, ModelTest 3.7 [57], and MRModelTest 2.3 [58] software in MrMTgui to determine the best alternative model (GTR + I + G) according to the Akaike information criterion (AIC). The BI tree analysis was performed using four Markov Chain Monte Carlo (MCMC) chains simultaneously running for 2 million generations, with a sampling frequency of every 1000 generations. In the first burn-in, 25% of trees were discarded, and convergence for independent operation was evaluated using the mean standard deviation of the splitting frequency (<0.01). All parameters for effective sample size (ESS) were checked using Tracer v.1.7 [59]. For ML tree building with IQ-TREE, the same dataset was used. We used ModelFinder [60] to select the best alternative model (TIM2+F+R10) for the ML tree based on the Bayesian Information Criterion (BIC). The consensus tree was reconstructed, and 1000 ultrafast likelihood bootstrap replicates were utilized. Finally, we edited the phylogenetic tree using FigTree v1.4.3 [61].
Both P. macrodactylus and P. tenuidactylus have 11 gene overlaps in their complete mitogenomes, as well as 16 and 17 gene gaps, respectively ( Figure 1, Table 2). The largest intergenic spaces of the two newly sequenced mitogenomes are 439 bp and 506 bp, respectively, located between the CR and trnI genes. Additionally, there are two relatively large intergenic regions, measuring 331 bp and 242 bp, respectively, located between the 12S rRNA and CR. Additionally, the maximum gene overlap in both mitogenomes is of 40 bp, between trnL1 and 16S rRNA ( Table 2). tRNA genes (trnL2, trnK, trnD, trnG, trnA, trnR, trnN, trnS1, trnE, trnT, trnS2, trnI, trnM and trnW), while the other 14 genes were located on the light chain ( Table 2). The CR was located between 12S rRNA and trnI in both of them, with a length of 180 bp for P. macrodactylus and 200 bp for P. tenuidactylus (Table 2).
Both P. macrodactylus and P. tenuidactylus have 11 gene overlaps in their complete mitogenomes, as well as 16 and 17 gene gaps, respectively ( Figure 1, Table 2). The largest intergenic spaces of the two newly sequenced mitogenomes are 439 bp and 506 bp, respectively, located between the CR and trnI genes. Additionally, there are two relatively large intergenic regions, measuring 331 bp and 242 bp, respectively, located between the 12S rRNA and CR. Additionally, the maximum gene overlap in both mitogenomes is of 40 bp, between trnL1 and 16S rRNA ( Table 2).

Protein-Coding Genes and Codon Usage
The total lengths of the PCGs in the P. macrodactylus and P. tenuidactylus mitogenomes were 11,101 bp and 11,127 bp, respectively. The A+T contents were 65.96% and 71.89% for P. macrodactylus and P. tenuidactylus, respectively, with AT-skews of −0.156 and −0.173 ( Figure 2B), indicating a clear bias towards T. The lengths of individual PCGs in both Palaemon species were consistent, except for the nad4 gene, as was their overlap. The longest PCG in both species was the nad5 gene, at 1713 bp, while the shortest was the atp8 gene, at 159 bp ( Table 2). In both Palaemon species, the atp8 and atp6 genes overlapped by seven nucleotides, atp6 and cox3 overlapped by one nucleotide, and nad6 and cytb overlapped by one nucleotide. The nad4 and nad4l genes of P. tenuidactylus overlapped by seven nucleotides (Table 2). Upon comparing the initiation and termination codons of all PCGs in the two Palaemon species, we identified four initiation codons and five termination codons ( Table 2). Most PCGs in both mitogenomes started with ATG, ATT, and ATA, except for the atp8 gene in P. tenuidactylus, which started with ATC. The cox1, cox2, atp6, cox3, nad4l, and cytb genes in both mitogenomes, as well as the atp8 gene in P. macrodactylus and the nad4 gene in P. tenuidactylus, all initiated with ATG. The nad3, nad5, nad2, and nad6 genes in both mitogenomes started with ATT, while the nad1 gene in both mitogenomes and the nad4 gene in P. macrodactylus started with ATA. The majority of the PGCs of the two mitogenomes were terminated with TAA and TAG, except for the cox1, cox2, cytb genes in P. macrodactylus, which stopped with CTA, ACT and ATT, respectively. Meanwhile, the cox1, cox2, cytb genes of P. tenuidactylus stopped with single T. Incomplete termination codons are a remarkably common phenomenon in mitochondrial genes of vertebrates and invertebrates [62].
Using MEGA-X, the amino acid content ( Figure 3A) and RSCU ( Figure 3B) of the two Palaemon mitogenomes were analyzed. The analysis revealed a relatively similar composition of amino acids in the PCGs for both species. Among the amino acids, Leu1, Lys, and Phe were the most frequently observed, while Arg and Cys were the least common ones. Analysis of codon preference showed that P. macrodactylus had 34 preferred codons (RSCU ≥ 1) out of the 13 PCGs, whereas P. tenuidactylus had 33 preferred codons out of the same number of genes. In the mitogenome of P. macrodactylus, the most frequently used codons, in descending order, were UUA (Leu), GGA (Gly), GUA (Val), and CGA (Arg). In the genome of P. tenuidactylus, the most frequently used codons, in descending order, were UUA (Leu), GUA (Val), ACU (Thr), and UCU (Ser). Both species had the lowest RSCU values for the codon GCG (Ala). In P. macrodactylus, except for the codons UGU (Cys), CGU (Arg), and AGU (Ser), all codons with A or U as the third base had RSCU values greater than 1. Similarly, in P. tenuidactylus, except for the codons CCA (Pro), GCA (Ala), AGU (Ser), and CGU (Arg), all codons with A or U as the third base had RSCU values greater than 1. Both species exhibited a preference for using the bases A and T in their codon usages.

Transfer and Ribosomal RNAs
In line with other Caridea mitogenomes, the mitogenomes of the two Palaemon species contained 22 tRNA genes. The total lengths of tRNAs in the mitogenomes of P. macrodactylus and P. tenuidactylus were 1442 bp and 1411 bp, respectively. The length of tRNAs in these species ranged from 61 to 68 bp ( Table 2). Apart from the trnD and trnR genes, the lengths of the other tRNA genes were consistent between the two species. All of the tRNA genes exhibited a high AT content, with P. macrodactylus having an AT content of 66.99% and P. tenuidactylus having an AT content of 73.78% (Figure 2A). The tRNA genes of P. macrodactylus and P. tenuidactylus displayed positive AT-skew (0.022 and 0.025, respectively) and GCskew (0.138 and 0.145, respectively) ( Figure 2B). The secondary cloverleaf structure of the 22 tRNAs from these species was examined. In both species, the trnS1 gene was unable to form a secondary structure due to the lack of dihydrouracil (DHU) arms, which is a common phenomenon in metazoans [63]. Similarly, trnA, trnF, trnM, and trnT were also unable to form a secondary structure due to the lack of a TΨC loop. The phenomenon of tRNA gene loss of the TψC loop has also been observed in the genomes of some previous metazoans [13,19,64]. However, the remaining tRNAs of two species were capable of folding into a typical cloverleaf structure (Figure 4). The secondary structure of tRNA allows for base mismatches, and all 22 tRNAs in the two Palaemon species exhibited four types of base mismatches. Among them, the G-U mismatch was the most common, with P. macrodactylus having 36 G-U mismatches and P. tenuidactylus having 35 G-U mismatches. Both Palaemon species had four A-A mismatches, three U-U mismatches, and five A-C mismatches. Except for trnF, trnH, trnL1, trnT, trnV, trnW, and trnY, 15 tRNAs showed consistent patterns of base mismatches in both species.
The total lengths of the 16S rRNA and 12S rRNA genes were found to be comparable between two species, as P. macrodactylus and P. tenuidactylus had total lengths of 1337 bp

Transfer and Ribosomal RNAs
In line with other Caridea mitogenomes, the mitogenomes of the two Palaemon species contained 22 tRNA genes. The total lengths of tRNAs in the mitogenomes of P. macrodactylus and P. tenuidactylus were 1442 bp and 1411 bp, respectively. The length of tRNAs in these species ranged from 61 to 68 bp ( Table 2). Apart from the trnD and trnR genes, the lengths of the other tRNA genes were consistent between the two species. All of the tRNA genes exhibited a high AT content, with P. macrodactylus having an AT content of 66.99% and P. tenuidactylus having an AT content of 73.78% (Figure 2A). The tRNA genes of P. macrodactylus and P. tenuidactylus displayed positive AT-skew (0.022 and 0.025, respectively) and GC-skew (0.138 and 0.145, respectively) ( Figure 2B). The secondary cloverleaf structure of the 22 tRNAs from these species was examined. In both species, the trnS1 gene was unable to form a secondary structure due to the lack of dihydrouracil (DHU) arms, which is a common phenomenon in metazoans [63]. Similarly, trnA, trnF, trnM, and trnT were also unable to form a secondary structure due to the lack of a TΨC loop. The phenomenon of tRNA gene loss of the TψC loop has also been observed in the genomes of some previous metazoans [13,19,64]. However, the remaining tRNAs of two species were capable of folding into a typical cloverleaf structure (Figure 4). The secondary structure of tRNA allows for base mismatches, and all 22 tRNAs in the two Palaemon species exhibited four types of base mismatches. Among them, the G-U mismatch was the most common, with P. macrodactylus having 36 G-U mismatches and P. tenuidactylus having 35 G-U mismatches. Both Palaemon species had four A-A mismatches, three U-U mismatches, and five A-C mismatches. Except for trnF, trnH, trnL1, trnT, trnV, trnW, and trnY, 15 tRNAs showed consistent patterns of base mismatches in both species.

Gene Rearrangement of the Family Palaemonidae
This study compares the gene arrangement patterns of 20 species in the family Palaemonidae with the ancestral gene arrangement pattern of Decapoda mitochondrial genomes. Among the five genera and twenty species in Palaemonidae (Table 1), five gene arrangement patterns were identified, with four patterns showing variation compared to the ancestral Decapoda ( Figure 5). In the two newly sequenced species (P. macrodactylus and P. tenuidactylus) of this study, the gene arrangement order was consistent with the majority of Palaemon species, but transpositions of trnT and trnP were observed compared to the ancestral gene sequence. This transposition of gene blocks is rare in shrimp but common in crabs. Previous studies have suggested that the transposition of trnT and trnP may be a shared phenomenon among Palaemon species [1,13]. However, with the enrichment of the GenBank database, it was discovered that the gene order in P. modestus is consistent with the ancestral sequence, providing new references for the gene arrangement patterns in Palaemon species. The gene arrangement sequences of four species in the genus Macrobrachium (M. nipponense, M. rosenbergii, M. bullatum, M. lanchesteri) were consistent with the ancestral sequence, indicating a conservative pattern in the evolution of Macrobrachium species [14]. A. brevicarpalis showed transposition events in 16S rRNA and trnV. The mitochondrial genome of H. picta exhibited a novel order where the gene fragment (nad1-trnL1-16S rRNA-trnV-12S rRNA-trnI-trnQ) was moved from downstream of trnS2 to the position downstream of nad4l [13]. A. australis exhibited the rare absence of The total lengths of the 16S rRNA and 12S rRNA genes were found to be comparable between two species, as P. macrodactylus and P. tenuidactylus had total lengths of 1337 bp and 1308 bp for 16S rRNA, and 803 bp and 801 bp for 12S rRNA, respectively ( Table 2). The 16S rRNA and 12S rRNA genes of two species were situated between trnL1 and trnI, separated by trnV. These genes exhibited high AT contents, with P. macrodactylus having an AT content of 73.04% and P. tenuidactylus having an AT content of 75.74% (Figure 2A). Additionally, both the rRNA genes of P. macrodactylus and P. tenuidactylus demonstrated negative AT-skew (−0.059 and −0.015, respectively) and positive GC-skew (0.285 and 0.398, respectively) ( Figure 2B).

Gene Rearrangement of the Family Palaemonidae
This study compares the gene arrangement patterns of 20 species in the family Palaemonidae with the ancestral gene arrangement pattern of Decapoda mitochondrial genomes. Among the five genera and twenty species in Palaemonidae (Table 1), five gene arrangement patterns were identified, with four patterns showing variation compared to the ancestral Decapoda ( Figure 5). In the two newly sequenced species (P. macrodactylus and P. tenuidactylus) of this study, the gene arrangement order was consistent with the majority of Palaemon species, but transpositions of trnT and trnP were observed compared to the ancestral gene sequence. This transposition of gene blocks is rare in shrimp but common in crabs. Previous studies have suggested that the transposition of trnT and trnP may be a shared phenomenon among Palaemon species [1,13]. However, with the enrichment of the GenBank database, it was discovered that the gene order in P. modestus is consistent with the ancestral sequence, providing new references for the gene arrangement patterns in Palaemon species. The gene arrangement sequences of four species in the genus Macrobrachium (M. nipponense, M. rosenbergii, M. bullatum, M. lanchesteri) were consistent with the ancestral sequence, indicating a conservative pattern in the evolution of Macrobrachium species [14]. A. brevicarpalis showed transposition events in 16S rRNA and trnV. The mitochondrial genome of H. picta exhibited a novel order where the gene fragment (nad1-trnL1-16S rRNA-trnV-12S rRNA-trnI-trnQ) was moved from downstream of trnS2 to the position downstream of nad4l [13]. A. australis exhibited the rare absence of the trnL2 gene in its mitogenome, along with transposition of the trnL1 gene with 16S rRNA gene, and transposition of the trnW gene with the gene block (trnC-trnY) [1].
Genes 2023, 14,1499 12 of 18 the trnL2 gene in its mitogenome, along with transposition of the trnL1 gene with 16S rRNA gene, and transposition of the trnW gene with the gene block (trnC-trnY) [1]. Among the 20 species in Palaemonidae, 15 species showed gene rearrangements compared to the ancestral gene sequence, indicating that gene arrangement is not conserved in Palaemonidae species. However, further systematic analysis is hindered by the limited availability of data for certain genera in the GenBank, such as Ancylocaris, Hymenocera, and Anchistus. More mitochondrial sequences of Palaemonidae species are needed in the future to explore the evolutionary relationships within this family. Figure 5. Linear representation of the mitochondrial gene arrangement of the ancestral mitogenome of pancrustaceans and Palaemonid species. In this study, the two newly sequenced species with gene rearrangements are marked with red star, and the rearranged gene blocks are signed by red gridlines and compared with the gene arrangement of ancestral Caridea.

Phylogenetic Relationships
Two Caridea phylogenetic trees were constructed using the sequences of 13 PCGs from the mitochondrial genomes, employing BI and ML methods ( Figure 6). The analysis included 89 species of Caridea shrimp, with a focus on P. macrodactylus and P. tenuidactylus, and using S. crassicornis (MF379621) and M. dalei (NC_029457) as outgroups for reference ( Table 2). The topologies of the phylogenetic trees constructed using the two methods showed slight differences, primarily manifested in the varying relationships between certain families. There were also subtle disparities in the support values of some branch nodes in both trees. The support values obtained by BI were generally higher than those obtained by ML, with most nodes having a support value of 1. The ML tree revealed the internal phylogenetic relationships among Caridea families as ((((Acanthephyridae + Oplophoridae) + Alvinocarididae) + Nematocarcinidae) + Atyidae) + Pandalidae) + Palaemonidae + Alpheidae) + (((Hippolytidae + Rhynchocinetidae) + Lysmatidae) + Thoridae)))))))), whereas the phylogenetic relationships inferred from the BI tree differed from the ML tree only in the case of four families: Hippolytidae, Rhynchocinetidae, Lysmatidae, and Thoridae. In the BI tree, the relationships of these four families were (((Hippolytidae + Rhynchocinetidae) + Thoridae) + Lysmatidae))). In order to Among the 20 species in Palaemonidae, 15 species showed gene rearrangements compared to the ancestral gene sequence, indicating that gene arrangement is not conserved in Palaemonidae species. However, further systematic analysis is hindered by the limited availability of data for certain genera in the GenBank, such as Ancylocaris, Hymenocera, and Anchistus. More mitochondrial sequences of Palaemonidae species are needed in the future to explore the evolutionary relationships within this family.

Phylogenetic Relationships
Two Caridea phylogenetic trees were constructed using the sequences of 13 PCGs from the mitochondrial genomes, employing BI and ML methods ( Figure 6). The analysis included 89 species of Caridea shrimp, with a focus on P. macrodactylus and P. tenuidactylus, and using S. crassicornis (MF379621) and M. dalei (NC_029457) as outgroups for reference ( Table 2). The topologies of the phylogenetic trees constructed using the two methods showed slight differences, primarily manifested in the varying relationships between certain families. There were also subtle disparities in the support values of some branch nodes in both trees. The support values obtained by BI were generally higher than those obtained by ML, with most nodes having a support value of 1. The ML tree revealed the internal phylogenetic relationships among Caridea families as ((((Acanthephyridae + Oplophoridae) + Alvinocarididae) + Nematocarcinidae) + Atyidae) + Pandalidae) + Palaemonidae + Alpheidae) + (((Hippolytidae + Rhynchocinetidae) + Lysmatidae) + Thoridae)))))))), whereas the phylogenetic relationships inferred from the BI tree differed from the ML tree only in the case of four families: Hippolytidae, Rhynchocinetidae, Lysmatidae, and Thoridae. In the BI tree, the relationships of these four families were (((Hippolytidae + Rhynchocinetidae) + Thoridae) + Lysmatidae))). In order to better explore the phylogenetic relationships among the families within Caridea, we compiled the research findings of previous studies and represented the simplified phylogenetic relationships of these families in Figure 7, including A [13], B [14], C [15,16], D&E [18], F [19], G [17], H [20] (Figure 7).
Genes 2023, 14, 1499 14 of 18 to the distinct computational methods used [65]. However, our research results regarding the fourth major branch do not align with Cronin et al.'s study [19], where the family Rhynchocinetidae formed a separate branch, while the families Hippolytidae, Lysmatidae, and Thoridae were most closely related ( Figure 7F). Considering the aforementioned research findings, the phylogenetic relationships within Caridea still pose certain questions due to the uneven representation of Caridea species data in GenBank. Some families, such as Rhynchocinetidae, Hippolytidae, Thoridae, Nematocarcinidae, and Oplophoridae, have limited mitochondrial genomic data. Future studies on Caridea phylogeny should incorporate more species from these relevant families to validate and support previous research findings. Figure 6. The phylogenetic tree based on 13 PCGs was inferred using Bayesian inference (BI) and maximum likelihood (ML) methods. The species with the same branches in the ML tree and the BI Figure 6. The phylogenetic tree based on 13 PCGs was inferred using Bayesian inference (BI) and maximum likelihood (ML) methods. The species with the same branches in the ML tree and the BI tree are replaced with black horizontal lines, and only species with different branches are displayed. The number at each branch is the bootstrap probability, and the two newly sequenced species are marked with red stars. Nodes in the ML tree with bootstrap support lower than 70 have been collapsed.   [13], (B) [14], (C) [15,16], (D,E) [18], (F) [19], (G) [17], (H) [20].

Conclusions
We sequenced the complete mitochondrial genomes of two Palaemon species and analyzed the fundamental characteristics of their gene sequences. We found that the genome size and nucleotide composition were similar to those in previous research findings. Among the 22 tRNA genes in these two species, the trnS1 gene was unable to form a secondary structure due to the absence of the DHU arm. Similarly, the trnA, trnF, trnM, and trnT genes were also unable to form secondary structures due to the absence of the TψC loop. However, the remaining tRNA genes in both species were able to fold into the typical cloverleaf structure. The gene arrangement in both Palaemon species underwent the same rearrangement pattern compared to the ancestral gene order of Decapoda, with a reversal occurring in the position of trnK-trnD. Additionally, a comparison of the gene arrangement patterns within Palaemonidae revealed a significant occurrence of extensive gene rearrangements. Phylogenetic analysis demonstrated that species from the 12 families formed separate clades, exhibiting a good level of monophyly. Our research results supported the division of these families into four major clades. Phylogenetic analysis also indicated that Acanthephyridae and Oplophoridae were sister groups, clustering with Alvinocarididae, while Alpheidae and Palaemonidae were sister groups. This study provides extensive information regarding the mitogenomes of Palaemon, laying a solid  [13], (B) [14], (C) [15,16], (D,E) [18], (F) [19], (G) [17], (H) [20].
The phylogenetic trees constructed using two methods revealed the evolutionary relationships within Caridea, at the species level, the newly sequenced Palaemon species showed a close relationship with P. gravieri, forming a clade of ((P. tenuidactylus + P. gravieri) + P. macrodactylus). Subsequently, the remaining Palaemon species clustered together, indicating a strong monophyly within the genus Palaemon. At the genus level, the observation revealed that Typhlatya and Caridina of Atyidae, as well as Plesionika of Pandalidae, were polyphyletic. Both Palaemon and Macrobrachium of Palaemonidae formed monophyletic groups. In the ML tree, their internal relationships were represented as ((Palaemon + Macrobrachium) + ((Hymenocera + Ancylocaris) + Anchistus)), while in the BI tree, they were depicted as ((Palaemon + Macrobrachium) + ((Anchistus + Ancylocaris) + Hymenocera)). Previous research conducted by Chow et al. employed four PCGs (H3, Enol, GAPDH, Nak) and three rRNA genes (16S, 12S, 18S) to construct ML and BI phylogenetic trees [6]. The topologies of the two trees were different, as their analysis included more Palaemonidae species. Their results suggested that Macrobrachium displayed a polyphyletic pattern, indicating the need for further verification regarding its monophyly. Additionally, their trees only included the four genera relevant to this study, and the inferred relationships among these genera were consistent with the present research. The discrepancies in topology between the two trees may arise from the different computational methods employed [65].
At the family level, species from each family formed a distinct clade, demonstrating the good monophyly of these families. These families then formed four major branches. The first major branch consisted of five families, with Acanthephyridae and Oplophoridae forming a sister group, subsequently clustering with Alvinocarididae, then with Nematocarcinidae, and finally merging with Atyidae to form a larger branch. This result is consistent with previous studies that used 13 PCGs to construct phylogenetic trees [18,19] ( Figure 7D,F). It is worth noting that the topologies of the BI and ML trees also differed slightly from those in Chak et al.'s study [18] (Figure 7D,E). Their BI tree supported the relationships among these five families, whereas the ML tree only supported the relationship of (((Acanthephyridae + Oplophoridae) + Alvinocarididae) + Nematocarcinidae), with Atyidae in a separate lineage. Furthermore, Wang et al.'s results also supported the relationships of these four families [17] ( Figure 7G). However, there is still some controversy regarding the evolutionary position of Atyidae. Specifically, our results conflict with those of Li et al. [20], who suggested that Atyidae represent basal lineages within Caridea based on five nuclear genes (Enolase, H3, NaK, PEPCK, 18S rRNA) [15]. Similarly, Bracken et al. inferred that Atyidae represent basal lineages within Caridea based on both mitochondrial and nuclear genes [66]. These differences may be due to the heterogeneity of the samples used. The second major branch represented the family Pandalidae, which was also supported by previous studies as monophyletic [13,14,[17][18][19] (Figure 7A,B,D,G). The third major branch consisted of the families Palaemonidae and Alpheidae, which formed a sister group. This relationship was also validated in previous studies [13][14][15][16][17][18][19] (Figure 7A-G). However, Li et al.'s study [20] suggested a closer affinity between the families Alpheidae and Hippolytidae, indicating differences possibly due to the heterogeneity of the samples used ( Figure 7H). Meanwhile, Li et al. [20] suggested that the family Palaemonidae was not a monophyletic group. Their study indicated that members of the families Hymenoceridae and Gnathophyllidae were clustered within the Palaemonidae. However, according to the latest records from WoRMS, the families Hymenoceridae and Gnathophyllidae have been updated to Palaemonidae (https://www.marinespecies.org/aphia.php?p=taxdetails&id=106788 (accessed on 22 May 2023)). The fourth major branch in both phylogenetic trees comprised the families Hippolytidae, Rhynchocinetidae, Lysmatidae, and Thoridae. Both results supported the closest relationship between Hippolytidae and Rhynchocinetidae. In the ML tree, these two families were closest to Lysmatidae, while in the BI tree, they were closest to Thoridae. This difference may be attributed to the distinct computational methods used [65]. However, our research results regarding the fourth major branch do not align with Cronin et al.'s study [19], where the family Rhynchocinetidae formed a separate branch, while the families Hippolytidae, Lysmatidae, and Thoridae were most closely related ( Figure 7F).
Considering the aforementioned research findings, the phylogenetic relationships within Caridea still pose certain questions due to the uneven representation of Caridea species data in GenBank. Some families, such as Rhynchocinetidae, Hippolytidae, Thoridae, Nematocarcinidae, and Oplophoridae, have limited mitochondrial genomic data. Future studies on Caridea phylogeny should incorporate more species from these relevant families to validate and support previous research findings.

Conclusions
We sequenced the complete mitochondrial genomes of two Palaemon species and analyzed the fundamental characteristics of their gene sequences. We found that the genome size and nucleotide composition were similar to those in previous research findings. Among the 22 tRNA genes in these two species, the trnS1 gene was unable to form a secondary structure due to the absence of the DHU arm. Similarly, the trnA, trnF, trnM, and trnT genes were also unable to form secondary structures due to the absence of the TψC loop. However, the remaining tRNA genes in both species were able to fold into the typical cloverleaf structure. The gene arrangement in both Palaemon species underwent the same rearrangement pattern compared to the ancestral gene order of Decapoda, with a reversal occurring in the position of trnK-trnD. Additionally, a comparison of the gene arrangement patterns within Palaemonidae revealed a significant occurrence of extensive gene rearrangements. Phylogenetic analysis demonstrated that species from the 12 families formed separate clades, exhibiting a good level of monophyly. Our research results supported the division of these families into four major clades. Phylogenetic analysis also indicated that Acanthephyridae and Oplophoridae were sister groups, clustering with Alvinocarididae, while Alpheidae and Palaemonidae were sister groups. This study provides extensive information regarding the mitogenomes of Palaemon, laying a solid foundation for future research into genetic variation, systematic evolution, and breeding of Palaemon using mitogenomes.