http://orcid.org/0000-0003-0987-8445
Figures
Abstract
Population structure of three ecotypes of Oncorhynchus nerka (sea-type Sockeye Salmon, lake-type Sockeye Salmon, and Kokanee) in the Fraser River and Columbia River drainages was examined with microsatellite variation, with the main focus as to whether Kokanee population structure within the Fraser River drainage suggested either a monophyletic or polyphyletic origin of the ecotype within the drainage. Variation at 14 microsatellite loci was surveyed for sea-type and lake-type Sockeye Salmon and Kokanee sampled from 121 populations in the two river drainages. An index of genetic differentiation, FST, over all populations and loci was 0.087, with individual locus values ranging from 0.031 to 0.172. Standardized to an ecotype sample size of 275 individuals, the least genetically diverse ecotype was sea-type Sockeye Salmon with 203 alleles, whereas Kokanee displayed the greatest number of alleles (260 alleles), with lake-type Sockeye Salmon intermediate (241 alleles). Kokanee populations from the Columbia River drainage (Okanagan Lake, Kootenay Lake), the South Thompson River (a major Fraser River tributary) drainage populations, and the mid-Fraser River populations all clustered together in a neighbor-joining analysis, indicative of a monophyletic origin of the Kokanee ecotype in these regions, likely reflecting the origin of salmon radiating from a refuge after the last glaciation period. However, upstream of the mid-Fraser River populations, there were closer relationships between the lake-type Sockeye Salmon ecotype and the Kokanee ecotype, indicative of the Kokanee ecotype evolving independently from the lake-type Sockeye Salmon ecotype in parallel radiation. Kokanee population structure within the entire Fraser River drainage suggested a polyphyletic origin of the ecotype within the drainage. Studies employing geographically restricted population sampling may not outline accurately the phylogenetic history of salmonid ecotypes.
Citation: Beacham TD, Withler RE (2017) Population structure of sea-type and lake-type sockeye salmon and kokanee in the Fraser River and Columbia River drainages. PLoS ONE 12(9): e0183713. https://doi.org/10.1371/journal.pone.0183713
Editor: Tzen-Yuh Chiang, National Cheng Kung University, TAIWAN
Received: March 28, 2017; Accepted: August 9, 2017; Published: September 8, 2017
Copyright: © 2017 Beacham, Withler. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Allele frequencies for all populations surveyed in the study are available via DRYAD doi identified as: doi:10.5061/dryad.3g824. Data files: Baseline Allele Frequencies.
Funding: Funding was provided by Fisheries and Oceans Canada and the Province of British Columbia. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The Pacific salmon species Sockeye Salmon Oncorhynchus nerka is characterized by three main ecotypes that are distinguished by differences in life history in fresh water. The “lake-type” ecotype typically spawns in lakes, or in tributaries associated with lakes, their offspring rear in these nursery lakes for at least one year before migrating to the ocean [1], and it is generally the most widespread and abundant life history type. However, where lake-rearing habitat is inaccessible or unavailable, the second ecotype can be common, where Sockeye Salmon spawn in tributaries or mainstem side channels, and the juveniles rear for several months in estuarine waters (“sea-type”) or at least one year (“river-type”) in the river environment before migrating to the ocean [2, 3]. Sea-type and river-type Sockeye Salmon are similar in that the juveniles both rear in river habitats prior to smolt migration to the ocean, but sea-type juveniles do not spend a winter in fresh water, and thus lack a freshwater annulus. The river-type form has been considered to be a special case of the sea-type form because neither type rear in lakes [4]. At maturity, both ecotypes undertake an anadromous migration, returning from the ocean to spawn in fresh water in their natal rivers. The third ecotype is commonly known as Kokanee, in which individuals in this ecotype are non-anadromous and complete their life cycle entirely in fresh water [5]. Within each ecotype, there can be differentiation with respect to spawning locations, with adults spawning on beaches within lakes, or in tributary rivers and streams [6, 7, 8, 9]. There is also evidence to suggest that in salmonids alternative migratory tactics co-exist within populations, and that all individuals may potentially adopt any of the alternative phenotypes or ecotypes [10].
The evolutionary relationships among the lake-type, sea-type, and Kokanee ecotypes have been a matter of continuing interest. The components of a “recurrent evolution” hypothesis for Oncorhynchus nerka have been outlined previously [4]. The basic components of the hypothesis included the following three main assumptions. The sea-type ecotype was considered a genetically-diverse ancestral form with poorly genetically differentiated populations, and straying by this ecotype allowed new habitats to be colonized after glacial retreat. Once the lake habitat became accessible and productive, genetically differentiated lake-type Sockeye Salmon evolved repeatedly from the sea-type ecotype in parallel adaptive radiations. When the lake environment became sufficiently productive, then the fitness of nonanadromous individuals was postulated to become at least equivalent to that of anadromous individuals, and consequently populations of the Kokanee ecotype evolved independently from the lake-type Sockeye Salmon ecotype in a parallel adaptive radiation.
An alternative perspective on Kokanee evolution was provided by [11], with allozyme frequencies for Kokanee populations from a portion of the Fraser River and Columbia River drainages more similar to each other than either was to allozyme frequencies for their respective sympatric Sockeye Salmon populations. In this perspective, it is assumed that Kokanee populations in the Fraser River and Columbia may share a common monophyletic origin relative to their sympatric Sockeye Salmon counterparts, and that present day population structure of Kokanee reflects radiation from a glacial refugium and gene exchange between these two river basins, rather than independent parallel evolution from the lake-type Sockeye Salmon ecotype within each basin. The key question to evaluate in the current study was whether the Kokanee ecotype in the Fraser River and Columbia River drainages evolved independently in parallel adaptive radiation from the lake-type Sockeye Salmon ecotype [4] or whether the Kokanee ecotype in both river drainages share a common monophyletic origin [11]. Earlier studies using allozymes showed distinct differences between the sea-type and lake-type ecotype population structure. Differentiation among lake-type populations was attributed to strong homing fidelity to their natal streams, whereas a lack of differentiation among sea-type populations was interpreted as reflecting high rates of straying among populations [3, 12]. Later studies with microsatellites on the population structure of the lake-type and sea-type ecotypes indicated that a regional structuring of populations was observed, with populations typically clustered within lakes and river drainages [13]. Within British Columbia, there was evidence of genetic differentiation among sea-type populations inhabiting different river drainages, with those in the Alsek River distinct from those in the Stikine and Taku rivers in northern British Columbia, and with those in the Nass River and Fraser River distinct from those in the Alsek, Taku, and Stikine river drainages [14].
In the current study, we outline the results of a survey of microsatellite variation of the sea-type Sockeye Salmon, lake-type Sockeye Salmon, and Kokanee ecotypes of O. nerka populations in the Fraser River and Columbia River drainages, and evaluate the recurrent evolution hypothesis outlined by [4] with reference to the three main assumptions of O. nerka evolution. Comparisons are conducted between the level of genetic diversity observed for two sea-type populations in the lower Fraser River drainage relative to that observed between Sockeye Salmon and Kokanee populations within the Fraser River and Columbia River drainages, with the analysis conducted by comparisons of heterozygosity as well as comparing the number of alleles observed in each ecotype standardized to a common sample size. Next, we evaluate whether genetically differentiated lake-type Sockeye Salmon have evolved repeatedly from the sea-type ecotype in parallel adaptive radiations. If so, then genetic differences between sea-type and lake-type ecotypes within a geographic region should be less than differences among regions within the lake-type ecotype, with the analysis conducted by comparisons of genetic differentiation (FST) among ecotypes and regions. Finally, we evaluate whether the Kokanee ecotype evolved independently from the lake-type Sockeye Salmon ecotype by comparing the level of differentiation within and between ecotypes from the same geographic region. If Kokanee populations in the Fraser River and Columbia River drainages share a common monophyletic origin, then differentiation between the ecotypes should be greater than differentiation among regions within an ecotype, with the evaluation conducted through gene diversity and cluster analysis.
Results
Variation within populations
Variation was observed in the number of alleles at the 14 microsatellite loci surveyed in the study. The fewest number of alleles was observed at Oki1a (4 alleles sea-type Sockeye Salmon ecotype, 8 alleles lake-type Sockeye Salmon and Kokanee ecotypes), and the greatest number of alleles was observed at Oki29 (23 alleles sea-type Sockeye Salmon ecotype, 36 alleles lake-type Sockeye Salmon ecotype, 37 alleles Kokanee ecotype). The number of alleles observed displayed considerable variation among the three ecotypes of O. nerka. After standardization to a common sample size, the sea-type Sockeye Salmon ecotype displayed considerably fewer alleles (203 alleles, P<0.05) across the 14 microsatellite loci than did either the lake-type Sockeye Salmon (241 alleles) or Kokanee (260 alleles) ecotypes (Table 1). The largest differences in number of alleles between sea-type Sockeye Salmon and Kokanee ecotypes were observed at Oki16 (14 alleles), Ots100 (8 alleles), Oki6 (6 alleles), and Oki29 (6 alleles). Average expected heterozygosity in the sea-type ecotype was 0.61 (observed 0.62), which was marginally lower than that of the lake-type ecotype (0.68, observed 0.67) and the Kokanee ecotype (0.68, observed 0.67).
Distribution of genetic variance
Gene diversity analysis of the 14 microsatellites surveyed was used to evaluate the distribution of genetic variation partitioned between the lake-type Sockeye Salmon and Kokanee ecotypes, among regions within ecotypes (10 regions Sockeye Salmon, 7 regions Kokanee, Table 2), among populations within regions (77 Sockeye Salmon populations, 42 Kokanee populations), and within populations. The amount of variation within populations ranged from 80.5% (Ots100) to 96.4% (Oki10), averaging 89.7%. Variation between the two ecotypes accounted for 2.04% of observed variation, which was not significant (P>0.05). However, variation among regions within ecotypes accounted for 5.36% of observed variation (P<0.01), and was the largest source of variation after within-population variation (Table 3). Differentiation between the two ecotypes was only 38% (2.04/5.36) of the magnitude of variation among regions within ecotypes. Variation among populations within regions was the next largest source of variation, and accounted for 2.93% of total observed variation. For populations in the Fraser River and Columbia River drainages, regional differences contributed more to differentiation of allele frequencies than ecotypes (lake-type Sockeye Salmon and Kokanee) or population sources of variation.
Population structure
Substantial allelic frequency differentiation was observed among all three ecotypes of O. nerka examined, with the largest average FST value observed between the sea-type Sockeye Salmon and Kokanee ecotypes (FST = 0.170), next between the sea-type and lake-type Sockeye Salmon ecotypes (FST = 0.140), and finally between the lake-type Sockeye Salmon and Kokanee ecotypes (FST = 0.115) (Table 4). FST values per locus over all 121 populations were: Oki10 0.031, Ots103 0.049, Oki1b 0.072, One8 o.074, Ots108 0.075, Oki1a 0.080, Oki29 0.083, Ots3 0.088, Ots2 0.098, Omy77 0.107, Ots107 0.104, Ots100 0.135, Oki6 0.141, and Oki16 0.172, with an overall FST value of 0.087. Higher allelic frequency differentiation was observed among regional groups of Kokanee populations (average FST = 0.132) compared with regional groups of lake-type Sockeye Salmon populations (average FST = 0.087). The largest differentiation among populations within an ecotype and region was observed between the sea-type Sockeye Salmon Harrison River and Widgeon Slough populations (FST = 0.176). Populations within ecotypes and geographically similar locations (the diagonal of Table 4) displayed less differentiation than ecotype and regional comparisons, with generally significant genetic differentiation among regional stock comparisons within ecotypes.
Population structure of the ecotypes was a function both of the ecotype and the region evaluated. For example, Kokanee populations from the Columbia River drainage (Okanagan Lake, Kootenay Lake), the South Thompson River drainage populations, and the mid-Fraser River populations all clustered together in the neighbor-joining tree (Fig 1). However, upstream of the mid-Fraser River populations, there were closer relationships between the lake-type Sockeye Salmon ecotype and the Kokanee ecotype, with the Nechako River Kokanee populations most similar to the Stuart River Sockeye Salmon populations. The sea-type Sockeye Salmon Harrison River population clustered with other lake-type Sockeye Salmon populations in the Harrison River drainage, and the sea-type Sockeye Salmon Widgeon Slough population in the lower Fraser River drainage was most similar to Kokanee populations in the lower Fraser River drainage.
Bootstrap values at major tree nodes indicate the percentages of 500 trees for which the populations beyond the node clustered together. Note the dendrogram proceeds vertically from one page to the next. Sockeye Salmon population names are in black, Kokanee population names are in blue. Harrison and Widgeon Slough are two sea-type Sockeye Salmon populations.
Discussion
With respect to the three main assumptions (1. sea-type ecotype ancestral with weakly differentiated populations; 2. genetically differentiated lake-type ecotype evolved repeatedly from the sea-type ecotype in parallel adaptive radiations; 3. Kokanee ecotype repeatedly evolved independently from the lake-type ecotype in a parallel adaptive radiation) of the evolution of O.nerka ecotypes outlined by [4], the results of our study can be summarized as follows. No evidence was found to support the hypothesis that the sea-type ecotype was comprised of weakly differentiated populations, nor was evidence available to suggest that the lake-type ecotype evolved repeatedly from the sea-type ecotype. Close genetic relationships between the lake-type ecotype and the Kokanee ecotype in the upper mid Fraser River drainage suggested that the lake-type ecotype could have been the ancestral form, but existing lake-type ecotypes may also have been derived from the Kokanee ecotype.
If the Kokanee ecotype is a result of parallel evolution, then pair-wise genetic distances between sympatric lake-type Sockeye Salmon and Kokanee should be less than among populations of the same ecotype in different lakes. However, since there can be contemporary gene flow between the ecotypes in some cases, it can be difficult to distinguish between parallel independent evolution of the two ecotypes versus a monophyletic origin of one ecotype with contemporary gene flow between ecotypes. Available genetic evidence indicates that there can be substantial differentiation between lake-type Sockeye Salmon and Kokanee within the same lake. Differentiation between the ecotypes in Takla Lake in the Fraser River drainage was substantially larger than variation among populations within ecotypes or among sampling years within populations [15]. Within one lake, the Kokanee ecotype was reported to be distinct from the lake-type ecotype, while in another lake, little differentiation was observed [16]. A survey of genetic variation between the lake-type and Kokanee ecotypes across a broad geographic range suggested parallel evolution between the lake-type and Kokanee ecotypes [17]. Significant genetic differentiation was observed between sympatric Sockeye Salmon and Kokanee in three separate localities [11]. The authors suggested that these two Kokanee populations may share a common monophyletic origin, relative to their sympatric Sockeye Salmon counterparts. They pointed out the close geographic proximity of the two river systems, < 20 km at some point, and suggested that stream capture may have occurred in recent geological time. They noted that, during the deglaciation of British Columbia, the Fraser and Columbia rivers were connected through a series of glacial lakes that formed in the Okanagan Valley [18]. Given the opportunities for gene exchange between these two river basins, gene flow may account for the genetic similarity between these Sockeye Salmon and Kokanee populations. In the current study, when one examines the population structure of the Kokanee ecotype as depicted in the dendrogram, the simplest explanation is that there was a monophyletic radiation of the ecotype in a portion of the Fraser River and Columbia River drainages, as shown by the clustering of populations in the Okanagan River, Kootenay River (Columbia River tributaries), and South Thompson River (Fraser River tributary). These results support those of [17], who suggested that Kokanee populations in these two river drainages may share a common monophyletic origin, relative to their sympatric Sockeye Salmon counterparts. However, when the entire Fraser River drainage is evaluated, the dendrogram suggests that there have been several independent evolutionary derivations of the Kokanee ecotype. Differentiation between the two ecotypes (2.04% of total observed variation) was only 38% of the magnitude of variation among regions within ecotypes (5.36% of total observed variation), suggestive of a polyphyletic origin of the ecotypes. Population structure within the lake-type ecotype is stable over time, as differentiation among river drainages and populations within river drainages has been reported to be approximately 19 times greater than that of annual variation within populations [13].
Has the Kokanee ecotype been derived repeatedly from the lake-type Sockeye Salmon ecotype? If so, then one may expect differences in genetic characters and possibly morphological characters between the ecotypes within a region to be less than differences among regions within an ecotype, particularly if there is contemporary gene flow between the ecotypes within a lake [19]. Differences in gill raker number between the lake-type and Kokanee ecotypes in Takla Lake in the Fraser River drainage have been examined [15, 20]. Both studies reported mean gill raker counts of 39.5–39.7 gillrakers for the Kokanee ecotype, and 36.2–36.5 gill rakers for the lake-type Sockeye Salmon ecotype, with the difference of three gill rakers the largest known to occur between sympatric ecotypes [10]. In a broad survey of variation in the number of gill rakers of Sockeye Salmon in North America, the mean number of gill rakers for Fraser River Sockeye Salmon was 36.5, with regional variation in the ecotype ranging from 34.4 (Adak Island) to 37.1 (Bristol Bay) [21]. Greater differences in gill raker number were reported between the sympatric ecotypes within Takla Lake than within the lake-type Sockeye Salmon ecotype over thousands of km in geographic distance. The distribution of gill raker phenotypes between the ecotypes did not support a finding of less differentiation between the ecotypes within a region than differences among regions within an ecotype. However, gill raker number may be subject to selection, and as such may not provide a reliable indicator on the plylogeny of the ecotypes.
There can be uncertainity in our study as to which ecotype was surveyed in a particular lake. For example, O. nerka from the Stave, Coquitlam, and Alouette lakes were defined as the Kokanee ecotype, largely because dams constructed in the watersheds blocked access for anadromous (lake-type) Sockeye Salmon to the lakes in the drainages. A dam downstream from Stave Lake was completed in 1912, with the Coquitlam Lake dam completed in 1914, and the Alouette Lake dam completed in 1927. The lake-type ecotype had been present in the watersheds prior to the construction of the dams, but once the dams were completed, access to the lakes was blocked, and the life cycle of O. nerka upstream from the dams was completed entirely in fresh water, hence the designation of the Kokanee ecotype in our study. However, experimental water releases past the dams on the Alouette and Coquitlam rivers in 2005 and 2006 resulted in juveniles passing the dams, which resulted in return migrations of anadromous adults in 2007 and 2008 after nearly 90 years of an entirely freshwater life cycle [22]. These three populations were among the most genetically atypical of either the lake-type Sockeye Salmon or Kokanee ecotypes surveyed (Fig 1). Designation of these populations as the lake-type Sockeye Salmon ecotype would not have altered the basic conclusions of the study.
Earlier surveys of allozyme variation in O. nerka indicated that there was little genetic differentiation in populations of the sea-type ecotype, even though populations were sampled across a broad geographic range of about 2,000 km [12]. If this were the consistent finding across studies, then indeed that this would suggest that there is a substantial amount of straying among populations of the sea-type ecotype, and this level of straying may lead to colonization of new habitats from which the lake-type and Kokanee ecotypes could evolve. However, lack of genetic differentiation among populations of the sea-type ecotype does not appear to be the general pattern of population structure in O. nerka [13, 23]. The sea-type ecotype is more common in northern rivers in glaciated regions where lake habitat is absent or non-productive [2]. In the Alsek River drainage in northern British Columbia, all sea/river type populations are distinct from all other populations in northern British Columbia [14], which does not support the concept of limited differentiation of populations of the sea-type ecotype across a broad geographic range. Furthermore, there was observed genetic differentiation among sea/river ecotype populations across British Columbia [13, 14], such that there is little support for a notion of limited differentiation among populations of this ecotype over a wide geographic area. In the current study, the pairwise FST value between the Harrison Rapids (River) and Widgeon Slough populations, both populations located in the lower Fraser River drainage, was 0.176, among the largest observed in the study. In British Columbia, population structure of O. nerka sea/river ecotype populations does not appear to support the notion of one large metapopulation, with relatively genetically undifferentiated populations.
Within a drainage, there are sometimes, but not always, genetic similarities between sea/river populations and lake-type populations. This association has been demonstrated for populations within the Alsek River drainage in northern British Columbia, where all 16 populations surveyed, including both ecotypes, clustered together in a dendrogram analysis of population structure [13]. This was also evident in the current study, where the sea/river type Harrison Rapids population clustered with all other lake-type populations in the Harrison River drainage. Within a drainage, there may be cases where a sea/river type population, located lower in the river drainage, is genetically similar to a dominant lake-type population higher in the drainage. This exact situation occurs in the Skeena River drainage in northern British Columbia. The Halliday Slough population, located below Babine Lake in the Skeena River and to which the population has no lake access, is similar genetically to the dominant Babine Lake population (unpublished data), which comprises 85% of Skeena River drainage escapement [24]. The 10 populations sampled in the Babine Lake complex are quite distinct genetically from the other 17 populations sampled in the drainage[24], but the Halliday Slough population is the only population ever surveyed in the drainage which displayed genetic similarity to the complex of Babine Lake populations. As there are no known fry migrants from the lake, it seems plausible that this population was initially founded from the lake-type ecotype that was migrating upstream to Babine Lake but simply ran out of energy reserves to complete migration, spawning in the available habitat in Halliday Slough, giving rise to the small present-day population. In this example, the sea-type ecotype was not the ancestral form, and lake-type Sockeye Salmon in Babine Lake have not evolved from the sea-type or river-type ecotype in parallel adaptive radiations. Rather, the reverse situation likely occurred, with the sea-type or river-type ecotype arising from the lake-type ecotype.
Kokanee populations in the Columbia River and South Thompson River may share a common monophyletic origin. With respect to the entire Fraser River drainage, if one excludes Kokanee populations in the lower Fraser River drainage, as these were lake-type ecotypes turned Kokanee ecotypes by dam construction [22] or otherwise altered by transplantation, and if one assumes a monophyletic origin of the Kokanee ecotype surveyed in the current study in the Columbia and South Thompson rivers, then the question arises as to why there are genetic similarities between lake-type Sockeye Salmon and Kokanee ecotypes in the middle portion of the Fraser River drainage (Stuart River lake-type and Nechako River Kokanee; Quesnel Lake lake-type and Kokanee) as noted in the current study. Geologic evidence from sites upstream of Texas Creek in the middle Fraser River drainage suggested that the Fraser River was dammed by avalanche debris and that one of these events occurred about 1200 yrs ago [25]. This may have prevented the upstream migration of salmon, since widespread collapse of First Nation culture occurred here at about that time [25]. If so, then the anadromous lake-type ecotype would have likely been exterminated from all lakes upstream of the postulated Texas Creek slide, and these lakes would have remained inaccessible until the presumed catastrophic destruction of the dam. It is possible that the closer relationship between lake-type Sockeye Salmon and Kokanee in the middle portion of the Fraser River drainage was a result of the existing Kokanee ecotype giving rise to the current lake-type Sockeye Salmon in this portion of the drainage, similar to the Kokanee ecotype in the dammed lakes in the lower portion of the Fraser River drainage giving rise to the newly developed lake-type anadromous Sockeye Salmon ecotype in the region.
In summary, the level of genetic differentiation observed among and within sea-type Sockeye Salmon, lake-type Sockeye Salmon, and Kokanee ecotypes suggested that there was little straying among populations within ecotypes, and limited introgression among ecotypes. Available evidence provides little support for the concept of the sea-type ecotype being the non-differentiated ancestral form of O. nerka owing to considerable genetic differentiation among populations within this ecotype. With ocean access blocked and later subsequently restored, the Kokanee ecotype can give rise to the lake-type Sockeye Salmon ecotype. Kokanee population structure within the Fraser River drainage suggested a polyphyletic origin of the ecotype within the drainage. The previous conclusions drawn from other studies about relationships among and within O. nerka ecotypes has been dependent upon the class of genetic variants employed (allozymes, minisatellies, and microsatellites) and the geographic scale of the survey of populations undertaken. Conclusions from studies employing older genetic technologies and restricted population sampling have not been confirmed by more recent surveys of variation within the ecotypes. Studies employing geographically restricted population sampling may not outline accurately the phylogenetic history of salmonid ecotypes, and larger-scale geographic sampling of populations is preferable in order to survey all variation that may be present within an ecotype.
Methods
Collection of DNA samples and laboratory analysis
Authorization to collect samples in the study was provided by a scientific license issued under the provisions of the Fisheries Act passed by the Canadian Parliament in 1985 and last amended in 2016. Under the Act, the scientific license was issued by Fisheries and Oceans Canada in order to allow Departmental staff to collect samples in the course of their work. As there is no requirement for an Institutional Animal Care and Use Committee (IACUC) or equivalent under the Act, sampling protocols were neither vetted nor approved by an IACUC. Sockeye Salmon and Kokanee are not an endangered or protected species in Canada. Fin clips or operculum punches were collected from recently dead or moribund adult fish on the spawning grounds in all populations surveyed in the study. Samples of Sockeye Salmon from the United States of America portion of the Columbia River drainage were provided by National Marine Fisheries Service (NMFS) staff of the National Oceanic and Atmospheric Administration following NMFS sampling protocols. DNA was extracted from the tissue samples as described by [26]. The study included a survey of microsatellite variation for over 26,000 fish from 121 populations in the Fraser River and Columbia River drainages (Fig 2). The specific populations, collection years, and sample sizes included in the survey are outlined in Table 3. PCR products at 14 microsatellite loci: Ots2, Ots3 [27], Ots100, Ots103, Ots107, and Ots108 [28, 29], Oki1a, Oki1b, Oki6, Oki10, Oki16, and Oki29 [30, 31], One8 [32], and Omy77 [33] were size fractionated on denaturing polyacrylamide gels and allele sizes initially determined with the ABI 377 automated DNA sequencer. Allele sizes were determined with Genescan 3.1 and Genotyper 2.5 software (PE Biosystems, Foster City, CA). Later in the study, microsatellites were size fractionated in an ABI 3730 capillary DNA sequencer, and genotypes were scored by GeneMapper software 3.0 (Applied Biosystems, Foster City, CA) using an internal lane sizing standard. Allele identification between the two sequencers was standardized by analyzing approximately 600 individuals on both platforms and converting the sizing in the gel-based data set to match that obtained from the capillary-based set.
The specific populations in each drainage are outlined in Table 3. Triangles on map indicate resident Kokanee populations, circles are anadromous Sockeye Salmon populations.
Data analysis
All annual samples available for a location were combined to estimate population allele frequencies, as was recommended by [34], as variation among Sockeye Salmon populations within drainages and among drainages was 19 times greater than variation among sampling years within populations [12]. The genotypic frequencies at each locus generally conformed to those expected under Hardy-Weinberg equilibrium [12]. FST estimates [35] for each locus over all populations were calculated with FSTAT version 2.9.3.2 [36]. The significance of the multilocus FST value over all samples was determined by jackknifing over loci. Cavalli-Sforza and Edwards chord distance (CSE) [37] was used to estimate genetic distances among all populations. An unrooted neighbor-joining tree based upon CSE was generated using NJPLOT [38]. Bootstrap support for the major nodes in the tree was evaluated with the CONSENSE program from PHYLIP based upon 500 replicate trees [39]. FSTAT was used to measure the ‘allelic richness’ (allelic diversity standardized to a sample size of 275 fish, all populations sampled within each ecotype combined) for each ecotype evaluated. Testing of significance was conducted by excluding populations with fewer than 20 individuals sampled, keeping populations separate within ecotypes, standardizing to a sample size of 27 fish, and employing a variance ratio (F-test) to test for differences among ecotypes. Computation of the number of alleles observed per locus was carried out with FSTAT. The distribution of genetic variation in lake-type Sockeye Salmon and Kokanee ecotypes was evaluated between ecotypes, among regions within ecotypes, and among populations within regions. Estimation of variance components of ecotype differentiation, among regions within ecotypes, and among populations within regions was determined with GDA [40]. A variance ratio was used to test significance of the different hierarchical levels. Allele frequencies for all populations surveyed in the study are available via DRYAD doi identified as: data package title: Data from: Population structure of sea-type and lake-type Sockeye Salmon and Kokanee in the Fraser River and Columbia River drainages. Provisional DOI: doi:10.5061/dryad.3g824 Data files: Baseline Allele Frequencies.
Acknowledgments
A very substantial effort was undertaken to obtain samples from Sockeye Salmon and Kokanee sampled in this study. In the Fraser River, we thank various staff of the Fisheries and Oceans Canada (DFO) for baseline sample collection, as well as First Nations staff. Samples of Kokanee were obtained by staff of DFO, as well as by staff of the British Columbia Ministry of Environment (MOE) and Ministry of Forests, Lands, and Natural Resource Operations (MFLNRO). L. Fitzpatrick drafted the map. C. Wallace developed the dendrogram. B. McIntosh and C. MacConnachie conducted the laboratory analysis of microsatellite variation.
References
- 1.
Burgner RL. Life history of sockeye salmon Oncorhynchus nerka. In: Groot C. and Margolis L. (editors.) Pacific salmon life histories. Vancouver, British Columbia. University of British Columbia Press. 1991. pp. 3–117.
- 2.
Wood CC, Riddell BE, Rutherford DT. Alternative juvenile life histories of sockeye salmon (Oncorhynchus nerka) and their contribution to production in the Stikine River, northern British Columbia. In: Smith, H. D., Margolis, L., and Wood, C.C. (editors) Sockeye salmon (Oncorhynchus nerka) population biology and future management. Can. Spec. Pub. Fish. Aquat. Sci. 1987;96: 12–24.
- 3. Wood CC. Life history variation and population structure in sockeye salmon. Am. Fish. Soc. Symp. 1995;17: 195–216.
- 4. Wood CC, Bickham JW, Nelson RJ, Foote CJ, Patton JC. Recurrent evolution of life history ecotypes in sockeye salmon: implications for conservation and future evolution. Evol. Appl. 2008;1:207–221. pmid:25567627
- 5. Nelson JS. Distribution and nomenclature of North American kokanee, Oncorhynchus nerka. J. Fish. Res. Board Can. 1968; 25:409–414.
- 6. Frazer KK, Russello MA. Lack of parallel genetic patterns underlying the repeated ecological divergence of beach and stream-spawning kokanee salmon. J. Evol. Biol. 2013; 26:2606–2621. pmid:24118176
- 7. Lemay MA, Russello MA. Genetic evidence for ecological divergence in kokanee salmon. Mol. Ecol. 2015; 24:798–811. pmid:25580953
- 8. Larson WA, Limborg MT, McKinney GJ, Schindler DE, Seeb JE, Seeb LW. Genomic islands of divergence linked to ecotypic variation in sockeye salmon. Mol. Ecol. 2017; 26:554–570. pmid:27864910
- 9. Veale AJ, Russello MA. An ancient selective sweep linked to reproductive life history evolution in sockeye salmon. 2017 Nature Sci. Rep. 7:1747.
- 10. Dodson JJ, Aubin-Horth N, Theriault V, Paez DJ. The evolutionary ecology of alternative migratory tactics in salmonid fishes. Biol. Rev. Cam. Phil. Soc. 2013; 88, 602–625.
- 11. Winans GA, Aebersold PB, Waples RS. Allozyme variability of Oncorhynchus nerka in the Pacific Northwest, with special consideration to populations of Redfish Lake, Idaho. Trans. Am. Fish. Soc. 1996; 125: 645–663.
- 12. Gustafson RG, Winans GA. Distribution and population genetic structure of river- and sea-type sockeye salmon in western North America. Ecol. Freshwater Fish 1999;8: 181–193.
- 13. Beacham TD, McIntosh B, MacConnachie C, Miller KM, Withler RE, and Varnavskaya NV. Pacific Rim population structure of sockeye salmon as determined from microsatellite analysis. Trans. Am. Fish. Soc. 2006;135: 174–187.
- 14. Beacham TD, McIntosh B, MacConnachie C. Population structure of lake-type and river-type sockeye salmon in transboundary rivers of northern British Columbia, Canada. J. Fish Biol. 2004;65: 389–402.
- 15. Wood CC, Foote CJ. Evidence for sympatric genetic divergence of anadromous and nonanadromous morphs of sockeye salmon (Oncorhynchus nerka). Evolution 1996;50:1265–1279. pmid:28565300
- 16. Taylor EB, Foote CJ, Wood CC. Molecular genetic evidence for parallel life-history evolution within a Pacific salmon (sockeye salmon and kokanee, Oncorhynchus nerka). Evolution 1996;50:401–416. pmid:28568856
- 17. Nichols KM, Kozfkay CC, Narum SR. Genomic signatures among Oncorhynchus nerka ecotypes to inform conservation and management of endangered Sockeye Salmon. Evolutionary Applications 2016;9:1285–1300. pmid:27877206
- 18.
McPhail JD, Lindsey CC. Zoogeography of the freshwater fishes of Cascadia (the Columbia system and rivers north to the Stikine). Hocutt i/i C. H. and Wiley E. O., editors. The zoogeography of North American freshwater fishes. John Wiley and Sons, New York. 1986. pp 615–637.
- 19. Veale AJ, Russello MA. Sockeye salmon repatriation leads to population re-establishment and rapid introgression with native kokanee. Evol. Appl. 2016;9: 1301–1311. pmid:27877207
- 20. Foote CJ, Moore K, Stenberg K, Craig KJ, Wenburg JK, Wood CC. Genetic differentiation in gill raker number and length in sympatric anadromous and nonanadromous morphs of sockeye salmon, Oncorhynchus nerka. Envir. Biol. Fish. 1999;54: 263–274.
- 21. Beacham TD. Variation in number of vertebrae and gill rakers of sockeye salmon (Oncorhynchus nerka) in North America. Env. Biol. Fish. 1985; 14: 97–105.
- 22. Godbout L, Wood CC, Withler RE, Latham S, Nelson RJ, Wetzel L, Barnett-Johnson R, Grove MJ, Schmitt AK, McKeegan KD. Sockeye salmon (Oncorhynchus nerka) return after an absence of nearly 90 years: a case of reversion to anadromy. Can. J. Fish. Aquat. Sci. 2011; 68: 1590–1602.
- 23. Beacham TD, Lapointe M, Candy JR, McIntosh B, MacConnachie C, Tabata A, Kaukinen K, Deng L, Miller KM, Withler RE. Stock identification of Fraser River sockeye salmon (Oncorhynchus nerka) using microsatellites and major histocompatibility complex variation. Trans. Am. Fish. Soc. 2004;133: 1106–1126.
- 24. Beacham TD, Cox-Rogers S, MacConnachie C, McIntosh B, Wallace CG. Population structure and run timing of sockeye salmon in the Skeena River, British Columbia. N. Amer. J. Fish. Manage. 2014;34: 335–348.
- 25. Ryder JM, Bovis MJ, Church M. Rock avalanches at Texas Creek, British Columbia. Can. J. Earth Sci. 1990;27: 1316–1329.
- 26. Withler RE, Le KD, Nelson RJ, Miller KM, Beacham TD. Intact genetic structure and high levels of genetic diversity in bottlenecked sockeye salmon, Oncorhynchus nerka, populations of the Fraser River, British Columbia, Canada. Can. J. Fish. Aquat. Sci. 2000.;57: 1985–1998.
- 27. Banks MA, Blouin MS, Baldwin BA, Rashbrook VK, Fitzgerald HA, Blankenship SM, Hedgecock D. Isolation and inheritance of novel microsatellites in chinook salmon (Oncorhynchus tshawytscha). J. Hered. 1999;90: 281–288.
- 28. Beacham TD, Margolis L, Nelson RJ. A comparison of methods of stock identification for sockeye salmon (Oncorhynchus nerka) in Barkley Sound, British Columbia. N. Pac. Anad. Fish Comm. Bull. 1998;1: 227–239.
- 29. Nelson RJ, Beacham TD. Isolation and cross species amplification of microsatellite loci useful for study of Pacific salmon. Animal Genet. 1999;30: 228–229. pmid:10442992
- 30. Smith CT, Koop BF, Nelson RJ. Isolation and characterization of coho salmon (Oncorhynchus kisutch) microsatellites and their use in other salmonids. Mol. Ecol. 1998;7: 1613–1621.
- 31. Nelson RJ, Wood CC, Cooper G, Smith C, Koop B. Population structure of sockeye salmon of the central coast of British Colulmbia: Implications for recovery planning. N. Amer. J. Fish. Manage. 2003;23: 703–720.
- 32. Scribner KT, Gust JR, Fields RL. Isolation and characterization of novel salmon microsatellite loci: cross-species amplification and population genetic applications. Can. J. Fish. Aquat. Sci. 1996;53: 833–841.
- 33. Morris DB, Richard KR,Wright JM. Microsatellites from rainbow trout (Oncorhynchus mykiss) and their use for genetic study of salmonids. Can. J.Fish. Aquat. Sci. 1996;53: 120–126.
- 34. Waples RS. Temporal changes of allele frequency in Pacific salmon populations: implications for mixed-stock fishery analysis. Can. J. Fish. Aquat. Sci. 1990;47: 968–976.
- 35. Weir BS, Cockerham CC. Estimating F-statistics for the analysis of population structure. Evolution 1984;38: 1358–1370. pmid:28563791
- 36. Goudet J. FSTAT A program for IBM PC compatibles to calculate Weir and Cockerham’s (1984) estimators of F-statistics (version 1.2). J. Heredity 1995;86: 485–486.
- 37. Cavalli-Sforza LL, Edwards AWF. Phylogenetic analysis: models and estimation procedures. Am. J. Human Genet. 1967;19: 233–257.
- 38. Perriere G, Gouy M. WWW-Query: an on-line retrieval system for biological sequence banks. Biochimie 1996;78:364–369. pmid:8905155
- 39.
Felsenstein J. PHYLIP: Phylogeny Inference Package. University of Washington, Seattle. 1993.
- 40.
Lewis PO, Zaykin D. Genetic Data Analysis: Computer program for the analysis of allelic data. Version 1.0 (d16c). Free program distributed by the authors over the internet from http://lewis.eeb.uconn.edu/lewishome/software.html. 2001.