Advancing Conservation Strategies for Native Eastern Highlands-Strain Walleye Sander vitreus in West Virginia: Insights from Genomic Investigations and Broodstock Screening

Abstract

Walleye, Sander vitreus, has several distinct genetic lineages throughout North America as a consequence of Pleistocene glaciation. Stocking walleye across genetic boundaries in the mid-20th century has led to the introduction of non-native strains that persist to this day. In West Virginia, the identification of the native Eastern Highlands strain led the West Virginia Division of Natural Resources (WVDNR) to employ broodstock screening to assist in the conservation of the native strain. To develop a baseline native ancestry prevalence in walleye populations throughout the state, 1532 broodstock were sampled across 17 sampling locations over a 6-year period. To evaluate the effectiveness of the current broodstock two-SNP qPCR assay protocol and identify whether more SNPs need to be implemented, 284 walleye were sequenced and ancestry-genotyped across 42 fixed SNPs between the two strains. When comparing the current protocol to the older microsatellite protocol, advancement in the ability to identify native-strain individuals was observed. Genotyping previously assigned walleye broodstock across multiple fixed SNPs revealed that the current ancestry assignment protocol, on average, assigned individuals that display 96% Eastern Highlands native ancestry to the native strain and accurately identified >93% of all pure Eastern Highlands walleye. Throughout the state of West Virginia, the New and Kanawha River systems contained a high prevalence of native ancestry, with the Ohio River and sampled impoundments displaying varying levels of ancestry. SNPs with >98% prevalence in individuals assigned to the Eastern Highlands strain were identified during the course of the study and can be implemented in future screening protocols. Our results highlight the utility of genomic approaches as tools to assist fisheries management goals and their capability to accurately identify native ancestry to assist in conservation efforts.

1. Introduction

The stocking of fishes into waterways is a common management technique employed throughout the United States to meet conservation needs and enhance sport fisheries [1,2]. The goals and objectives of different stocking plans can vary widely, including reintroductions, population recovery, population supplementation, and a reduction in angling pressure on heavily targeted species [1,2,3]. The criteria for a successful stocking program vary as much as the goals that direct the stocking. Managers and anglers may consider increased angler catch per unit effort (CPUE) a success, while a broader public may be more invested in the supplementation of depleted stocks of species with high public interest [1,4,5]. While utilizing stocking can be met with success, there are variable amounts of success and debates as to its benefits and impacts [6]. Introductions of non-native fish can lead to introgression, and genetic swamping in the native population can occur [7,8], which can result in a loss of genetic diversity [9]. Maintaining the genetic diversity of valuable game fish is often incorporated into management goals. Ensuring that management agencies have proper stocking guidance to conserve genetic diversity can help maintain the native genetic integrity of managed fish species [10,11].

Stocking individuals of the same species to supplement a declining population or enhance an established fishery is common, but the secondary effect on the established population varies from species to species and population to population. The introgression of the native Brook Trout, Salvelinus fontinalis, with the hatchery strain stocked throughout Wisconsin was mostly confined to areas impacted by anthropogenic disturbances, with potential evidence of the ability of native populations to purge hatchery-derived alleles [12]. Atlantic Salmon, Salmo salar, in the southern reaches of their range were stocked with Atlantic Salmon from Scotland to aid in restoring the population; however, it was revealed that heavy stocking resulted in introgression that led to a shift in allelic variation and lowered the effective population size (N_e) [9]. The impacts of stocking individuals that are genetically distinct from the native population can lead to introgression that results in lowered fitness, increased susceptibility to disease, and the introduction of unfit phenotypes [13,14,15].

Walleye, Sander vitreus, is an important recreational and ecological species throughout its range. As an apex piscivorous predator, walleye have an important role in top-down influences on local fish assemblages and thus have been extensively studied to quantify their influence within local waterbodies [16,17,18]. Walleye are often targeted by anglers and support important recreational and economic fisheries in local systems throughout their range [19]. Recreational fisheries and angler site choice in Michigan have been shown to be driven by biomass availability of walleye and Brook Trout, highlighting that increasing abundance of walleye and Brook Trout in these streams would return the most value to the fishery [20]. Due to this strong impact on local economies and ecosystems, walleye have been extensively stocked to meet angler pressure and offset stock reductions [21]. This can be observed in Wisconsin, where more walleye fisheries are becoming dependent on supplemental stocking than ever before [22].

Walleye have distinct lineages in North America that arose as a result of Pleistocene glaciation and the retreating ice sheets. There is believed to be a pre-glacial connection between the Eastern Highland region, which stretches from West Virginia, Virginia, Ohio, and northern Kentucky and extends southward following the Appalachian Mountains, and the Interior Highlands in the central United States, including parts of western Arkansas, Missouri, and eastern Oklahoma [23]. Glaciation altered this previous connection, and as the glaciers began to retreat, fish in unaffected glacial refugia began to recolonize areas previously glaciated, such as the Great Lakes. Colonizers of these newly available systems then became isolated from their region of origin, became susceptible to evolutionary forces such as mutation and genetic drift that alter allele frequencies, and became genetically distinct after ~12–14 thousand years [24,25]. This resulting spatial isolation resulted in clear genetic differentiation between Great Lakes-strain walleye and the Eastern Highlands strain, which is highlighted in the observance of a remnant native population in Claytor Lake, VA, a dammed reservoir within the New River system. A genetically distinct population of river-spawning walleye was observed that possessed a unique mitochondrial haplotype that varied from the lake-dwelling individuals [26]. This mitochondrial haplotype was similar to those observed in the Ohio River, indicating a shared ancestry between the New River and Ohio River individuals. Continued genetic investigations led to the discovery of the Eastern Highlands strain [27,28,29].

The native Eastern Highlands-strain walleye has been stocked across Virginia and West Virginia to establish native populations and supplement existing populations using a genetic-marker-assisted restoration program [30]. Broodstock fish were collected from the wild and held in hatcheries while genetic testing was conducted to identify whether the individual was native or Great Lakes-derived. The markers originally used to screen these broodstock individuals were derived from the initial discovery in the New River, where 94% of individuals bearing the native mitochondrial haplotype possessed a homozygous 99 bp genotype at microsatellite locus SVI-17 and 77% of native individuals had a 78 bp allele at locus SVI-17 [30,31]. For an individual to be considered a pure native-strain walleye, it had to possess both alleles at loci SVI-17 and SVI-33. Due to the fact the alleles were only correlative and not completely diagnostic, native walleye could have been passed over due to having a different set of alleles at these loci. To develop a quicker diagnostic protocol, a quantitative PCR (qPCR) SNP (i.e., single nucleotide polymorphism) assay was developed by using next-generation sequencing to identify fixed SNP differences between known Great Lakes-strain walleye and native Eastern Highlands walleye [31]. Johnson et al. (2023) found 57 fixed SNPs between the two strains, with two of these loci being developed into qPCR SNP assays for broodstock screening. Due to the small number of SNPs in the current qPCR assay, only pure individuals of each strain and their F1 hybrid can be detected, proving to be efficient for rapid broodstock screening; however, they do not allow for investigations into potential backcrossing between the strains and advanced-generation ancestry assignment.

The discovery of the native Eastern Highlands walleye strain led to the development of the Ohio River Fisheries Management Plan to conserve the recently recognized strain. Supplemental stocking of walleye in West Virginia is conducted annually by the West Virginia Division of Natural Resources (WVDNR), and in 2023 alone, a total of 86,892 fingerlings and advanced fingerlings were stocked throughout the state. The vast majority of the broodstock used for fingerling propagation were identified as the Eastern Highlands strain, with a small proportion having been identified as hybrids (Jim Hedrick, WVDNR, personal communication). Studies of walleye genetics in West Virginia have mostly been focused on the Ohio River, with limited studies having been conducted in the interior portions of the state [27,31,32]. To be able to identify areas of high native Eastern Highlands ancestry, identify systems that display high introgression of the Great Lakes strain, and inform future stocking management directions, a baseline of the strain composition must be developed for walleye populations throughout the state of West Virginia.

In this study, several objectives were set to assist in the conservation of the native Eastern Highlands strain in West Virginia. The first objective was to compare the strain ancestry assignments made by the recently developed two-SNP qPCR assay that is used to identify Eastern Highlands strain, Great Lakes strain, and their F1 hybrid to those of the microsatellite diagnostic protocol that was previously used until 2021. Our second objective was to evaluate the efficacy of the two-SNP qPCR panel for the identification of pure and hybrid individuals by comparing it to the full panel of 57 diagnostic SNPs identified in Johnson et al. (2023). Lastly, the strain assignments of all walleye broodstock collected throughout the state of West Virginia during a 6-year span were assessed to calculate the proportion of native walleye in different populations, observe how the proportion of native walleye in these populations has changed through time, and establish an informed ancestry baseline of each population to guide future management directions.

2. Methods

2.1. Sample Collection

Walleye broodstock were sampled throughout the state of West Virginia between the years of 2019 and 2024 (Figure 1). Broodstock were sampled during the spawning run, typically late February–middle of March in each of the six years. All sampling was performed by the West Virginia Division of Natural Resources (WVDNR) following a standard protocol. Fin clips were taken from individual walleye and sent to the Wild Genomics Lab at West Virginia University for genetic analysis.

2.2. Laboratory Methods

DNA was extracted from fin clips using the Promega Wizard^® SV 96 Genomic DNA Purification System (©Promega, Madison, WI, USA). DNA extracts were quantified on a Nanodrop spectrophotometer (©ThermoScientific, Wilmington, DE, USA) and standardized to a concentration of 20 ng/µL for sequencing and to 10 ng/µL for broodstock screening using the qPCR and microsatellite protocols.

Microsatellite analysis was used for strain identification in 2019 and 2020, prior to the development of SNP assays in 2021 that were used for 2021–2024 walleye broodstock screening. Microsatellite DNA variability was investigated following a protocol adapted from Borer et al. (1999) [33], with amplified products verified using capillary electrophoresis on an ABI 3500xl Genetic Analyzer and an internal size standard (GeneScanTM 600 LIZ^TM Size Standard v2.0Dx, Applied Biosystems, Walthem, MA, USA). Two polymorphic, putatively diagnostic loci (Svi17 and Svi33) [26] were examined, with alleles previously used to diagnose native walleye in West Virginia. Native Eastern Highlands walleye were identified only if they were homozygous for both the 99 bp allele at locus SVI7 and the 78 bp allele at locus SVI33; hybrids were identified if they were heterozygous at one or both of the sites and non-native if they possessed neither of these alleles. Each reaction consisted of 2 µL of nanopure water, 5 µL of 2x Qiagen Multiplex PCR Master Mix, 1 µL of 2 µM primer mix, and 2 µL of standardized DNA that underwent the following thermocycling conditions: 95° for 15 min, 30 cycles of 94° for 30 s, 57° for 90 s, and 72° for 60 s, with a final step of 60° for 30 min. Samples were then sent to the WVU Genomics Core Facility (CTSI Grant #U54 GM104942) for fragment analysis using a LIZ600 size standard (©ThermoScientific, Wilmington, DE, USA). Allele peaks were identified and manually confirmed using GeneMarker™ Genotyping Software V 2.6 by SoftGenetics.

All walleye broodstock samples collected for the current study were run using a two-SNP assay consisting of SNPs 5164 and 14,317 (Johnson et al., 2023;

Table 1. Details of the two TaqMan^TM SNP assays used to identify native, Great Lakes-strain, and F₁-hybrid walleye. For each probe, the diagnostic SNP is highlighted in red. In the assay, the Great Lakes probe was VIC-labeled; the native probe was FAM-labeled.

Assay	Forward Primer	Reverse Primer	Great Lakes Probe	Native Probe
5164	TGCAGCCTCAAATACCTTGGG	TCTGCTGCGCCGATTCTC	CAATCTCCCACTCCATTG	ATCTCCCACACCATTG
14,317	GCGGTTGGCCATCAGTGA	TCCTGGACGCCTGGGA	CTCAGGAGATCAGATGC	TCAGGAGACCAGATGC

) using qPCR thermal cycling (©Bio-Rad CFX Connect). Each reaction was run with a 10 µL total volume containing 5 µL of TaqMan™ 2x Genotyping Master Mix (©ThermoScientific), 2.5 µL of nanopure water, 2 µL of sample DNA standardized to 10 ng/µL, and 0.5 µL of the 20 SNP assay. Individual runs contained two no-template controls, a positive native Eastern Highlands-strain sample, and a positive Great Lakes-strain sample to confirm strain calls. The qPCR thermal cycler program was run at 95° for 10 min, followed by 40 cycles of 95° for 15 s and 60° for one minute. For both assays, the Great Lakes strain was VIC-labeled, and the Eastern Highland strain was FAM-labeled.

To confirm the effectiveness of the current two-SNP qPCR protocol in assigning true native Eastern Highlands ancestry, 284 previously screened walleye broodstock were sequenced using a double-digest restriction-site-associated DNA (ddRAD) genotype-by-sequencing (GBS) protocol adapted from Poland et al. (2012) [34]. In short, samples were digested with two restriction enzymes (PstI and MspI); unique barcoded adaptors (0.1 µM) and common reverse (10 µM) adaptors in 1x adaptor buffer were added to each sample and ligated onto the cut ends. Samples were then pooled, with fragments between 250 and 450 bp selected using a Pippin Prep (Sage Science, Beverly, MA, USA), with PCR amplification conducted on the resulting fragments. The quality and quantity of the resulting product were assessed using an Agilent 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA). Samples were then sequenced using an Illumina NextSeq 2000 (Illumina, San Diego, CA, USA), generating paired-end reads that were then analyzed for ancestry assignment.

2.3. Genomic Analysis

Paired-end sequences were aligned to the walleye reference genome (NCBI WGS accession: JAKUJW01) using Bowtie 2 software [35]. The resulting SAM files were then converted to BAM files using SAMtools software version 1.20 [36]. Aligned sequences were then processed using STACKS version 2.65 [37]. STACKS was used to filter the sequenced data and to call genotypes at the previously identified 57 fixed SNP differences between the Great Lakes and Eastern Highlands strains [31]. Sequences were cleaned (-c), low-quality reads discarded (-q), and individuals separated based on barcode sequences using process_radtags. Low-quality reads were discarded if bases were uncalled or if the raw Phred score dropped below 10 using a sliding window of 15%. Processed sequences were then analyzed using the ref_map.pl pipeline. The resulting catalog was manually searched to identify the 57 fixed loci and then called with the populations program by whitelisting (-W) the identified loci. The resulting catalog was then analyzed in BioEdit [38], where alleles at each locus for each individual were manually scored and recorded. At each locus, ancestry was genotyped based on whether the SNP was homozygous for the known Great Lakes-strain allele (i.e., pure Great Lakes), homozygous for the known Eastern Highlands-strain allele (i.e., pure native), or heterozygous, indicating a hybrid of the two strains. The resulting file was then analyzed in NEWHYBRIDS [39] using the EasyParallel GUI [40] to assign individuals to Eastern Highlands ancestry, Great Lakes ancestry, F1 hybrids, F2 hybrids, or backcrossed hybrids, resulting in 6 potential classifications of further ancestry assignments. The data were analyzed using 10,000 burn-in iterations and 10,000 sweeps for six runs. Individuals with fewer than 10 genotyped SNPs and loci found in fewer than 100 individuals were removed from ancestry assignment calculations, resulting in 239 individuals and 42 SNPs.

3. Results

3.1. Comparison of Microsatellite and SNP-Assay Strain Identification Calls

In 2019 and 2020, a total of 388 walleye were tested with the microsatellite diagnostic protocol and compared to ancestry assignments using the two-SNP assay (

Table 2. The total numbers of Eastern Highlands-strain, Great Lakes-strain, and hybrid individuals identified by the microsatellite panel and the SNP assay in 2019 and 2020 are presented. The percentage of matching calls by the other protocol is presented in parentheses.

	Eastern Highlands	Hybrid	Great Lakes
Microsatellite panel	208 (87.0%)	171 (21.0%)	9 (77.7%)
SNP assay	312 (58.0%)	61 (59.0%)	15 (46.6%)

). The microsatellite panel was corroborated by the two-SNP qPCR assay in 87% of the individuals assigned to the Eastern Highlands strain, but the microsatellite protocol was only able to identify 58% of the total Eastern Highlands broodstock identified by the two-SNP assay during the two-year sampling period. Of the individuals called Eastern Highlands by the microsatellite panel, 11.5% were found to be hybrids by the SNP assay, and 1.5% were found to be Great Lakes individuals. The microsatellite panel vastly underperformed in its ability to distinguish between hybrid individuals and Eastern Highlands individuals; of the 171 hybrids identified by the microsatellite panel, 130 (76.0%) were found to be the native Eastern Highlands strain by the SNP assay.

3.2. Comparison of Two-SNP Strain Assignment to The Use of Additional Diagnostic SNPs

Of the 239 walleye sequenced for further ancestry assignment, 21 were found to be the Great Lakes strain by the two-SNP assay protocol, 181 were identified as the Eastern Highlands strain, and 37 were found to be F1 hybrids between the two strains. On average, each individual was genotyped at 31.9 loci with a median of 34, and the average number of individuals genotyped for each of the 42 SNPs was 181.7. Among the 181 Eastern Highlands-strain walleye identified by the two-SNP assay, the average percentage of Eastern Highlands ancestry using all SNPs was found to be 92.6%, with a range of 54.3% to 100%.

When comparing the probability of six ancestry assignments to previous strain assignments based on the current screening protocol (

Table 3. The 239 walleye sequenced for ancestry assignment are presented by strain identification using the two-SNP assay, and their further ancestry assignment probability percentage (%) assignment when utilizing 48 SNPs is presented. The number in parentheses represents the number of individuals assigned to each of the 3 strain classifications based on the two-SNP assay.

Strain	Probability	Eastern Highlands	Great Lakes	F1 Hybrid	F2 Hybrid	EH × Bx	GL × Bx
Eastern Highlands (181)	>99.999	81.2	0.0	0.0	0.5	2.2	0.0
	>95.00	87.2	0.0	0.0	1.1	5.5	0.0
	>80.00	89.5	0.0	0.0	1.1	6.6	0.0
F1 hybrid (37)	>99.999	21.6	0.0	0.0	35.1	0.0	0.0
	>95.00	29.7	0.0	0.0	35.1	8.1	0.0
	>80.00	32.4	0.0	2.7	40.5	13.5	0.0
Great Lakes (21)	>99.999	0.0	90.4	0.0	4.7	0.0	0.0
	>95.00	0.0	90.4	0.0	4.7	0.0	0.0
	>80.00	0.0	90.4	0.0	4.7	0.0	0.0

), it was found that among the 181 previously called Eastern Highlands individuals, 147 (81.2%) had > 99.999%, 158 (87.2%) had > 95.000%, and 161 (88.95%) had > 80.00% probability of being pure native Eastern Highlands walleye. The second most common ancestry assignment was a backcross between an F1 hybrid and a pure Eastern Highlands walleye, with 12 individuals (6.6%) having > 90.00% probability of belonging to this classification. Two individuals (1.1%) were found to be F2 hybrids with probabilities > 99.50%. No Great Lakes-strain individuals, F1 hybrids, or F1 backcrosses with the Great Lakes strain were identified in the previously identified native Eastern Highland broodstock.

Among the 37 previously identified F1 hybrids, 11 individuals (29.7%) displayed >98.500% probability of being Eastern Highlands individuals, and 1 individual displayed an 85.6% probability of belonging to this genotype classification. The most common assignment among previously called F1 hybrids was an F2 hybrid, with 13 individuals (35.1%) displaying >99.100% probability of belonging to this classification. Four individuals displayed >91.000% probability of originating from an F1 individual backcrossed with a pure Eastern Highlands individual, and one individual displayed a 94.797% probability of being an F1 hybrid. No Great Lakes-strain individuals or F1 backcrosses with the Great Lakes strain were identified among the previously identified hybrids.

Among the 21 Great Lakes-strain individuals, 19 (90.48%) were found to display >99.999% probability of belonging to the pure Great Lakes genotype classification. One individual was found to be an F2 hybrid (99.76% probability), and one individual was found to be either an F1 hybrid backcrossed to the Great Lakes strain or an F2 hybrid, with 53.97 and 46.03% probability of each classification, respectively.

The use of one diagnostic SNP may be sufficient for rapid broodstock screening, as of the over 1500 individuals sampled with the two-SNP qPCR assay, only 1 individual had different ancestry assignments at the two SNPs (Table 1). By effectively using one known variant SNP between the two strains, we found that the current qPCR SNP protocol accurately assigned 93.06% of all 173 identified native Eastern Highlands walleye upon utilizing more diagnostic SNPs while also being 88.95% accurate when assigning an individual Eastern Highlands-strain ancestry. When investigating what individuals the current protocol misassigned as Eastern Highlands, 17 of the 20 (85%) were found to be backcrosses with the Eastern Highlands strain, and 3 individuals were likely to be F2 hybrids. The current protocol also correctly identified each of the 19 Great Lakes individuals and misassigned 1 individual that was found to be an F2 hybrid and 1 individual that was found to likely be a backcross of the Great Lakes strain. Among the 37 previously identified hybrids, 15 (40.5%) were found to be F2 hybrids, 12 (32.4%) were found to be Eastern Highlands individuals, 5 (13.4%) were found to be backcrosses of the Eastern Highlands strain, and 1 F1 hybrid was identified, with the remaining displaying mixtures of probability of either being an F2 or backcross with an Eastern Highlands individual upon further ancestry assignment classification.

3.3. Ancestry Assignment of Sequenced Walleye

Of the 239 sequenced walleye, 173 (72.4%) were found to display >80% probability of belonging to the pure Eastern Highlands-strain genotype assignment. Three genotype classifications were found to be prevalent in the remaining 66 individuals, with 19 identified as Great Lakes-strain individuals, 18 found to be F2 hybrids, and 17 classified as backcrosses with the Eastern Highlands strain. Only one F1 hybrid was identified in the current study, and the remaining individuals all had below 80% probability of belonging to a single genotype classification, displaying mixed probabilities of being the Eastern Highlands strain, F2 hybrids, backcrosses with the Eastern Highlands strain, or the Great Lakes strain.

The 173 identified Eastern Highlands individuals were further investigated for genomic insights to improve the current two-SNP qPCR assay protocol. The average number of SNPs genotyped per Eastern Highlands individual was 32.5 with a median of 34, with a total of 5624 genotyped SNPs across these 173 highly probable native walleye. Of these genotyped SNPs, 5119 (91.02%) contained the Eastern Highlands allele, 145 (2.58%) SNPs contained the Great Lakes allele, and 360 (6.4%) were heterozygous. On average, per individual, 91.2% of the genotyped SNPs display the Eastern Highlands allele, 2.76% display the Great Lakes allele, and 6.04% display heterozygosity.

Among the 42 genotyped SNPs within the 173 of Eastern Highlands walleye, the average prevalence of the Eastern Highlands allele was 90.7% with a median of 93.2% and a range of 64.2–100%. Of the 42 SNPs genotyped, 11 (26.1%) had zero Great Lakes alleles observed, with an average of 126.7 individuals being genotyped across these 11 SNPs. One SNP, 98337, had a 100% prevalence of Eastern Highlands alleles in the pure Eastern Highland individuals and was genotyped in 143 of the 173 (83.2%) individuals. SNP 98,337 was also found to be fixed for the Great Lakes allele in the 16 Great Lakes-strain individuals in which the SNP was genotyped. Two additional SNPs were highly prevalent for the Eastern Highlands allele; SNP 47,040 was genotyped in 137 Eastern Highlands individuals and displayed the native allele in all but 4 individuals, which were found to be heterozygous. SNP 65,462 was genotyped in 110 individuals and displayed the native Eastern Highlands allele in all but 4 individuals, which were all also found to be heterozygous.

3.4. Evaluation of the Prevalence of Native Eastern Highlands Walleye in West Virginia

Over a 6-year time period, 5 of the 16 sampling locations consisted of over 80% native walleye broodstock (

Table 4. The percentage of native Eastern Highlands-strain walleye in West Virginia is presented by sampling location broken down by sampling year. The sum percentage of native ancestry through the duration of the project is presented in the final column. The number of walleye tested in each year and the sum total of walleye tested are presented in parentheses. N/A indicates that no individuals were collected from the sampling location during the given year.

Location	2019	2020	2021	2022	2023	2024	Total
Charles Fork Lake	94.1% (17)	90.0% (30)	88.2% (17)	100% (18)	81.8% (11)	N/A	91.4% (93)
Cheat Lake	N/A	20% (20)	N/A	N/A	N/A	N/A	20.0% (20)
Dog Run Lake	100% (25)	N/A	100% (11)	100% (24)	100% (9)	100% (41)	100% (110)
Kanawha River Marmet Pool	88.9% (36)	83.9% (62)	77.8% (81)	89.3% (47)	82.8% (35)	83.9% (62)	83.3% (323)
Kanawha River Kanawha Falls	83.7% (43)	87.5% (40)	86.4% (37)	84.5% (58)	100% (10)	92.6% (27)	86.9% (215)
Monongahela River	N/A	30.0% (20)	N/A	N/A	N/A	N/A	30.0% (20)
New River	78.8% (33)	92.1% (38)	100% (30)	94.2% (35)	93.0% (57)	100% (36)	93.0% (229)
Stonecoal Reservoir	N/A	N/A	N/A	1.5% (64)	N/A	N/A	1.5% (64)
Summersville Lake	N/A	54.5% (11)	66.6% (12)	N/A	N/A	N/A	60.1% (23)
Sutton Lake	N/A	N/A	61.5% (26)	N/A	N/A	N/A	61.5% (26)
Tygart Lake	N/A	0% (20)	N/A	N/A	N/A	N/A	0% (20)
Ohio River Pike Island Pool	N/A	66.6% (24)	88.9% (18)	N/A	N/A	N/A	76.2% (42)
Ohio River Hannibal Pool	N/A	58.3% (12)	60.0% (10)	66.6% (6)	N/A	N/A	60.7% (28)
Ohio River Willow Island Pool	55.9% (34)	20.0% (5)	69.7% (43)	65.2% (23)	28.6% (7)	86.6% (15)	63.8% (127)
Ohio River Racine Pool	N/A	54.5% (11)	N/A	68.9% (29)	57.1% (7)	75% (12)	67.8% (59)
Ohio River R.C. Byrd Pool	50.0% (30)	63.1% (19)	100% (1)	50.0% (2)	42.8% (7)	N/A	54.2% (59)
Ohio River Greenup Pool	53.8% (13)	33.3% (12)	82.3% (17)	50.0% (10)	68.1% (22)	N/A	60.8% (74)

). Tygart Lake had the lowest proportion of Eastern Highlands walleye with 0%, followed by Stonecoal Lake with 1.5%, Cheat Lake with 20%, and the Monongahela River with 30%. The Kanawha River and New River systems both contained a very high proportion of native walleye broodstock, with averages of 85.1% and 93.0%, respectively, over the six-year sampling period. The Ohio River system had a lower prevalence of Eastern Highlands walleye in comparison, with native ancestry ranging from 54.2% to 76.2% across the sampled pools. Overall, the Ohio River contained an average of 63.9% native Eastern Highlands walleye across the system. Dog Run Lake (100%) and Charles Fork Lake (91.4%), two native broodstock lakes established by the WVDNR to conserve Eastern Highlands genetic diversity, both contain a high prevalence of native ancestry.

4. Discussion

Insights gained from the results of this study highlight the benefits of next-generation sequencing and demonstrate how it can be used in a complementary role to inform fisheries management decisions. Our results highlight the effectiveness of utilizing genomics to assist in restoring native genetic diversity, the capability of effectively utilizing one known variant SNP to differentiate between native and non-native strains, and improvements in diagnostic protocols to enhance ongoing restoration efforts. The previously used microsatellite protocol was similarly effective in accuracy when calling a native strain compared to the SNP protocol (87 and 88.95%, respectively), but the ability of the SNP protocol to correctly call 93% of all identified native walleye compared to the microsatellites 58% shows vast improvement in screening effectiveness. Even when investigating the individuals that the SNP protocol misidentifies, 3 were found to be probable F2 hybrids, while the remaining 17 were found to be probable F1 backcrosses with the Eastern Highlands strain, the second highest prevalence of native ancestry possible among the six genotype classifications. By genotyping sampled walleye broodstock throughout the state of West Virginia, we were also able to identify SNP 98,337 as the primary candidate SNP to utilize in further screening efforts due to it being the only SNP found to be truly fixed for the Great Lakes and Eastern Highlands alleles in the pure individuals of each strain.

After the discovery of the Eastern Highlands strain due to parallel genetic investigations [26,27,28,29], the WVDNR used two microsatellite loci that were correlative with a diagnostic mitochondrial haplotype during broodstock screening to conserve the newly discovered native strain throughout the state [26,41]. When investigating how well these microsatellite markers diagnose true native-strain walleye, we found that the microsatellite protocol identified only 58% of the native Eastern Highlands broodstock screened in a two-year time period compared to the current two-SNP qPCR assay. By advancing to a protocol that used observed fixed differences between the two genetic strains [31] rather than a protocol based on mitochondrial haplotype correlations [26], we found that the two-SNP protocol correctly identifies 93% of all individuals displaying >80% probability of being the pure Eastern Highlands strain. With the high prevalence of this pure Eastern Highlands assignment and the second most common assignment being individuals of an F1 backcrossed with a pure Eastern Highlands individual, our genetic screening protocol has sufficient resolution to conserve and restore the native Eastern Highlands strain. When looking at pure ancestry based on allele frequency rather than probability assignment, we found that only two previously identified native Eastern Highlands individuals contained less than 70% native ancestry when using the full panel of diagnostic SNPs. These two individuals came from Summersville Reservoir and the R.C. Byrd Pool of the Ohio River, two sampling locations found to have a high degree of mixed ancestry within their systems (Table 2) [27,32].

The quantification of the prevalence of Eastern Highlands-strain walleye throughout West Virginia is novel with respect to the sampling effort and the use of new diagnostic protocols but represents an expansion of previous work on the prevalence of genetic ancestry in the region. A recent study utilizing microsatellite variation found that the Ohio River and New River displayed significant population structure between the two systems (F_ST = 0.03), but both systems were genetically distinct from the Great Lakes (F_ST = 0.06 and 0.11, respectively) [42]. Although the walleye were not tested for strain lineage, the lowered population structure between the Ohio River and the Great Lakes compared to the New River is substantiated in the current study, with the Ohio River displaying a lower prevalence of Eastern Highlands walleye than the New River (Table 2). Previous work had found that the New River and Kanawha River systems displayed limited introgression with the Great Lakes strain in their waters, while the Ohio River, in comparison, was more introgressed [27,32,41]. This is supported by the current results, with the Kanawha River displaying roughly 85% native ancestry assignment throughout the system and the New River displaying 93%. A trend has been previously observed within the Ohio River that more upstream pools (Pike Island and Hannibal Pools) contained a higher degree of native genetic ancestry than more downstream pools (R.C. Byrd and Greenup Pools) [27,32]. This trend remains observable in the current study, with R.C. Byrd Pool displaying the lowest percentage of native individuals among Ohio River pools and Pike Island Pool displaying the highest. This trend is likely due to the previous stockings of Great Lakes-derived walleye into streams of the Ohio River from management agencies throughout the region [27]. Further research may need to be conducted to determine whether environmental factors are playing a role in the distribution and reproductive success of both strains throughout the Ohio River system. Tygart Lake, the only system in the current study to have 0% Eastern Highlands ancestry, was originally stocked with Great Lakes-derived walleye following reintroduction into the reservoir, with our results and previous research showing no native ancestry remaining in the system [31,41].

The conservation of genetic strains via the use of stocking and hatchery supplementation has been documented across a wide array of species, and genetic investigations have often provided insight to improve stocking practices, conserve native populations, and enhance fisheries. In Brown Trout, Salmo trutta, it has been found that the native Brown Trout’s genetic diversity significantly increased downstream, while captive-reared individuals had increasing genetic diversity upstream [43]. Atlantic Salmon, Salmo salar, in the southern reaches of their range were stocked with Atlantic Salmon from the north to restore the population; however, heavy stocking resulted in introgression that led to a shift in allelic variation and lowered the effective population size (N_e) [9]. Quantifying the distribution of native and introduced strains contributes to key baseline information that supports conservation and fisheries management. This has already been implemented with Brook Trout in North Carolina, where over 9000 Brook Trout were genotyped, and a genetic baseline was established, with little to no introgression from a hatchery strain occurring in most populations sampled [10]. This finding is key in establishing native populations, but it also informs on the effectiveness of stocking these hatchery-reared Brook Trout strains in North Carolina, guiding managers in future efforts.

Updating stocking records based on genetic investigations has been implemented in other states with walleye as well. Throughout a sample area that consisted of the Interior Highlands throughout Arkansas and Missouri, genetic differences were detected using mitochondrial and microsatellite techniques [23]. Berkman et al. (2023) found that the Black River watershed contained an almost exclusive group of genetically distinct Highlands walleye that had remained isolated and uninfluenced by historical stockings of Great Lakes walleye in the region. The influences of stocking and the identification of genetically distinct, naturally recruiting walleye populations have also been observed in Wisconsin [44], leading to the delineation of four genetic groups that likely arose due to historical glacial activity in the region and the observation of genetic impacts in populations that had been previously stocked across genetic boundaries.

Similar genomic methods to the current study have also been used to conserve an imperiled southern walleye strain [45]. Zhao et al. (2020) developed an SNP panel to readily identify southern and northern walleye lineages, resulting in evidence of an anthropogenic hybridization zone as a consequence of previous stockings within the southern walleye range. Similar approaches have been used within the Great Lakes with the development of panels to identify potential local adaptations and to delineate stocks to guide management practices [46]. While the costs of large-scale genotyping have become more affordable, the adaptation of these genomic and genetic approaches to guide and assist fisheries management practices is well underway.

Our results highlight the advancement in capabilities to conserve native genetic ancestry in the Eastern Highlands region by showing the differences between the microsatellite and two-SNP qPCR protocols and then are substantiated by the high prevalence of Eastern Highlands ancestry when more SNPs are used to infer ancestry. These results show that the current protocol is advancing the conservation of the Eastern Highlands strain and shows the effectiveness of using genomic approaches to assist fisheries management. By using the newly developed protocol to test six years of walleye broodstock collection by the WVDNR, we were able to assess the prevalence of Eastern Highlands ancestry in walleye populations throughout the state to assist in future management directions.