Circumpolar genetic population structure of polar cod, Boreogadus saida

Abstract

Polar cod, Boreogadus saida, is an important part of Arctic and boreal marine ecosystems. Knowledge of polar cod population genetic structuring can provide insight into how the species may respond to environmental change, and allow for establishment of meaningful management units. To examine population genetic structure of B. saida, we analysed nine microsatellite DNA loci in 2269 fish collected at 19 locations across the species’ range. Genetic differentiation was detected (FST = 0.01, p < 0.01), which had concordance with geography. A Canada East group consisted of fish collected from Resolute Bay to the Gulf of St. Lawrence. Three additional groups were observed. These consisted of a Canada West group containing specimens from the Canadian Beaufort Sea and Amundsen Gulf, a Europe group containing specimens from the Greenland Sea, Iceland and the Laptev Sea, and a US group consisting of specimens collected in the North Bering, Chukchi and Western Beaufort seas. Very little genetic differentiation was detected within the identified groups. Physical distance, geophysical structure and oceanography all appeared to have the potential to influence levels of genetic divergence. The population genetic structure detected is likely to be important for the response of B. saida to environmental change, and should be considered in management of human activities that may impact this foundational species.

Introduction

Polar cod, Boreogadus saida, is a small, widespread and abundant gadid fish that occupies a central role in Arctic ecosystems as a key prey item for marine mammals, sea birds and other fishes (Bradstreet et al. 1986; Welch et al. 1992; Loseto et al. 2009; Kuletz et al. 2015). B. saida occurs in the waters of the Arctic Ocean and surrounding seas. In the Pacific Ocean, B. saida is found as far south as Norton Sound (Western Alaska) and the Sea of Okhotsk, and in the Atlantic Ocean, as far south as the Gulf of St. Lawrence (Canada). The species occupies a wide range of habitats including coastal shelves, semi-enclosed brackish bays, deep basins, and under landfast and pack ice (Drolet et al. 1991; David et al. 2016; De Robertis et al. 2016). In general, B. saida is believed to be iteroparous (Hop et al. 1995, Nahrgang et al. 2014) with a maximum life span of 7 to 8 years, reaching sexual maturity in 2 to 3 years (Hop and Gjøsæter 2013). Nahrgang et al. (2014) suggest that males may reach maturity earlier and have shorter life spans. Broadcast spawning occurs in late winter/early spring under the ice with females producing positively buoyant eggs (Spencer et al. 2019) that drift from two to four months before hatching (Graham and Hop 1995). Polar cod less than one year in age feed mainly on the nauplii and copepodite stages of calanoid copepods, with a strong preference for Calanus glacialis (Michaud et al. 1996; Falardeau et al. 2014; Bouchard et al. 2016; Bouchard and Fortier 2020). In the fall, most young of the year individuals are thought to transition to deeper water as they age (Geoffroy et al. 2011; Majewski et al. 2015) although part of them may recruit to pack ice (Gradinger and Bluhm 2004; Melnikov and Chernova 2013; David et al. 2016).

The ecology and physiology of polar cod has been a topic of increased research focus given its crucial ecological role, and the potential for climate change-driven effects on its abundance and distribution. Recent studies tracking individual fish movements are starting to provide insights into the distances individual fish travel. Acoustically tagged individual B. saida were found to travel 192 km in 67 to 215 days at an average speed of 1.4 km/day (Kessel et al. 2015). The underlying biological driver of such movements is not known, but it would have clear implications for population structure if associated with spawning. A striking aspect of the biology of the species is large dense mobile schools that form in surface waters during the ice break up and the summer (Crawford and Jorgenson 1993). The purpose of this schooling behaviour is not known, but again, it would have obvious implications for genetic structure if contributing the mixed reproduction involving different genetic populations.

Increasing temperatures of Arctic waters will put pressure on cold-adapted B. saida populations (Bouchard et al. 2017). The species appears to have a low degree of adaptability to temperature increase (Drost et al. 2016, Laurel et al. 2018) and the way in which B. saida responds to increasing temperatures will have attendant effects on other species. Recent increases in sea surface temperature and decrease in sea ice have reduced their availability to upper trophic level predators (Divoky et al 2015). Northward range shifts of populations found in the south could result in replacement of northern populations, or provide an influx of genes and gene complexes that allow for adaptation to warmer temperatures and southern ice dynamics. Both of these processes could prevent potential local extirpation of B. saida in northern areas that are affected by climate driven changes. Molecular genetic surveys can provide information regarding south-north geneflow and the potential for genetically unique, locally adapted southern populations. Once genetic populations and their boundaries are identified, a range of assessment and management options more explicitly focused upon relevant biological units become possible including monitoring of populations that make transboundary migrations. Molecular population genetic tools also allow for the assessment of the genetic diversity within populations and the identification of “evolutionary significant units” for management of long-term viability of the species.

An early study using random amplified polymorphic DNA markers on B. saida from the North Atlantic Ocean revealed no differentiation at the population level, leading to the conclusion that the species could be panmictic (Fevolden et al. 1999). More recently, analysis of the mitochondrial DNA of B. saida from waters around Greenland identified two lineages but no strong population differentiation (Pálsson et al. 2009). In the Greenland Sea B. saida, microsatellite DNA analysis detected weak genetic differentiation between fish from different fjords, and between populations living inside and outside fjords (Madsen et al. 2016). Using mitochondrial and microsatellite DNA analysis, Wilson et al. (2019) found general panmixis from the Western Beaufort Sea to the Eastern Chukchi Sea, but with suggestion of small scale geographic partitioning in the area immediately west of the Mackenzie River. Wilson et al. (in review) took a mitochondrial genome approach to study polar cod from the Bering, Chukchi and Beaufort seas and found no population structuring but found higher genetic diversity in polar cod in the northern samples. Taken together the studies above suggest that B. saida, like other pelagic marine fish including Atlantic cod, Gadus morhua (O'Leary et al. 2007) and Mallotus villosus (Præbel et al. 2008; Kenchington et al. 2015) show genetic differentiation over very large spatial scales, with the potential for differentiation on local geographic scales. Microsatellite DNA markers have the potential to reveal genetic heterogeneity at both regional and global scales. Here we characterize genetic heterogeneity of B. saida with microsatellite DNA analysis over a larger geographical range than previous studies and, where possible, examine regional scale differentiation.

Materials and methods

Sample collection

Gadids visually identified as B. saida were collected by ten institutions in 19 locations across most of the range of the species (Fig. 1, Table 1). One sample was collected from foraging Mandt’s Black Guillemot (Cepphus grylle mandtii) at nesting sites (no. 7), while other collections were made from research vessels employing trawl nets. In some instances, samples were collected relatively close together. Annual temporal replicates were collected in seven locations. Typically, a fin clip was cut from a freshly caught specimen and placed in 95% ethanol with the volume of the tissue not exceeding 25% of the final volume.

Fig. 1
figure1

Map showing locations of sample collection as well as prevailing surface currents. Dots placed on the map show the location of collection with colour-coding reflecting groupings made by discriminant analysis of principal components. Depiction of cold (purple) and warm (red) surface currents is from AMAP (1998) and Macdonald and Bewers (1996)

Table 1 Location and year of collection of Boreogadus saida specimens analysed after removal of Arctogadus glacialis

Genotyping

A total of 2587 gadid specimens were genotyped at ten microsatellite loci: Gmo8, Gmo32, Gmo34, Gmo127, Bsa6, Bsa7, Bsa14, Bsa15, Bsa60, and Bsa101 (Nelson et al. 2013). For genotyping, DNA was extracted using a chelex-100-based method (Nelson et al. 1998) and then amplified in polymerase chain reactions (PCR) following loci-specific methods detailed in Nelson et al. (2013). The size (in base pairs) of PCR products were determined with an Applied Biosystems 3730 sequencer with ROX1000 as a size standard. Alleles were scored visually using GENEMAPPER software (Applied Biosystems, Foster City, CA). Allele frequencies, observed and expected heterozygosity, and FIS for each sample by locus are found in the supplementary material.

Removal of Arctogadus glacialis

Discriminating between B. saida and Arctogadus glacialis based on morphological features can be difficult especially for individuals less than one year in age. To identify A. glacialis among the B. saida samples, we first employed a test using the microsatellite marker Gmo8 that has been shown to differentiate between these species with high accuracy (Madsen et al. 2009). However, comparing the Gmo8 test to sequencing of the cytochrome b (cytb) mitochondrial gene (Bouchard et al. 2013) indicated that the Gmo8 test is not 100% reliable outside the geographical range for which it was developed. Thus we used a second test to screen out A. glacialis. Under the PCR conditions of the study, we found that the locus Bsa15 amplified in B. saida but not in A. glacialis. This was verified by sequencing cytb (GenBank accession numbers MT345688-MT345757) as per Bouchard et al. (2013). Specimens that did not yield a PCR product for locus Bsa15 were considered to be A. glacialis and set aside for the present analysis. Table 1 gives the numbers of B. saida specimens identified by the above criteria, carried forward in the analysis.

Data analysis

The “exact probability” test as found in Guo and Thompson (1992) was implemented by the software package GENEPOP v4.7.2 using default values (Rousset 2008), to test for adherence to Hardy–Weinberg equilibrium (HWE) in each sample with the value of α adjusted for the number of loci examined (0.05/10 = 0.005) after Rice (1989). During the course of these tests, it was noticed that locus Gmo32 deviated from HWE expectations in multiple samples, and this locus was removed from further analyses. GENEPOP v4.7.2 was also used to perform pairwise differentiation tests as per Raymond and Rousset (1995) to determine if samples collected in different years (temporal replicates) or close together in same location could be pooled. The α for this test (rejection of the null hypothesis of genetic homogeneity) was 0.05/9 because Gmo32 was removed from the analysis. Pooled samples were then again subject to the “exact probability” test” of GENEPOP v4.7.2 to test whether pooled samples (populations) conformed to HWE expectations.

Pairwise FST analysis between all populations was carried out (Table 3) with GENETIX v4.5.2 (Belkhir et al. 2004) as per Weir and Cockerham (1984) with 100 bootstrap resamplings over loci to derive p values. GENETIX v4.5.2 was also used to calculate locus-by-locus observed and expected heterozygosity, and FIS according to Robertson and Hill (1984). To examine the relationship among the different samples, a Discriminant Analysis of Principal Components (DAPC) plot was produced with adegenet v 2.1.1 (Jombart 2008) in the R data analysis environment (R Core Team 2013).

The ARLEQUIN v 3.5 software package (Schneider et al. 2010) was used to calculate AMOVA as described in Weir (1996), using the four groups as designated by DAPC analysis (Hudson Strait held separate) with 1000 permutations of the data used to generate pairwise distances. ARLEQUIN v3.5 was also used to produce summary statistics, gene diversity Nei (1987), mean number of alleles, and FIS (Robertson and Hill 1984) for each sample across loci (Table 2). Examination of the data set for loci potentially under selection was done with BayeScan 2.1 first described in Foll and Gaggiotti (2008). A Mantel test was performed between matrices of geographical distance and pairwise FST (Table 3) between all stations with adegenet v 2.1.1 (Jombart 2008) in the R data analysis environment (R Core Team 2013) with 9999 replicates. Distance in kilometers between all sampling locations was accomplished by plotting the locations in Google Earth and manually measuring the shortest ocean distance between the two points.

Table 2 Summary statistics for each population across loci
Table 3 Pairwise FST between Boreogadus saida samples

Results

Final data set: removal of Arctogadus glacialis, removal of locus in Hardy–Weinberg disequilibrium, sample pooling, and basic indices

A total of 300 individuals were identified as A. glacialis by genetic tests and removed from subsequent analyses, leaving a total of 2287 specimens identified as B. saida in the data set. The null hypothesis of genetic homogeneity (pairwise differentiation test with an adjusted α of 0.05/9 = 0.0056 accounting for the removal of Gmo32) could not be rejected for any temporal or geographical replicates; thus all geographically proximate and temporal replicate samples were pooled. HWE was again tested on the pooled sample and none of the pooled samples were observed to deviate from HWE (α = 0.0056). This left a total of 19 sample sets judged to represent the areas from which they were collected. Allele frequencies as well as observed and expected heterozygosity and FIS for each locus per sample are given in the supplementary materials. Table 2 shows summary statistics across loci for each sample. Gene diversity ranged from 0.39 ± 0.22 (n = 9) in the North Bering Sea sample to 0.54 ± 0.30 (n = 9) in the Labrador Sea sample and mean gene diversity was 0.50 ± 0.02 (n = 19).

The average number of alleles across all loci and samples was 7.4 ± 2.8 (n = 171) and was the highest in the Amundsen Gulf sample at 11.3 ± 3.6 (n = 9) and the lowest in the sample from Hudson Strait at 5.3 ± 2.1 (n = 9). In outlier analysis, locus Bsa60 gave an FST of 0.03, while for the remainder of the loci it ranged from 0.001 to 0.009.

Population structure

The circumpolar population structure of B. saida corresponded relatively well with geography as examined via DAPC analysis which suggests four sub-groups (Fig. 2). The largest group contained nine populations, all of them in eastern Canada: Resolute Bay, Lancaster Sound, Oliver Sound, Baffin Bay, Frobisher Bay, Labrador Sea, Bonavista Bay, Trinity Bay, and the Gulf of St. Lawrence (FigS. 1, 2). The sample collected in Hudson Strait is geographically within the Canada East area, but could not be placed definitively within any group by DAPC analysis. The balance of the samples consisted of three groups: a US group comprised samples collected in the Northern Bering, Chukchi and Western Beaufort seas; a Europe group made up of specimens from Greenland, Iceland and the Laptev Sea; and a Canada West group made up of specimens collected in the Canadian Beaufort Sea (near Herschel Island and Tuktoyaktuk), and from Amundsen Gulf. Across the entire sample set, there was a correlation between geographical distance and FST as judged by Mantel test (p = 0.0006).

Fig. 2
figure2

Discriminant Analysis of Principal Components plot. The plot shows the genetic relationship among Boreogadus saida samples. Each sample is represented by a single dot with those grouping together colour coded similarly

Between samples within each of the Canada East, US, Europe, and Canada West groups, only two instances of intra-group differentiation as judged by pairwise FST were detected (Table 3). These were within the Canada East group, between the specimens collected in both Bonavista Bay and the Gulf of St. Lawrence, both compared to Baffin Bay. As judged by FST analysis, the sample from Hudson Strait showed a pattern of differentiation, similar to that of the Canada East group members. There were multiple instances of statistically significant FST values between samples from different groups. Out of 155 pairwise FST values between samples in the four different groups (omitting the Hudson Strait sample), all but nine were statistically significant (p < 0.05) with FST ranging from 0.002 to 0.045. Consistent with the results of pairwise genetic differentiation tests between samples from different DAPC-defined groups, AMOVA analysed for the US, Europe, Canada East, and Canada West groups gave an overall FST of 0.014 (p < 0.001). In this analysis the balance of the genetic variation was within populations (p < 0.001).

Pairwise FST values among the six samples representing the North Bering-Chukchi-Beaufort Sea-Amundsen Gulf corridor indicated population differentiation within this region (Table 3). For all three samples collected in US waters, FST values with samples from Amundsen Gulf (Canada) were significant. For the samples from the North Bering and Chukchi seas (US), FST values were significant with the Tuktoyaktuk (Canada) sample, whereas the Herschel Island (Canada) sample was differentiated only from the Chukchi Sea sample.

Discussion

Boreogadus saida is not a panmictic species: the role of currents, sea ice and bathymetry

This study examined B. saida genetic population structure over an area larger than previously examined and provides an overarching framework to explore species-wide evolutionary history and population connectivity. Genetic differentiation among populations of B. saida was detected across the Arctic, the North Pacific and the North Atlantic. The level of differentiation observed was at levels expected for marine fishes, such as Atlantic cod Gadus morhua (Ruzzante et al. 1998, 1999; Beacham et al. 2002; Rose et al. 2011, Andre et al. 2016) and capelin Mallotus villosus (Præbel et al. 2008; Kenchington et al. 2015). Passive dispersal during early life stages and migrations during the adult stage, likely contribute to significant gene flow and reduce genetic structuring in polar cod. Although indirect evidence from otolith chemistry suggest polar cod eggs and larvae can drift long distances (Bouchard et al. 2015), accurate dispersal patterns still need to be described. Tagged individuals were found 192 km away from their release location (Kessel et al. 2017), but the patterns and nature (active or passive) of migration along horizontal scales remain mostly unknown. Recent description of site fidelity in the species is noteworthy (Madsen et al. 2016, Kessel et al. 2017) and may underpin population genetic structure in polar cod. Detailed understanding of larval dispersal, adult migration, and habitat use at all life stages would help in assessing the importance of different environmental factors on the population structure.

Our results indicate the existence of four groups that are concordant with geography. A significant phylogenetic break delimited a Canada East group from the US, Europe and Canada West groups. The Canada East group consisted of samples collected from Resolute Bay in the central Canadian Archipelago to the Gulf of St. Lawrence, a distance of over 4000 km. Within this geographically large group, very little genetic differentiation was detected and among 45 pairwise FST, only two were statistically significant. This apparent lack of genetic heterogeneity seems surprising given the large expanse of habitats that span 26 degrees of latitude, with associated differences in photoperiod, and oceanography. It is unlikely that individual fish migrate between these two locations. It seems more likely that a lack of barriers to migration allows for a series of reproductively connected and genetically homogeneous populations along the coast of eastern Canada and into the Canadian Arctic Archipelago. The prevailing North to South coastal current running along the west coast of Baffin Bay and down the east coast of Canada (LeBlond 1980; Fissel et al. 1982; Tang et al. 2004) may promote dispersal from the north. If north to south, rather than south to north, geneflow is the predominant homogenizing factor, “rescue” of northern populations by southern populations may not be likely.

Across the entire data set, a relationship between FST and geographical distance was observed; thus it appears that isolation by distance is a factor in genetic divergence. But it is not possible to use distance as the sole predictive criteria. Within the Canada East group, we detected little genetic heterogeneity over a relatively large expanse of 4000 km. In contrast, between the US and Canada West groups, we saw differentiation over a scale of less than 1000 km. This suggests that although isolation by distance plays a role in the genetic structure of polar cod, geography and oceanography can also have a significant role.

Prevailing currents flow southward along the east coast of Greenland around the southern cape and then northward along Western Greenland into Baffin Bay (LeBlond 1980; Fissel et al. 1982; Tang et al., 2004; Münchow et al. 2015; also see Fig. 1). This prevailing current system could potentially act to connect the polar cod populations encompassed by the Europe group to those in the Canada East group. However, this is not observed here. Based on the observations made in this study, and the idea that prevailing currents serve to promote genetic connectivity, we can speculate that along the route taken by the prevailing currents from east to west Greenland there exists a barrier to migration of adults or passive transport of eggs and larvae. Future genetic analysis of specimens collected along this pathway could identify the barrier to geneflow between the Canada East group and the Europe group that we detect here.

The Europe group includes specimens collected across a wide geographic range, stretching across approximately 4000 km over which we have relatively sparse sampling coverage. Possibly due to the low spatial coverage, our results do not suggest further genetic subdivision within this group, and similar to the Canada East group, the Europe group shows genetic homogeneity over a large spatial scale. More sample coverage is needed to reach any definitive conclusions regarding genetic heterogeneity across this area.

The US group stretches from the Northern Bering Sea to the Western Beaufort Sea, a distance of approximately 1000 km. Across this region the prevailing northward current (Coachman et al.1975) transports a large quantity of phytoplankton (Springer & McRoy 1993) and zooplankton (Springer et al. 1989) into the Chukchi Sea. A portion of this water heads eastward to the Beaufort Sea (Pickart et al. 2005) to where the US Beaufort specimens were collected. This dominant oceanographic feature may be facilitating south to north gene flow via transport of polar cod eggs and larvae, as well as facilitating northward migration of adults and juveniles.

A genetic break was detected in the Western Arctic region between the US group and the Canada West group. Two of the Canadian samples, Tuktoyaktuk and Amundsen Gulf, were collected east of the Mackenzie River with the third, Herschel Island, located 140 km to the west of the Mackenzie River and 780 km away from the nearest US Beaufort Sea collection site. The sample from Herschel Island appears to be more closely related to the members of the US group as it has a statistically significant FST with just one of the three samples of the US group, whereas the Tuktoyaktuk sample is differentiated from two of the three US samples, and the Amundsen Gulf sample (furthest east) is differentiated from all three. The plume of the Mackenzie River may be reducing passive dispersal of buoyant polar cod eggs and larvae thus contributing to the genetic divergence of populations from Herschel Island, Tuktoyaktuk, and Amundsen Gulf from those in the US group. Herschel Island being west of the mouth of the Mackenzie River may be less isolated from the populations to the west by this mechanism.

The observed phylogenetic isolation between the Canada West group and the Canada East group suggests that the Canadian Arctic Archipelago (CAA) may represent a geographical barrier for polar cod as it appears to be for seabirds (Harkness 2017) and historically for bowhead whale (Balaena mysticetus) (Dyke et al. 1996). Most likely, heavy sea ice conditions in M’Clure Strait/Viscount Melville Sound compromise larval survival in this area, acting as a barrier to dispersal in the northern part of the CAA (Bouchard et al. 2018). In the southern CAA, Queen Maud Gulf and Victoria Strait may also present a barrier due to a combination of shallow waters, cold temperatures, slow circulation and scarcity of lipid-rich Calanus copepods that forms suboptimal habitat for all life stages of polar cod (Bouchard et al. 2018). It is not clear why the sample from Hudson Strait did not group with the rest of the geographically bounded Canada East group; it is the smallest sample carried forward for analysis, and it is possible that the individuals collected do not accurately reflect the population from which it was drawn. It is also possible this sample was drawn from a genetically unique population. This issue will remain unresolved until more specimens from this location can be examined, or a more powerful marker set can be applied.

Previous work on the genetics of polar cod have been carried out over smaller spatial scale and in some cases with different markers types, and are not directly comparable to the present study (Fevolden et al. 1999, Pálsson et al. 2009, Madsen et al. 2016). Regardless, these studies are not in conflict in with the results described here. Madsen et al. (2016) detects genetic differentiation over a spatial scale similar to that which encompasses the US and Canada West groups supporting the interpretation that genetic differentiation of the species is present over regional spatial scales. Wilson et al. (in review) studied polar cod from the Bering, Chukchi and Beaufort seas with a mitochondrial genome approach and found no population structuring but found higher genetic diversity in polar cod from the northern samples. This is consistent with our findings of the lowest gene diversity in polar cod collected in the Bering Sea and would be consistent with Lewontin’s (1974) hypothesis that genetic diversity decreases at the periphery of a species’ range. Wilson et al. (in review) examined only one sample east of the Mackenzie River which may be why they did not observe genetic differentiation as we did over a similar spatial scale. Alternatively, this could be due to a difference in the genetic markers used (mitochondrial vs microsatellite); at this point, we cannot rule out either of these possibilities.

It is also important to keep in mind the limitations of the microsatellite data set used here. FST outlier analysis, pointed to one locus Bsa60, as being potentially under selection although this remains inconclusive due to the low number of loci (Beaumont and Nichols 1996, Excoffier et al. 2009) that limits the power of this test. This should be kept in mind when exploring the potential for examination of the number of migrants (Nem) which assumes genetic neutrality of the genetic markers used. It is also possible that adaptation could be driven and maintained by selection for traits that confer a local selective advantage, even in face of gene flow that homogenizes neutral genetic variation (Fitzpatrick et al. 2014). Studies with greater numbers of markers and deeper genomic coverage are needed to explore the possibility of functionally distinct populations that are undetectable by examining the marker set used in this study. Another consideration here is sample size, which ranged from 26 to 252. Hale et al. (2012) argue based on simulations that 25 to 30 individuals are sufficient to characterize allele frequencies of microsatellite alleles, and thus all samples in our data set are within this guideline. Regardless of potential limitations of our data set, we detect a consistent, non-random distribution of genetic heterogeneity concordant with geography. Examination of the genetic difference between temporal replicates is proposed by Waples (1989), as a way to judge the relevance of geographical population structure. Further bolstering the population structuring we observe here, we found no difference between temporal replicates collected in the same location.

Toward a more complete understanding of polar cod genetics

Future studies on the population structure of B. saida would benefit from including samples collected in areas not covered in the present study. Analysis of specimens collected in the Canadian Arctic Archipelago would potentially illuminate potential barriers to geneflow between the Canada West group and Canada East. Also of interest would be examination of more specimens from the Siberian seas from the Chukchi Sea towards the Northeast Atlantic Ocean, to explore whether Siberian River outflow influences population structure as proposed here for the Mackenzie River. The recruitment source of B.saida found in the Arctic basins could be examined by analysis of specimens collected further offshore than studied here. This work relied on samples collected from a variety of sources with different levels of documentation. Collections made in concert with documentation of environmental parameters and morphological characteristics would allow for examination of the relationships between habitat, phenotype and genetic differentiation. The development and application of higher density markers, such as single-nucleotide polymorphism markers, would allow a deeper exploration of the differentiation patterns observed in the present study (Glover et al. 2010).

Conclusion

We detected population differentiation of B. saida over circumpolar and regional scales. The present analysis provides context for population monitoring and future assessments of population structure, and for the design of other types of research. Moreover, our findings can be used for the design of a large-scale spatial template for management plans for this important species and the marine areas in which it is present. Genetic analyses presented indicate that patterns of gene flow are complex and potentially influenced by geography, river discharge, prevailing currents, bathymetry, water properties and prey availability. We note that the direction of gene flow has important implications for expansion of southern alleles, and in some regions such as the Bering and Chukchi seas, prevailing currents may be promoting northward range expansions, while along the east coast of Canada the opposite may be true.

References

  1. AMAP (1998) AMAP Assessment Report: Arctic Pollution Issues. Arctic Monitoring and Assessment Programme (AMAP), Oslo

    Google Scholar 

  2. Andre C, Svedäng H, Knutsen H, Dahle G, Jonsson P, Ring A-K, Sköld M, Jorde P-E (2016) Population structure in Atlantic cod in the eastern North Sea-Skagerrak-Kattegat: early life stage dispersal and adult migration. BMC Res Notes 9:63. https://doi.org/10.1186/s13104-016-1878-9

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  3. Beacham TD, Brattey J, Miller KM, Le KD, Withler RE (2002) Multiple stock structure of Atlantic cod (Gadus morhua) off Newfoundland and Labrador determined from genetic variation. ICES J Mar Sci 59:650–665

    Google Scholar 

  4. Beaumont MA, Nichols RA (1996) Evaluating loci for use in the genetic analysis of population structure. Proc R Soc Lond. https://doi.org/10.1098/rspb.1996.0237

    Article  Google Scholar 

  5. Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F (2004) GENETIX 4.05, logiciel sous Windows TM pour la génétique des populations. Laboratoire Génome, populations, interactions. CNRS UMR 5000, Université de Montpellier II, Montpellier

    Google Scholar 

  6. Bouchard C, Fortier L (2020) The importance of Calanus glacialis for the feeding success of young polar cod: a circumpolar synthesis. Polar Biol. https://doi.org/10.1007/s00300-020-02643-0

    Article  Google Scholar 

  7. Bouchard C, Robert D, Nelson RJ, Fortier L (2013) The nucleus of the lapillar otolith discriminates the early life stages of Boreogadus saida and Arctogadus glacialis. Polar Biol 36:1537–1542

    Google Scholar 

  8. Bouchard C, Thorrold SR, Fortier L (2015) Spatial segregation, dispersion and migration in early stages of polar cod Boreogadus saida revealed by otolith chemistry. Mar Biol 162:855–868

    CAS  Google Scholar 

  9. Bouchard C, Mollard S, Suzuki K, Robert D, Fortier L (2016) Contrasting the early life histories of sympatric Arctic gadids Boreogadus saida and Arctogadus glacialis in the Canadian Beaufort Sea. Polar Biol 39:1005–1022

    Google Scholar 

  10. Bouchard C, Geoffroy M, LeBlanc M, Majewski A, Gauthier S, Walkusz W, Reist JD, Fortier L (2017) Climate warming enhances polar cod recruitment, at least transiently. Prog Oceanogr 156:121–12

    Google Scholar 

  11. Bouchard C, Geoffroy M, LeBlanc M, Fortier L (2018) Larval and adult fish assemblages along the Northwest Passage: the shallow Kitikmeot and the ice-covered Parry Channel as potential barriers to dispersal. Arctic Sci. 4:781–793. https://doi.org/10.1139/AS-2018-0003

    Article  Google Scholar 

  12. Bradstreet MSW, Finley KJ, Sekerak AD, Griffiths WB, Evans CR, Fabijan MF, Stallard, HE (1986) Aspects of the biology of Arctic cod (Boreogadus saida) in Arctic marine food chains. Canadian Technical Report of Fisheries and Aquatic Sciences 1491. Winnipeg, Manitoba: Department of Fisheries and Oceans, Central and Arctic Region

  13. Coachman LK, Aagaard K, Tripp RB (1975) Bering Strait. The regional and physical oceanography. University of Washington Press, Seattle

    Google Scholar 

  14. Crawford RE, Jorgenson JK (1993) Schooling behaviour of arctic cod, Boreogadus saida, in relation to drifting pack ice. Environ Biol Fishes 36:345–357

    Google Scholar 

  15. David C, Lange B, Krumpen T, Schaafsma F, van Franeker JA, Flores H (2016) Under-ice distribution of polar cod Boreogadus saida in the central Arctic Ocean and their association with sea-ice habitat properties. Polar Biol 39:981–994. https://doi.org/10.1007/s00300-015-1774-0

    Article  Google Scholar 

  16. De Robertis A, Taylor K, Wilson C, Farley E (2016) Abundance and distribution of Arctic cod (Boreogadus saida) and other pelagic fishes over the U.S. continental shelf of the Northern Bering and Chukchi seas. Deep Sea Res Part II Top Stud Oceanogr. https://doi.org/10.1016/j.dsr2.2016.03.002

    Article  Google Scholar 

  17. Divoky GJ, Lukacs PM, Druckenmiller ML (2015) Effects of recent decreases in Arctic sea ice on an ice-associated marine bird. Prog Oceanogr. https://doi.org/10.1016/j.pocean.2015.05.010

    Article  Google Scholar 

  18. Drolet R, Fortier L, Ponton D, Gilbert M (1991) Production of fish larvae and their prey in subarctic southeastern Hudson Bay. Mar Ecol Progr Ser 77:105–118. https://doi.org/10.3354/meps077105

    Article  Google Scholar 

  19. Drost HE, Lo M, Carmack EC, Farrell AP (2016) Acclimation potential of Arctic cod (Boreogadus saida) from the rapidly warming Arctic Ocean. J Exp Biol 219:3114–3125. https://doi.org/10.1242/jeb.140194

    CAS  Article  PubMed  Google Scholar 

  20. Dyke AS, Hooper J, Savelle JM (1996) A history of sea ice in the Canadian Arctic Archipelago based on postglacial remains of the bowhead whale (Balaena mysticetus). Arctic 49:235–255

    Google Scholar 

  21. Falardeau M, Robert D, Fortier L (2014) Could the planktonic stages of polar cod and Pacific sand lance compete for food in the warming Beaufort Sea? ICES J Mar Sci 71:1956–1965. https://doi.org/10.1093/icesjms/fst221

    Article  Google Scholar 

  22. Fevolden SE, Martinez I, Christiansen JS (1999) RAPD and scnDNA analyses of polar cod, Boreogadus saida (Pisces, Gadidae), in the North Atlantic. Sarsia 84:99–103

    Google Scholar 

  23. Fissel DB, Lemon DD, Birch JR (1982) Major features of the summer near-surface circulation of western Baffin-Bay, 1978 and 1979. Arctic 35:180–200

    Google Scholar 

  24. Fitzpatrick SW, Gerberich JC, Kronenberger JA, Angeloni LM, Funk WC (2014) Locally adapted traits maintained in the face of high gene. Ecol Lett 18:37–47

    PubMed  Google Scholar 

  25. Foll M, Gaggiotti OE (2008) A genome scan method to identify selected loci appropriate for both dominant and codominant markers: a Bayesian perspective. Genetics 180:977–993

    PubMed  PubMed Central  Google Scholar 

  26. Excoffier L, Hofer T, Foll M (2009) Detecting loci under selection in a hierarchically structured population. Heredity 103:285–298

    CAS  PubMed  Google Scholar 

  27. Geoffroy M, Robert D, Darnis G, Fortier L (2011) The aggregation of polar cod (Boreogadus saida) in the deep Atlantic layer of ice-covered Amundsen Gulf (Beaufort Sea) in winter. Polar Biol 34:1959–1971. https://doi.org/10.1007/s00300-011-1019-9

    Article  Google Scholar 

  28. Glover KA, Hansen MM, Lien S, Als TD, Høyheim B, Skaala O (2010) A comparison of SNP and STR loci for delineating population structure and performing individual genetic assignment. BMC Genet. https://doi.org/10.1186/1471-2156-11-2

    Article  PubMed  PubMed Central  Google Scholar 

  29. Gradinger RR, Bluhm BA (2004) In-situ observations on the distribution and behavior of amphipods and Arctic cod (Boreogadus saida) under the sea ice of the High Arctic Canada Basin. Polar Biol 27:595–603. https://doi.org/10.1007/s00300-004-0630-4

    Article  Google Scholar 

  30. Graham M, Hop H (1995) Aspects of reproduction and larval biology of Arctic cod (Boreogadus saida). Arctic 48:130–135

    Google Scholar 

  31. Guo SW, Thompson EA (1992) Performing the exact test of Hardy-Weinberg proportions for multiple alleles. Biometrics 48:361–372

    CAS  PubMed  Google Scholar 

  32. Hale ML, Burg TM, Steeves TE (2012) Sampling for microsatellite-based population genetic studies: 25 to 30 individuals per population is enough to accurately estimate allele frequencies. PLoS ONE. https://doi.org/10.1371/journal.pone.0045170

    Article  PubMed  PubMed Central  Google Scholar 

  33. Harkness BAS (2017) An assessment of population genetic structure in guillemots (Cepphus). MS Thesis, Queen’s University, Kingston, Ontario

  34. Hop H, Gjøsæter H (2013) Polar cod (Boreogadus saida) and capelin (Mallotus villosus) as key species in marine food webs of the Arctic and the Barents Sea. Mar Biol Res 9:878–894

    Google Scholar 

  35. Hop H, Graham M, Trudeau VL (1995) Spawning energetics of Arctic cod (Boreogadus saida) in relation to seasonal development of the ovary and plasma sex steroid levels. Can J Fish Aquat Sci 52:541–550

    CAS  Google Scholar 

  36. Jombart T (2008) adegenet: an R package for the multivariate analysis of genetic markers. Bioinformatics 24:1403–1405. https://doi.org/10.1093/bioinformatics/btn129

    CAS  Article  PubMed  Google Scholar 

  37. Kenchington EL, Nakashima BS, Taggart CT, Hamilton LC (2015) Genetic structure of capelin (Mallotus villosus) in the northwest Atlantic Ocean. PLoS ONE. https://doi.org/10.1371/journal.pone.0122315

    Article  PubMed  PubMed Central  Google Scholar 

  38. Kessel ST, Hussey NE, Crawford RE, Yurkowski DJ, Webber DM, Dick TA, Fisk AT (2017) First documented large-scale horizontal movements of individual Arctic cod (Boreogadus saida). Can J Fish Aquat Sci 74:292–296

    Google Scholar 

  39. Kessel ST, Hussey NE, Crawford RE, Yurkowski DJ, O’Neill CV, Fisk AT (2015) Distinct patterns of Arctic cod (Boreogadus saida) presence and absence in a shallow high Arctic embayment, revealed across open-water and ice-covered periods through acoustic telemetry. Polar Biol 39:1057–1068. https://doi.org/10.1007/s00300-015-1723-y

    Article  Google Scholar 

  40. Kuletz KJ, Ferguson MC, Hurley B, Gall AE, Labunski EA, Morgan TC (2015) Seasonal spatial patterns in seabird and marine mammal distribution in the Pacific Arctic: identifying biologically important pelagic areas. Prog Oceanogr 136:175–200

    Google Scholar 

  41. Laurel BJ, Copeman LA, Spencer M, Iseri P (2018) Comparative effects of temperature on rates of development and survival of eggs and yolk-sac larvae of Arctic cod (Boreogadus saida) and walleye pollock (Gadus chalcogrammus). ICES J Mar Sci 75:2403–2412. https://doi.org/10.1093/icesjms/fsy042

    Article  Google Scholar 

  42. LeBlond PH (1980) On the surface circulation in some channels of the Canadian Arctic Archipelago. Arctic 33:189–197

    Google Scholar 

  43. Lewontin RC (1974) The genetic basis of evolutionary change. Colombia University Press, New York

    Google Scholar 

  44. Loseto LL, Stern GA, Connelly TL, Deibel D, Gemmill B, Prokopowicz A, Fortier L, Ferguson SH (2009) Summer diet of beluga whales inferred by fatty acid analysis of the eastern Beaufort Sea food web. J Exp Mar Biol Ecol 374:12–18

    CAS  Google Scholar 

  45. Macdonald RW, Bewers JM (1996) Contaminants in the arctic marine environment: priorities for protection. ICES J Mar Sci 53:537–563

    Google Scholar 

  46. Madsen M, Fevolden S-E, Christiansen JS (2009) A simple molecular approach to distinguish between two Arctic gadoid fishes Arctogadus glacialis (Peters, 1874) and Boreogadus saida (Lepechin, 1774). Polar Biol 32:937–939

    Google Scholar 

  47. Madsen ML, Nelson RJ, Fevolden S-E, Christiansen JS, Præbel K (2016) Population genetic analysis of Euro-Arctic polar cod Boreogadus saida suggests fjord and oceanic structuring. Polar Biol 39:969–980

    Google Scholar 

  48. Majewski AR, Walkusz W, Lynn BR, Atchison S, Eert J, Reist JD (2015) Distribution and diet of demersal Arctic Cod, (Boreogadus saida), in relation to habitat characteristics in the Canadian Beaufort Sea. Polar Biol 39:1087–1098. https://doi.org/10.1007/s00300-015-1857-y

    Article  Google Scholar 

  49. Melnikov IA, Chernova NV (2013) Characteristics of under-ice swarming of polar cod Boreogadus saida (Gadidae) in the central Arctic Ocean. J Ichthyol 53:7–15. https://doi.org/10.1134/S0032945213010086

    Article  Google Scholar 

  50. Michaud J, Fortier L, Rowe P, Ramseier R (1996) Feeding success and survivorship of Arctic cod larvae, Boreogadus saida, in the Northeast Water Polynya (Greenland Sea). Fish Oceanogr 5:120–135. https://doi.org/10.1111/j.1365-2419.1996.tb00111.x

    Article  Google Scholar 

  51. Münchow A, Falkner KK, Melling H (2015) Baffin island and west Greenland current systems in northern Baffin Bay. Prog Oceanogr 132:305–317

    Google Scholar 

  52. Nahrgang J, Varpe Ø, Korshunova E, Murzina S, Hallanger IG et al (2014) Gender specific reproductive strategies of an arctic key species (Boreogadus saida) and implications of climate change. PLoS One 9:e98452. https://doi.org/10.1371/journal.pone.0098452.s008

    Article  PubMed  PubMed Central  Google Scholar 

  53. Nei M (1987) Molecular evolutionary genetics. Columbia University Press, New York

    Google Scholar 

  54. Nelson RJ, Beacham TD, Small MP (1998) Microsatellite analysis of the population structure of a Vancouver Island sockeye salmon (Oncorhynchus nerka) stock complex using nondenaturing gel electrophoresis. Mol Mar Biol Biotech 7:312–319

    CAS  Google Scholar 

  55. Nelson RJ, Bouchard C, Madsen M, Præbel K, Rondeau E, von Schalburg K, Leong J, Jantzen SZ, Puckett S, Messmer A, Fevolden S-E, Koop BF (2013) Microsatellite loci for genetic analysis of the arctic gadids Boreogadus saida and Arctogadus glacialis. Conserv Genet Resour. https://doi.org/10.1007/s12686-012-9824-1

    Article  Google Scholar 

  56. O'Leary DB, Coughlan J, Dillane E, McCarthy TV, Cross TF (2007) Microsatellite variation in cod Gadus morhua throughout its geographic range. J Fish Biol 70:310–335

    CAS  Google Scholar 

  57. Pálsson S, Källman T, Paulsen J, Árnason E (2009) An assessment of mitochondrial variation in Arctic gadoids. Polar Biol 32:471–479. https://doi.org/10.1007/s00300-008-0542-9

    Article  Google Scholar 

  58. Pickart RS, Weingartner TJ, Zimmermann S, Torres DJ, Pratt LJ (2005) Flow of winter transformed Pacific water into the western Arctic. Deep-Sea Res II 52:3175–3198

    Google Scholar 

  59. Præbel K, Westgaard JI, Fevolden SE, Christiansen JS (2008) Circumpolar genetic population structure of capelin Mallotus villosus. Mar Ecol Prog Ser 360:189–199. https://doi.org/10.3354/meps07363

    Article  Google Scholar 

  60. R Core Team (2013) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria https://www.R-project.org/

  61. Raymond M, Rousset F (1995) An exact test of population differentiation. Evolution 49:1283–1286

    Google Scholar 

  62. Rice WR (1989) Analyzing tables of statistical tests. Evolution 43:223–225

    PubMed  Google Scholar 

  63. Robertson A, Hill WG (1984) Deviations from Hardy-Weinberg proportions: sampling variances and use in estimation of inbreeding coefficients. Genetics 107:703–718

    CAS  PubMed  PubMed Central  Google Scholar 

  64. Rose GA, Nelson RJ, Mello LGS (2011) Isolation or metapopulation: whence and whither the Smith Sound cod? Can J Fish Aquat Sci 68:152–169

    Google Scholar 

  65. Rousset F (2008) Genepop’007: a complete reimplementation of the Genepop software for Windows and Linux. Mol Ecol Res 8:103–106

    Google Scholar 

  66. Ruzzante DE, Taggart CT, Cook D (1999) A review of the evidence for genetic structure of cod (Gadus morhua) populations in the Northwest Atlantic and population affinities of larval cod off Newfoundland and the Gulf f St. Lawrence. Fish Res 43:79

    Google Scholar 

  67. Ruzzante D, Taggart CT, Cook D (1998) A nuclear DNA basis for shelf- and bank-scale population structure in northwest Atlantic cod (Gadus morhua): Labrador to Georges Bank. Mol Ecol 7:1663–1680

    Google Scholar 

  68. Schneider L, Stefan G, Excoffier L (2010) Arlequin (version 3.5.1.2): an integrated software package for population genetics data analysis. Evol Bioinform Online 1:47–50

    Google Scholar 

  69. Spencer ML, Vestfals CD, Mueter FJ, Laurel BJ (2019) Ontogenetic changes in the buoyancy and salinity tolerance of eggs and larvae of polar cod (Boreogadus saida) and other gadids. Polar Biol. https://doi.org/10.1007/s00300-020-02620-7

    Article  Google Scholar 

  70. Springer AM, McRoy CP (1993) The paradox of pelagic food webs in the northern Bering Sea. III. Patterns of primary production. Cont Shelf Res 13:575–599

    Google Scholar 

  71. Springer AM, McRoy CP, Turko KR (1989) The paradox of pelagic food webs in the northern Bering Sea. II. Zooplankton communities. Cont Shelf Res 9:359–386

    Google Scholar 

  72. Tang CCL, Ross CK, Yao T, Petrie B, DeTracey BM, Dunlap E (2004) The circulation, water masses and sea-ice of Baffin Bay. Progr Oceanogr 63:183–228

    Google Scholar 

  73. Waples RS (1989) Temporal variation in allele frequencies; testing the right hypothesis. Evolution 43:1236–1251. https://doi.org/10.1111/j.1558-5646.1989.tb02571.x

    Article  PubMed  Google Scholar 

  74. Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution 38:1358–1370

    CAS  PubMed  Google Scholar 

  75. Weir BS (1996) Genetic data analysis II: methods for discrete population genetic data. Sinauer Assoc., Inc., Sunderland

    Google Scholar 

  76. Welch H, Bergmann M, Siferd T, Martin K, Curtis M, Crawford R, Conover R, Hop H (1992) Energy flow through the marine ecosystem of the Lancaster sound region, Arctic Canada. Arctic 45:343–357

    Google Scholar 

  77. Wilson RE, Sage GK, Wedemeyer K, Sonsthagen SA, Menning DM, Gravley MC, Sexson MG, Nelson RJ, Talbot SL (2019) Micro-geographic population genetic structure within Arctic cod (Boreogadus saida) in Beaufort Sea of Alaska ICES. J Mar Sci. https://doi.org/10.1093/icesjms/fsz041

    Article  Google Scholar 

  78. Wilson RE, Sonsthagen SA, Gharrett AJ, Majewski A, Wedemeyer K, Nelson RJ, Talbot SL (in review) Mitochondrial genome diversity and population mitogenomics of Arctic cod (Boreogadus saida) and Arctic dwelling gadids

Download references

Acknowledgements

We acknowledge those who, beside the authors, have contributed samples and expertise to this study: J.S. Christiansen, S-E. Fevolden, D. Archambault, T. Siferd, B. Norcross, K. Dunton and S. Pálsson. We thank the numerous institutions that have contributed funding for this study, including the North Pacific Research Board, the Natural Sciences and Engineering Research Council (Canada), ArcticNet, the W. Garfield Weston Foundation, the TUNU-Programme (IPY ID: 318), UiT the Arctic University of Norway, Fisheries and Oceans Canada, Crown-Indigenous Relations and Northern Affairs Canada, and the Canada First Research Excellence Fund. The authors are grateful for the insightful and constructive reviews of Marvin Choquet, Kimberly Howland and an anonymous reviewer which greatly improved this manuscript.

Author information

Affiliations

Authors

Corresponding author

Correspondence to R. J. Nelson.

Ethics declarations

Conflict of interest

There is no known conflict of interest of any of the authors in the publication of this manuscript or its contents.

Ethical approval

Fish collections were obtained opportunistically rather than expressly for the study described. Applicable permits held by contributors of samples in Canadian waters were from Fisheries and Oceans Canada, and for the specimens contributed by the Arctic Net program see https://arcticnet.ulaval.ca/expeditions. For collections in the Greenland Sea, permits from the Government of Greenland were held for TUNU-Programme (IPY ID: 318). Fish were handled according to permitting conditions.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This article belongs to the special issue on the “Arctic Gadids in a Changing Climate”, coordinated by Franz Mueter, Haakon Hop, Benjamin Laurel, Caroline Bouchard, and Brenda Norcross.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 89 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Nelson, R.J., Bouchard, C., Fortier, L. et al. Circumpolar genetic population structure of polar cod, Boreogadus saida. Polar Biol 43, 951–961 (2020). https://doi.org/10.1007/s00300-020-02660-z

Download citation

Keywords

  • Polar cod
  • Arctic cod
  • Boreogadus saida
  • Arctogadus glacialis
  • Population genetics
  • Biogeographical barriers