Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Aug 28;35(17):e70537. doi: 10.1111/mec.70537

Phylogenetically Diverse Introgression Drives Subtle Population Structure in Pacific Rockfishes

Nathan T B Sykes 1, R Nicolas Lou 2, Matthew R Siegle 3, Peter H Sudmant 2, Wesley A Larson 4, Gregory L Owens 1,✉
PMCID: PMC13525399  PMID: 42666042

ABSTRACT

Genomic methods have shown that admixture and introgression are common across animal taxa. Pacific rockfishes, genus Sebastes, are a group of commercially important species that primarily inhabit inshore, shelf and slope habitats along the North American west coast. Among these, Copper and Quillback Rockfishes are closely related species known to hybridize, particularly within the Salish Sea in North America's Pacific Northwest. Here, we investigate genetic population structure and introgression patterns in Copper and Quillback Rockfish from Alaska to California. Using whole‐genome resequencing across a broad geographic range, we seek to (1) compare population structure between these species, and (2) assess how introgression affects population structure patterns. Our analyses reveal that Copper exhibits much higher levels of population differentiation compared to Quillback, especially separating Salish Sea samples from all other populations. In contrast, Quillback populations appear to be nearly panmictic, with lower overall differentiation. Surprisingly, we detected signatures of introgression from seven other rockfish species in Copper and 10 species in Quillback. This introgression was highly regional, suggesting hybridization depended on geographic or environmental context. Yelloweye Rockfish introgression in the Salish Sea drives the strongest signal of regional population structure in Quillback. These findings provide novel insights into the range‐wide genetic structure of these species and highlight that hybridization in Sebastes is phylogenetically broader than previously appreciated.

1. Introduction

Marine fishes tend toward subtle population structure, even panmixia. This is due in part to their large population sizes and high dispersal capacity, as well as the relative continuity of marine habitat, reducing barriers to gene flow (Nielsen et al. 2009; Waples 1998). Some species may exhibit population differentiation reflecting their unique life history, fragmented habitat, or capacity for dispersal. Anadromous Pacific salmon, for example, have population structure reflecting their spawning site fidelity (Xuereb et al. 2022), run timing (Narum et al. 2023), or life cycle (Tarpey et al. 2018). In the absence of strong reproductive isolation based on geography, though, intraspecific differentiation is harder to detect. High rates of migration and corresponding high gene flow can lead to largely homogenized populations, both at subcontinental and oceanic scales (Díaz‐Arce et al. 2024; Natasha et al. 2022; Tovar Verba et al. 2022). This presents a challenge for fisheries managers, who seek insight into patterns of dispersal to establish conservation units. Genetic analyses, such as those which leverage sequence divergence within allozymes or microsatellites, have historically been applied to answer these questions, but are limited to only a handful of genetic markers (Buonaccorsi et al. 2004; Gilbert‐Horvath et al. 2006; Gao et al. 2016). However, subtle population structure may be challenging to identify with small numbers of markers. Genomic techniques provide researchers with improved resolution to study population structure in marine fishes (Bernatchez et al. 2017; Longo et al. 2020), but many species have not been analysed with these techniques, leaving fisheries managers to rely on decades‐old genetic analyses in many cases. Given the increasing effectiveness and affordability of genomic approaches, it is essential to close this knowledge gap, particularly for managed species.

North Eastern Pacific rockfishes, genus Sebastes, comprise many economically, socially and culturally important species that inhabit inshore, shelf and slope habitats along the North American West coast (Love et al. 2002; Hyde and Vetter 2007). Among inshore rockfishes, Copper ( S. caurinus ) and Quillback ( S. maliger ) Rockfishes are two closely related species that are known to hybridize, particularly within the Salish Sea (Seeb 1998; Schwenke et al. 2018; Wray et al. 2024). Both species are broadly distributed from Alaska to California, although Copper Rockfish exhibit a more southerly distribution extending into Baja California.

In Canada and the United States, rockfish stocks are often managed along political boundaries (between countries and states) and at regional scales whereby fish in the Salish Sea, for example, are considered a separate stock from those of the greater coast for many species. Although for Copper and Quillback Rockfish, the latter delineation was originally made to reflect exploitation history and not genomic differentiation between the regions, earlier work suggests that the management units reflect some degree of reproductive isolation (Buonaccorsi et al. 2002, 2005; Seeb 1998). The trend is similar in Yelloweye Rockfish ( S. ruberrimus ), which in the Salish Sea are genetically distinct from the outer coast (Siegle et al. 2013; Andrews et al. 2018). Retention of Copper and Quillback Rockfish in the Salish Sea is prohibited in US waters and highly limited in Canadian waters. Copper Rockfish outside of the Salish Sea from Alaska to California are managed based on the following spatial boundaries: Alaska, British Columbia, Oregon and Washington and North and South California, while for Quillback, Washington and Oregon are further divided into separate management regions (Pacific Fishery Management Council 2023).

Pacific rockfishes exhibit life history characteristics that can result in spatially discrete population structure. Small adult home ranges, courtship rituals, internal fertilization, parturition of live offspring, oceanographic effects on larval dispersal, and nearshore early life‐history stages may all reasonably conspire to limit gene flow over large distances and promote genetic differentiation (Buonaccorsi et al. 2002, 2005; Hannah and Rankin 2011; Helmstetter et al. 2016; Love et al. 2002; Matthews 1990; Tolimieri et al. 2009). However, rockfish can also display very low genetic differentiation across large spatial scales, likely due to their relatively long pelagic larval dispersal (lasting up to several months) that can facilitate gene flow over significant distances (Siegle et al. 2013). Previous examinations of population structure in Copper and Quillback Rockfish have been technologically or regionally limited. Allozyme and microsatellite‐based genetic analyses suggest that both species exhibit significant population structure only when Puget Sound is considered (Buonaccorsi et al. 2002, 2005; Seeb 1998). Another study that examined differentiation at microsatellite loci in Copper Rockfish found replicate divergence between pairs of inlet and coast samples on the west coast of Vancouver Island (Dick et al. 2014). Most recently, a genomic study focusing on the Georgia Basin and Puget Sound identified two distinct populations in Quillback, one largely in southern Puget Sound and another in the Georgia Basin (Wray et al. 2024). They also uncovered three discrete populations in Copper Rockfish, again largely separating Puget Sound from the Georgia Basin. Together, this could be taken as an indication that Puget Sound and the Georgia Basin harbour unique genetic diversity, but these findings have not been put in the context of the rest of the species range.

Despite the relatively recent radiation of rockfish species, hybridization is thought to be uncommon in the genus (Love et al. 2002). Documented examples of hybridization include Quillback, Copper and Brown Rockfish ( S. auriculatus ), which often have admixed ancestry in Puget Sound (Seeb 1998; Buonaccorsi et al. 2002; Buonaccorsi et al. 2005). The relaxation of reproductive barriers in Puget Sound is thought to be caused by a combination of anoxic conditions, reduced habitat availability, and/or differences in population size between species (Seeb 1998; Buonaccorsi et al. 2002; Buonaccorsi et al. 2005; Wray et al. 2024; Schwenke et al. 2018). Hybridization in Sebastes has also been noted between the closely related sister species, Black‐and‐Yellow and Gopher Rockfish ( S. chrysomelas and S. carnatus ) (Buonaccorsi et al. 2005), fox and banded jacopever ( S. vulpes and S. zonatus ) (Muto et al. 2013), deepwater and Norway redfish ( S. mentella and S. viviparus ) (Artamonova et al. 2013). The extent to which hybridization has played a role in speciation or adaptation in the genus is currently an open question.

This study leverages whole genome resequencing of Copper and Quillback Rockfish to describe their genetic diversity and population structure throughout their native range. Through this international sampling effort, we seek to: (1) compare population structure between each species, (2) quantify introgression and its effect on population structure. Results from the genomic analyses conducted in this study provide novel insight into their patterns of dispersal, differentiation and introgression in these two species.

2. Methods

2.1. Sample Collection

Samples were collected by various means, including in annual longline surveys conducted by Fisheries and Oceans Canada (DFO) and the US National Oceanic and Atmospheric Administration (NOAA), and in individual rod‐and‐reel catch efforts conducted by authors and collaborators (Anderson et al. 2019). In total, 113 Copper ( S. caurinus ) and 200 Quillback ( S. maliger ) Rockfish were collected between June 1994 and August 2022. Samples were sorted into eight regional locations: Alaska (AK), Hecate Straight (HS), Queen Charlotte Sound (QCS), Western Vancouver Island (WVI), Georgia Basin (GB), Puget Sound (PS), Washington and Oregon (WA/OR) and California (CA; Figure 1B). Here we consider Georgia Basin to include all of the Salish Sea, except for Puget Sound. Throughout the paper, we refer to samples in Puget Sound and the Georgia Basin as ‘Salish Sea’ and all others collectively as ‘coastal’.

FIGURE 1.

FIGURE 1

Open in a new tab

Sampling and interspecific differentiation. (A) Underwater photographs of each species by Andy Murch at Big Fish Expeditions. (B) Sampling distribution and geographic bins. The number in parentheses represents the samples retained after filtering from each population: Alaska (AK), Hecate Strait (HS), Queen Charlotte Sound (QCS), Georgia Basin (GB), Western Vancouver Island (WVI), Puget Sound (PS), Washington and Oregon (WA/OR) and California (CA). (C) Unrooted consensus tree of all individuals retained in the data set, coloured by species. (D) Mean nucleotide diversity (π) of chromosomes in both species. (E) Principal component analysis of all samples, coloured by species. (F) Combined OHANA qpas analysis of all samples for K = 2. Samples are ordered by ancestry.

2.2. Library Prep and Sequencing

All whole genome resequencing library preparation was performed using methods similar to Baym et al. (2015), Therkildsen and Palumbi (2017) and Euclide et al. (2023). Genomic DNA input was normalized to 10 ng per individual. Per Euclide et al. (2023), sample purification, product normalization and pooling were conducted with SequalPrep plates (ThermoFisher Scientific: Waltham, MA, USA) rather than AMPure XP beads (Beckman Coulter: Brea, CA, USA). Prior to sequencing, the final pooled library was visualized on a 2% agarose E‐gel (ThermoFisher Scientific) and quantified with the Qubit HS dsDNA Assay Kit (ThermoFisher Scientific).

Sequencing of an initial batch of 113 Copper and 152 Quillback Rockfish occurred at Novogene (Sacramento, CA, USA) on Illumina NovaSeq S4. An additional 48 Quillback samples were also sequenced on a Novaseq with S4 chemistry by the University of Oregon's Genomic & Cell Characterization Core Facility (GC3F; Eugene, OR, USA). Samples were sequenced on lanes containing 96 rockfish each and were multiplexed with samples for other projects. All samples were sequenced with a target depth of 6×.

When selecting samples to sequence, we optimized geographic spread. This means that for most samples, there is only one fish in a specific coordinate. To explore the geography of populations we used two schemes for grouping samples. First, we grouped samples by broad geographic regions into ‘regional samples’: Alaska, Hecate Strait, Queen Charlotte Sound, West Coast of Vancouver Island, Georgia Basin (excluding Puget Sound), Puget Sound, Washington and Oregon and California (Figure 1B). As a second approach, we took all collection coordinates in decimal degrees and rounded to the nearest whole number for latitude and longitude. All samples within a single rounded value were classified as a ‘coordinate samples’. Given the size of the areas represented in both population types, we use them for illustration and do not claim that they represent randomly mating populations. Eight Quillback and five Copper Rockfish coordinate samples included only a single fish and were excluded.

2.3. Alignment, Variant Calling and Quality Control

Alignment and initial quality control were done using the Snakemake pipeline grenepipe v0.10.0 (Czech and Exposito‐Alonso 2021). All tools were run using default settings unless specified otherwise. First, adapter sequences were trimmed using fastp v0.20.0 (Chen 2023). Reads were then aligned to a Korean rockfish ( S. schlegelii ) genome acquired from the Chinese National GeneBank Database (He et al. 2019; CNA0000824) using bwa‐mem v2.2.1 (Vasimuddin et al. 2019). This reference genome was chosen since it is an outgroup of all sequenced species and prevents reference bias in interspecies comparisons. Overlapping read pairs were clipped using BamUtil clipOverlap v1.0.15 (Jun et al. 2015), and duplicated reads were removed using picard MarkDuplicates v2.23.0 (Broad Institute 2019) with the parameters ‐Xmx80g REMOVE_DUPLICATES=true, VALIDATION_STRINGENCY=SILENT. FastQC v0.11.9 (Andrews 2010), Qualimap v2.2.2a (Okonechnikov et al. 2016), picard CollectMultipleMetrics v2.23.0 (Broad Institute 2019) and MultiQC v1.10.1 (Ewels et al. 2016) were used to assess the quality of the data throughout the pipeline. To test for contamination, we estimated heterozygosity of the mitochondrial genome using ANGSD (v0.937) (Korneliussen et al. 2014). Although heteroplasmy is possible, high mitochondrial heterozygosity is a sign that DNA from multiple fish were incorporated into single sample sequencing library.

Variants were called using freebayes v1.3.6 (Garrison and Marth 2012), for which we skipped sites with overall read depth in excess of 20× the total number of samples since average depth was 6.04×, and set both minimum mapping and base quality to 20 (‐g 6260 ‐m 20 ‐q 20). Monomorphic sites were retained in the unfiltered dataset for estimates of genetic diversity. We then used vcftools v0.1.16 (Danecek et al. 2011) to extract individual and site‐based statistics for the remaining samples and excluded some poor‐quality samples (> 25% missing data). In total, 18 Copper and six Quillback were excluded from analysis due to missing data. The final total number of samples retained were 95 Copper and 194 Quillback (Figure 1B; Table S1).

We then filtered the data for downstream analyses using both bcftools v1.20 (Danecek et al. 2021) and vcftools (Danecek et al. 2011); we retained only biallelic SNPs with a minor allele frequency > 5%, with a maximum 25% missing data, a minimum genotype read depth of five, a maximum genotype read depth of 50, excess heterozygosity (p > 0.007), and QUAL ≥ 20. This filtering regime was applied to the whole dataset for interspecific analyses, as well as to each species separately for intraspecific analyses. The above computations and most downstream analyses were parallelized using GNU parallel (Tange 2023).

2.4. Population Structure, Differentiation and Genetic Diversity

For both inter‐ and intraspecific population structure analyses, we performed principal component analyses (PCAs) using EMU v1.6.0 (Meisner et al. 2021) and plotted using R v4.2.2 (R Core Team 2025). For EMU, we used four PCs for the expectation maximization step, a maximum of 500 iterations and retained PCs equal to 80% of the number of samples, or 10 minus the number of samples, whichever is lower, because retaining all possible PCs resulted in program failure. This method was chosen to handle the variable levels of missing data caused by modest sequence coverage. We assessed ancestry proportions of our samples with the qpas function in OHANA (Cheng et al. 2017) using genotype likelihoods, again to account for lower genotype confidence. To generate a phylogeny of all samples in the dataset, we thinned and converted the full filtered dataset using a custom perl v5.32.1 (Wall et al. 2020) script, which retained 5% of sites. We then used iqtree2 v2.3.6 (Minh et al. 2020) with 1000 Ultrafast bootstraps (Minh et al. 2013, 2020) to generate a consensus tree (‐m ‘GTR + ASC’ ‐st DNA ‐B 1000).

Next, we sought to characterize the genomic differentiation both between and within the study species. To achieve this, we used a custom perl script to calculate F ST for each position in the genome and summarized the results on a per‐chromosome basis in R (Weir and Cockerham 1984; Samuk et al. 2017). This analysis was applied to both species together and separately. Intraspecific F ST was calculated between regional and coordinate samples. To test for isolation by distance, we plotted F ST versus geographic distance for all coordinates. Geographic distance was calculated as the minimum distance in water between the center of each coordinate sample location (van Etten 2017; Pebesma and Bivand 2023). In the case when the center was on land, the water position nearest to the center was used instead. Since we found a major genetic barrier between the populations inside and outside of the Salish Sea, we grouped comparisons by inside‐inside, inside‐outside and outside‐outside. To test for significance, we used a Mantel test and partial Mantel test, including inside/outside comparison type as a covariate using the vegan package in R (Sokal 1979; Oksanen et al. 2013). Lastly, we used pixy v1.2.7 (Korunes and Samuk 2021) to estimate nucleotide diversity (π) in both species, and individually for each sample. Given that differentiation was highest in comparisons made between fish in the Salish Sea and coastal fish, another pairwise F ST calculation was conducted between all Georgia Basin and Puget Sound samples and all others. We also subset samples inside and outside of the Salish Sea for both species and ran EMU and OHANA qpas to test for further population structure.

2.5. Reconciling Introgression and Population Structure

Previous studies suggested that population structure in rockfish can be caused by introgression (Wray et al. 2024), so we conducted formal genome‐wide tests of introgression using Patterson's D, also known as the ABBA‐BABA test, in Dsuite (Malinsky et al. 2021). This test selects two closely related species or populations (P1 and P2) and asks if one shares more derived alleles with a third species (P3). Without gene flow, both P1 and P2 should share equal numbers of derived alleles with P3 and D will be zero. If there is gene flow between P2 and P3, then D will be positive; conversely, gene flow between P1 and P3 will result in a negative D. Importantly, if P1 or P2 has received gene flow from another source, this can also affect D (Tricou et al. 2022).

For our initial test, we assigned individuals to one of four groups based on the interspecific OHANA qpas analysis; pure types from each species represented two of the groups, while ≥ 0.5% of minor ancestry resulted in assignment to an admixed group for each species. We then used the Dtrios command to test whether samples identified as admixed were closer to their congener than non‐admixed samples. The Korean rockfish reference genome was included as a sample in the vcf and specified as the outgroup.

Surprisingly, we found that the Quillback Rockfish samples identified as admixed using OHANA qpas shared fewer derived alleles with Copper Rockfish than non‐admixed samples. To explore these results and reconcile population structure and introgression, we conducted individual ABBA‐BABA tests on all samples with the following scheme. Each individual sample was placed in the P2 position, and all other samples of the same species were put into the P1 position in a one‐vs‐all test. All samples of the other species (Quillback or Copper) were placed in the P3 position (Figure 3C,D). The resulting scores represent a relative amount of allele sharing compared to all other samples, and we name this statistic D rel because it is the D score relative to all other samples. Importantly, D rel does not reflect the absolute amount of introgression but the relative introgression compared to the average of the rest of the species. A positive D rel score means that the individual sample is more admixed than the average, and a negative score means that it is less admixed. If all samples have 10% admixed ancestry, then all individual D rel scores would be zero because they do not differ in proportion of derived alleles shared with the possible donor.

FIGURE 3.

FIGURE 3

Open in a new tab

Introgression explains population structure in Rockfish. (A) D rel for Copper testing for Quillback introgression. Individuals with > 2.5% admixture based on interspecies ADMIXTURE are highlighted. (B) D rel for Quillback testing for Copper introgression. Individuals with > 2.5% admixture based on interspecies ADMIXTURE are highlighted. (C, D) Conceptual diagram showing how introgression can cause positive or negative D rel observed in A and B. Dotted lines indicate groupings for P1, P2 and P3 from the D rel test. Coloured dots indicate admixed individuals. In Panel C, introgression shares derived alleles from the donor species (Blue) into the recipient species (Yellow) leading to positive or negative D scores depending on whether the tested P2 sample contains (+) or does not contain (−) introgression. In Panel D, introgression from a more distantly related species (Grey) contributes ancestral alleles leading to positive or negative D scores depending on whether the tested P2 sample contains (−) or does not contain (+) introgression. The donor also donates distinct derived alleles, but these are ignored by the statistic because they are not shared with the tested donor species. (E) Simulation models of admixture. (F) D rel scores for simulated individuals for no admixture (Model 1), 5% admixture (Model 2), 10% admixture (Model 3), 10% outgroup admixture (Model 4). Each population tested had 10 samples.

We emphasize that D rel is not a new analysis method, but instead a specific approach of applying Patterson's D to individual samples. To validate that D rel scores work as predicted, we simulated four scenarios. In the first model, which includes no admixture, Species 1 and 2 diverge and evolve for 10,000 generations with a population size of 10,000 each. At this point, Species 1 splits into two populations (Pop 1 and Pop 2) with a population size of 5000 and further evolve for 500 more generations. We then calculate D rel for all samples of Species 1 testing for admixture from Species 2. As a reminder, when calculating D rel, each sample is tested in a one‐vs‐all model, so it shows the relative difference in allele sharing. The next two models include a single pulse of admixture from Species 2 into Pop 2 representing 5% and 10% of the ancestry. The last model introduces Species 3, which diverged 10,000 generations before Species 1 and 2, and contributes 10% admixture to Pop 2. All simulations were coded in Slim (v5.2) using a simulated genome size of 100 Mbp, a recombination rate of 1e−8 and tree sequences (Haller et al. 2019, 2025). Trees were capitated with an ancestral population size of 10,000 using tskit (0.6.4), pyslim (1.1.0) and msprime (1.3.4), and mutations were overlayed with a mutation rate 1e−7 for 10 samples per population (Kelleher et al. 2016; Baumdicker et al. 2022; Wong et al. 2024; Gopalan et al. 2025).

2.6. Genus‐Wide Search for Introgression Donors

Our D rel analysis suggested that Quillback Rockfish contained introgression from a source other than Copper Rockfish. To explore this, we selected publicly available whole‐genome sequencing data for 51 species of Northeast Pacific Rockfish from across the genus (Table S2). Each sample was down‐sampled to 240 million paired‐end reads if it had greater read depth to reduce processing time. Samples were processed as before, but SNP‐calling was done in two datasets: Quillback Rockfish samples plus genus‐wide set, and Copper Rockfish samples plus genus‐wide set. After variant calling, we retained only biallelic SNPs with a minor allele frequency > 2%, with a maximum of 20% missing data, a minimum read depth of five, a maximum read depth of 60, and QUAL ≥ 20. We then repeated our individual D rel analysis, isolating each Quillback/Copper Rockfish sample and testing it against all others. We focused on comparisons in the following arrangement: P1: All other Quillback/Copper samples, P2: Target Quillback/Copper sample, P3: other Sebastes species. To interpret the introgression signals, we organized donor species by phylogeny (Kolora et al. 2021). Based on the similarity of scores and the phylogenetic tree, we divided donors into three groups. Group 1 was the most distantly related from Quillback and Copper and included largely the subgenera Sebastosomus and Allosebastes. Group 2 was more closely related to Quillback and Copper and largely included the subgenus Sebastomus. Lastly, Group 3 included Copper, Quillback, and their closest relatives largely in subgenus Pteropodus.

To describe how admixed samples were identified, we first have to describe the patterning observed in D rel scores. Firstly, strong positive D rel scores in one species were associated with positive, but less strong, D rel scores in closely related species within the same group. Secondly, strong positive D rel scores in Groups 1 or 2 were associated with broad negative D rel scores for Group 3, and inversely positive scores in Group 1 were associated with negative scores in Groups 1 and 2. Lastly, because we use a one‐vs‐all model, positive scores in some samples correspond to negative scores in others. The combination of these effects means that individual D rel scores are not informative of introgression. For example, one sample had significantly positive D rel for 37 different donor species, but it is highly implausible that it contained introgression from each. Our approach to identify likely introgression focused on D rel scores that were disproportionately high for the phylogenetic position of the donor. We considered a sample to be introgressed from a donor if the D rel score was 0.05 higher than the median D rel score for other species within the donor's group, the donor had > 1% admixture based on the f4‐ratio, and the D rel score was significant based on jackknife bootstrap (p < 0.0001). If multiple donors were candidates in a single group, we selected the donor species with the highest D rel score, except in Copper Rockfish where we selected both Quillback and Brown Rockfish introgression in the same sample. We chose this exception because Quillback and Brown Rockfish introgression appeared to be phylogenetically distinct, despite being in the same group. The thresholds described here were chosen to select a set of likely admixed samples and may not select all admixed samples (see discussion).

3. Results

3.1. Interspecific Analyses

Sequencing yielded an average read depth per sample of 5.25 in Copper, 6.83 in Quillback. All sequence data is available on the SRA (Table S1). After filtering, 1,047,615 SNPs were retained in the interspecific dataset, whereas 350,889 SNPs were retained in the Copper dataset and 1,447,179 SNPs in the Quillback dataset. In the interspecific PCA, the first principal component (PC1) separated samples along species lines and explained 32.35% of the total variance (percent variance explained; PVE). PC2 and PC3 separated Copper samples while PC4 separated Quillback samples (Figure 1E). We found slightly higher diversity in Quillback (π = 0.002) over Copper (π = 0.00188) Rockfish (Figure 1D). We found slightly increased π for individuals below 5× depth, but consistently higher diversity in Quillback Rockfish across the range of sequence depths (Figure 1). An unrooted consensus tree of the full sample set revealed a pattern consistent with the PCA, whereby Quillback nodes radiate more from a central point than do the more differentiated Copper (Figure 1C). Interspecific OHANA qpas results indicate that there were 34 out of 194 Quillback with > 0.5% Copper ancestry (up to a maximum of 4% admixed ancestry), and that there were 39 Copper with > 0.5% Quillback ancestry out of 95 tested (up to a maximum of 15% admixed ancestry) (Figure 1F). Overall interspecific F ST was calculated to be 0.333.

3.2. Intraspecific Population Structure and Differentiation

Intraspecific principal component analyses reveal that Copper have a more defined population structure than Quillback Rockfish (Figure 2A). The first principal component in Copper (7.05 PVE) divides fish in the Salish Sea from the coastal populations, whereas PC2 (2.91 PVE) separates Puget Sound and some Georgia Basin samples. The coastal Copper samples separated along PC3 and PC4 into three groups roughly corresponding to California, Washington/Oregon and Hecate Strait.

FIGURE 2.

FIGURE 2

Open in a new tab

Intraspecific population structure and differentiation. (A) Principal component analyses Copper Rockfish, individual points coloured by geographic population: Alaska (AK), Hecate Strait (HS), Queen Charlotte Sound (QCS), Georgia Basin (GB), Western Vancouver Island (WVI), Puget Sound (PS), Washington and Oregon (WA/OR) and California (CA). (B) Principal component analyses Quillback Rockfish, individual points coloured by geographic population. (C) Heatmap of genetic differentiation (F ST) between geographic populations in both species. (D) Genetic differentiation versus marine distance for coordinate samples. Inside includes Georgia Basin and Puget Sound populations and outside includes all others. (E) Intraspecific ancestry proportions from OHANA qpas for K = 2 in Copper and Quillback Rockfish. Samples are sorted by latitude. See Figure S4 for Quillback specific F ST graphs.

Quillback PCAs are less informative, as most samples tend toward a single dense cluster (Figure 2B). The first PC largely separates out a subset of the Georgia Basin, Puget Sound and West Vancouver Island samples, with outliers concentrating in the narrow straits at the North end of the Salish Sea, while outliers in PC2 are largely from Puget Sound and the Georgia Basin. Subsequent PCs have little geographic patterning.

Within Copper Rockfish, F ST between the regional samples was highest inside vs. outside of the Salish Sea comparisons (Figure 2C). Values were highest between Puget Sound and Hecate Strait (F ST = 0.078), and lowest between California and Washington/Oregon (F ST = 0.0095). The pattern in Quillback was similar, but population differentiation was much lower—the highest F ST value was between Queen Charlotte Sound and Puget Sound (F ST = 0.011), while the lowest, highlighting the very low differentiation of coastal Quillback, was between California and Alaska (F ST = 0.0003) (Figure 2C, Figure S4).

Analysis of pairwise differentiation by coordinate samples instead of regional samples showed a similar trend (Figure 2D). In Copper Rockfish, mean pairwise F ST between coastal sampling locations was 0.008, between locations in the Salish Sea was 0.033, and in pairwise comparisons between Salish Sea and outer coastal locations was 0.058. In Quillback Rockfish, these values were much lower: mean F ST of comparisons between coastal locations was 0.003, within the Salish Sea was 0.002 and between these regions was 0.004.

OHANA qpas population structure analyses revealed regional population structure in Copper Rockfish (Figure 2E, Figure S2). For example, at K = 4, the four ancestry components largely separated Hecate Strait, the Salish Sea, Washington/Oregon and California. In contrast, Quillback Rockfish population structure analyses found two ancestry groups largely limited to the Georgia Basin and Puget Sound when K = 3, and higher K values were not geographically informative (Figure S3).

When analyzing only inside or outside of the Salish Sea, we saw some evidence of further regional structure in Copper Rockfish. For example, at K = 3, coastal Copper samples largely, although not perfectly, separate into Californian, Washington/Oregon and Hecate Strait clusters (Figure S5). In the southern half of the range, PC2 roughly follows latitude (Figure S6). For Salish Sea Copper Rockfish, PCA and OHANA qpas identified regional population structure in the southern Georgia Basin and Puget Sound (Figure S7). Coastal Quillback show a subtle signal of regional clustering in PCA, but no clear patterning in OHANA qpas (Figure S8). Finally, Salish Sea Quillback showed regional population structure between Puget Sound and the Georgia Basin (Figure S9).

Both full and partial (controlling for comparisons inside vs. outside of the Salish Sea) Mantel test revealed significant isolation by distance, with a moderate positive correlation between genetic and geographic distance in Copper (Full: r = 0.53, p = 0.0002, 10,000 permutations. Partial: r = 0.56, p = 0.0001, 10,000 permutations). Mantel tests were also significant when restricted to coastal comparisons (Full: r = 0.67, p = 0.0001, 10,000 permutations), but not to comparisons within the Salish Sea (Full: r = 0.26, p = 0.35, 119 permutations). For Quillback, most Mantel tests were positive, but non‐significant (Full: r = 0.14, p = 0.15, 10,000 permutations. Partial: r = 0.16, p = 0.11, 10,000 permutations), including when limited to comparisons outside (Full: r = 0.11, p = 0.22, 10,000 permutations). The one exception is when testing inside of the Salish Sea, which was significant (Full: r = 0.60, p = 0.018, 719 permutations).

3.3. Introgression of Quillback Into Copper, and of an Unknown Species Into Quillback

A basic interspecies OHANA qpas analysis (Figure 1G) initially suggested bidirectional introgression between Copper and Quillback. We thus tested this using Patterson's D, or the ABBA‐BABA test, to ask if ‘admixed’ samples were genetically closer to their congener than ‘non‐admixed’ samples. As predicted, admixed Copper were closer to Quillback than non‐admixed (D = 0.06; p = 2.3e−16), however, surprisingly, non‐admixed Quillback were closer to Copper than admixed Quillback (D = −0.08; p = 2.3e−16). This suggests a more complex introgression scenario.

Simulations of admixture show that D rel scores are generally zero in the absence of admixture and consistently positive for admixed populations, as predicted (Figure 3E,F). Importantly, admixture from an outgroup results in negative D rel scores for the admixed populations when testing from the incorrect donor, and positive D rel scores for the unadmixed population.

Using D rel scores for Copper Rockfish samples, we found that most individuals were slightly negative, except for Georgia Basin and Puget Sound samples which were strongly positive (Figure 3A). This pattern is consistent with Quillback Rockfish introgression largely localized to the Georgia Basin and Puget Sound (Figure 3C). In contrast, Quillback Rockfish individuals were generally positive, except for a few strongly negative scores from samples in Vancouver Island, the Georgia Basin and Puget Sound (Figure 3B). This suggests that there is introgression from a non‐Copper source in these regions driving negative D‐scores and consequently, generally positive D‐scores in all other samples (Figure 3D). We found only one sample, located in the Georgia Basin, with a disproportionately high D‐score indicating likely true Copper introgression into Quillback.

3.4. Genus‐Wide Survey of Introgression Donors

Under our D rel analysis framework, we found diverse sources of introgression in both Copper and Quillback Rockfish. For Copper Rockfish, we corroborated our previous analysis and found that some Salish Sea samples have Quillback admixture, but we also found evidence for admixture from Brown and Yelloweye (Figure 4). In coastal waters, we found admixture from Black, Yellowtail, Vermillion and Canary Rockfish in California and Black and Yelloweye Rockfish in Washington/Oregon. For Quillback, we found that most samples had positive scores for Group 1 and 2 donors and negative scores for Group 3, or the reverse (Figure 5). The most extreme values occurred for Yelloweye Rockfish as the donor, suggesting the pattern is driven by the presence or absence of Yelloweye Rockfish introgression. A total of 28 samples were identified as having Yelloweye admixture, largely in the Salish Sea. Rare signals of admixture were also found from Aurora ( S. aurora ), Blue ( S. mystinus ), Canary ( S. pinniger ), Greenblotched ( S. rosenblatti ), Light Dusky ( S. variabilis ), Northern ( S. polyspinis ), Redstripe ( S. proriger ), Rougheye ( S. aleutianus ) and Widow ( S. entomelas ) Rockfish. Admixture was geographically structured; in some cases, admixture was only found in one geographic region (Figure 6A). Furthermore, we found that admixture from Brown and Yelloweye Rockfish explained variation in a top principal component axis for Copper and Quillback Rockfish, respectively (Figure 6B). Lastly, we found that mitochondrial heterozygosity was very near zero for all samples, and samples identified as admixed did not have higher mitochondrial heterozygosity, suggesting that inadvertent sample mixing is unlikely to drive the patterns observed (Figure S10).

FIGURE 4.

FIGURE 4

Open in a new tab

Introgression into Copper from diverse sources. (A) D rel scores for all Copper Rockfish samples organized by the phylogeny of possible donors. Red values indicate greater sharing of derived alleles and the donor than the rest of the Copper Rockfish samples. Outlined squares indicate outlier values and circles indicate the top outlier for each group (except for Quillback and Brown Rockfish, where both are selected). Donor phylogenetic groups are shown at the top. Target samples are organized by similarity. D rel scores > 0.3 or < −0.3 are visualized as ±0.3 to aid visualization. Geographic region of each sample is colour‐coded on the right most column. POP means Pacific Ocean Perch ( S. alutus ). (B) f 4‐ratio scores for identified donor which indicates the proportion of admixed ancestry.

FIGURE 5.

FIGURE 5

Open in a new tab

Introgression into Quillback from diverse sources. (A) D rel scores for all Copper Rockfish samples organized by the phylogeny of possible donors. Red values indicate greater sharing of derived alleles and the donor than the rest of the Copper Rockfish samples. Outlined squares indicate outlier values and circles indicate the top outlier for each group. Donor phylogenetic groups are shown at the top. Target samples are organized by similarity of scores. D rel scores > 0.3 or −0.3 are visualized as ±0.3 to aid visualization. Geographic region of each sample is colour‐coded on the right most column. POP means Pacific Ocean Perch ( S. alutus ). (B) f 4‐ratio scores for identified donor which indicates the proportion of admixed ancestry.

FIGURE 6.

FIGURE 6

Open in a new tab

Summary of admixture in Copper and Quillback Rockfish. (A) Species highlighted are admixture donors in more than one sample in the region. Copper Rockfish appears as yellow, and Quillback Rockfish as blue. (B) Principal component two for Copper Rockfish compared with f 4‐ratio scores, representing admixture with Brown Rockfish. (C) Principal component one for Quillback Rockfish compared with f 4‐ratio scores, representing admixture with Yelloweye Rockfish.

Overall, there is a significantly higher proportion of admixed samples in the Salish vs. coastal waters for Quillback (52% vs. 8%, Welch's Two Sample t‐test, t(77.4) = 6.47, p = 8e‐9) but not Copper Rockfish (31% vs. 13%, Welch's Two Sample t‐test, t(41.8) = 1.79, p = 0.08).

4. Discussion

In this study, we observed significant differences in the distribution of genetic variation between Copper and Quillback Rockfish, both within the genome and between sampling locations, despite similar levels of genetic diversity. The subtle overall population structure in both species is in line with genetic and genomic analyses in other Pacific rockfish species (An et al. 2012; Buonaccorsi et al. 2004; Gao et al. 2016; Zhu et al. 2024). Higher differentiation between coast and inlet sampling locations is also consistent with previous genetic studies in this genus (Buonaccorsi et al. 2002, 2005; Dick et al. 2014), but the significant interspecific differences in range‐wide population structure are described for the first time here. Lastly, we showed that both Copper and Quillback Rockfish have introgression from multiple sources. Interestingly, introgression from Yelloweye Rockfish drove regional population structure in Quillback Rockfish and introgression from Quillback Rockfish contributed to structure in Copper Rockfish (Figure 6).

4.1. Introgression and Population Structure in Copper Rockfish

Hybridization between Copper and Quillback Rockfish has resulted in largely unidirectional geneflow from Quillback into Copper Rockfish. The low level of mixed ancestry in the sample set is consistent with long‐term, low levels of hybridization, primarily restricted to the Salish Sea and Puget Sound. This largely supports previous studies using fewer markers that found introgression in the Salish Sea and Puget Sound (Seeb 1998; Schwenke et al. 2018; Wray et al. 2024). Both Schwenke et al. (2018) and Wray et al. (2024) found that Quillback Rockfish admixture into Copper Rockfish was focused in, but not limited to, South Puget Sound. Our analysis found that Quillback introgression spanned the Salish Sea, and there was a greater proportion of Brown introgression in Puget Sound, although our sampling in Puget Sound is very limited. Furthermore, admixed Copper Rockfish are found along the coast, suggesting that hybridization is not limited to Puget Sound, nor limited to closely related congeners. The regional variation in admixture donor cannot be simply explained by range overlaps as most species have large ranges. For example, although Vermilion Rockfish ( S. miniatus ) is only an admixture donor in California, it co‐occurs with Copper Rockfish from Northern BC to California. It's possible that features like local population density, depth, continental slope or benthic environment led to greater sympatry and hybridization between particular species. Further work is needed to determine if all admixture from a single species can be traced back to a single admixture event, or if hybridization between the same species has occurred multiple times in different locations.

Copper Rockfish are significantly more differentiated throughout their range than Quillback Rockfish, despite similar levels of genetic diversity. Our results support previous work suggesting that the population structure of Copper Rockfish in the Salish Sea is partially driven by hybridization with Quillback (Schwenke et al. 2018; Wray et al. 2024). For example, the second principal component of the Copper Rockfish PCA likely represents Brown Rockfish admixture, which is limited to Puget Sound and the Georgia Basin. Although Salish Sea samples have more Quillback Rockfish ancestry than other regions, it does not appear that this is fully driving the population structure signal seen in principal component one, suggesting dispersal limitations and drift are also playing a role. This is not unprecedented in marine species, as local introgression has been shown to be the main cause of intraspecific differentiation in mussels (Fraïsse et al. 2016), but this finding highlights the role that introgression may play in regional genetic variation, especially in species with otherwise low genetic differentiation.

4.2. Introgression and Population Structure in Quillback Rockfish

The very low levels of genetic differentiation between Quillback Rockfish populations more resemble the norm for marine fishes with long pelagic larval durations (PLDs), which tend toward subtle, if any, population structure (Waples 1998). Outside of the Salish Sea, we find some signals of subtle population structure, although all F ST scores were very close to zero (0–0.004). The large differences in the magnitude of population structure between Quillback and Copper Rockfish in coastal waters along the Pacific coast are highly intriguing. The contrasting patterns of population structure are unlikely the result of differences in generation time or PLDs as both species have generally similar life spans and PLDs (Love et al. 2002). Additionally, Copper and Quillback Rockfish appear to occupy very similar ecological niches based on diets, growth patterns and distributions (Murie 1995). Juvenile Copper and Quillback Rockfish also settle in similar nearshore complex habitats containing understory kelp, although they do show some subtle differences in depth preferences as adults, with Copper Rockfish preferring slightly shallower water (Palsson et al. 2009). Quillback Rockfish have slightly higher diversity, suggesting they have a higher effective population size. A higher effective population size slows the rate at which F ST increases between populations, which could contribute to the lack of population structure in the species (Reynolds et al. 1983). We further hypothesize that cryptic ecological processes that limit the dispersal potential of Copper Rockfish compared to Quillback Rockfish may explain the variation in population structure that we observed but additional research is necessary to investigate this.

Initial admixture tests suggested Copper Rockfish introgression in Quillback Rockfish, but using a wide array of possible donors we showed that this signal was not Copper‐specific and instead likely was driven by Yelloweye Rockfish introgression in the Salish Sea. Yelloweye Rockfish is not an expected candidate for hybridization with Quillback Rockfish because they are not in the same clade and diverged approximately 7 mya (Kolora et al. 2021; Love et al. 2002). Interestingly, admixture seems to be geographically structured. Yelloweye, Canary and Rougheye Rockfish admixture is primarily in the Salish Sea and Widow and Redstripe admixture in the outer coast of Vancouver Island. Surprisingly, we see a signal of admixture with Northern Rockfish in Puget Sound, where Northern Rockfish do not occur. Since the admixture is only 4% of the genome, the admixed fish is multiple generations from the initial hybridization event and may have migrated long distances over generations. Alternatively, the Northern Rockfish signal may have originated from another closely related species and be misattributed based on the incomplete sampling of possible donors.

Overall, hybridization is increased in more complex and degraded estuarine habitats but is not exclusive to these areas. This may be due to opportunity if species are brought into contact due to the environmental gradients, or necessity if reduced population density leads to challenges finding conspecific mates. The low levels of admixture we observe here could be due to temporary or human influence‐induced breakdowns of reproductive barriers. Depending on the amount of hybridization, the relative population sizes and the amount of selection on introgressed ancestry, signals of admixture can persist in a population for thousands of generations. Alternatively, admixture amounts from 1% to 15% could occur as early as backcross generation two, so scattered admixture signatures need not represent population‐wide patterns. The lack of population structure outside of region‐specific introgression suggests that current migration rates are not high enough to fully homogenize the species or that natural selection prevents the spread of introgression in some areas. Future work should attempt to date hybridization events and disentangle the relative genetic contributions of multiple species.

4.3. Approaches to Identifying Admixture

We approached introgression using our newly described relative D because of the number of possible species involved and the lack of clear allopatric non‐admixed populations. Although other approaches attempt to disentangle multiple introgression events (e.g., TreeMix, PhyloNet), they are designed for smaller number of species and admixture events and would not function well with 51 species and more than a dozen introgression events (Hibbins and Hahn 2022). Similarly, approaches like ADMIXTURE or OHANA work better with more reference samples and would struggle computationally with 51 possible donors (Cheng et al. 2017). This approach is conceptually very similar to F branch approach in Dsuite which tests all possible trios and partitions admixture timing into a phylogenetic context using a tree (Malinsky et al. 2021). D rel works better in cases where admixture is sample‐specific, not population‐specific, or when relationships between samples is not clear (i.e., samples have a star topology). In these cases, testing all possible trios can quickly become intractable if samples are not grouped into populations, and the tree topology between samples is not informative of admixture timing.

Our approach directly highlights that significant D‐scores can occur even when not using the true donor species in the test. This means it is possible that in some cases, our suggested donors are not accurate because we did not include the true donor species in our reference set. We see in some cases that admixture signals are clustered in a clade, but are not exceptionally high in any donor species. We recognize our current approach to identifying true donor signal based on scores in phylogenetic groups is ad hoc and should be generalized in future work. The traditional mechanism of assessing significance for D is a block jackknife, where data is resampled in blocks along the genome. This is effective at distinguishing a null model of no introgression versus a model with admixture, but does not test between multiple models of admixture. In our case, it is overly permissive and suggests introgression from multiple donors in almost all cases. We believe our approach is helpful for identifying admixture when there are many possible donors and no clear sympatric/allopatric divide.

Our work shows how admixture from unsampled or ghost sources can influence admixture estimates. In our case, unadmixed Quillback Rockfish look more like Copper Rockfish because Yelloweye‐admixed Quillback Rockfish have extra ancestral alleles from the Yelloweye Rockfish introgression, thereby comparatively increasing derived allele sharing between unadmixed Quillback and Copper Rockfish. The strength of this mirror signal depends on the phylogenetic relatedness of each species involved in the test. One approach to understanding these patterns is through building admixture graphs, like in admixtools; although they work best with smaller numbers of samples or populations and admixture events, and are currently not designed to identify outlier samples from populations (Maier and Patterson 2024).

Future work should include as many Sebastes species as possible, although even with an exhaustive approach, extinct species may still leave a genetic signal and cannot be included. Our measures of introgression amount, the f 4‐ratio, assume only one introgression event and it is not clear how all other events will influence this value. Lastly, our approach can only identify the largest introgression contributor per group and either ignores or is confounded by secondary introgression. For example, a balanced score between two species may be because both species have contributed equally. D rel also depends on both the admixture amount and the phylogenetic distance between donor and source, further complicating interpretation. Taken together, we think our approach is effective at identifying likely admixed samples and their donor from a broad pool of options. Further work could focus on likely donors identified here and use more targeted approaches to disentangle admixture patterns.

5. Conclusions and Management Implications

Like many other inshore rockfishes, Copper and Quillback Rockfish are currently managed as two separate populations in BC and WA: inside waters of the Salish Sea and outside coastal waters (Fisheries and Oceans Canada 2024; NMFS 2023). Our findings suggest that the current management regime separating inside and outside of the Salish Sea is likely representative of the biological reality in Copper Rockfish. In terms of coastal Copper populations, current US stocks are assessed in northern and southern populations divided by Point Conception in California. Our work finds that Point Conception is roughly a transition point between southern and northern population groups; the boundary is not absolute, and samples from both groups are found on both sides of the transition. For outside Quillback populations, our data indicate very little structure and near panmixia, although Yelloweye admixture is concentrated in the Salish Sea. Although this may represent neutral or deleterious breakdowns in reproductive isolation, it is also possible that admixture is adaptive and used for local adaptation. Genetic data for Quillback were unavailable when management boundaries were constructed; therefore, the boundaries were constructed based on what was known in Copper Rockfish. Constructing boundaries based on the closely related Copper was reasonable for Quillback at the time, but our study illustrates that the two species have substantially different patterns of population structure. Therefore, the current boundaries are not well supported based on genetic data alone but may be justified based on other factors such as demography, fisheries exploitation, or habitat availability. Our work also highlights that in species with little or no regional population structure, variation in introgression can be the major axis of variation. Importantly, when the true source of introgression is not included, population structure programs can misrepresent the source. Finally, our work suggests that hybridization in Sebastes is far broader than previously thought and can occur between species not thought to occupy similar niches or generally co‐occur. Genomic‐based studies can detect hybridization that occurred hundreds or thousands of years ago that leave no obvious morphological signal; thus, we are getting a fuller picture of how genes have been exchanged in this rapidly speciating genus. Future studies should explore the genes that are being shared between species and their potentially adaptive role. Lastly, studies in Sebastes should appreciate that reproductive barriers are more porous than previously thought, and a wide variety of hybrids are possible.

Author Contributions

Nathan T. B. Sykes: conceptualization, data curation, formal analysis, methodology, software, visualization, writing – original draft preparation and writing – review and editing. R. Nicolas Lou: data curation, formal analysis, methodology, software and writing – review and editing. Matthew R. Siegle: conceptualization, methodology, resources and writing – review and editing. Peter H. Sudmant: conceptualization, funding acquisition, resources and writing – review and editing. Wesley A. Larson: conceptualization, funding acquisition, resources and writing – review and editing. Gregory L. Owens: conceptualization, data curation, formal analysis, funding acquisition, methodology, project administration, supervision, software, visualization, writing – original draft preparation and writing – review and editing.

Funding

This work was supported by an NSERC Discovery grant (RGPIN‐2021‐02482) and funding from Fisheries and Ocean Canada to GLO. This work was also supported by NIH National Institute of General Medicine award R35GM142916 to PHS and NPRB project 2112 to PHS. Computational work was supported by the Digital Research Alliance of Canada, the Canadian Foundation for Innovation and the British Columbia Knowledge Development Fund.

Ethics Statement

All tissue samples used in this study were collected as part of regular fisheries monitoring by Fisheries and Ocean Canada and the National Oceanic and Atmospheric Administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Sequence diversity for individual samples compared with sequence depth.

Figure S2: Copper Rockfish population structure with Ohana qpas for k = 2 to 6. The phylogenetic trees represent similarity between ancestry components.

Figure S3: Quillback Rockfish population structure with Ohana qpas for k = 2 to 6. The phylogenetic trees represent similarity between ancestry components.

Figure S4: mec70537‐sup‐0001‐FigureS1‐S10@Supplementary Figures.pdf. F ST in Quillback Rockfish. (A) A heatmap of F ST between regions. (B) F ST between geographic coordinate samples, with linear trendline added for visualization. Inside and outside refers to coordinates inside and outside the Salish Sea.

Figure S5: PCA and ancestry analysis for Copper Rockfish outside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 5. Phylogeny represents similarity between ancestry components.

Figure S6: PC2 and latitude for Copper Rockfish outside of the Salish Sea. The dotted line indicates Point Conception, an oceanographic transition point.

Figure S7: PCA and ancestry analysis for Copper Rockfish inside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 5. Phylogeny represents similarity between ancestry components.

Figure S8: PCA and ancestry analysis for Quillback Rockfish outside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 4. Phylogeny represents similarity between ancestry components.

Figure S9: PCA and ancestry analysis for Quillback Rockfish inside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 4. Phylogeny represents similarity between ancestry components.

Figure S10: Mitochondrial heterozygosity and total introgression for (A) Copper and (B) Quillback Rockfish. Introgression amounts were calculated as the total f 4‐ratio for all identified donors.

MEC-35-e70537-s001.pdf (2.4MB, pdf)

Table S1: Species, collection location and Sequence Read Archive accession ID for all samples sequenced.

Table S2: Species and accession ID for previous published samples used in admixture analyses.

MEC-35-e70537-s002.xlsx (31KB, xlsx)

Acknowledgements

We would like to thank Schon Hardy from Fisheries and Oceans Canada, and John Hyde, Krista Nichols, and Matt Craig from the National Oceanic and Atmospheric Administration for sharing tissue samples necessary for this work. We thank Katie D'Amelio and Kirby Karpan from the National Oceanic and Atmospheric Administration Alaska Fisheries Science Center for processing tissue samples and preparing libraries for whole genome sequencing. We thank Krista Nichols for helpful feedback on the manuscript.

Data Availability Statement

All sequence data generated for this project is on the SRA listed in Table S1 and in SRA project PRJNA1269622. Additional samples used are listed in Table S2. All code used for this project is available at https://github.com/owensgl/copper_quillback.

References

  1. An, H. , Kim M. J., Park K., et al. 2012. “Genetic Diversity and Population Structure in the Heavily Exploited Korean Rockfish, Sebastes schlegelii , in Korea.” Journal of the World Aquaculture Society 43, no. 1: 73–83. 10.1111/j.1749-7345.2011.00544.x. [DOI] [Google Scholar]
  2. Anderson, S. C. , Keppel E. A., and Edwards A. M.. 2019. “A Reproducible Data Synopsis for Over 100 Species of British Columbia Groundfish.” DFO Can. Sci. Advis. Sec. Res. Doc. 2019/041. 321 p.
  3. Andrews, K. S. , Nichols K. M., Elz A., et al. 2018. “Cooperative Research Sheds Light on Population Structure and Listing Status of Threatened and Endangered Rockfish Species.” Conservation Genetics 19, no. 4: 865–878. 10.1007/s10592-018-1060-0. [DOI] [Google Scholar]
  4. Andrews, S. 2010. “FASTQC: A Quality Control Tool for High Throughput Sequence Data.” http://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
  5. Artamonova, V. S. , Makhrov A. A., Karabanov D. P., Rolskiy A. Y., Bakay Y. I., and Popov V. I.. 2013. “Hybridization of Beaked Redfish ( Sebastes mentella ) With Small Redfish ( Sebastes viviparus ) and Diversification of Redfish (Actinopterygii: Scorpaeniformes) in the Irminger Sea.” Journal of Natural History 47, no. 25–28: 1791–1801. [Google Scholar]
  6. Baumdicker, F. , Bisschop G., Goldstein D., et al. 2022. “Efficient Ancestry and Mutation Simulation With Msprime 1.0.” Genetics 220, no. 3: iyab229. 10.1093/genetics/iyab229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Baym, M. , Kryazhimskiy S., Lieberman T. D., Chung H., Desai M. M., and Kishony R.. 2015. “Inexpensive Multiplexed Library Preparation for Megabase‐Sized Genomes.” PLoS One 10, no. 5: e0128036. 10.1371/journal.pone.0128036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bernatchez, L. , Wellenreuther M., Araneda C., et al. 2017. “Harnessing the Power of Genomics to Secure the Future of Seafood.” Trends in Ecology & Evolution 32, no. 9: 665–680. 10.1016/j.tree.2017.06.010. [DOI] [PubMed] [Google Scholar]
  9. Broad Institute . 2019. “Picard Toolkit.” Broad Institute GitHub Repository. https://github.com/broadinstitute/picard.
  10. Buonaccorsi, V. P. , Kimbrell C. A., Lynn E. A., and Vetter R. D.. 2002. “Population Structure of Copper Rockfish ( Sebastes caurinus ) Reflects Postglacial Colonization and Contemporary Patterns of Larval Dispersal.” Canadian Journal of Fisheries and Aquatic Sciences 59, no. 8: 1374–1384. 10.1139/f02-101. [DOI] [Google Scholar]
  11. Buonaccorsi, V. P. , Kimbrell C. A., Lynn E. A., and Vetter R. D.. 2005. “Limited Realized Dispersal and Introgressive Hybridization Influence Genetic Structure and Conservation Strategies for Brown Rockfish, Sebastes auriculatus .” Conservation Genetics 6, no. 5: 697–713. 10.1007/s10592-005-9029-1. [DOI] [Google Scholar]
  12. Buonaccorsi, V. P. , Westerman M., Stannard J., Kimbrell C., Lynn E., and Vetter R. D.. 2004. “Molecular Genetic Structure Suggests Limited Larval Dispersal in Grass Rockfish, Sebastes rastrelliger .” Marine Biology 145, no. 4: 779–788. 10.1007/s00227-004-1362-2. [DOI] [Google Scholar]
  13. Chen, S. 2023. “Ultrafast One‐Pass FASTQ Data Preprocessing, Quality Control, and Deduplication Using Fastp.” iMeta 2, no. 2: e107. 10.1002/imt2.107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cheng, J. Y. , Mailund T., and Nielsen R.. 2017. “Fast Admixture Analysis and Population Tree Estimation for SNP and NGS Data.” Bioinformatics 33, no. 14: 2148–2155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Czech, L. , and Exposito‐Alonso M.. 2021. “Grenepipe: A Flexible, Scalable, and Reproducible Pipeline to Automate Variant and Frequency Calling From Sequence Reads.” arXiv (arXiv:2103.15167). 10.48550/arXiv.2103.15167. [DOI] [PMC free article] [PubMed]
  16. Danecek, P. , Auton A., Abecasis G., et al. 2011. “The Variant Call Format and VCFtools.” Bioinformatics 27, no. 15: 2156–2158. 10.1093/bioinformatics/btr330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Danecek, P. , Bonfield J. K., Liddle J., et al. 2021. “Twelve Years of SAMtools and BCFtools.” GigaScience 10, no. 2: giab008. 10.1093/gigascience/giab008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Díaz‐Arce, N. , Gagnaire P.‐A., Richardson D. E., et al. 2024. “Unidirectional Trans‐Atlantic Gene Flow and a Mixed Spawning Area Shape the Genetic Connectivity of Atlantic Bluefin Tuna.” Molecular Ecology 33, no. 1: e17188. 10.1111/mec.17188. [DOI] [PubMed] [Google Scholar]
  19. Dick, S. , Shurin J. B., and Taylor E. B.. 2014. “Replicate Divergence Between and Within Sounds in a Marine Fish: The Copper Rockfish ( Sebastes caurinus ).” Molecular Ecology 23, no. 3: 575–590. 10.1111/mec.12630. [DOI] [PubMed] [Google Scholar]
  20. Euclide, P. T. , Larson W. A., Shi Y., et al. 2023. “Conserved Islands of Divergence Associated With Adaptive Variation in Sockeye Salmon Are Maintained by Multiple Mechanisms.” Molecular Ecology 33: e17126. 10.1111/mec.17126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ewels, P. , Magnusson M., Lundin S., and Käller M.. 2016. “MultiQC: Summarize Analysis Results for Multiple Tools and Samples in a Single Report.” Bioinformatics 32, no. 19: 3047–3048. 10.1093/bioinformatics/btw354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fisheries and Oceans Canada (DFO) . 2024. “Groundfish Integrated Fisheries Management Plan 2024.” https://waves‐vagues.dfo‐mpo.gc.ca/library‐bibliotheque/41255732.pdf.
  23. Fraïsse, C. , Belkhir K., Welch J. J., and Bierne N.. 2016. “Local Interspecies Introgression Is the Main Cause of Extreme Levels of Intraspecific Differentiation in Mussels.” Molecular Ecology 25, no. 1: 269–286. 10.1111/mec.13299. [DOI] [PubMed] [Google Scholar]
  24. Gao, T. , Han Z., Zhang X., Luo J., Yanagimoto T., and Zhang H.. 2016. “Population Genetic Differentiation of the Black Rockfish Sebastes schlegelii Revealed by Microsatellites.” Biochemical Systematics and Ecology 68: 170–177. 10.1016/j.bse.2016.07.013. [DOI] [Google Scholar]
  25. Garrison, E. , and Marth G.. 2012. “Haplotype‐Based Variant Detection From Short‐Read Sequencing.” arXiv arXiv:1207.3907. 10.48550/arXiv.1207.3907. [DOI]
  26. Gilbert‐Horvath, E. A. , Larson R. J., and Garza J. C.. 2006. “Temporal Recruitment Patterns and Gene Flow in Kelp Rockfish ( Sebastes atrovirens ).” Molecular Ecology 15, no. 12: 3801–3815. 10.1111/j.1365-294X.2006.03033.x. [DOI] [PubMed] [Google Scholar]
  27. Gopalan, S. , Rodrigues M. F., Ralph P. L., and Haller B. C.. 2025. “Bridging Forward‐In‐Time and Coalescent Simulations Using Pyslim.” bioRxiv: The Preprint Server for Biology. 10.1101/2025.09.30.679676. [DOI]
  28. Haller, B. C. , Galloway J., Kelleher J., Messer P. W., and Ralph P. L.. 2019. “Tree‐Sequence Recording in SLiM Opens New Horizons for Forward‐Time Simulation of Whole Genomes.” Molecular Ecology Resources 19, no. 2: 552–566. 10.1111/1755-0998.12968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Haller, B. C. , Ralph P. L., and Messer P. W.. 2025. “SLiM 5: Eco‐Evolutionary Simulations Across Multiple Chromosomes and Full Genomes.” Molecular Biology and Evolution 43, no. 1: msaf313. 10.1093/molbev/msaf313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hannah, R. W. , and Rankin P. S.. 2011. “Site Fidelity and Movement of Eight Species of Pacific Rockfish at a High‐Relief Rocky Reef on the Oregon Coast.” North American Journal of Fisheries Management 31, no. 3: 483–494. 10.1080/02755947.2011.591239. [DOI] [Google Scholar]
  31. He, Y. , Chang Y., Bao L., et al. 2019. “A Chromosome‐Level Genome of Black Rockfish, Sebastes schlegelii , Provides Insights Into the Evolution of Live Birth.” Molecular Ecology Resources 19, no. 5: 1309–1321. 10.1111/1755-0998.13034. [DOI] [PubMed] [Google Scholar]
  32. Helmstetter, A. J. , Papadopulos A. S. T., Igea J., Van Dooren T. J. M., Leroi A. M., and Savolainen V.. 2016. “Viviparity Stimulates Diversification in an Order of Fish.” Nature Communications 7, no. 1: 11271. 10.1038/ncomms11271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Hibbins, M. S. , and Hahn M. W.. 2022. “Phylogenomic Approaches to Detecting and Characterizing Introgression.” Genetics 220, no. 2: iyab173. 10.1093/genetics/iyab173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hyde, J. R. , and Vetter R. D.. 2007. “The Origin, Evolution, and Diversification of Rockfishes of the Genus Sebastes (Cuvier).” Molecular Phylogenetics and Evolution 44, no. 2: 790–811. 10.1016/j.ympev.2006.12.026. [DOI] [PubMed] [Google Scholar]
  35. Jun, G. , Wing M. K., Abecasis G. R., and Kang H. M.. 2015. “An Efficient and Scalable Analysis Framework for Variant Extraction and Refinement From Population Scale DNA Sequence Data.” Genome Research 25: 918–925. 10.1101/gr.176552.114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kelleher, J. , Etheridge A. M., and McVean G.. 2016. “Efficient Coalescent Simulation and Genealogical Analysis for Large Sample Sizes.” PLoS Computational Biology 12, no. 5: e1004842. 10.1371/journal.pcbi.1004842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Kolora, S. R. R. , Owens G. L., Vazquez J. M., et al. 2021. “Origins and Evolution of Extreme Life Span in Pacific Ocean Rockfishes.” Science 374, no. 6569: 842–847. 10.1126/science.abg5332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Korneliussen, T. S. , Albrechtsen A., and Nielsen R.. 2014. “ANGSD: Analysis of Next Generation Sequencing Data.” BMC Bioinformatics 15, no. 1: 356. 10.1186/s12859-014-0356-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Korunes, K. L. , and Samuk K.. 2021. “ Pixy: Unbiased Estimation of Nucleotide Diversity and Divergence in the Presence of Missing Data.” Molecular Ecology Resources 21, no. 4: 1359–1368. 10.1111/1755-0998.13326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Longo, G. C. , Lam L., Basnett B., et al. 2020. “Strong Population Differentiation in Lingcod ( Ophiodon elongatus ) is Driven by a Small Portion of the Genome.” Evolutionary Applications 13, no. 10: 2536–2554. 10.1111/eva.13037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Love, M. S. , Yoklavich M., and Thorsteinson L. K.. 2002. The Rockfishes of the Northeast Pacific. University of California Press. [Google Scholar]
  42. Maier, R. , and Patterson N.. 2024. “Admixtools: Inferring Demographic History From Genetic Data.” R Package Version 2.0.4. https://github.com/uqrmaie1/admixtools.
  43. Malinsky, M. , Matschiner M., and Svardal H.. 2021. “Dsuite—Fast D‐Statistics and Related Admixture Evidence From VCF Files.” Molecular Ecology Resources 21, no. 2: 584–595. 10.1111/1755-0998.13265. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Matthews, K. R. 1990. “An Experimental Study of the Habitat Preferences and Movement Patterns of Copper, Quillback, and Brown Rockfishes (Sebastes spp.).” Environmental Biology of Fishes 29, no. 3: 161–178. 10.1007/BF00002217. [DOI] [Google Scholar]
  45. Meisner, J. , Liu S., Huang M., and Albrechtsen A.. 2021. “Large‐Scale Inference of Population Structure in Presence of Missingness Using PCA.” Bioinformatics 37, no. 13: 1868–1875. 10.1093/bioinformatics/btab027. [DOI] [PubMed] [Google Scholar]
  46. Minh, B. Q. , Nguyen M. A. T., and von Haeseler A.. 2013. “Ultrafast Approximation for Phylogenetic Bootstrap.” Molecular Biology and Evolution 30, no. 5: 1188–1195. 10.1093/molbev/mst024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Minh, B. Q. , Schmidt H. A., Chernomor O., et al. 2020. “IQ‐TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era.” Molecular Biology and Evolution 37, no. 5: 1530–1534. 10.1093/molbev/msaa015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Murie, D. J. 1995. “Comparative Feeding Ecology of Two Sympatric Rockfish Congeners, Sebastes caurinus (Copper Rockfish) and S. maliger (Quillback Rockfish).” Marine Biology 124, no. 3: 341–353. [Google Scholar]
  49. Muto, N. , Kai Y., Noda T., and Nakabo T.. 2013. “Extensive Hybridization and Associated Geographic Trends Between Two Rockfishes Sebastes vulpes and S. zonatus (Teleostei: Scorpaeniformes: Sebastidae).” Journal of Evolutionary Biology 26, no. 8: 1750–1762. 10.1111/jeb.12175. [DOI] [PubMed] [Google Scholar]
  50. Narum, S. R. , Horn R., Willis S., Koch I., and Hess J.. 2023. “Genetic Variation Associated With Adult Migration Timing in Lineages of Steelhead and Chinook Salmon in the Columbia River.” Evolutionary Applications 17, no. 2: e13626. 10.1111/eva.13626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Natasha, J. , Stockwell B. L., Marie A. D., et al. 2022. “No Population Structure of Bigeye Tunas ( Thunnus obesus ) in the Western and Central Pacific Ocean Indicated by Single Nucleotide Polymorphisms.” Frontiers in Marine Science 9: 799684. 10.3389/fmars.2022.799684. [DOI] [Google Scholar]
  52. Nielsen, E. E. , Hemmer‐Hansen J., Larsen P. F., and Bekkevold D.. 2009. “Population Genomics of Marine Fishes: Identifying Adaptive Variation in Space and Time.” Molecular Ecology 18, no. 15: 3128–3150. 10.1111/j.1365-294X.2009.04272.x. [DOI] [PubMed] [Google Scholar]
  53. NMFS . 2023. “Magnuson‐Stevens Act Provisions; Fisheries Off West Coast States; Pacific Coast Groundfish Fishery Management Plan; Amendment 31.” 88 Fed. Reg. 57,400 (Aug. 23, 2023). https://www.govinfo.gov/content/pkg/FR‐2023‐08‐23/pdf/2023‐18089.pdf.
  54. Okonechnikov, K. , Conesa A., and García‐Alcalde F.. 2016. “Qualimap 2: Advanced Multi‐Sample Quality Control for High‐Throughput Sequencing Data.” Bioinformatics 32, no. 2: 292–294. 10.1093/bioinformatics/btv566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Oksanen, J. , Blanchet F. G., Kindt R., et al. 2013. “Package ‘vegan’.” Community Ecology Package, Version, 2(9) 1, 295.
  56. Pacific Fishery Management Council . 2023. “Stock Assessment Review (STAR) Panel Report for Copper Rockfish in California, Rex Sole, and Shortspine Thornyhead.” https://www.pcouncil.org/documents/2023/06/star‐panel‐1‐report‐for‐copper‐rockfish‐in‐ca‐rex‐sole‐and‐shortspine‐thornyhead.pdf/.
  57. Palsson, W. A. , Tsou T. S., Bargmann G. G., et al. 2009. “The Biology and Assessment of Rockfishes in Puget Sound.” Washington Department of Fish and Wildlife Report FPT‐09‐04.
  58. Pebesma, E. , and Bivand R.. 2023. Spatial Data Science: With Applications in R. Chapman and Hall/CRC. 10.1201/9780429459016. [DOI] [Google Scholar]
  59. R Core Team . 2025. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. https://www.R‐project.org/. [Google Scholar]
  60. Reynolds, J. , Weir B. S., and Cockerham C. C.. 1983. “Estimation of the Coancestry Coefficient: Basis for a Short‐Term Genetic Distance.” Genetics 105, no. 3: 767–779. 10.1093/genetics/105.3.767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Samuk, K. , Owens G. L., Delmore K. E., Miller S. E., Rennison D. J., and Schluter D.. 2017. “Gene Flow and Selection Interact to Promote Adaptive Divergence in Regions of Low Recombination.” Molecular Ecology 26, no. 17: 4378–4390. 10.1111/mec.14226. [DOI] [PubMed] [Google Scholar]
  62. Schwenke, P. L. , Park L. K., and Hauser L.. 2018. “Introgression Among Three Rockfish Species (Sebastes spp.) in the Salish Sea, Northeast Pacific Ocean.” PLoS One 13, no. 3: e0194068. 10.1371/journal.pone.0194068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Seeb, L. 1998. “Gene Flow and Introgression Within and Among Three Species of Rockfishes, Sebastes auriculatus , S. caurinus, and S. maliger .” Journal of Heredity 89, no. 5: 393–403. 10.1093/jhered/89.5.393. [DOI] [Google Scholar]
  64. Siegle, M. R. , Taylor E. B., Miller K. M., Withler R. E., and Yamanaka K. L.. 2013. “Subtle Population Genetic Structure in Yelloweye Rockfish ( Sebastes ruberrimus ) is Consistent With a Major Oceanographic Division in British Columbia, Canada.” PLoS One 8, no. 8: e71083. 10.1371/journal.pone.0071083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Sokal, R. R. 1979. “Testing Statistical Significance of Geographic Variation Patterns.” Systematic Zoology 28, no. 2: 227–232. 10.2307/2412528. [DOI] [Google Scholar]
  66. Tange, O. 2023. “GNU Parallel 20230522 (‘Charles’).” Zenodo, 42–47.
  67. Tarpey, C. M. , Seeb J. E., McKinney G. J., et al. 2018. “Single‐Nucleotide Polymorphism Data Describe Contemporary Population Structure and Diversity in Allochronic Lineages of Pink Salmon ( Oncorhynchus gorbuscha ).” Canadian Journal of Fisheries and Aquatic Sciences 75, no. 6: 987–997. 10.1139/cjfas-2017-0023. [DOI] [Google Scholar]
  68. Therkildsen, N. O. , and Palumbi S. R.. 2017. “Practical Low‐Coverage Genomewide Sequencing of Hundreds of Individually Barcoded Samples for Population and Evolutionary Genomics in Nonmodel Species.” Molecular Ecology Resources 17, no. 2: 194–208. 10.1111/1755-0998.12593. [DOI] [PubMed] [Google Scholar]
  69. Tolimieri, N. , Andrews K., Williams G., Katz S., and Levin P. S.. 2009. “Home Range Size and Patterns of Space Use by Lingcod, Copper Rockfish and Quillback Rockfish in Relation to Diel and Tidal Cycles.” Marine Ecology Progress Series 380: 229–243. 10.3354/meps07930. [DOI] [Google Scholar]
  70. Tovar Verba, J. , Stow A., Bein B., et al. 2022. “Low Population Genetic Structure Is Consistent With High Habitat Connectivity in a Commercially Important Fish Species ( Lutjanus jocu ).” Marine Biology 170, no. 1: 5. 10.1007/s00227-022-04149-1. [DOI] [Google Scholar]
  71. Tricou, T. , Tannier E., and de Vienne D. M.. 2022. “Ghost Lineages Highly Influence the Interpretation of Introgression Tests.” Systematic Biology 71, no. 5: 1147–1158. 10.1093/sysbio/syac011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. van Etten, J. 2017. “R Package Gdistance: Distances and Routes on Geographical Grids.” Journal of Statistical Software 76, no. 13: 1–21. 10.18637/jss.v076.i13.36568334 [DOI] [Google Scholar]
  73. Vasimuddin, M. , Misra S., Li H., and Aluru S.. 2019. “Efficient Architecture‐Aware Acceleration of BWA‐MEM for Multicore Systems.” IEEE International Parallel and Distributed Processing Symposium (IPDPS) 314–324. 10.1109/IPDPS.2019.00041. [DOI]
  74. Wall, L. , Christiansen T., and Orwant J.. 2020. “Perl v5.32.1.” https://www.perl.org.
  75. Waples, R. 1998. “Separating the Wheat From the Chaff: Patterns of Genetic Differentiation in High Gene Flow Species.” Journal of Heredity 89, no. 5: 438–450. 10.1093/jhered/89.5.438. [DOI] [Google Scholar]
  76. Weir, B. S. , and Cockerham C. C.. 1984. “Estimating F‐Statistics for the Analysis of Population Structure.” Evolution 38, no. 6: 1358–1370. 10.2307/2408641. [DOI] [PubMed] [Google Scholar]
  77. Wong, Y. , Ignatieva A., Koskela J., Gorjanc G., Wohns A. W., and Kelleher J.. 2024. “A General and Efficient Representation of Ancestral Recombination Graphs.” Genetics 228, no. 1: iyae100. 10.1093/genetics/iyae100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Wray, A. , Petrou E., Nichols K. M., et al. 2024. “Contrasting Effect of Hybridization on Genetic Differentiation in Three Rockfish Species With Similar Life History.” Evolutionary Applications 17, no. 7: e13749. 10.1111/eva.13749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Xuereb, A. , Rougemont Q., Dallaire X., et al. 2022. “Re‐Evaluating Coho Salmon ( Oncorhynchus kisutch ) Conservation Units in Canada Using Genomic Data.” Evolutionary Applications 15, no. 11: 1925–1944. 10.1111/eva.13489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Zhu, B. , Gao T., He Y., Qu Y., and Zhang X.. 2024. “Population Genomics of Commercial Fish Sebastes schlegelii of the Bohai and Yellow Seas (China) Using a Large SNP Panel From GBS.” Genes 15, no. 5: 5. 10.3390/genes15050534. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: Sequence diversity for individual samples compared with sequence depth.

Figure S2: Copper Rockfish population structure with Ohana qpas for k = 2 to 6. The phylogenetic trees represent similarity between ancestry components.

Figure S3: Quillback Rockfish population structure with Ohana qpas for k = 2 to 6. The phylogenetic trees represent similarity between ancestry components.

Figure S4: mec70537‐sup‐0001‐FigureS1‐S10@Supplementary Figures.pdf. F ST in Quillback Rockfish. (A) A heatmap of F ST between regions. (B) F ST between geographic coordinate samples, with linear trendline added for visualization. Inside and outside refers to coordinates inside and outside the Salish Sea.

Figure S5: PCA and ancestry analysis for Copper Rockfish outside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 5. Phylogeny represents similarity between ancestry components.

Figure S6: PC2 and latitude for Copper Rockfish outside of the Salish Sea. The dotted line indicates Point Conception, an oceanographic transition point.

Figure S7: PCA and ancestry analysis for Copper Rockfish inside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 5. Phylogeny represents similarity between ancestry components.

Figure S8: PCA and ancestry analysis for Quillback Rockfish outside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 4. Phylogeny represents similarity between ancestry components.

Figure S9: PCA and ancestry analysis for Quillback Rockfish inside of the Salish Sea. (A) The first two principal components coloured by sample region. (B) The first two principal components coloured by proportion of missing data. (C) Ancestry from Ohana qpas for K = 2 to 4. Phylogeny represents similarity between ancestry components.

Figure S10: Mitochondrial heterozygosity and total introgression for (A) Copper and (B) Quillback Rockfish. Introgression amounts were calculated as the total f 4‐ratio for all identified donors.

MEC-35-e70537-s001.pdf (2.4MB, pdf)

Table S1: Species, collection location and Sequence Read Archive accession ID for all samples sequenced.

Table S2: Species and accession ID for previous published samples used in admixture analyses.

MEC-35-e70537-s002.xlsx (31KB, xlsx)

Data Availability Statement

All sequence data generated for this project is on the SRA listed in Table S1 and in SRA project PRJNA1269622. Additional samples used are listed in Table S2. All code used for this project is available at https://github.com/owensgl/copper_quillback.

Articles from Molecular Ecology are provided here courtesy of Wiley

RESOURCES