Black rockfish (Sebastes melanops Girard, 1856), Dark rockfish (Sebastes ciliatus Tilesius, 1813) and Dusky rockfish (Sebastes variabilis Pallas, 1814) are members of Sebastes (Cuvier, 1829), a diverse genus of marine fishes comprising more than 110 species (Fig. 1). These commercially important rockfishes are found in the North Pacific Ocean, with sympatric geographic ranges. Sebastes melanops schools over high relief rocky outcrops from 0-366 m, S. ciliatus schools over high relief on rocky reefs and in kelp forests from 5-160 m and S. variabilis schools over high-relief sea floors from 6-675 m (Butler et al. 2012).

Figure 1.  

Photographs of Black rockfish (Sebastes melanops), Dark rockfish (Sebastes ciliatus) and Dusky rockfish (Sebastes variabilis). Photos taken by Mark C. Belk.

Although S. ciliatus and S. variabilis were described separately in the early 1800s, they have long been considered a single variable species under the name S. ciliatus (Jordan and Gilbert 1881, Eigenmann and Beeson 1894). However, the presence of two colour morphs within S. ciliatus, with associated ecological differences, led to speculation that S. ciliatus consisted of a dark, shallow-water morph (S. ciliatus) and a light, deep-water morph (S. variabilis; Eschmeyer and Herald (1983), Kessler (1985)). Orr and Blackburn (2004) officially resurrected S. variabilis from S. ciliatus using morphological and meristic data, but molecular analyses have produced conflicting results. Genetic differences were identified in S. ciliatus using allozymes (Tsuyuki et al. 1965, Seeb 1986) and microsatellites (Orr and Blackburn 2004); however, it is unclear if these differences resulted from species-level separation or population-level differences resulting from geographic separation of the samples. Tsuyuki et al. (1965) did not provide specific location data for their nine samples of S. ciliatus and the samples analysed by Seeb (1986) did not come from sympatric populations. Conversely, mitochondrial DNA, specifically NADH dehydrogenase subunits, was not significantly different (Orr and Blackburn 2004). We report the complete mitogenomes of S. melanops, S. ciliatus, and S. variabilis to provide new insight into the taxonomic relationships amongst these species. We aimed to determine if the lack of resolution in mitochondrial DNA was limited by the small portion of mitochondrial DNA examined in previous studies.

Using hook-and-line sampling, we collected three rockfish specimens (Sebastes melanops, S. ciliatus and S. variabilis) from Frederick Sound, near Admiralty Island (57.307504, -134.133069) in 2018 and two rockfish specimens (S. ciliatus and S. variabilis) near Excursion Inlet, Alaska (58.3159, -135.4592) in 2019 (IACUC-approved protocol #15-0602). Upon capture, we euthanised specimens with tricaine methanesulfonate (MS-222, MilliporeSigma, St. Louis, MO, USA), excised liver samples and placed the samples in RNAlater (MilliporeSigma, St. Louis, MO, USA). Samples were flash-frozen at -20°C, transported to Brigham Young University and stored at -80°C. Samples were catalogued in the Monte L. Bean Life Science Museum under accession numbers: S. melanops (BYU:1003048), S. ciliatus (BYU:1003050, BYU:267108) and S. variabilis (BYU:1003082, BYU:267107), (Table 1). Morphological vouchers were retained for S. ciliatus and S. variabilis collected in 2018 (BYU:1003050 and BYU:1003082, respectively). Total DNA was extracted from 40 mg liver samples using a DNeasy Blood & Tissue kit (Qiagen, Hilden, Germany). DNA concentration was measured by a Qubit Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). Libraries of each sample were created according to the Illumina Library prep protocol (Illumina 1003806 Rev. A) and sequenced on the Illumina HiSeq 2500 (Illumina, San Diego, CA, USA; Paired End 150 bp) at Brigham Young University’s DNA Sequencing Center (Provo, Utah, USA). We used FastQC (Andrews 2010) to assess quality of raw reads. Mitogenomes were assembled with Geneious v. 2021.2 (Biomatters Ltd., Auckland, New Zealand) using S. fasciatus as a reference genome (KX897946) and annotated with MitoAnnotator (Iwasaki et al. 2013). Raw reads and assembled genomes were deposited in NCBI's Sequence Read Archive (raw data) and GenBank nucleotide (assembled mitogenomes) databases (Table 1). Mitogenome maps were generated using Circos (Krzywinski et al. 2009). We included 35 rockfish mitogenomes, representing 27 species, in our phylogenetic analysis. We used MAFFT v. 7.475 (Katoh et al. 2002) to generate multiple sequence alignments for each of the 13 protein-coding genes and concatenated the alignments. We generated a Maximum Likelihood phylogeny with W-IQ-Tree (Trifinopoulos et al. 2016).

The complete mitochondrial genome of Sebastes melanops (OK048741) was 16,405 bp in length, S. ciliatus (MZ420215, OK048740) were both 16,400 bp in length and S. variabilis (OK048743, OK048742) were 16,400 and 16,401 bp, respectively, in length. Consistent with previous studies (Zhang et al. 2012, Sandel et al. 2018, Campbell et al. 2022), the control region’s length was highly variable because of repetitive DNA sequences (Fig. 2). The complete mitogenomes of S. ciliatus and S. variabilis were ~ 0.5% divergent. In comparison, the complete mitogenome of S. melanops was between 6.2% and 10.6% divergent with other members in its clade. In our phylogeny, S. melanops is sister to a clade including S. rubrivinctus, S. nigrocinctus, S. umbrosus and S. oculatus, whereas S. ciliatus and S. variabilis are sister to a clade including S. norvegicus, S. viviparus, S. mentella and S. fasciatus. We were unable to resolve the phylogenetic relationship between S. ciliatus and S. variabilis (Fig. 3).

Figure 2.  

Mitogenome map of Sebastes melanops. Outer circle illustrates order of genes, tRNAs, rRNAs and control region. Inner circle represents GC content with darker shades indicating higher GC content. Sebastes melanop's mitogenome consists of 13 protein-coding genes, 22 tRNAs, two rRNAs and one control region. Order is identical in S. ciliatus and S variabilis (mitogenome maps not displayed).

Figure 3.  

Phylogenetic tree inferred by Maximum Likelihood using W-IQ-Tree. Thirty-five Sebastes mitogenomes, representing 27 species, were used in the phylogeny. Ultrafast bootstrap values > 95 are not displayed and Sebastiscus tertius (MT117231) was used as an outgroup, but is not shown.

Previous molecular analyses of allozymes, microsatellites and mitochondrial DNA have produced inconsistent results about the relationship of Sebastes ciliatus and S. variabilis (Tsuyuki et al. 1965, Seeb 1986, Orr and Blackburn 2004). However, in these studies, it is unclear if these differences result from species-level or population-level differences. In addition, for the studies that used mitochondrial DNA, only a small portion of mitochondrial DNA was examined (Orr and Blackburn 2004). By using samples of S. ciliatus and S. variabilis from sympatric populations, as well as generating whole mitochondrial genomes, we provide new insight into the status of these species. Consistent with previous studies using partial mitochondrial sequences, we found minimal sequence divergence between S. ciliatus and S. variabilis. Assuming S. ciliatus and S. variabilis are distinct sister species, we would expect greater sequence divergence, as well as a monophyletic relationship in our phylogeny, with higher bootstrap values. Further research is needed. This research should include specimens from a wider range of locations across their geographical ranges, with both allopatric and sympatric populations, a suite of genetic markers (nuclear and mitochondrial), as well as ecological and morphological characteristics. Such information will be essential in resolving the complicated relationships between these two putative species.

We thank Scott and Jody Jorgenson at Pybus Point Lodge, Mark and Kristina Warner and staff at Doc Warner’s lodge, the Roger and Victoria Sant Foundation, the College of Life Sciences, the Department of Biology at Brigham Young University and the Monte L. Bean Life Science Museum for supporting this study.

Department of Biology, Brigham Young University, Provo, UT, United States of America

All samples were collected under Brigham Young Universities’ IACUC-approved protocol #15-0602 for Dennis K. Shiozawa.

Conceptualisation: P.C.S., A.L.K., D.K.S. & R.P.E.; Data Curation: P.C.S., A.L.K., & J.R.C.; Formal Analysis: P.C.S. & J.R.C; Funding Acquisition: D.K.S., M.C.B., & R.P.E.; Investigation: P.C.S., A.L.K., J.R.C., D.K.S., & R.P.E.; Methodology: P.C.S., A.L.K., J.R.C., D.K.S. & R.P.E.; Project Administration: R.P.E.; Resources: D.K.S., M.C.B., & R.P.E.; Supervision: D.K.S., M.C.B., & R.P.E.; Validation: D.K.S., M.C.B., & R.P.E.; Visualisation: P.C.S. & J.R.C.; Writing – Original Draft: A.L.K.; Writing – Reviewing & Editing: P.C.S., J.R.C., D.K.S., M.C.B., & R.P.E.

The authors declare no potential conflict of interest and the authors alone are accountable for the content and composition of the paper.