Morphological evidence supports splitting of species in the North Atlantic Sebastes spp. complex

Ingrid Marie Bruvold; Agneta Hansen; Arve Lynghammar; Hannes Höffle; Tanja Hanebrekke; Caroline Aas Tranang; Kjell Nedreaas; Einar Nilssen; Atal Saha; Torild Johansen

doi:10.1371/journal.pone.0316988

Abstract

The redfishes (genus Sebastes) are long-lived, commercial species in the North Atlantic. Excessive harvest through decades has led to a decline in the mature population of golden redfish (Sebastes norvegicus) in Norwegian waters, which is currently considered severely depleted. Accumulating genetic evidence suggests a more complex structure within this genus in the North Atlantic, which has recently inspired the hypotheses of cryptic species within S. norvegicus. Despite apparent genetic divergence between two types, they have yet to be verified morphologically. The morphology of genetically assigned fishes from Norwegian and Greenland waters was investigated using traditional morphometric methods, applying Linear Discriminant Analysis and Random Forest classification procedures to identify and evaluate the performance of descriptive characters. Combined with non-parametric meristic analysis, the results show that features such as beak length and eye diameter provide sufficient discrimination between the proposed cryptic species as well as separating them from the sympatric species S. mentella and S. viviparus. These findings support the presence of an additional redfish species in the North Atlantic, distinguishable both by morphological and genetic characters. This needs to be taken into consideration in future monitoring and management strategies for North Atlantic redfish.

Citation: Bruvold IM, Hansen A, Lynghammar A, Höffle H, Hanebrekke T, Tranang CA, et al. (2025) Morphological evidence supports splitting of species in the North Atlantic Sebastes spp. complex. PLoS ONE 20(2): e0316988. https://doi.org/10.1371/journal.pone.0316988

Editor: Peter Eklöv, Uppsala Universitet, SWEDEN

Received: May 31, 2024; Accepted: December 19, 2024; Published: February 6, 2025

Copyright: © 2025 Bruvold et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: We have published the dataset used, which can be freely accessed at the following DOI: https://doi.org/10.21335/NMDC-1910512771

Funding: This work was funded through the Institute of Marine Research financed by the Ministry of Trade, Industry and Fisheries via the projects “Stock complexes in the Barents Sea” (Barents Sea program) and “Deepwater species” (Program for the Norwegian Sea), as well as UiT – the Arctic University of Norway. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The genus Sebastes (commonly known as redfishes) is a diverse group of mainly demersal fishes represented by approximately 110 species across the Pacific and Atlantic Oceans [1]. Only four species are recognized in the North Atlantic: Sebastes norvegicus (Ascanius 1772, formerly S. marinus) and S. mentella Travin 1951 are distributed across the North Atlantic from the east coast of North America to Novaya Zemlya (Russia), while the sister species S. fasciatus Storer 1854 and S. viviparus Krøyer 1845 are found in the western and eastern parts of the North Atlantic, respectively [2,3].

The evolutionary origin of the Sebastes species has been a topic of great interest among biologists due to the large diversity in morphology and life history traits within the genus [4]. Their diversification from ancestral Pacific populations into the four North Atlantic species is a relatively recent event on the geological time scale. Sebastes viviparus branched off from the basal lineage less than a million years ago followed by S. fasciatus, S. norvegicus, and S. mentella [5]. The phylogenetic relationship between the species has been problematic to establish. They share highly similar external morphology, and the classification of S. mentella and S. norvegicus as separate species was still a topic of discussion in the 1960’s [6]. The North Atlantic Sebastes species are long-lived, slow growing, and mature at a late age [7]. Whilst lifespans of over 60 years have been recorded for S. norvegicus, S. mentella can live for over 70 years [8,9]. Sebastes viviparus is the smallest species of the three [10], reaching a maximum age of 40 years [11].

Correct species identification is vital for sustainable fishery management. Historically, both S. norvegicus and S. mentella have been commercially important to countries such as Norway, Germany, Russia, Greenland, and Iceland, with the latter performing the majority of redfish harvest in the North Atlantic [12]. Generation times of Sebastes species can exceed a decade [13], making them particularly vulnerable to overfishing. Direct fishery for both species in Norwegian waters has been unsustainable, causing population declines over the past decades. Consequently, direct fishery for S. mentella and S. norvegicus was prohibited in Norwegian waters from 2003 and 2015, respectively [14]. The Northeast Atlantic stock of S. mentella has since been rebuilt while S. norvegicus is still considered to be severely depleted with a reduced spawning stock and poor recruitment [15]. Since the fishery for S. mentella resumed, morphological misclassification between the superficially similar species has become an issue, potentially driving underestimated bycatch rates of S. norvegicus [16].

All Sebastes species share an ovoviviparous reproductive mode with internal fertilization [13,17] which has been suggested to be a contributing factor for complex mating behaviors. Introgressive hybridization, where genetic flow between hybrid offspring and parent species occurs through backcrossing, has been documented within the North Atlantic Sebastes genus [18–21] particularly between S. mentella and S. fasciatus (see [22]) and between S. mentella and S. viviparus (see [23]). Interspecific hybridization and its effect on genetic population structure is suspected to be extensive [23] but is largely unexplored in Norwegian waters. This is further complicated by the limited knowledge of Sebastes migration patterns and reproductive habits. The complex biology of redfish including reproductive behaviors and specific habitat preference could potentially facilitate the observed rapid genetic divergence in multiple lineages [5,24,25]. This is thought to be of a sympatric or parapatric nature, leaving morphological traits largely retained [5,26,27], which has inspired hypotheses of potentially frequent cryptic speciation.

Among the four established North Atlantic Sebastes species, cryptic species and incipient speciation have been suspected within S. norvegicus [19,26,28–30] and S. mentella [20,31], respectively. Research on the genetic variation within S. mentella has previously disclosed three S. mentella ecomorphs, delineated by depth and geographic location into ‘deep pelagic’, ‘shallow pelagic’ and ‘slope’ morphs [20,31,32]. These are partially supported by color variation, parasite infection rate [33] and morphometrics [34].Within S. norvegicus, a giant form was proposed in the 1960’s [6] and later identified with molecular and morphological studies from East Greenland waters [19,20] and the Reykjanes Ridge [35]. Two additional cryptic species within S. norvegicus in Greenland, Iceland, and Faroe Island waters were genetically identified by Schmidt [26], later referred to as S. norvegicus type A and type B by Saha et al. [19]. Little is known about the distributions of the two types in Norwegian waters except for the registration of four individuals of type A from tissue sample analysis [19,20].

The genetic divergence between S. norvegicus types A and B was found to be similar to the divergence between S. mentella and S. viviparus with low levels of hybridization [26], exceeding the genetic divergence found between the S. mentella ecomorphs [19,20]. The two S. norvegicus types A and B have previously not been described morphologically, but Single Nucleotide Polymorphism (SNP) markers have been developed to differentiate between the types based on Saha et al. [19,36] and Johansen et al. [37].

The present study builds upon the hypothesized S. norvegicus cryptic species derived from genetic findings [19] as well as observations of morphological variations described in Nedreaas and Nævdal [28,29]. Using a combination of morphometric and meristic methods with molecular support, this study aims to reveal visual diagnostic characters useful for identifying genetically delimited S. norvegicus types A and B, as well as separate them from S. mentella and S. viviparus. The following research questions are in focus: i) Are there morphological differences between the genetically assigned types S. norvegicus-A and S. norvegicus-B? ii) How are they distributed by area and depth? These questions are vital for facilitating correct identification at sea, providing a basis for conservation purposes.

Materials and methods

Sample collection

In total, we considered 1,170 specimens of Sebastes including 992 archived tissue samples from IMR (see below) and 178 whole frozen fish. The samples were collected in spring and fall of 2016 to 2020 from Norway, Greenland and Iceland (Fig 1) by research and commercial vessels using trawls, gillnets, and longlines (Table 1). Most specimens were visually classified to species at sea. The morphological study included S. norvegicus and S. mentella from Norwegian and Greenland waters, as well as S. viviparus and undetermined Sebastes from Norwegian waters. Prior to analysis, the specimens were genetically assigned to species or types using three diagnostic Single Nucleotide Polymorphism markers developed at the Institute of Marine Research, Norway [37]. For genetic analysis, fin clips and gill filaments were sampled from all fish and stored in 96% ethanol. Otoliths were collected and stored dry in paper envelopes. The archived samples included fin clips, gill samples, and life history information of Sebastes specimens from Norwegian, Greenlandic, and Icelandic waters.

Download:

Fig 1. Sampling locations of the 1,170 Sebastes spp. collected from three regions: Greenland, Iceland, and Norway.

Both whole fish (▲) collected for morphological analysis and archived samples (● ) from the Institute of Marine Research were assigned to species at sea and by genetic analysis in the lab. Point size indicates relative number of individuals (see legend). Map created with ArcGIS® software by Esri. Bathymetry sources: Esri, GEBCO, NOAA, National Geographic, DeLorme, HERE, Geonames.org, and other contributors [38].

https://doi.org/10.1371/journal.pone.0316988.g001

Download:

Table 1. Overview of the 1,170 sampled Sebastes spp. collected for morphological analysis (whole fish) and genetic analysis (archived material) from three regions. All fish were assigned to species at sea and later genetically assigned using tissue samples. When necessary, total length (TL) was estimated from standard length using a conversion formula [39] shown in S1 Table n = sample size, BT = bottom trawl, PT = pelagic trawl.

https://doi.org/10.1371/journal.pone.0316988.t001

Genetic analysis

DNA was extracted from gill filaments or fin clips using the E-Z 96 Tissue omega DNA Kit (Omega Bio-Teck, Inc.) following the manufacturer’s protocol. The three diagnostic SNPs (Table 2) were developed from ddRAD sequencing of 500 Sebastes spp. from the Northeast Atlantic [37], previously identified to species by molecular genetic markers in earlier projects [19,32,35,40]. The SNP markers were selected to diagnostically differentiate between the three common Northeast Atlantic Sebastes, as well as identifying the proposed cryptic species S. norvegicus-A and B observed by Saha and colleagues [19]. The markers SEB29 and SEB39 identified S. viviparus and S. mentella respectively, while SEB25 separated S. norvegicus-A from S. norvegicus-B. A set of TaqMan SNP Genotyping assays (Thermo Fisher Scientific, Waltham, USA) was designed for fast identification of species by Johansen et al. [37] and for the present study.

Download:

Table 2. In total, 1,171 Sebastes spp. were analyzed for the three diagnostic SNPs. Heterozygous individuals were identified for SEB29 (n = 1) and SEB25 (n = 33). See text for details.

https://doi.org/10.1371/journal.pone.0316988.t002

The three diagnostic SNP markers were used to genetically assign 1,170 fish into species, independent of morphological identification. An unrooted Neighbour-Joining tree was constructed to visualize the segregation of the Sebastes species based on these SNP markers (S1 Fig). The extent of divergence among Sebastes spp. was quantified by the chord distance (D_CE [41]). Pair-wise distances were measured to construct an unrooted-phylogram using the Neighbour-Joining (NJ) algorithm [42] available in Populations [43]. We performed 1,000 bootstraps on loci to estimate confidence of nodes in the tree. The phylip format tree generated by Populations was finally viewed in Fig Tree 1.4.2 (http://tree.bio.ed.ac.uk/software/figtree/ Fig Tree v1.4.2).

Morphological and meristic analysis

Morphological analysis was conducted on 178 specimens of Sebastes from Greenland and Norway (Table 1) which were defrosted in freshwater overnight prior to examination. All fish were measured to total and standard lengths to the nearest 0.1 mm below on a measuring board. Measurements (n = 23, Fig 2) were taken to the nearest 0.1 mm with digital calipers, and meristic counts (n = 9, Fig 3) were recorded on the left side of the fish when possible, following Garabana [44] and Power and Ni [45]. This included the number of gill rakers (Fig 3B) which were counted on the first gill arch. Angles of the preopercular spines were recorded using a coding system ([44], Fig 3C). All measurements and counts were made by the same person to minimize observational errors. Only a subset of the variables was recorded for the reference samples from Greenland as some variables did not show variations, such as the number of rays in the dorsal, pelvic, and caudal fins.

Download:

Fig 2. Interlandmark distances measured in traditional morphometrics.

Width between opercula (AN, not shown) was measured on the dorsal side. SL = Standard Length, TL = Total Length. See S5 Table for explanation of abbreviations.

https://doi.org/10.1371/journal.pone.0316988.g002

Download:

Fig 3. Features included in meristic analysis including a) fin ray counts for the spiny dorsal fin (RD1), soft rays in dorsal fin (RD2), caudal fin (RCF), pectoral fin (RPF), anal fin ray (RAF).

Details in b) shows branchial arch with gill rakers (left) and gill filaments (right), and c) illustrates third preopercular spine (PS3, yellow circle) and fifth preopercular spine (PS5, red circle) with recorded angles following the coding system described by Garabana [44].

https://doi.org/10.1371/journal.pone.0316988.g003

To identify the optimal combinations of variables for classification under varying species and type compositions, Linear Discriminant Analysis (LDA) was performed for two subsets of data [46]. This is a supervised multivariate approach to classification, accounting for prior genetic assignment. Norwegian specimens of S. norvegicus-A, S. norvegicus-B, S. mentella, and S. viviparus were analyzed together, and specimens of S. norvegicus types A and B from Norwegian and Greenland waters were compared in a separate analysis. As the assumptions of LDA are often not met for morphometric data [46], a non-parametric Random Forest analysis [47] was performed on the dataset consisting of Norwegian specimens as well as the total dataset including Greenland fish. All analyses were performed using the software R version 4.0.4 [48].

Potential outliers were identified in residuals plotted from regression of all variables against the standard length for each group of genetically assigned fish and removed if deemed illegitimate. When possible, individuals with missing measurements were used in the combined analysis for practical application described below, or otherwise removed from analysis. Measurements were expressed as a fraction of standard length for each individual. Multivariate normality was evaluated by computing Mardia’s skewness and kurtosis (R package ‘MVN’ [49]) and examining quantile-quantile plots, and the Fligner-Killeen’s test was used to assess homogeneity of variances prior to statistical analyses.

Prior to LDA, dataset dimensionality was reduced through recursive feature elimination. This provides a backward selection of variables based on feature ranking to minimize overfitting and consequentially poor generalization ability of the model [50]. The reduced dataset was partitioned into a training (70%) and testing (30%) subset which were subjected to LDA for classification. These results were compared to the Random Forest procedure through which relative variable importance with permutation was also calculated [51]. Averaged model prediction accuracy of group membership was estimated and compared between LDA and Random Forest using 10-fold cross-validation with 5 repeats, as well as the out-of-bag error measurement calculated by Random Forest.

For individuals with indeterminate genetic results, we predicted their morphological group assignment based on morphometric measurements as part of the classification procedure in LDA.

Meristic counts were analyzed using a Kruskal-Wallis test assessing mean counts across groups [52] and a subsequent post-hoc Dunn test (package ‘dunn.test’ [53]) identifying between-group differences. The Dunn test was adjusted for comparisons of multiple groups with the Benjamini-Hochberg procedure [54]. Individuals with missing measurements were removed from the dataset but used in the combined analysis for practical application.

To evaluate the performance of morphometric and meristic variables together, a combined analysis was conducted with Random Forest including the most important variables identified in morphological analyses to classify specimens from Norwegian waters (n = 107, S2 Table). This also included individuals removed from the initial morphometric and meristic analyses due to missing measurements.

Age estimation

The left sagittal otolith was chosen for age determination and prepared using the break-and-burn technique [55]. The otolith was broken in half through the nucleus using a scalpel and held over the flame of an alcohol burner for a few seconds until reaching the desired brown color [56]. The otolith was then mounted in plasticine and the burned surface was coated with mineral oil to enhance the growth zones. Age was determined by two experienced technicians counting the hyaline zones (winter zones) under a stereomicroscope with reflected light.

Results

Genetic assignment and age determination

Of the three diagnostic SNPs, SEB29 and SEB39 identified S. viviparus and S. mentella, respectively. SEB25 separated S. norvegicus-A from S. norvegicus-B (Table 2). Of the 1,170 genetically assigned specimens, 140 fish were assigned to S. norvegicus-A, 781 to S. norvegicus-B, 144 to S. mentella, and 71 to S. viviparus (Table 2, Table 3) while 34 specimens were heterozygous for the SNPs used. Age ranged from two to 63 years for all fish measuring from 110 to 760 mm total length. The morphologically examined fish were determined to be between eight and 60 years old with total length ranging from 200 to 630 mm, excluding any juveniles from the morphological analyses.

Download:

Table 3. Classification matrix depicting morphological classification at sea and genetic assignment for the 1,170 Sebastes spp. from three regions. Numbers in parenthesis show the number of individuals included in morphological analysis.

https://doi.org/10.1371/journal.pone.0316988.t003

Most specimens classified by visual inspection as S. norvegicus at sea were genetically assigned to S. norvegicus-B. Genetically identified S. norvegicus-A were initially classified as either S. norvegicus or S. mentella at sea, but more frequently as the latter. The majority of S. mentella specimens collected in Greenland waters were genetically assigned to S. norvegicus-A, while all S. mentella specimens from Icelandic waters were assigned to S. norvegicus-B, S. norvegicus-A, or resulted heterozygous, with two individuals assigned to S. viviparus. None of the specimens from Icelandic waters were genetically assigned to S. mentella.

Of all redfish analyzed, only 2.9% (n = 33) of the fish were heterozygous for the SEB25 and could not be assigned either to S. norvegicus-A or S. norvegicus-B. The heterozygous individuals were observed among the redfish from Norway, Greenland, and Iceland, both in the archived tissue samples (n = 29) and among whole fish for morphometric analysis (n = 4). One redfish from Norwegian waters was heterozygous for the SEB39, the marker separating S. mentella and S. viviparus. All heterozygote individuals were excluded from the morphological analyses apart from the combined analysis for practical application.

Among the whole fish for morphological analysis from Norwegian waters, 16 were S. norvegicus-A, 44 S. norvegicus-B, 27 S. mentella and 22 S. viviparus. This included the undetermined whole fish (see Table 1), of which 16 were assigned to S. norvegicus-A, one was assigned to S. norvegicus-B. In addition, two fish were heterozygous for the SEB25. Of the whole fish from Greenland, 26 fish were assigned to S. norvegicus-A, 38 to S. norvegicus-B, and one individual was assigned to S. mentella, while two fish were heterozygous for SEB25.

Morphometric analysis

Out of 178 whole redfish, a total of 99 Norwegian and 30 Greenland specimens were retained in analyses after removing three outlier individuals and eleven specimens with missing measurements. Four individuals were heterozygous for the SEB25, and the remaining 31 individuals were only included in the combined analysis for practical application, as they were missing measurements needed for the full analyses. The morphometric data was found to deviate from multivariate normality and homogeneity of variances. As the main focus of the morphometric analyses was on delimitating the Norwegian specimens, the Greenland specimens were only used as reference sample.

Among the morphometric variables measured on Norwegian specimens, the recursive feature elimination model selected eight variables for an optimal accuracy. This included eye diameter (DO), beak length (LB), pectoral fin length (LP), distance from snout to preopercular spine (LPO), pelvic fin length (LAV), distance snout to edge of operculum (LC), caudal peduncle height (CP), and distance from the anterior insertion of the dorsal fin to the anterior insertion of the anal fin (AD). The first two linear discriminants in the LDA explained 96.4% of the total variation among specimens from Norwegian waters (Fig 4A) which fully separated S. norvegicus-B and S. norvegicus-A from S. mentella and S. viviparus with a cross-validation classification accuracy of 0.95 (Kappa = 0.93). All S. viviparus were correctly classified, and classification accuracy was generally high for the genetically assigned S. mentella, S. norvegicus-A, and S. norvegicus-B (S3 Table). By reducing the number of variables for practical application, a dataset containing only the top three characters DO, LB, and LP were provided to the LDA. This model achieved a high degree of separation, with the first and second discriminants explaining 99% of the variation (Fig 4B) with a minor decrease in Kappa to 0.91. Cross-validation classification achieved an accuracy of 0.93 (Kappa = 0.90). When comparing both analyses, LB contributed the most explanatory power to the first discriminant and varied to the greatest degree between groups, as indicated by arrow length. The strongest influence on the second discriminant was CP, while DO and LPO influenced both axes. The group orientations in the biplots suggest that specimens of S. norvegicus-A had larger eye diameters as well as longer beaks compared to S. norvegicus-B while S. mentella and S. viviparus had the relatively longest and shortest beak lengths respectively. Specimens of S. viviparus had longer pectoral fins than the remaining groups.

Download:

Fig 4. Morphometric ordination of Sebastes specimens from Norwegian waters.

Panel a) shows biplot from Linear Discriminant Analysis (LDA) based on 8 variables selected by Recursive Feature Elimination, while b) shows biplot for the top three variables eye diameter (DO), beak length (LB), and pectoral fin length (LP). Arrow lengths and directions indicate vector loadings, showing how variables influence the linear discriminants and to what extent. Ellipses show 95% confidence interval for each group.

https://doi.org/10.1371/journal.pone.0316988.g004

Norway vs. Greenland.

The variation between S. norvegicus-B and S. norvegicus-A was explored further in an analysis of the specimens from Norwegian and Greenland waters. Here, recursive feature elimination selected 16 variables to provide the highest classification accuracy. Although the first two linear discriminants of the LDA accounted for 92% of the variation, overlap between groups in the ordination plot (Fig 5) was reflected in a cross-validation classification accuracy of 0.89 (Kappa = 0.82). The relative ordination of the groups suggested that S. norvegicus-A from Norwegian and Greenland waters had relatively longer beaks and larger eye diameter than S. norvegicus-B from both Norway and Greenland. The distance between eye and post opercular spine contributed the most to the separation of types A and B within regions. Assigning the heterozygous Norwegian (n = 2) and Greenland (n = 2) whole specimens with the morphological models placed the individuals together with the Norwegian and Greenland S. norvegicus-A respectively.

Download:

Fig 5. Morphometric ordination based on Linear Discriminant analysis of S. norvegicus-A and S. norvegicus-B from Norwegian (N) and Greenland (G) waters including 16 variables selected by Recursive Feature Elimination (seeS5 Table for explanation of abbreviations).

Arrow lengths and directions indicate vector loadings, showing how variables influence the linear discriminants and to what extent. Ellipses show 95% confidence interval for each group.

https://doi.org/10.1371/journal.pone.0316988.g005

Between-group variable importance was calculated with permutation using Random Forest, which showed that some variables were consistently ranked high across group comparisons. Results were produced for a reduced dataset containing only Norwegian specimens (S2 Fig), as well as the full dataset with Greenland samples (S3 Fig). Between the Norwegian S. norvegicus-B and S. norvegicus-A, the variables DO, LB, and LPO were ranked highest in variable importance suggesting that these three characters contributed the most to discrimination. The CP was the third most important character for separating both S. norvegicus-A and B from S. mentella after DO and LB. Compared with S. viviparus, DO, LAV, and LP were important characters for identifying S. norvegicus-A. In all comparisons involving S. mentella, the model frequently ranked CP as an important variable. A consistently high variable importance was attributed to LP and LAV in regarding comparisons with S. viviparus. Between the Greenland specimens of S. norvegicus types A and B (S2 Fig), the characters DO, LP, and snout to ventral fin length had the highest values for mean decrease in accuracy. Eye diameter was consistently highest ranked in all comparisons including Greenland specimens of S. norvegicus-A and S. norvegicus-B. Between the two regions Greenland and Norway, DO and LP were highly ranked for S. norvegicus-A while DO and LMS were highly ranked for S. norvegicus-B.

A 10-fold cross-validation repeated 5 times gave an accuracy of 0.98 (Kappa = 0.98) for the dataset containing Norwegian samples (S4 Table). All specimens were accurately classified apart from one S. viviparus classified as S. norvegicus-A. When including the Greenland specimens, the accuracy dropped to 0.85 (Kappa = 0.81). Here, 16 out of 93 fish were misclassified largely within S. norvegicus-A and S. norvegicus-B. The out-of-bag estimates of classification error (overall model error, S4 Table) for the Norwegian dataset and the full dataset including Greenland specimens were 1.41% and 17.2% respectively.

Meristics

For the meristic counts, 59 Greenland and 87 Norwegian redfish (S2 Table) were analyzed. The Kruskal-Wallis test revealed significant differences between groups for all variables, except for the number of fin rays in the spiny dorsal fin (Table 4). The two species S. mentella and S. viviparus could be clearly separated by examining the preopercular spines. For S. mentella the PS5 was frequently found to point at a forward angle, while all specimens of S. viviparus consistently displayed backwards pointing preopercular spines. However, for the Norwegian S. norvegicus-A and S. norvegicus-B, patterns of preopercular spine angles did not differ as they typically exhibited a down-backwards pointed angle of the third preopercular spine (PS3) and a downwards pointing angle of the fifth preopercular spine (PS5). Sebastes norvegicus-A and S. norvegicus-B both showed on average more gill rakers (GR) than S. mentella, and the number of gill rakers was typically highest in S. viviparus.

Download:

Table 4. Summary of test statistics from Kruskal-Wallis (K-W) test, which tests for differences between groups, and multiple comparison Dunn test on meristic variables. Variables tested include number of pectoral fin rays (RPF), fin rays in the spiny dorsal fin (RD1), soft fin rays in the dorsal fin (RD2), caudal fin rays (RCF), anal fin rays (RAF), no. of gill rakers (GR) and the angles of the 3^rd preopercular spine (PS3) and 5^th preopercular spine (PS5) for samples from Norway (N) and Greenland (G). For the Dunn test, ranked means between groups are significantly different (*) if they do not share a letter for a given variable.

https://doi.org/10.1371/journal.pone.0316988.t004

The Greenland S. norvegicus types A and B were found to follow the same pattern of preopercular spine angles as the specimens collected in Norwegian waters (Table 4). However, S. norvegicus-B from Greenland waters had fewer gill rakers on average than S. norvegicus-B from Norwegian waters, and S. norvegicus-A overall. This group was also differentiated by a higher average number of pectoral and anal fin rays. The remaining characters could not be used to separate S. norvegicus-A and S. norvegicus-B. A one-tailed T-test showed that for the Norwegian and Greenland S. norvegicus-B, mean standard length was significantly higher in Greenland specimens (p < 0.05).

Combined morphometric and meristic analysis.

The highest ranked characters identified in morphometric and meristic analysis were combined in a Random Forest classification procedure, which included DO, LB, LP, RAF, GR, PS3, and PS5. The dataset including only Norwegian specimens had a classification accuracy of 0.95 (Kappa = 0.93, out-of-bag = 5.8%). Variable importance showed that LB, DO and PS5 were generally of highest importance for all groups. RAF was of lower importance, but more relevant for S. mentella and S. viviparus. Including Greenland specimens resulted in a lower classification accuracy of 0.81 (Kappa = 0.76, out-of-bag = 24.1%). The majority of misclassifications were among S. norvegicus-B specimens, where the model could not perfectly separate specimens of S. norvegicus types between regions. Here, LB was less important among S. norvegicus-A and S. norvegicus-B specimens compared to the Norwegian dataset, while DO remained highly important. GR was particularly important for distinguishing between Norwegian and Greenland S. norvegicus-B.

Geographical and depth distribution

Sampled individuals of S. norvegicus-A and S. norvegicus-B were partially overlapping in the Norwegian and Barents Sea (Fig 6). Sebastes norvegicus-A was collected between 62.7°N to 72°N at depths between 266 to 581 meters, while S. norvegicus-B was collected even further north (62.7°N to 73.9°N) at 90–706 meters. A mixture of both types as well as S. mentella was present in multiple trawl catches. Regarding seasonality, S. norvegicus-B were observed in both spring (n = 400) and fall (n = 305). The S. norvegicus-A, however, were collected in spring (n = 109) with only a single specimen registered in fall. In spring, several female specimens of both types A and B were found eggs with developing larvae, showing maturation stages equivalent to shortly before, during, or after the larval extrusion period. Mature specimens were also recorded in samples from Greenland and Iceland.

Download:

Fig 6. Locations of redfish sampled in Norwegian waters genetically identified as S. norvegicus-A (left), S. norvegicus-B (middle), and S. mentella (right).

The point size indicates relative number of individuals (see legend) partitioned between whole fish (▲) for morphometric analysis and archived tissue samples (● ). Map created with ArcGIS® software by Esri. Bathymetry sources: Esri, GEBCO, NOAA, National Geographic, DeLorme, HERE, Geonames.org, and other contributors [38].

https://doi.org/10.1371/journal.pone.0316988.g006

Discussion

The differences found by both the genetic markers and by morphological examination supports that S. norvegicus-A and S. norvegicus-B represent separate species. Beak length, eye diameter, caudal peduncle height, and pectoral fin length were consistently emphasized as the most contributing factors to morphometric distinction between the specimens. These characters have historically been among the most important features to differentiate between the North Atlantic Sebastes species [44,45].

Previous studies found genetic divergence between S. norvegicus-A and S. norvegicus-B, similar to that between existing species [36,37]. In the present study we found this reflected in morphometric differentiation, where beak length and eye size contributed the most to morphological separation. Multivariate analyses revealed that the classification accuracy of S. norvegicus-A and S. norvegicus-B was comparable to the classification accuracy between the recognized species S. mentella and S. viviparus. The longer beak length in S. norvegicus-A compared to S. norvegicus-B matches a short description of ten individuals caught in Greenland waters by Schmidt [26]. These specimens were genetically distinct from S. norvegicus-B and S. mentella, and could potentially represent S. norvegicus-A.

High morphological classification success was achieved even based on very few characters (Fig 4B, Kappa = 0.91). Whilst the models separated specimens in Norwegian waters relatively well, slightly higher error rate was observed when the Greenlandic samples were included, which may reflect a more similar morphology between S. norvegicus-A and S. norvegicus-B in Greenland waters. This could also be influenced by significant differences in length. Despite the closer genetic relationship between the two S. norvegicus types, the majority of S. norvegicus-A were initially misclassified at sea as S. mentella, which can likely be explained by the rather distinct beaks found on both S. mentella and S. norvegicus-A, and less pronounced in S. norvegicus-B.

Meristic characters were successful in separating the S. norvegicus-A and S. norvegicus-B from S. mentella and S. viviparus. The preopercular spine angles were found to be similar between S. norvegicus types A and B, with the 5^th preopercular spine pointing downwards as opposed to the forwards angle often seen in S. mentella, while S. viviparus consistently displayed backwards pointing spines. This reinforces the historical practice of identification based on preopercular spine angles. Although no consistent meristic differences were observed between S. norvegicus-A and S. norvegicus-B in Norwegian waters, S. norvegicus-B from Greenland showed significantly fewer gill rakers. As the number of gill rakers has been found to be negatively correlated with size in redfish [44], this may be caused by unequal length distributions between the Greenland and Norwegian specimens examined with mean lengths of 502 and 399 mm respectively (p < 0.05) which is not accounted for in the meristic analysis. However, fewer gill rakers have also been observed for the giant type of S. norvegicus caught on the Reykjanes ridge close to Iceland [6,40]. Further diet analyses could be conducted to determine whether differences in gill raker numbers are linked to potential adaptive behaviors or diet between S. norvegicus-A and B, as well as the the giant type.

Due to intraspecific morphological variation, a combination of morphometric characters is required for accurate classification at sea – in line with what Power and Ni [45] suggested to separate S. mentella and S. norvegicus. For morphologically separating S. norvegicus-A and S. norvegicus-B, we recommend using a combination of eye diameter, beak length, and caudal peduncle height. Larger eye diameters and longer beaks are more pronounced among the Norwegian specimens of S. norvegicus-A, whereas S. norvegicus-A from Greenland have slightly smaller eyes and beaks in comparison. The two S. norvegicus types can also be distinguished from S. mentella using the same morphometric characters, where the eye and beak are largest in S. mentella while caudal peduncle height is comparatively the narrowest. As color has previously been identified as an important feature distinguishing Sebastes spp. [33], fresh specimens of S. norvegicus types and S. mentella should be further examined to identify possible color variations. Based on previous literature, it is likely that both S. norvegicus-A and S. norvegicus-B have orange, golden red coloration distinct from the pink color of S. mentella [28,29]. Regardless of color, our classification results show that the S. norvegicus-A and S. norvegicus-B can be separated based on morphometric characters only.

Geographical overlap in the distribution of S. norvegicus-A and S. norvegicus-B in Norwegian waters along the continental shelf break and in the Barents Sea as well as the presence of both types in single hauls could suggest that they exist in sympatry in line with the findings of Schmidt [26] and Saha et al. [19]. While S. norvegicus-B has been observed across the North Atlantic, S. norvegicus-A has previously been found in the area around Faroe Islands, Greenland, and Iceland [26] and until the present work, only four S. norvegicus-A have been detected in Norwegian waters [19]. However, it is very likely that it was S. norvegicus-A that Nedreaas and Nævdal [28,29] also noted on the Norwegian shelf in the late 1980s. Among the samples in our study, S. norvegicus-B was considerably more abundant than S. norvegicus-A. The S. norvegicus type A described by Saha et al. [19] is equivalent to the S. norvegicus-A presented in this study [37], thereby expanding its known range to include the Norwegian shelf and the Barents Sea.

While depth and substrate preferences have contributed to establishing barriers for several cryptic North Pacific Sebastes spp. [5], partially overlapping distribution of S. norvegicus-A and S. norvegicus-B indicates that this case of divergence relies on additional mechanisms for reproductive isolation. The life history of Sebastes provides great potential for speciation events either partially in sympatry, or for periods of reproductive isolation allowing for allopatric speciation where they remain geographically separated [5]. Observations of females carrying larvae of both S. norvegicus-A and S. norvegicus-B coincide in time and space with the described larval release of Sebastes along the coast of Norway [9,13]. It remains unclear whether the historical registrations of S. norvegicus larval release can be attributed to S. norvegicus-A, S. norvegicus-B, S. mentella, or a mixture of all three depending on the accuracy of morphological identification. Larval release could contribute to reproductive barriers, as shifts in timing or area can prevent the mixing of larvae between populations through altered dispersal [57]. Redfish eggs are not fertilized immediately after copulation but delayed due to storage of sperm [7]. Therefore, potential observations of reproductive overlap do not necessarily provide sufficient information about the timing of copulation or larval extrusion to make inferences about reproductive habits or gene flow between S. norvegicus-A and S. norvegicus-B. Furthermore, most S. norvegicus-A specimens were collected in spring, with only one specimen collected in autumn located in the northern part of the Norwegian shelf. Previous studies have suggested that seasonal migrations can cause different populations to aggregate in certain areas [58]. It is unclear whether our sampled specimens for morphological analysis and the archived material represent a single or multiple populations of S. norvegicus-A as we only used the three diagnostic SNP markers to ID them to species or types. Considering that historical records of all aspects of biology and ecology of S. norvegicus have been made based on the assumption that only a single S. norvegicus species is represented, an effort should be made to delimit and describe the biology and ecology of S. norvegicus-A and S. norvegicus-B separately in more detail.

The identification of Sebastes species by external morphology in the field is challenging [44]. Cryptic speciation and development of reproductive isolation could be masked by a less rapid evolution in external morphology facilitated by homogenous environmental conditions that favor the retention of the current biological expression in redfish [59]. It can be challenging to set the criteria for recognizing and establishing boundaries between species, but instances of cryptic speciation have commonly been uncovered among Pacific Sebastes species through morphological and genetic studies [60–64]. The traditional definition of a biological species is based on the reproductive isolation of a group of organisms, with hybrid offspring incapable of reproducing [65]. However, this narrow definition does not necessarily cover the range of genetic, ecological, behavioral, physiological, morphological, and evolutionary traits of separately evolving metapopulations [66]. The rapid evolutionary divergence observed and assumed gene flow between North Atlantic Sebastes spp. support a theory of recent speciation [5], where evidence of hybridization and introgression show that reproductive isolation between the established species is incomplete [26]. A shift in morphology while occupying the same range could be driven by ecological divergence, indicated by a potential association between the traits found to be the most different between the types A and B (such as eye size, caudal peduncle height, and beak length) with locomotion and dietary differences [67].

Genetic assignment of the fish showed that a small fraction of specimens from all three areas were heterozygous for the SEB25 separating S. norvegicus-A and S. norvegicus-B, and one fish displayed heterozygosity for the SEB39 marker identifying S. viviparus and S. mentella. The appearance of heterozygous individuals for otherwise diagnostic markers could indicate that there are genetic variations within the groups or possible hybridization between them. However, the morphological assignment of the four heterozygotes in SEB25 in the morphological data suggested that the individuals belonged to S. norvegicus-A. The heterozygous individuals could also indicate ongoing gene flow between S. norvegicus-A and S. norvegicus-B. As genetic assignment was inconclusive for 34 (3%) of the S. norvegicus individuals, additional markers may be required to capture the full genetic variation within and between the types.

Conclusions

Here, we reveal that both S. norvegicus-A and S. norvegicus-B are identifiable by morphological characters in Norwegian, Greenland, and Icelandic waters. In particular, eye diameter, beak length, and caudal peduncle height are important characters for delimitation of adult S. norvegicus types as well as separation from S. mentella and S. viviparus, for which color and meristic characters can also be useful. These traits linked to S. norvegicus-A and S. norvegicus-B coincide with previous observations of morphological variation, providing an explanation to specimens of S. norvegicus resembling S. mentella. As morphological elements are partially overlapping, characters such as fresh color variations should be further investigated to assist in accurate visual classification on the individual level.

Geographical and depth distribution of samples show that S. norvegicus-A and S. norvegicus-B are overlapping both in catch locations and in depth, hence strengthening the notion of sympatric speciation. We should explore the potential for practical application of the characters as a basis for identification in the field both for adult specimens as well as for juvenile fish study for potential ontogenetic shifts. In addition, the implications of additional Sebastes species should be considered in future revisions of conservation practices to ensure proper management. This is especially relevant for the long-lived Sebastes species, which are particularly susceptible to overfishing.

Supporting information

S1 Table. Metrics used for conversion from standard length to total length used with the formula [64].

https://doi.org/10.1371/journal.pone.0316988.s001

(DOCX)

S2 Table. Data on length, age and sex of Sebastes spp. included in morphometric analysis divided by collection area.

NA=Not available. Asterisk indicates specimens not included in training models for morphometric analysis but classified as part of the testing dataset.

https://doi.org/10.1371/journal.pone.0316988.s002

(DOCX)

S3 Table. Confusion matrix produced by cross-validated linear discriminant analysis models showing percentage prediction of specimens to group based on morphometric measurements against a priori genetic assignment.

Darker grey values indicate higher percentage of predicted specimens.

https://doi.org/10.1371/journal.pone.0316988.s003

(DOCX)

S4 Table. Confusion matrix produced by random forest classifier based on cross-validation showing number of specimens predicted to group based on morphometric measurements against a priori genetic assignment.

Darker grey values indicate higher number of predicted specimens. OOB = Out-of-bag metric produced by Random Forest describing overall error rate.

https://doi.org/10.1371/journal.pone.0316988.s004

(DOCX)

S5 Table. List of abbreviations and explanations for morphometric variables measured.

https://doi.org/10.1371/journal.pone.0316988.s005

(DOCX)

S1 Fig. Unrooted Neighbour-Joining tree visualizing the Sebastes species based on three Single Nucleotide Polymorphism markers.

https://doi.org/10.1371/journal.pone.0316988.s006

(DOCX)

S2 Fig. Pairwise plots showing relative morphometric variable importance between Norwegian Sebastes spp. ranked by random forest permutation.

https://doi.org/10.1371/journal.pone.0316988.s007

(DOCX)

S3 Fig. Pairwise plots showing relative morphometric variable importance between Norwegian Sebastes spp. ranked by random forest permutation.

https://doi.org/10.1371/journal.pone.0316988.s008

(DOCX)

Acknowledgments

Thank you to Kristin Windsland and Lise Heggebakken for assisting in ageing of the fish, and to numerous fishermen and crew from IMR surveys for collection and identification of fish at sea. We would also like to thank Peter Nick Psomadakis for providing helpful comments.

References

1. Nelson J. Fishes World. 4th ed., New York: John Wiley & Sons; 2006, p. 321–5.
2. Templeman W. Redfish distribution in the North Atlantic. Bull Fish Res Board Can. 1959;120:120–73.
- View Article
- Google Scholar
3. Mecklenburg CW, Lynghammar A, Johannesen E, Byrkjedal I, Christiansen J, Dolgov A, et al. Marine fishes of the Arctic region. Iceland: Circumpolar Biodiversity Monitoring Program; 2018.
4. Kendall A. An historical review of Sebastes taxonomy and systematics. Mar Fish Rev. 2000;62.
- View Article
- Google Scholar
5. Hyde JR, Vetter RD. The origin, evolution, and diversification of rockfishes of the genus Sebastes (Cuvier). Mol Phylogenet Evol. 2007;44(2):790–811. pmid:17320419
- View Article
- PubMed/NCBI
- Google Scholar
6. Kotthaus A. Contribution to the race problem in redfish. Rapp P-V Reun Cons Int Explor Mer. 1961;150:42–4.
- View Article
- Google Scholar
7. Raitt DFS, Hall WB. On the Fecundity of the Redfish, Sebastes marinus (L.). ICES J Mar Sci. 1967;31(2):237–45.
- View Article
- Google Scholar
8. Campana SE, Zwanenburg KCT, Smith JN. 210Pb/226Ra determination of longevity in redfish. Can J Fish Aquat Sci. 1990;47(1):163–5.
- View Article
- Google Scholar
9. Drevetnyak K, Nedreaas KH, Planque B. Redfish. In: Jakobsen T, Ozhigin VK, editors. Barents Sea Ecosyst. Resour. Manag. Half Century Russ.-Nor. Coop., Trondheim: Tapir; 2011, p. 825.
10. Andriyashev AP. Fishes of the northern seas of the U.S.S.R. vol. 53. Jerusalem: Israel program for scientific translations; 1964.
11. Muus BJ, Nielsen JG. Sea fish. Scand. Fish. Year-Book, Denmark: Hedehusene; 1999, p. 168.
12. ICES. Golden redfish (Sebastes norvegicus) in subareas 5, 6, 12, and 14 (Iceland and Faroes grounds, West of Scotland, North of Azores, East of Greenland). ICES Stock Annexes: Recurrent Advice. ICES Stock Annexes. 2019.
- View Article
- Google Scholar
13. Sorokin VP. The redfish: Gametogenesis and migrations of the Sebastes marinus (L.) and Sebastes mentella Travin. Rapp P-V Reun Cons Int Explor Mer. 1961;150245–50.
- View Article
- Google Scholar
14. ICES. Report of the Arctic Fisheries Working Group (AFWG). ICES Expert Group reports; 2018. https://doi.org/10.17895/ices.pub.5608
15. ICES. Stock Annex: Golden redfish (Sebastes norvegicus) in subareas 1 and 2 (Northeast Arctic). ICES Stock Annexes; 2018. https://doi.org/10.17895/ices.pub.18622400.v1
16. ICES. Arctic Fisheries Working Group (AFWG). ICES Scientific Reports; 2021. https://doi.org/10.17895/ices.pub.8196
17. Barsukov VV, Litvinenko NI, Serebryakov VP. Manual for the identification of redfish species in the North Atlantic and adjacent areas. Kalingrad, USSR: AtlantNIRO; 1984.
18. Rehbein H. Differentiation of fish species by PCR-based DNA analysis of nuclear genes. Eur Food Res Technol. 2013;236(6):979–90.
- View Article
- Google Scholar
19. Saha A, Hauser L, Hedeholm R, Planque B, Fevolden S-E, Boje J, et al. Cryptic Sebastes norvegicus species in Greenland waters revealed by microsatellites. ICES J Mar Sci. 2017;74(8):2148–58.
- View Article
- Google Scholar
20. Saha A, Johansen T, Hedeholm R, Nielsen EE, Westgaard J-I, Hauser L, et al. Geographic extent of introgression in Sebastes mentella and its effect on genetic population structure. Evol Appl 2017;10:77–90.
- View Article
- Google Scholar
21. Schwenke PL, Park LK, Hauser L. Introgression among three rockfish species (Sebastes spp.) in the Salish Sea, northeast Pacific Ocean. PLoS One. 2018;13(3):e0194068. pmid:29566070
- View Article
- PubMed/NCBI
- Google Scholar
22. Roques S, Sévigny JM, Bernatchez L. Evidence for broadscale introgressive hybridization between two redfish (genus Sebastes) in the North-west Atlantic: a rare marine example. Mol Ecol. 2001;10(1):149–65. pmid:11251794
- View Article
- PubMed/NCBI
- Google Scholar
23. Artamonova VS, Makhrov AA, Karabanov DP, Rolskiy AYu, Bakay YuI, Popov VI. Hybridization of beaked redfish (Sebastes mentella) with small redfish (Sebastes viviparus) and diversification of redfish (Actinopterygii: Scorpaeniformes) in the Irminger Sea. J Nat His. 2013;47(25–28):1791–801.
- View Article
- Google Scholar
24. Stefánsson MÖ, Reinert J, Sigurðsson Þ, Kristinsson K, Nedreaas K, Pampoulie C. Depth as a potential driver of genetic structure of Sebastes mentella across the North Atlantic Ocean. ICES J Mar Sci. 2009;66(4):680–90.
- View Article
- Google Scholar
25. Shum P, Pampoulie C, Kristinsson K, Mariani S. Three-dimensional post-glacial expansion and diversification of an exploited oceanic fish. Mol Ecol. 2015;24(14):3652–67. pmid:26073046
- View Article
- PubMed/NCBI
- Google Scholar
26. Schmidt C. Molecular genetic studies on species and population structure of North Atlantic redfish (genus Sebastes; Cuvier 1829). PhD Thesis. University of Hamburg, 2005.
27. Cadrin SX, Bernreuther M, Daníelsdóttir AK, Hjörleifsson E, Johansen T, Kerr L, et al. Population structure of beaked redfish, Sebastes mentella: evidence of divergence associated with different habitats. ICES J Mar Sci. 2010;67(8):1617–30.
- View Article
- Google Scholar
28. Nedreaas K, Naevdal G. Studies of Northeast Atlantic species of redfish (genus Sebastes) by protein polymorphism. ICES J Mar Sci. 1989;46(1):76–93.
- View Article
- Google Scholar
29. Nedreaas K, Naevdal G. Genetic studies of redfish (Sebastes spp.) along the continental slopes from Norway to East Greenland. ICES J Mar Sci. 1991;48(2):173–86.
- View Article
- Google Scholar
30. Nedreaas K. Genetic studies of redfish (Sebastes spp.) from Icelandic and Greenland waters. ICES J Mar Sci. 1994;51(4):461–7.
- View Article
- Google Scholar
31. Daníelsdóttir AK, Gíslason D, Kristinsson K, Stefánsson M, Johansen T, Pampoulie C. Population structure of deep‐sea and oceanic phenotypes of deepwater redfish in the irminger sea and icelandic continental slope: are they cryptic species? Trans Am Fish Soc. 2008;137(6):1723–40.
- View Article
- Google Scholar
32. Johansen T, Danı́elsdóttir AK, Meland K, Nævdal G. Studies of the genetic relationship between deep-sea and oceanic Sebastes mentella in the Irminger Sea. Fish Res. 2000;49(2):179–92.
- View Article
- Google Scholar
33. Magnusson J, Magnusson JV. Oceanic redfish (Sebastes mentella) in the Irminger Sea and adjacent waters. Scie Mar Espana. 1995.
- View Article
- Google Scholar
34. Trella K, Podolska M, Nedreaas K, Janusz J. Discrimination of the redfish (Sebastes mentella ) stock components in the Irminger Sea and adjacent waters based on meristics, morphometry and biological characteristics. J Appl Ichthyol. 2012;29(2):341–51.
- View Article
- Google Scholar
35. Johansen T, Nævdal G, Daniélsdóttir AK, Hareide NR. Genetic characterisation of giant Sebastes in the deep water slopes in the Irminger Sea. Fish Res. 2000;45(3):207–16.
- View Article
- Google Scholar
36. Saha A, Kent M, Hauser L, Drinan DP, Nielsen EE, Westgaard J-I, et al. Hierarchical genetic structure in an evolving species complex: Insights from genome wide ddRAD data in Sebastes mentella. PLoS One. 2021;16(5):e0251976. pmid:34043665
- View Article
- PubMed/NCBI
- Google Scholar
37. Johansen T, Hanebrekke T, Besnier F, Westgaard J-I, Bruvold IM, Dahle G, et al. Development of SNP for Sebastes species identification with special focus on the cryptic species complex of Sebastes norvegicus. Ecol Evol. 2025;15(1):e70767.
- View Article
- Google Scholar
38. Esri. Topographic [Web Map. “The Ocean Basemap” Oct 4, 2018. [Cited 2024 March 08] Available from: https://www.arcgis.com/home/item.html?id=67ab7f7c535c4687b6518e6d2343e8a2
39. Binohlan C, Froese R, Pauly D, Reyes R. The length-length table in FishBase 2011. FishBase. 2011.
- View Article
- Google Scholar
40. Johansen T, Danielsdottir AK, Naevdal G. Genetic variation of Sebastes viviparus Kroyer in the North Atlantic. J Appl Ichthyol. 2002;18(3):177–80.
- View Article
- Google Scholar
41. Cavalli-Sforza LL, Edwards AW. Phylogenetic analysis. Models and estimation procedures. Am J Hum Genet. 1967;19(3 Pt 1):233–57. pmid:6026583
- View Article
- PubMed/NCBI
- Google Scholar
42. Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987;4(4):406–25. pmid:3447015
- View Article
- PubMed/NCBI
- Google Scholar
43. Langella O. Population Genetic Software (Individuals or Populations Distances, Phylogenetic Trees). 2002.
44. Garabana D. The genus Sebastes Cuvier, 1829 (Pisces, Scorpaenidae) in the North Atlantic: species and stock discrimination using traditional and geometric morphometrics. PhD Thesis. University of Vigo, 2005.
45. Power DJ, Ni I. Morphometric Differences between Golden Redfish (Sebastes marinus) and Beaked Redfishes (S. mentellaandS. fasciatus). J Northw Atl Fish Sci. 1985;6:1–7.
- View Article
- Google Scholar
46. Doyle D, Gammell MP, Nash R. Morphometric methods for the analysis and classification of gastropods: a comparison using Littorina littorea. J Molluscan Stud. 2018;84(2):190–7.
- View Article
- Google Scholar
47. Breiman L. Random Forests. Mach Learn. 2001;45(1):5–32.
- View Article
- Google Scholar
48. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing; 2022.
49. Korkmaz S, Göksülük D, Zararsiz G. MVN: An R package for assessing multivariate normality. 2014.
50. Pappu V, Pardalos P. High-dimensional data classification. In: Aleskerov F, Goldengorin B, Pardalos P, editors. Clust. Orders Trees Methods Appl. New York: Springer; 2014, p. 119–50. https://doi.org/10.1007/978-1-4939-0742-7_8
51. Hong Han, Xiaoling Guo, Hua Yu. Variable selection using Mean Decrease Accuracy and Mean Decrease Gini based on Random Forest. 2016 7th IEEE International Conference on Software Engineering and Service Science (ICSESS). 2016:219–24. https://doi.org/10.1109/icsess.2016.7883053
52. Kruskal WH, Wallis WA. Use of Ranks in One-Criterion Variance Analysis. J Am Stat Assoc. 1952;47(260):583–621.
- View Article
- Google Scholar
53. Dinno A. Dunn’s test of multiple comparisons using rank sums. 2017.
54. Dinno A. Nonparametric Pairwise Multiple Comparisons in Independent Groups using Dunn’s Test. Sata J. 2015;15(1):292–300.
- View Article
- Google Scholar
55. Christensen JM. Burning of otoliths, a technique for age determination of soles and other fish. ICES Journal of Marine Science. 1964;29(1):73–81.
- View Article
- Google Scholar
56. Campana SE, Valentin AE, MacLellan SE, Groot JB. Image-enhanced burnt otoliths, bomb radiocarbon and the growth dynamics of redfish (Sebastes mentella and S. fasciatus) off the eastern coast of Canada. Mar Freshwater Res. 2016;67(7):925.
- View Article
- Google Scholar
57. Bernardi G. Speciation in fishes. Mol Ecol. 2013;22(22):5487–502. pmid:24118417
- View Article
- PubMed/NCBI
- Google Scholar
58. Saborido-Rey F. Geographic variation of Sebastes mentella in the Northeast Arctic derived from a morphometric approach. ICES J Mar Sci. 2000;57(4):965–75.
- View Article
- Google Scholar
59. Colborn J, Crabtree RE, Shaklee JB, Pfeiler E, Bowen BW. The evolutionary enigma of bonefishes (Albula spp.): cryptic species and ancient separations in a globally distributed shorefish. Evolution. 2007;55(4):807–20.
- View Article
- Google Scholar
60. Rocha-Olivares A, Rosenblatt RH, Vetter RD. Cryptic species of rockfishes (Sebastes: Scorpaenidae) in the southern hemisphere inferred from mitochondrial lineages. J Hered. 1999;90(3):404–11. pmid:10355124
- View Article
- PubMed/NCBI
- Google Scholar
61. Gharrett AJ, Matala AP, Peterson EL, Gray AK, Li Z, Heifetz J. Two genetically distinct forms of rougheye rockfish are different species. Trans Am Fish Soc. 2005;134(1):242–60.
- View Article
- Google Scholar
62. Hyde JR, Kimbrell CA, Budrick JE, Lynn EA, Vetter RD. Cryptic speciation in the vermilion rockfish (Sebastes miniatus) and the role of bathymetry in the speciation process. Mol Ecol. 2008;17(4):1122–36. pmid:18261052
- View Article
- PubMed/NCBI
- Google Scholar
63. Hess JE, Chittaro P, Elz A, Gilbert-Horvath EA, Simon V, Garza JC. Cryptic population structure in the severely depleted cowcod,Sebastes levis. Can J Fish Aquat Sci. 2014;71(1):81–92.
- View Article
- Google Scholar
64. Frable BW, Wagman DW, Frierson TN, Aguilar A, Sidlauskas BL. A new species of Sebastes (Scorpaeniformes: Sebastidae) from the northeastern Pacific, with a redescription of the blue rockfish, S. mystinus (Jordan and Gilbert, 1881). Fish Bull. 2015;113(4):355–77.
- View Article
- Google Scholar
65. Mayr E. Systematics and the origin of species. New York: Columbia University Press; 1942.
66. De Queiroz K. Species concepts and species delimitation. Syst Biol. 2007;56(6):879–86. pmid:18027281
- View Article
- PubMed/NCBI
- Google Scholar
67. Myers EMV, Anderson MJ, Eme D, Liggins L, Roberts CD. Changes in key traits versus depth and latitude suggest energy-efficient locomotion, opportunistic feeding and light lead to adaptive morphologies of marine fishes. J Anim Ecol. 2020;89(2):309–22. pmid:31646627
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Nelson J. Fishes World. 4th ed., New York: John Wiley & Sons; 2006, p. 321–5.

[ref2] 2. Templeman W. Redfish distribution in the North Atlantic. Bull Fish Res Board Can. 1959;120:120–73.
View Article
Google Scholar

[3] View Article

[4] Google Scholar

[ref3] 3. Mecklenburg CW, Lynghammar A, Johannesen E, Byrkjedal I, Christiansen J, Dolgov A, et al. Marine fishes of the Arctic region. Iceland: Circumpolar Biodiversity Monitoring Program; 2018.

[ref4] 4. Kendall A. An historical review of Sebastes taxonomy and systematics. Mar Fish Rev. 2000;62.
View Article
Google Scholar

[7] View Article

[8] Google Scholar

[ref5] 5. Hyde JR, Vetter RD. The origin, evolution, and diversification of rockfishes of the genus Sebastes (Cuvier). Mol Phylogenet Evol. 2007;44(2):790–811. pmid:17320419
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref6] 6. Kotthaus A. Contribution to the race problem in redfish. Rapp P-V Reun Cons Int Explor Mer. 1961;150:42–4.
View Article
Google Scholar

[14] View Article

[15] Google Scholar

[ref7] 7. Raitt DFS, Hall WB. On the Fecundity of the Redfish, Sebastes marinus (L.). ICES J Mar Sci. 1967;31(2):237–45.
View Article
Google Scholar

[17] View Article

[18] Google Scholar

[ref8] 8. Campana SE, Zwanenburg KCT, Smith JN. 210Pb/226Ra determination of longevity in redfish. Can J Fish Aquat Sci. 1990;47(1):163–5.
View Article
Google Scholar

[20] View Article

[21] Google Scholar

[ref9] 9. Drevetnyak K, Nedreaas KH, Planque B. Redfish. In: Jakobsen T, Ozhigin VK, editors. Barents Sea Ecosyst. Resour. Manag. Half Century Russ.-Nor. Coop., Trondheim: Tapir; 2011, p. 825.

[ref10] 10. Andriyashev AP. Fishes of the northern seas of the U.S.S.R. vol. 53. Jerusalem: Israel program for scientific translations; 1964.

[ref11] 11. Muus BJ, Nielsen JG. Sea fish. Scand. Fish. Year-Book, Denmark: Hedehusene; 1999, p. 168.

[ref12] 12. ICES. Golden redfish (Sebastes norvegicus) in subareas 5, 6, 12, and 14 (Iceland and Faroes grounds, West of Scotland, North of Azores, East of Greenland). ICES Stock Annexes: Recurrent Advice. ICES Stock Annexes. 2019.
View Article
Google Scholar

[26] View Article

[27] Google Scholar

[ref13] 13. Sorokin VP. The redfish: Gametogenesis and migrations of the Sebastes marinus (L.) and Sebastes mentella Travin. Rapp P-V Reun Cons Int Explor Mer. 1961;150245–50.
View Article
Google Scholar

[29] View Article

[30] Google Scholar

[ref14] 14. ICES. Report of the Arctic Fisheries Working Group (AFWG). ICES Expert Group reports; 2018. https://doi.org/10.17895/ices.pub.5608

[ref15] 15. ICES. Stock Annex: Golden redfish (Sebastes norvegicus) in subareas 1 and 2 (Northeast Arctic). ICES Stock Annexes; 2018. https://doi.org/10.17895/ices.pub.18622400.v1

[ref16] 16. ICES. Arctic Fisheries Working Group (AFWG). ICES Scientific Reports; 2021. https://doi.org/10.17895/ices.pub.8196

[ref17] 17. Barsukov VV, Litvinenko NI, Serebryakov VP. Manual for the identification of redfish species in the North Atlantic and adjacent areas. Kalingrad, USSR: AtlantNIRO; 1984.

[ref18] 18. Rehbein H. Differentiation of fish species by PCR-based DNA analysis of nuclear genes. Eur Food Res Technol. 2013;236(6):979–90.
View Article
Google Scholar

[36] View Article

[37] Google Scholar

[ref19] 19. Saha A, Hauser L, Hedeholm R, Planque B, Fevolden S-E, Boje J, et al. Cryptic Sebastes norvegicus species in Greenland waters revealed by microsatellites. ICES J Mar Sci. 2017;74(8):2148–58.
View Article
Google Scholar

[39] View Article

[40] Google Scholar

[ref20] 20. Saha A, Johansen T, Hedeholm R, Nielsen EE, Westgaard J-I, Hauser L, et al. Geographic extent of introgression in Sebastes mentella and its effect on genetic population structure. Evol Appl 2017;10:77–90.
View Article
Google Scholar

[42] View Article

[43] Google Scholar

[ref21] 21. Schwenke PL, Park LK, Hauser L. Introgression among three rockfish species (Sebastes spp.) in the Salish Sea, northeast Pacific Ocean. PLoS One. 2018;13(3):e0194068. pmid:29566070
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref22] 22. Roques S, Sévigny JM, Bernatchez L. Evidence for broadscale introgressive hybridization between two redfish (genus Sebastes) in the North-west Atlantic: a rare marine example. Mol Ecol. 2001;10(1):149–65. pmid:11251794
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref23] 23. Artamonova VS, Makhrov AA, Karabanov DP, Rolskiy AYu, Bakay YuI, Popov VI. Hybridization of beaked redfish (Sebastes mentella) with small redfish (Sebastes viviparus) and diversification of redfish (Actinopterygii: Scorpaeniformes) in the Irminger Sea. J Nat His. 2013;47(25–28):1791–801.
View Article
Google Scholar

[53] View Article

[54] Google Scholar

[ref24] 24. Stefánsson MÖ, Reinert J, Sigurðsson Þ, Kristinsson K, Nedreaas K, Pampoulie C. Depth as a potential driver of genetic structure of Sebastes mentella across the North Atlantic Ocean. ICES J Mar Sci. 2009;66(4):680–90.
View Article
Google Scholar

[56] View Article

[57] Google Scholar

[ref25] 25. Shum P, Pampoulie C, Kristinsson K, Mariani S. Three-dimensional post-glacial expansion and diversification of an exploited oceanic fish. Mol Ecol. 2015;24(14):3652–67. pmid:26073046
View Article
PubMed/NCBI
Google Scholar

[59] View Article

[60] PubMed/NCBI

[61] Google Scholar

[ref26] 26. Schmidt C. Molecular genetic studies on species and population structure of North Atlantic redfish (genus Sebastes; Cuvier 1829). PhD Thesis. University of Hamburg, 2005.

[ref27] 27. Cadrin SX, Bernreuther M, Daníelsdóttir AK, Hjörleifsson E, Johansen T, Kerr L, et al. Population structure of beaked redfish, Sebastes mentella: evidence of divergence associated with different habitats. ICES J Mar Sci. 2010;67(8):1617–30.
View Article
Google Scholar

[64] View Article

[65] Google Scholar

[ref28] 28. Nedreaas K, Naevdal G. Studies of Northeast Atlantic species of redfish (genus Sebastes) by protein polymorphism. ICES J Mar Sci. 1989;46(1):76–93.
View Article
Google Scholar

[67] View Article

[68] Google Scholar

[ref29] 29. Nedreaas K, Naevdal G. Genetic studies of redfish (Sebastes spp.) along the continental slopes from Norway to East Greenland. ICES J Mar Sci. 1991;48(2):173–86.
View Article
Google Scholar

[70] View Article

[71] Google Scholar

[ref30] 30. Nedreaas K. Genetic studies of redfish (Sebastes spp.) from Icelandic and Greenland waters. ICES J Mar Sci. 1994;51(4):461–7.
View Article
Google Scholar

[73] View Article

[74] Google Scholar

[ref31] 31. Daníelsdóttir AK, Gíslason D, Kristinsson K, Stefánsson M, Johansen T, Pampoulie C. Population structure of deep‐sea and oceanic phenotypes of deepwater redfish in the irminger sea and icelandic continental slope: are they cryptic species? Trans Am Fish Soc. 2008;137(6):1723–40.
View Article
Google Scholar

[76] View Article

[77] Google Scholar

[ref32] 32. Johansen T, Danı́elsdóttir AK, Meland K, Nævdal G. Studies of the genetic relationship between deep-sea and oceanic Sebastes mentella in the Irminger Sea. Fish Res. 2000;49(2):179–92.
View Article
Google Scholar

[79] View Article

[80] Google Scholar

[ref33] 33. Magnusson J, Magnusson JV. Oceanic redfish (Sebastes mentella) in the Irminger Sea and adjacent waters. Scie Mar Espana. 1995.
View Article
Google Scholar

[82] View Article

[83] Google Scholar

[ref34] 34. Trella K, Podolska M, Nedreaas K, Janusz J. Discrimination of the redfish (Sebastes mentella ) stock components in the Irminger Sea and adjacent waters based on meristics, morphometry and biological characteristics. J Appl Ichthyol. 2012;29(2):341–51.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref35] 35. Johansen T, Nævdal G, Daniélsdóttir AK, Hareide NR. Genetic characterisation of giant Sebastes in the deep water slopes in the Irminger Sea. Fish Res. 2000;45(3):207–16.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref36] 36. Saha A, Kent M, Hauser L, Drinan DP, Nielsen EE, Westgaard J-I, et al. Hierarchical genetic structure in an evolving species complex: Insights from genome wide ddRAD data in Sebastes mentella. PLoS One. 2021;16(5):e0251976. pmid:34043665
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref37] 37. Johansen T, Hanebrekke T, Besnier F, Westgaard J-I, Bruvold IM, Dahle G, et al. Development of SNP for Sebastes species identification with special focus on the cryptic species complex of Sebastes norvegicus. Ecol Evol. 2025;15(1):e70767.
View Article
Google Scholar

[95] View Article

[96] Google Scholar

[ref38] 38. Esri. Topographic [Web Map. “The Ocean Basemap” Oct 4, 2018. [Cited 2024 March 08] Available from: https://www.arcgis.com/home/item.html?id=67ab7f7c535c4687b6518e6d2343e8a2

[ref39] 39. Binohlan C, Froese R, Pauly D, Reyes R. The length-length table in FishBase 2011. FishBase. 2011.
View Article
Google Scholar

[99] View Article

[100] Google Scholar

[ref40] 40. Johansen T, Danielsdottir AK, Naevdal G. Genetic variation of Sebastes viviparus Kroyer in the North Atlantic. J Appl Ichthyol. 2002;18(3):177–80.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref41] 41. Cavalli-Sforza LL, Edwards AW. Phylogenetic analysis. Models and estimation procedures. Am J Hum Genet. 1967;19(3 Pt 1):233–57. pmid:6026583
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref42] 42. Saitou N, Nei M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol. 1987;4(4):406–25. pmid:3447015
View Article
PubMed/NCBI
Google Scholar

[109] View Article

[110] PubMed/NCBI

[111] Google Scholar

[ref43] 43. Langella O. Population Genetic Software (Individuals or Populations Distances, Phylogenetic Trees). 2002.

[ref44] 44. Garabana D. The genus Sebastes Cuvier, 1829 (Pisces, Scorpaenidae) in the North Atlantic: species and stock discrimination using traditional and geometric morphometrics. PhD Thesis. University of Vigo, 2005.

[ref45] 45. Power DJ, Ni I. Morphometric Differences between Golden Redfish (Sebastes marinus) and Beaked Redfishes (S. mentellaandS. fasciatus). J Northw Atl Fish Sci. 1985;6:1–7.
View Article
Google Scholar

[115] View Article

[116] Google Scholar

[ref46] 46. Doyle D, Gammell MP, Nash R. Morphometric methods for the analysis and classification of gastropods: a comparison using Littorina littorea. J Molluscan Stud. 2018;84(2):190–7.
View Article
Google Scholar

[118] View Article

[119] Google Scholar

[ref47] 47. Breiman L. Random Forests. Mach Learn. 2001;45(1):5–32.
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref48] 48. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing; 2022.

[ref49] 49. Korkmaz S, Göksülük D, Zararsiz G. MVN: An R package for assessing multivariate normality. 2014.

[ref50] 50. Pappu V, Pardalos P. High-dimensional data classification. In: Aleskerov F, Goldengorin B, Pardalos P, editors. Clust. Orders Trees Methods Appl. New York: Springer; 2014, p. 119–50. https://doi.org/10.1007/978-1-4939-0742-7_8

[ref51] 51. Hong Han, Xiaoling Guo, Hua Yu. Variable selection using Mean Decrease Accuracy and Mean Decrease Gini based on Random Forest. 2016 7th IEEE International Conference on Software Engineering and Service Science (ICSESS). 2016:219–24. https://doi.org/10.1109/icsess.2016.7883053

[ref52] 52. Kruskal WH, Wallis WA. Use of Ranks in One-Criterion Variance Analysis. J Am Stat Assoc. 1952;47(260):583–621.
View Article
Google Scholar

[128] View Article

[129] Google Scholar

[ref53] 53. Dinno A. Dunn’s test of multiple comparisons using rank sums. 2017.

[ref54] 54. Dinno A. Nonparametric Pairwise Multiple Comparisons in Independent Groups using Dunn’s Test. Sata J. 2015;15(1):292–300.
View Article
Google Scholar

[132] View Article

[133] Google Scholar

[ref55] 55. Christensen JM. Burning of otoliths, a technique for age determination of soles and other fish. ICES Journal of Marine Science. 1964;29(1):73–81.
View Article
Google Scholar

[135] View Article

[136] Google Scholar

[ref56] 56. Campana SE, Valentin AE, MacLellan SE, Groot JB. Image-enhanced burnt otoliths, bomb radiocarbon and the growth dynamics of redfish (Sebastes mentella and S. fasciatus) off the eastern coast of Canada. Mar Freshwater Res. 2016;67(7):925.
View Article
Google Scholar

[138] View Article

[139] Google Scholar

[ref57] 57. Bernardi G. Speciation in fishes. Mol Ecol. 2013;22(22):5487–502. pmid:24118417
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref58] 58. Saborido-Rey F. Geographic variation of Sebastes mentella in the Northeast Arctic derived from a morphometric approach. ICES J Mar Sci. 2000;57(4):965–75.
View Article
Google Scholar

[145] View Article

[146] Google Scholar

[ref59] 59. Colborn J, Crabtree RE, Shaklee JB, Pfeiler E, Bowen BW. The evolutionary enigma of bonefishes (Albula spp.): cryptic species and ancient separations in a globally distributed shorefish. Evolution. 2007;55(4):807–20.
View Article
Google Scholar

[148] View Article

[149] Google Scholar

[ref60] 60. Rocha-Olivares A, Rosenblatt RH, Vetter RD. Cryptic species of rockfishes (Sebastes: Scorpaenidae) in the southern hemisphere inferred from mitochondrial lineages. J Hered. 1999;90(3):404–11. pmid:10355124
View Article
PubMed/NCBI
Google Scholar

[151] View Article

[152] PubMed/NCBI

[153] Google Scholar

[ref61] 61. Gharrett AJ, Matala AP, Peterson EL, Gray AK, Li Z, Heifetz J. Two genetically distinct forms of rougheye rockfish are different species. Trans Am Fish Soc. 2005;134(1):242–60.
View Article
Google Scholar

[155] View Article

[156] Google Scholar

[ref62] 62. Hyde JR, Kimbrell CA, Budrick JE, Lynn EA, Vetter RD. Cryptic speciation in the vermilion rockfish (Sebastes miniatus) and the role of bathymetry in the speciation process. Mol Ecol. 2008;17(4):1122–36. pmid:18261052
View Article
PubMed/NCBI
Google Scholar

[158] View Article

[159] PubMed/NCBI

[160] Google Scholar

[ref63] 63. Hess JE, Chittaro P, Elz A, Gilbert-Horvath EA, Simon V, Garza JC. Cryptic population structure in the severely depleted cowcod,Sebastes levis. Can J Fish Aquat Sci. 2014;71(1):81–92.
View Article
Google Scholar

[162] View Article

[163] Google Scholar

[ref64] 64. Frable BW, Wagman DW, Frierson TN, Aguilar A, Sidlauskas BL. A new species of Sebastes (Scorpaeniformes: Sebastidae) from the northeastern Pacific, with a redescription of the blue rockfish, S. mystinus (Jordan and Gilbert, 1881). Fish Bull. 2015;113(4):355–77.
View Article
Google Scholar

[165] View Article

[166] Google Scholar

[ref65] 65. Mayr E. Systematics and the origin of species. New York: Columbia University Press; 1942.

[ref66] 66. De Queiroz K. Species concepts and species delimitation. Syst Biol. 2007;56(6):879–86. pmid:18027281
View Article
PubMed/NCBI
Google Scholar

[169] View Article

[170] PubMed/NCBI

[171] Google Scholar

[ref67] 67. Myers EMV, Anderson MJ, Eme D, Liggins L, Roberts CD. Changes in key traits versus depth and latitude suggest energy-efficient locomotion, opportunistic feeding and light lead to adaptive morphologies of marine fishes. J Anim Ecol. 2020;89(2):309–22. pmid:31646627
View Article
PubMed/NCBI
Google Scholar

[173] View Article

[174] PubMed/NCBI

[175] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Sample collection

Genetic analysis

Morphological and meristic analysis

Age estimation

Results

Genetic assignment and age determination

Morphometric analysis

Norway vs. Greenland.

Meristics

Combined morphometric and meristic analysis.

Geographical and depth distribution

Discussion

Conclusions

Supporting information

S1 Table. Metrics used for conversion from standard length to total length used with the formula [64].

S2 Table. Data on length, age and sex of Sebastes spp. included in morphometric analysis divided by collection area.

S3 Table. Confusion matrix produced by cross-validated linear discriminant analysis models showing percentage prediction of specimens to group based on morphometric measurements against a priori genetic assignment.

S4 Table. Confusion matrix produced by random forest classifier based on cross-validation showing number of specimens predicted to group based on morphometric measurements against a priori genetic assignment.

S5 Table. List of abbreviations and explanations for morphometric variables measured.

S1 Fig. Unrooted Neighbour-Joining tree visualizing the Sebastes species based on three Single Nucleotide Polymorphism markers.

S2 Fig. Pairwise plots showing relative morphometric variable importance between Norwegian Sebastes spp. ranked by random forest permutation.

S3 Fig. Pairwise plots showing relative morphometric variable importance between Norwegian Sebastes spp. ranked by random forest permutation.

Acknowledgments

References