Background

In the speciose and morphologically dissimilar crustacean order Decapoda [22], spiny and slipper lobsters (infraorder Achelata) characteristically lack chelae on the first, second, and third pair of pereopods and exhibit a remarkable type of larva, the long-lived phyllosomata [41]. In this clade, many species are the target of subsistence, artisanal, and/or industrial fisheries in all subtropical and tropical oceans and in a few warm temperate coasts given that they often attain large body sizes and high population densities [41].

Among spiny lobsters (Achelata: Palinuridae), the Easter Island Panulirus pascuensis (Reed, 1954) or ‘Ura’ in the Rapa Nui language, is a little known species native to the south eastern Pacific Ocean, distributed along the coasts of Rapa Nui or Easter Island, Pitcairn Island, and the Salas y Gómez Ridge [41]. In Easter Island, P. pascuensis inhabits a variety of habitats between 2 and 200 m depth and is the target of a small and profitable but probably overexploited fishery (see [10]). The life history of P. pascuensis is poorly known, specially when compared to that of the Caribbean spiny lobster P. argus ([1, 3], and references therein) and western rock lobster P. cygnus ([72] and references therein).

Only a few genetic and genomic resources have been developed for P. pascuensis [23, 57, 62]. Using two mitochondrial gene fragments (i.e., large-subunit ribosomal RNA [16S] and cytochrome oxidase subunit I (cox1), Ptacek et al. [62] determined that P. pascuensis belonged to the P. japonicus clade, confirming previous studies based on the analysis of morphological traits [41]. Also, a microsatellite panel (n = 9 polymorphic SSRs) developed by Díaz-Cabrera et al. [23] was used to examine connectivity between populations in Eastern Island and the Salas y Gómez Island. Panulirus pascuensis exhibits high larval retention in the two studied islands as well as quite low and asymmetric larval connectivity between islands (greater from Salas y Gómez to Eastern Island - [57].

This study is part of a comprehensive program for the development of genetic and genomic resources in P. pascuensis and other species targeted by fisheries in the south eastern Pacific Ocean [5]. In this study, we profited from a series of bioinformatic analyses to mine biological insight from low-pass short-read next generation sequencing datasets. Using a series of bioinformatic tools tailored for retrieving information from such low coverage datasets, we have estimated genome size and ploidy in P. pascuensis using a k-mer strategy. We also discovered, annotated, and quantified mobile elements in the nuclear genome of the studied species. We assembled the 45S rRNA nuclear DNA cassette and mitochondrial chromosome. We have provided a detailed analysis of the latter genomic element. We explored the phylogenetic position of P. pascuensis within the genus Panulirus using the signal retrieved from translated mitochondrial protein coding genes. Lastly, we discovered a large set of microsatellites. We expect that the information generated in this study will guide the assembly of a gold-standard nuclear genome for P. pascuensis in the near future. The newly assembled 45S rRNA nuclear DNA cassette and mitochondrial chromosome can support bioprospecting and biomonitoring of P. pascuensis using environmental DNA. The same elements can help to survey the public market place and detect mislabelling of this and other spiny lobsters. Overall, the genomic resources generated in this study will aid in supporting fisheries management and conservation strategies in this iconic spiny lobster that is likely experiencing overexploitation.

Methods

Panulirus pascuensis specimen and DNA extraction, library preparation, and sequencing

The specimen used for sequencing belonging to P. pascuensis was deposited at the Clemson University Crustacean Collection. A small tissue sample (approx. 5 mm3) was dissected from a pereiopod and immediately stored in sterile centrifuge tubes containing ethyl alcohol (95%) that was shipped to Iridian Genomes, Inc. (Bethesda, MD) for genomic DNA (gDNA) extraction and next generation sequencing (NGS). gDNA was extracted from the tissue sample with the DNeasy Blood and Tissue Kit (Qiagen, Germany) using the manufacturer’s protocol. Library preparation was performed using the Illumina TruSeq kit following the manufacturer’s instructions. NGS was conducted in a Illumina HiSeq X Ten system (Illumina, San Diego, CA, USA) using a 2 × 150 cycle. A total of 74,534,229 pairs (PE) reads were produced by Iridian Genomes and are available in the short read archive (SRA) repository (Bioproject: PRJNA996211, BioSample: SAMN36530443; SRA accession number: SRR25340) at NCBI’s GenBank.

Cleaning and decontamination of raw reads

Illumina adapters and low quality sequences (Phred scores < 20) were removed from the dataset using the program fastp v.0.20.1 with default parameters [15]. Next, the clean set of reads was ‘decontaminated’ from viral archaeal, bacterial, fungal, protozoan, and human reads with the program Kraken2 v2.1.2 [75] and the database kraken2-microbial-fatfree (https://lomanlab.github.io/mockcommunity/mc_databases.html).

Genome size of Panulirus pascuensis

We calculated the genome size of P. pascuensis using the high quality and decontaminated set of reads using the program KMC 3 v. 3.2.1 [51] with 11 different k-mer sizes,21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51 bp long following Baeza et al. [5]. Then, the pipeline REPeat SPECTra Estimation (RESPECT) v.1.0.0 [64] was used to analyze each k-mer frequency distribution retrieved from each k-mer size and calculate nuclear genome size in P. pascuensis.

Ploidy in Panulirus pascuensis

We estimated ploidy in P. pascuensis in the software Smudgeplot v0.2.5 [63] specifically using the k-mer frequency distribution obtained with k-mer size = 21 in the program KMC. Immediately before the analysis of heterozygous k-mer pairs in the program Smudgeplot, we visually examined the coverage of the used k-mer frequency distribution in the web server GenomeScope (http://qb.cshl.edu/genomescope/genomescope2.0 - [63]) and selected for the analysis high coverage k-mers ranging between 20 × and 140 × .

Repetitive elements in the genome of Panulirus pascuensis

To discover, annotate, and quantify repetitive elements in the nuclear genome of P. pascuensis, we first mapped with the software HISAT2 v2.2.1 [50] the high quality and decontaminated set of PE reads to a newly assembled mitochondrial genome of the same species (see below). Next, we exclusively used those reads that did not map to the newly assembled mitochondrial genome (n = 72,688,900 PE reads [99.81%]) for the analysis of the nuclear ‘repeatome’ in P. pascuensis using the pipeline dnaPipeTE v1.4c [31, 32]. As a first step, DnaPipeTE assembled repetitive elements using the program Trinity [33] and subsequently annotated them based on homology with the pipeline RepeatMasker [29]. Lastly, DnaPipeTE quantified the abundance of repetitive elements by mapping a random sample of the reads onto the assembled repetitive elements. DnaPipeTE was executed using two iterations of the assembler Trinity with independent sets of read (sampled at 0.15X) each time [32]. For the analysis, we used the Protostomia-specific database of transposable elements from the consortium Dfam [43, 47]. Finally, we estimated the repetitive elements landscape of P. pascuensis when requesting dnaPipeTE to calculate the divergence (blastn) between transposable elements copies in the genomes (estimated from reads) and their respective assembled consensus sequences [32].

Nuclear ribosomal operon in Panulirus pascuensis

The nuclear ribosomal cassette or operon encodes the large (28S or lsrDNA) and small (18S or ssrDNA) nuclear rRNA genes along the 5.8S rDNA gene, two internal transcribed spacers (ITS1 and ITS2 that flank the 5.8S rDNA gene), and two external transcribed spacers (5′ ETS and 3′ ETS) [2]. This genomic element in the genome of P. pascuensis was assembled using the program CAP3 [42] as implemented on the platform RepeatExplorer 2.3.8 (http://repeatexplorer.org/ - [59, 60]). The exact coding positions of the small and large nuclear rDNAs along the boundaries of the 5′ and 3′ ETS were determined using RNAmmer with default parameters [53]. In turn, the boundaries of the ITS1 and ITS2 along the exact coding positions of the 5.8S nuclear rDNA were determined with the program ITSx [6].

Mitochondrial genome of Panulirus pascuensis

The mitochondrial genome of P. pascuensis was assembled ‘de novo’ with the software GetOrganelle v1.6.4 [45]. The complete mitochondrial genome of the congeneric Caribbean spiny lobster P. argus (GenBank’s accession number MH068821- [1] was used as a ‘seed’ during the assembly process that was run with k-mer sizes of 21, 55, 85, and 115 bp. Next, the program MITOS2 [24] as implemented in the platform Galaxy [70] was employed for annotating the newly assembled mitochondrial genome. To depict the newly assembled mitochondrial genome as a circular map, we used the web server Genome Vx (http://wolfe.ucd.ie/GenomeVx/— [17]). Nucleotide composition of the complete mitochondrial genome and specific genes were estimated with the program MEGA X [52]. The Codon Usage Tool available in the web server Sequence Manipulation Suite (https://www.bioinformatics.org/sms2/codon_usage.html— [68]) was used to estimate the codon usage profile of all concatenated PCGs. We also estimated Relative Synonymous Codon Usage (RSCU) using the tool EZcodon available in the web server EZmito (http://ezmito.unisi.it/— [19]).

To analyze selective pressures in each PCG encoded in the newly assembled mitochondrial genome, the program KaKs_calculator 2.0 [73] was employed. Specifically, using the aforementioned program, we calculated for each PCG the number of nonsynonymous substitutions per nonsynonymous site (dN), the number of synonymous substitutions per synonymous site (dS), and the ratio dN/dS (= ω). The observed ω ratio is expected to be equal to 1, < 1, or > 1, if a particular PCG is exposed to neutral selection, purifying (negative), or diversifying (positive) selection, respectively. We used the congeneric Caribbean spiny lobster P. argus as an outgroup (GenBank accession number MH068821.1 - [1]) and the γ-MYN model to reflect mutation rate variability along the studied sequences during the analysis.

The secondary structure of each tRNA gene detected in the studied mitochondrial genome was predicted with the software MITFI [46], as implemented in MITOS2, and visualized using the on-line platform Forna (http://rna.tbi.univie.ac.at/forna/; [49]).

Lastly, the control region (CR) of the newly assembled mitochondrial genome was described in detail. We detected the presence of Simple Sequence Repeats (SSRs or microsatellites) and tandem repeats in this region using the web servers Microsatellites Repeats Finder (http://insilico.ehu.es/mini_tools/microsatellites/— [9]) and Tandem Repeats Finder (https://tandem.bu.edu/trf/trf.html— [7]), respectively. Furthermore, we detected the existence (or not) of ‘stem and loops’ or ‘hairpins’ in the CR using the the web server RNAfold (http://rna.tbi.univie.ac.at//cgi-bin/RNAWebSuite/RNAfold.cgi— [36]).

Phylogenetic position of Panulirus pascuensis

We examined the phylogenetic position of P. pascuensis in the genus Panulirus using the phylogenetic signal retrieved from PCGs (nucleotides) within a maximum likelihood (ML) framework. We conducted the phylogenetic analysis using the newly assembled mitochondrial genome plus those of 12 congeneric and 4 other cofamilial species whose annotated mitochondrial genomes were available in NCBI’s GenBank. As outgroups, we included 7 other species belonging to family Scyllaridae, one species belonging to the infraorder Caridea, and one species belonging to the infraorder Astacidea. First, each set of PCG nucleotide sequences were aligned with the program MUSCLE [25, 67] as implemented in the software MEGA X. Next, poorly aligned regions in each PCG alignment were trimmed using the program GBlocks [12, 14]. Then, all alignments were concatenated and submitted to the web server IQ-TREE version 1.6.10 for ML analysis [56]. IQ-TREE used the program ProtTest [20] to partition the dataset and select the best fitting models of sequence evolution for each partition. The robustness of the ML tree topology was assessed using the Shimodaira–Hasegawa approximate likelihood ratio test ([SH]-aLRT) and 1,000 (ultra-fast) bootstrap iterations.

Results and discussion

Out of 74,534,229 PE raw reads produced by the sequencing facility, a total of 73,387,237 (98.46%) high quality PE clean reads remained after low quality sequences and Illumina adapters removal using the software fastp. The pipeline Kraken2 classified a total of 2.29% (n = 1,703,443) of the clean reads as contaminants (Supplementary Materials, Fig. S1). None of the reads were classified as human and among those classified as microbial (2.24%), bacterial (0.719%), viral (0.0309%), fungal (0.328%), and protozoan (0.406%), no overrepresentation of a specific taxon (species) was observed. We considered the sequenced specimen and dataset devoid of any parasite and/or pathogen that might have biased our downstream analyses.

Genome size of Panulirus pascuensis

The haploid genome size (GS) estimated for Panulirus pascuensis using an in-silico k-mer approach ranged between 2,752,094,588 bp (2.75 Gbp) using a kmer size equal to 51 and 3,387,547,447 (3.39 Gbp) using a kmer size = 18. Increases in k-mer word size resulted in a decrease in the estimated genome size. The difference in estimate is roughly 19% of the maximum genome size estimate.

Genome size has been calculated only in 5 representatives of the family Paniluridae using flow cytometry or static cell fluorometry (Animal Genome Size Database (https://www.genomesize.com/) – [34] [consulted on 03 03 2024]) and varies between 3.08 Gb in the Pink spiny lobster Palinurus mauritanicus (estimated using flow cytometry - [21]) and 5.43 Gbp in the Caribbean spiny lobster Panulirus argus (estimated with static cell fluorometry - [44]). Our estimate of genome size using a k-mer approach in Panulirus pascuensis is within the range (when using k-mer size = 18) or somewhat lower (when using k-mer size = 21–51) than that reported for cofamilial species. The relatively small genome size estimated for the studied species might be explained by the relatively small number of reads used during the analysis or due to the relatively large portion of repetitive elements in the nuclear genome of this species (see below). Genome size estimations are biased downwards when using small datasets and/or when transposable elements account for a large portion of a genome [2].

Ploidy in Panulirus pascuensis

Panulirus pascuensis was determined to be diploid after analyzing the abundance of heterozygous k-mer pairs with the program Smudgeplot (Fig. 1). Decapod crustaceans, including spiny lobsters in the Infra-Order Achelata are assumed to be diploid even though research on ploidy is rare in this clade [48, 58]. Studies focusing on ploidy estimation are common in other eumetazoan clades and ancient polyploidization events appear to have driven the reshaping of genomes and spur diversification and evolutionary innovations in other pan-arthropod clades [48, 58]. A recent study has demonstrated that a crayfish, the marmorkrebs Procambarus virginalis, belonging to the closely related infraorder Achelata, is a triploid [61]. Originally, it was thought that a triploid and clonal marmorkrebs population descended as recently as 25 years ago from a single specimen of the slough crayfish P. fallax in the laboratory [35]. However, recent studies have shown that the two parental haplotypes of P. virginalis were inherited from natural populations of P. fallax and that P. fallax triploids are relatively common in nature [35]. A hybrid origin has recently been reported for a few subspecies of spiny lobsters (i.e., P. homarus rubellus) but nothing is known about the ploidy of hybrid specimens [27]. Research examining ploidy in the genus Panulirus and other decapod crustaceans, as already conducted in other arthropod clades (e.g., in Trichoptera - [39]), is needed to understand how often polyploidy occurs in nature and the genomic consequences of such events in the Achelata and beyond.

Fig. 1
figure 1

Relationship between coverage of heterozygous k-mer pairs and normalized minor k-mer coverage in Panulirus pascuensis

Full size image

Nuclear repetitive elements of Panulirus pascuensis

The nuclear repetitive genome content in P. pascuensis ranged between 44% using a kmer size equal to 51 and 74% (with kmer = 18) as estimated by the program RESPECT. Increases in k-mer word size resulted in decreases (30% difference) in the nuclear repetitive genome content in our analysis. In turn, the pipeline dnaPipeTE determined that 57.64% of the genome in P. pascuensis comprised repetitive elements (Fig. 2), a value within the range estimated by the program RESPECT. Overall, repetitive elements comprise at least a half and a maximum of three fourths of the nuclear genome of Panulirus pascuensis. Repetitive content in the long-legged spiny lobster P. longipes and the Caribbean spiny lobster P. argus is 56.28% [2] and 69.02% [4], respectively. No estimation of repetitive content is available for P. ornatus, a third species of spiny lobster with an assembled genome available [71].

Fig. 2
figure 2

Transposable elements genome composition (circle graph) and landscape in the genome of Panulirus pascuensis

Full size image

DnaPipeTE didn’t annotate almost a third (64.94%) of the repetitive elements present in the nuclear genome of P. pascuensis; these ‘unknown’ repetitive elements were not assigned to any known family from the database of transposable elements specific to the Protostomia developed by the Dfam consortium (Fig. 2). Taking into consideration only annotated repetitive elements, the most abundant were classified as LINEs (Long Interspersed Nuclear Elements, 22.81%). Less common repetitive elements included LTRs (Long Terminal Repeats, 2.88%), Satellite DNA (2.66%), and DNA transposons (2.45%), among a few others (Fig. 3). The ‘repeatome’ has been characterized only in two other spiny lobsters, P. argus and P. longipes, and the portion of unannotated repetitive elements as well as repetitive element content in those species is similar to that reported for P. pascuensis in this study [2, 4]. Both in P. argus and P. longipes, LINEs and Satellite DNA were the most abundant annotated repetitive elements [2, 4]. Also, LINEs and LTRs comprise a large proportion of the genome in the few other decapods in which mobile elements have been quantified (e.g., in the Chinese mitten crab Eriocheir japonica sinensis - [69].

Fig. 3
figure 3

Circular map of the mitochondrial genome of Panulirus pascuensis

Full size image

Lastly, the repetitive elements ‘landscape’ of P. pascuensis estimated with the program dnaPipeTE exhibited a right-skewed distribution suggesting an expansion (burst) of repetitive elements in the recent past (Fig. 2). Other than in P. pascuensis, no previous studies have examined the repetitive elements landscape in the Achelata and other decapod crustaceans.

We argue in favor of future studies characterizing the ‘repeatome’ in spiny lobsters. Such studies likely will result in the discovery of numerous new mobile genetic elements, given the relatively high number of unannotated repetitive elements observed in this and other spiny lobster genomes [2, 4]. The same studies will advance the understanding of the role that these elements have in driving genome size (e.g., [37]), genomic architecture (e.g., [65]) and evolutionary innovations (e.g., [74]) in the Achelata and other decapod crustaceans. Furthermore, Casacuberta and González [13] suggested that repetitive elements can affect the ability of their hosts to react to environmental insults. To what extent the ‘repeatome’ plays a role in determining the ability of spiny lobsters, decapod crustaceans, and other marine invertebrates to acclimate and adapt to pervasive global change needs to be examined by future studies.

Nuclear ribosomal operon in Panulirus pascuensis

The 45S rRNA DNA operon of P. pascuensis was partially assembled by the pipeline RepeatExplorer2. One contig, 2,226 bp long, encoded a partially assembled 5′ ETS (length = 105 bp), the entire ssrDNA (1,861 bp, GenBank accession number: PP417726), and a partial ITS1 (260 bp). A second contig, 6,714 bp long, encoded a partially assembled ITS1 (172 bp), the entire 5.8S rDNA (158 bp, GenBank accession number: PP416763), ITS2 (897 bp, fully assembled), the entire lsrDNA (4,938 bp, GenBank accession number: PP417725), and a partial 3′ ETS (549 bp). The two contigs matched nuclear ribosomal sequences available in GenBank that belonged to the genus Panulirus with E-values < 1 × 10 − 6.

During the last decades, fragments of the 45S rRNA DNA cassette have been employed to reveal phylogenetic relationships at multiple taxonomic levels in crustaceans, including representatives of the infraorder Achelata [11]. Yet, research focusing on the organization and genomic localization of the nuclear ribosomal RNA gene is rare in this and other clades of decapod crustacean. The study of Yu et al. [77] in the Chinese mitten crab Eriocheir sinensis is one of a few exceptions. Considering that the 45S rRNA DNA cassette can be assembled (either partially, nearly completely, or on its entirety) using low-coverage sequencing using the strategy used in this or other studies (see [1, 4]), we argue that this element, coupled with mitochondrial genomes (also assembled from low coverages sequencing data, see below), can be used to explore phylogenetic relationships among closely related species at a fraction of a cost, especially when compared to strategies like ultraconserved elements [28], anchored hybrid enrichment [54], or single nucleotide variants (SNVs) mining from assembled nuclear genomes [38].

Mitochondrial genome of Panulirus pascuensis

The program GetOrganelle assembled a complete (circularized) mitochondrial genome for the Eastern Island spiny lobster P. pascuensis with an average coverage of 129.4 × and 539.3 × per k-mer and base pair, respectively. The mitochondrial genome of P. pascuensis (GeneBank accession number OR612316) was 15,613 bp long, AT-rich (A + T content = 63.02%), and encoded for 13 protein-coding genes (PCGs), 22 transfer RNA (tRNA) genes, and two ribosomal RNA genes (12S ribosomal RNA [rrnS] and 16S ribosomal RNA [rrnL]). Nine PCGs and 14 tRNA genes were located on one of the strands while 4 PCGs, the two ribosomal RNA genes, and 8 tRNA genes were encoded in the opposite strand (Table 1, Fig. 3). A single long intergenic region 688 bp long in the studied mitochondrial genome was assumed to be the Control Region (CR) or D-loop (Fig. 3, Table 1). The length of the mitochondrial genome assembled for P. pascuensis is within the range previously reported for congeneric species. In the genus Panulirus, the length of the mitochondrial genome varies between 15,739 bp in P. argus [1] and 15,665 bp in P. homarus [76]. Also, mitochondrial synteny in P. pascuensis is identical to that previously reported for the genus Panulirus and the infraorder Achelata [1] and references therein). Indeed, gene composition in the mitochondrial genome of P. pascuensis corresponds to the assumed Pancrustacean (Hexapoda + Crustacea) ground pattern [40].

Table 1 Mitochondrial genome of Panulirus pascuensis. Arrangement and annotation
Full size table

In the mitochondrial genome of P. pascuensis, eight out of the 13 PCGs started and terminated with canonical mitochondrial crustacean codons (Table 1). Cox1 exhibited the alternative (putative) start codon ACG while nad1, nad3, and nad6 exhibited the alternative (putative) start codons ATT (Table 1). Non-canonical mitochondrial start codons have been reported before for other decapod crustaceans, including spiny lobsters [1] and references therein). One gene (cox3) terminated with TGA while cob ended with the incomplete stop codon T. Incomplete stop codons are commonly observed in the mitochondrial genome of decapod crustaceans, including spiny lobster [1] and references therein) and these truncated termination codons are assumed to be completed via post-transcriptional poly-adenylation [66].

In the mitochondrial PCGs of P. pascuensis, codons were not used evenly. The most frequently used codons were AT-rich, and included TTT (Phe, n = 206 times used), ATT (Ile, n = 204), TTA (Leu, n = 180), GTT, (Val, n = 113), ATA (Met, n = 113), TTG (Leu, n = 112), TCT, (Ser, n = 111), and TTC (Phe, n = 105). In turn, other than stop codons, codons least frequently used were GC-rich and included CGC (Arg, n = 9), ACG (Thr, n = 12), CGG (Arg, n = 12), TGC, (Cys, n = 13), AGC (Ser, n = 14) (Supplementary Materials, Table S1). The RSCU analysis also indicated that among synonymous codons, most frequently used codons were AT-rich (Fig. 4). Our results are in line with those reported for P. argus, the only species of spiny lobster for which codon usage has been estimated [1].

Fig. 4
figure 4

Relative synonymous usage in the 13 protein coding genes encoded in the mitochondrial genome of Panulirus pascuensis

Full size image

The analysis of selective pressures in the mitochondrial PCGs of P. pascuensis indicated that all the studied genes are evolving under purifying selection; all estimated KA/KS ratios exhibited values < 1 (Table 2). The KA/KS ratios estimated for cob, cox1, and cox2 (KA/KS = 0.0042, 0.0015, and 0.0056, respectively) were one or two orders of magnitude lower than those estimated for the rest of the PCGs (Table 2). The aforementioned differences in KA/KS ratios suggest strong evolutionary constraints in the cob, cox1, and cox2 genes (Table 2). In spiny lobsters, selective pressure in mitochondrial PCGs has rarely been explored [1]. Nonetheless, widespread negative selection in mitochondrial PCGs s] is well documented in eumetazoans, including other decapod crustaceans [1].

Table 2 Selective pressure analysis of the protein coding genes in the mitochondrial genome of Panulirus pascuensis
Full size table

In the studied mitochondrial genome, tRNA genes ranged in length between 63 pb (tRNA-D) and 73 bp (tRNA-E) and all of them featured a standard ‘cloverleaf’ secondary structure with the exception of tRNA-Ser1 that was missing the D-arm stem and featured a short loop compared to other tRNA genes (Supplementary Materials, Fig. S2). In eumetazoans, including decapod crustaceans, either tRNA-Ser1 or tRNA-Ser2 are often reported as truncated and whether or not truncated tRNA genes are functional remains to be addressed in most eumetazoans, including decapod crustaceans and spiny lobsters [8].

In the mitochondrial genome of P. pascuensis, the 688 bp long CR is located between tRNA-I and the 12 S ribosomal RNA (Fig. 3). The region was AT-rich (70.64%) having an overall base composition equal to A = 36.19%, T = 34.45%, C = 18.75%, and G = 10.61%. The analysis conducted in the online tool Microsatellite repeats finder revealed the presence of four SRRs in the studied CR, all of them were dinucleotide SRRs repeated between 3 and 5 times and 4 out of the 5 SRRs were AT-rich (Supplementary Materials, Table S2). In turn, the online tool Tandem Repeat Finder failed to detect tandem repeats in the studied CR, in line to that observed in the Caribbean spiny lobster [1] but in disagreement to that reported for the Chinese spiny lobster Panulirus stimpsoni [55]. Lastly, prediction (using minimum free energy [MFE] and Centroid optimization) of the secondary structure in RNAFold revealed stem-loop structures along the entirety of the CR (Supplementary Materials, Fig. S2), similarly to that observed in other spiny lobster mitochondrial genomes in which a detailed analysis of this region has been conducted [1]. Additional detailed studies focusing on the organization of the CR are needed to understand its function during mitochondrial transcription and replication.

Phylogenetic position of Panulirus pascuensis

The ML phylogenetic analysis was based on 26 terminals, 11,103 characters, and 5,974 parsimony-informative sites. The analysis recovered the infraorder Achelata and the families Scyllaridae and Paniluridae as monophyletic (in all three cases, bootstrap support [bs] values: 98 < bs > 100). In the latter family, the genus Panulirus was recovered as monophyletic (bv = 100) and comprised two fully supported (bv = 100) sister clades, Lineage 1 and Lineage 2 sensu George [30]. In our analysis, species belonging to Lineage 1 included P. argus, P. cygnus, P. echinatus, P. interruptus, P. japonicus, P. longipes, and P. pascuensis, while Lineage 2 was represented by P. homarus, P. ornatus, P. penicillatus, P. polyphagus, P. stimpson, and P. versicolor. In Lineage 1, the Caribbean spiny lobster P. argus had an early branching position; it was sister to all other species in this clade that clustered together into a single fully supported clade. In turn, P. echinatus was arranged sister to a fully supported clade (bv = 100) containing the remaining species of Lineage 1 except P. argus. Panulirus pascuensis, the focus of this study, was sister to a fully supported clade comprising P. cygnus and P. longipes. In Lineage 2, P. versicolor was sister to a moderately supported clade containing all other species in this lineage. The relationships among the species comprising the latter clade were poorly resolved other than a fully supported sister relationship between P. homarus and P. ornatus (Fig. 5).

Fig. 5
figure 5

Maximum likelihood phylogenetic hypothesis for the family Palinuridae and phylogenetic placement of Panulirus pascuensis. The phylogenetic tree was retrieved using the phylogenetic signal provided by (translated) protein coding genes. Numbers above branches near nodes represent bootstrap pseudo-replicates (N = 1,000) of the tree search. Depiction of the studied species Panulirus pascuensis by Brooke Fitzwater (used with permission)

Full size image

Overall, the recovered phylogenetic relationships within the Achelata support findings by previous studies using a larger set of species but with a smaller number of markers [62] or a smaller set of species using entire mitochondrial genomes [1]. Also, in line with previous studies in other infraorders of decapod crustaceans (e.g., Caridea - [26]; Dendrobranchiata - [18]; Anomura - [16] our study suggests that mitochondrial genomes can reliably recover phylogenetic relationships at and below the family level in the Achelata. We argue in favor of additional studies sequencing mitochondrial genomes from low-pass coverage datasets in spiny and slipper lobsters to advance our understanding of the evolutionary history of this remarkable clade of decapod crustaceans.

Conclusion

In this study, we have generated genomic resources for P. pascuensis, a spiny lobster for which little is known but that likely plays an important ecological role and that is probably overfished in Rapa Nui. We used low-coverage short-read sequencing to determine ploidy and genome size of P. pascuensis. Also, we identified, classified, and quantified mobile elements in the nuclear genome of the studied species. Importantly, the large size and number of mobile elements in the nuclear genome of P. pascuensis implies that long-reads (i.e., Oxford Nanopore Technology and/or Pacific Biosciences) plus chromosome conformation capture techniques (i.e., Hi-C) will be needed to assemble a high-quality (i.e., chromosome-level) genome in this spiny lobster. We have also assembled the ribosomal RNA cassette and mitochondrial genome of the studied species. Overall, the genomic resources generated in this study will aid in supporting fisheries management and conservation strategies in this iconic spiny lobster that is likely experiencing overexploitation.