Differentiating salmonid migratory ecotypes through stable isotope analysis of collagen: Archaeological and ecological applications

Abstract

The ability to distinguish between different migratory behaviours (e.g., anadromy and potamodromy) in fish can provide important insights into the ecology, evolution, and conservation of many aquatic species. We present a simple stable carbon isotope (δ13C) approach for distinguishing between sockeye (anadromous ocean migrants) and kokanee (potamodromous freshwater residents), two migratory ecotypes of Oncorhynchus nerka (Salmonidae) that is applicable throughout most of their range across coastal regions of the North Pacific Ocean. Analyses of kokanee (n = 239) and sockeye (n = 417) from 87 sites spanning the North Pacific (Russia to California) show that anadromous and potamodromous ecotypes are broadly distinguishable on the basis of the δ13C values of their scale and bone collagen. We present three case studies demonstrating how this approach can address questions in archaeology, archival, and conservation research. Relative to conventional methods for determining migratory status, which typically apply chemical analyses to otoliths or involve genetic analyses of tissues, the δ13C approach outlined here has the benefit of being non-lethal (when applied to scales), cost-effective, widely available commercially, and should be much more broadly accessible for addressing archaeological questions since the recovery of otoliths at archaeological sites is rare.

About the Authors:

Eric Guiry

Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected] (EG); [email protected] (TCAR)

Affiliations Department of Anthropology, Trent University, Peterborough, Ontario, Canada, School of Archaeology and Ancient History, University of Leicester, Leicester, United Kingdom, Department of Anthropology, University of British Columbia, Vancouver, British Columbia, Canada

Thomas C. A. Royle

Roles Data curation, Formal analysis, Funding acquisition, Investigation, Resources, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected] (EG); [email protected] (TCAR)

Affiliation: Department of Archaeology, Ancient DNA Laboratory, Simon Fraser University, Burnaby, British Columbia, Canada

ORCID logo http://orcid.org/0000-0001-8178-123X

R. G. Matson

Roles Resources, Writing – review & editing

Affiliation: Department of Anthropology, University of British Columbia, Vancouver, British Columbia, Canada

Hillary Ward

Roles Investigation, Resources, Writing – review & editing

Affiliation: Ministry of Forests, Lands, Natural Resource Operations and Rural Development, Government of British Columbia, Penticton, British Columbia, Canada

Tyler Weir

Roles Investigation, Resources, Writing – review & editing

Affiliation: Ministry of Forests, Lands, Natural Resource Operations and Rural Development, Government of British Columbia, Penticton, British Columbia, Canada

Nicholas Waber

Roles Formal analysis, Visualization

Affiliation: Department of Anthropology, University of British Columbia, Vancouver, British Columbia, Canada

Thomas J. Brown

Roles Formal analysis

Affiliation: Department of Anthropology, University of British Columbia, Vancouver, British Columbia, Canada

Brian P. V. Hunt

Roles Resources, Writing – review & editing

Affiliations Institute for the Oceans and Fisheries, University of British Columbia, Vancouver, British Columbia, Canada, Department of Earth, Ocean and Atmospheric Sciences, University of British Columbia, Vancouver, British Columbia, Canada, Hakai Institute, Heriot Bay, British Columbia, Canada

Michael H. H. Price

Roles Resources, Writing – review & editing

Affiliation: Department of Biological Sciences, Earth to Ocean Research Group, Simon Fraser University, British Columbia, Canada

Bruce P. Finney

Roles Resources, Writing – review & editing

Affiliations Department of Biological Sciences, Idaho State University, Pocatello, Idaho, United States of America, Department of Geosciences, Idaho State University, Pocatello, Idaho, United States of America

Masahide Kaeriyama

Roles Resources

Affiliation: Arctic Research Center, Hokkaido University, Sapporo, Hokkaido, Japan

Yuxue Qin

Roles Resources

Affiliation: School of Marine Science and Environmental Engineering, Dalian Ocean University, Dalian, Liaoning, China

Dongya Y. Yang

Roles Resources, Writing – review & editing

Affiliation: Department of Archaeology, Ancient DNA Laboratory, Simon Fraser University, Burnaby, British Columbia, Canada

Paul Szpak

Roles Resources, Writing – review & editing

Affiliation: Department of Anthropology, Trent University, Peterborough, Ontario, Canada

ORCID logo http://orcid.org/0000-0002-1364-6834

Introduction

Shifting migratory modes are an important behavioural and evolutionary characteristic of many economically and environmentally important fish species [1–3]. The flexibility to switch between life history strategies, such as migrating to the ocean and back (anadromy) [4], on the one hand, and living solely in freshwater (potamodromy) [5], on the other hand, can contribute to species’ adaptability to new environments and resilience under difficult and changing conditions [2, 6]. For this reason, there has been long standing, multidisciplinary interest in developing independent techniques for distinguishing between anadromous and potamodromous ecotypes [e.g., 7–9]. For instance, the ability to independently assess ecotype can allow for more detailed studies of how sympatric populations respond differently to changing environmental conditions through time. This may, in turn, help to guide future conservation policy and enhance environmental restoration efforts. Techniques for differentiating ecotypes could also be applied to specimens in natural history museum collections to validate key aspects of archival metadata or to study the historical ecology of endangered or extinct populations. Differentiating these ecotypes is important not only for biological and conservation research but also for archaeologists interested in understanding past fishing practices [e.g., 10–12].

Isotopic and elemental analyses have been at the forefront of efforts to distinguish between anadromous and potamodromous ecotypes. This is because there can be large differences between the isotopic and elemental composition of typical freshwater and marine environments that are passed along the food web to consumers, like fish, providing a useful marker for anadromous and potamodromous behaviours. Among these, the most commonly used approaches have been measuring the Sr/Ca and Sr/Ba ratios [8, 13] as well as the isotopic composition of radiogenic strontium (87Sr/86Sr) [14, 15], stable sulfur isotopes (δ34S) [7, 16], and stable oxygen isotopes (δ18O) [17] in fish tissues. These approaches, which are typically applied to fish otoliths, have been deployed by ecologists [8], palaeontologists [17, 18], and archaeologists [19, 20] for decades to establish whether fish capable of anadromy have migrated to the ocean and back as part of their life history or if they had remained in a freshwater environment their entire lives.

Despite their popularity and highly successful track record, there are several challenges associated with the application of these conventional approaches, the magnitude of which will depend on the research field. From an archaeological perspective, fish otoliths, in general, are not frequently recovered [21] due to their small size. Although researchers have successfully applied otolith-based techniques to archaeological specimens [e.g., 22, 23], their use will be circumscribed to the limited number of archaeological sites where preservation conditions and excavation practices have allowed for the recovery of fish otoliths in larger quantities [21]. Moreover, applying strontium-based analyses (either isotopic composition or elemental concentration) to bones, which are typically the most abundant fish remains recovered archaeologically, as an alternative to otoliths, is potentially complicated by the greater susceptibility of fish bone to diagenetic alterations (i.e., addition or loss of Sr and other elements in the burial environment [18, 24]). Tooth enamel, which is less porous than bone, and may be relatively more resistant to diagenetic contamination, has also been used [20]. However, not all fish species have suitable teeth and tooth bearing elements with intact teeth are often much less frequently recovered archeologically. From an ecological perspective using otoliths could also be problematic. Extracting otoliths is fatal and while otoliths in many cases can be collected after a spawning run, when some species of fish are already near the end of their lifecycle, many fish do not die after spawning. From a conservation perspective, research methods that minimize invasive procedures that kill fish have a wide range of other benefits [e.g., 25–27]. Finally, for ecologists and archaeologists, relatively high labour, financial, and other resource costs are associated with undertaking sectioning and elemental and/or isotopic analyses of otoliths, which may be prohibitive for smaller research or conservation budgets. While the analysis of otoliths provides a powerful and unparalleled tool for reconstructing fish life histories in intricate detail, a simpler approach for distinguishing between potamodromous and anadromous behavioural strategies may be sufficient to meet the needs of many fisheries management conservation programs or archaeological and ecological research questions.

Using stable carbon isotope (δ13C) compositions of collagen from 656 anadromous sockeye and potamodromous kokanee (both Oncorhynchus nerka, Salmonidae) from 87 sites across the northern Pacific Rim (Fig 1), we demonstrate a simple and cost-effective alternative approach for establishing the migratory life history of fish that are capable of anadromy. Because collagen can be extracted from both fish scales and bones, it offers an additional option for ecologists to sample fish non-lethally and to archaeologists for whom bone is often the only sample material available in great abundance. As collagen is also available in historical scale archives and preserved fish specimens curated at museums, this approach can also enable historical reconstructions of fish behaviours over the recent past. O. nerka provides an ideal test species to validate this approach because their pelagic planktivorous feeding behaviour mitigates potential issues with isotopic overlap between potamodromous and anadromous ecotypes (see Section, Stable carbon isotopes in aquatic environments). Moreover, throughout their wide geographic range, numerous freshwater resident O. nerka populations exist, allowing for a spatially comprehensive assessment of the effect of diverse environmental (differing carbon sources and productivity levels) and behavioural variability on kokanee stable carbon isotope compositions. To demonstrate the broader applicability of this approach we present three case studies in which stable carbon isotope analyses of O. nerka of unknown ecotype are used to distinguish between kokanee and sockeye at (1) archaeological sites, (2) in natural history collections, and (3) in modern conservation work. We also include stable nitrogen isotope (δ15N) compositions, which were measured in tandem with δ13C, to explore variation in freshwater and marine nitrogen inputs and cycling. In addition to stable isotope analyses, we also conducted ancient DNA (aDNA) analysis in order to confirm the species identity of a subset of the analyzed archaeological remains.

[Figure omitted. See PDF.]

Fig 1. Map showing collection locations for samples.

Known kokanee (circles) and sockeye (diamonds) as well as unknown (squares) samples are shaded in blue and red, respectively, according to the mean δ13C values.

https://doi.org/10.1371/journal.pone.0232180.g001

Context

Stable carbon isotopes in aquatic environments

Many previous studies have used δ13C of fish tissues as a basis for distinguishing between sympatric marine and freshwater ecotypes [e.g., 7, 28–30]. These studies have relied partly on the observation that food webs in marine environments can have elevated δ13C relative to their freshwater counterparts, which provides a useful marker for distinguishing between fish with anadromous and potamodromous life histories. All these applications, however, have been context-specific and often rely at least partly on additional evidence from other isotopic proxies, such as δ34S and δ15N measurements.

An important reason for why δ13C has not previously been deployed for use as a spatiotemporally broad-scale approach for distinguishing between freshwater and marine fish is that freshwater food webs can have an enormous range of isotopic variability [31]. This variability can create overlap between δ13C values for freshwater and marine conspecifics, making the utility of δ13C as a universal means of differentiating between potamodromous and anadromous ecotypes uncertain, particularly across broader geographical regions [31]. The stable carbon isotope composition of aquatic food webs is governed by a complex and highly interconnected set of processes, contributing to this variability (for reviews see [32–35]). While some environmental variables, such as temperature [36], may have similar impacts on δ13C of food webs in both marine and freshwater ecosystems, many factors have more potential to influence the isotopic composition of fish in freshwater environments due to their smaller size and constrained resource pools (for review see [31]). Despite these complexities, and potential for overlap between the stable carbon isotope compositions of fish inhabiting both marine and freshwater environments, it may still be possible to use δ13C (a simpler and more cost-effective alternative; see Section Introduction) to differentiate between marine and freshwater ecotypes within the same taxon provided the dietary behaviour of the species analyzed is sufficiently specialized (i.e., the species is not a dietary generalist with potential to feed across other major axes of systematic δ13C variations, such as nearshore and offshore areas).

Stable nitrogen isotopes in aquatic environments

Although not the focus of this study, δ15N was obtained in tandem with δ13C measurements, and we provide these isotopic compositions here to give additional insights into freshwater and marine environments. Stable nitrogen isotope compositions undergo a stepwise increase by roughly 3 to 5‰ at each trophic step within a food web [37] and have therefore been used to explore predator-prey relationships and trophic dynamics [38, 39]. However, the δ15N of key nitrogenous nutrients that form the baseline of both marine and freshwater environments is controlled by a wide range of variables (e.g., microbial activity, productivity, oxygenation, alkalinity) that can result in large shifts in ecosystems’ stable nitrogen isotope baselines, even over relatively small spatiotemporal spans [31, 40, 41]. As with δ13C, this is particularly the case in freshwater ecosystems where, in comparison with larger marine environments, the smaller overall pool of dissolved inorganic nitrogen available in a given local nitrogen cycle, may be more sensitive to fluctuations in key processes governing isotopic fractionations [31]. In this context, and given that our analyses focus on adults from a single species of fish (mostly feeding at a similar trophic level) from a very wide range of environments, we expect variation in δ15N to primarily reflect regional differences in nitrogen inputs and cycling [42].

Collagen and temporal resolution of diet

Biologically, collagen has excellent properties for assessing questions about animal life histories. Isotopic compositions of tissues can vary based on differing time scales that are specific to each tissue’s turnover rate [43]. For example, the muscle or organ tissues typically analyzed in ecological research will have isotopic compositions reflecting diet over a period of months or weeks [44,45]. In contrast, the slow turnover of collagen in bone, which remodels over the course of an organism’s lifetime, provides an average perspective on diet during the period over which a fish underwent the bulk of its growth [46, 47]. For this reason, the isotopic composition of bone collagen from shorter lived fish species should provide a lifetime average perspective on dietary intake (weighted towards the period when their growth-rate was fastest). From an isotopic and compositional perspective, scale collagen is directly comparable with bone collagen [48] and is also similar in that it has an isotopic composition reflecting a long-term average of diet because it is incrementally laid down during growth [49, 50]. For migratory fish like salmon, research comparing δ13C in scales from smolts (juveniles) departing their natal streams, and adults, returning to those same streams (after only one winter at sea), has shown that the diets of adults undergoing substantial growth quickly ‘overwrites’ the isotopic signal from their natal freshwater environments [51]. Therefore, in addition to being a readily available and potentially non-lethal source of samples, scale and bone collagen provides an ideal material for testing hypotheses about potamodromous versus anadromous life histories in fish because it gives a time-averaged perspective [52] that is less susceptible to short-term fluctuations resulting from seasonal change or occasional behavioural aberrations.

Species identification of archaeological Pacific salmonid remains

Throughout the northern Pacific Rim, Pacific salmonid (Oncorhynchus spp.) bones are commonly recovered from archaeological sites, often in large quantities [53–55]. At most sites in the region, most archaeologically recovered Pacific salmonid bones are vertebrae due to their high bone density relative to cranial and other post-cranial elements mitigating destructive taphonomic processes [56]. With the exception of a few cranial elements, most Pacific salmonid skeletal elements, including vertebrae, lack inter-specific morphological variation meaning it is not possible to assign species-level identifications to most archaeological Pacific salmonid remains through conventional comparative morphological approaches [57, 58]. Consequently, metric [59], morphometric [60], radiographic [61], and stable isotope [62, 63] methods have been developed for the taxonomic identification of archaeological Pacific salmonid vertebrae. However, these methods often yield inaccurate results or are unable to refine identifications to a particular species rather than groups of species [62–67]. Due to the limitations of other species identification approaches, both aDNA analysis [e.g., 12, 58, 59, 64–66, 68–70] and peptide mass fingerprinting [e.g., 71], more commonly known as zooarchaeology by mass spectrometry (ZooMS) [72] are increasingly being used to assign species identities to archaeological Pacific salmonid remains. Both of these approaches can assign reliable species-level identifications to Pacific salmonid remains, making them useful for taxonomic identification. For these reasons, we used aDNA analysis to confirm the species identities of our archaeological samples (see Section Ancient DNA analysis).

Methods

Sampling

Samples included in this study are summarized in Table 5 in S1 Text. We collected scale samples from 213 potamodromous kokanee and 43 anadromous sockeye (hereafter referred to as kokanee and sockeye, respectively) from museums and biologists for isotopic analysis, with a view to maximizing geographical coverage for our δ13C baseline. Fish were assigned to either sockeye, kokanee, or unknown groups using information provided in associated records (i.e., identifications made by the biologists collecting or cataloguing specimens) or collection location (i.e., a kokanee identification could be assigned to fish collected in watersheds not draining into the Pacific Ocean or lacking anadromous sockeye). Temporally, these samples span the last century (Fig 2), with all decades since the 1920s represented but more prominent representation from the last three decades. Where possible, we also sourced δ13C data for muscle (only lipid corrected δ13C used), scale, and bone from the literature [73–82], including 380 sockeye and 17 kokanee. For our three case studies we worked with museums, archaeologists, First Nations, and conservation biologists to source samples. In all cases, samples (n = 58) came from specimens with no associated information about ecotype (i.e., whether they are potamodromous or anadromous). No live fish were sampled in this study. All materials from modern specimens were sourced from fish collected for previous studies, donated or purchased from licensed commercial fishers, or collected as part of government conservation programs. No permits were required for the described study, which complied with all relevant regulations. Complete repository information for all samples is provided in Table 5 in S1 Text and includes the following: the Royal British Columbia Museum (Victoria, BC, Canada), the Beaty Biodiversity Museum (Vancouver, BC Canada), the Royal Ontario Museum (Toronto, ON, Canada), the British Columbia Ministry of Forests, Lands and Natural Resource Operations (Penticton, BC, Canada), and the Xeni Gwet’in First Nation (curated at the Department of Anthropology, University of British Columbia, Vancouver, BC, Canada).

[Figure omitted. See PDF.]

Fig 2. Number of samples versus collection date.

https://doi.org/10.1371/journal.pone.0232180.g002

Sample pre-treatment

In many cases, scales came from specimens curated in natural history collections that have been in long-term storage and treated with preservatives, typically formalin and ethanol. Therefore, scale collagen pre-treatments sought to maximize collagen purity and the removal of potential contaminants. Following Guiry and colleagues [83], scales were cleaned prior to isotopic analyses with a scalpel and sonicated in deionized water for 3 × 15 min to remove potential adhering contaminant materials (tissue, guanine). Cleaned scales were soaked in 1.0 M HCl for 2 min followed by additional rinses in an ultrasonic bath of deionized water 2 × 3–5 min. This acid pre-treatment removes the mineralized external plate from scales, leaving behind the underlying acid insoluble, collagen-rich fibrillar plate. This process should also help loosen any materials that may have settled on or become attached to the external surface of scale samples during their long-term storage. Scales were then air dried.

Bone collagen from modern and archaeological samples was extracted following established methods [84]. Bones surfaces were cleaned of debris and cut into small, ⁓3x3 mm chunks. Samples were then treated with a 2:1 chloroform-methanol solution in an ultrasonic bath (5–10 min each) to remove lipids [80, 85]. Following lipid removal, samples were demineralized by soaking in 0.5 M HCl. Samples were then rinsed to neutrality in deionized water. For archaeological samples, which may have become contaminated with humic acids and other base-soluble contaminants from their burial environment, an additional treatment was applied. These samples were soaked in 0.1 M NaOH multiple times in an ultrasonic bath, with the solution refreshed every 15 min until the solution remained clear, and then rinsed again in deionized water to neutrality. All samples were then solubilized in a 10−3 M HCl (pH ~3) solution in a heating oven (at 70 °C) for 48 h. Samples were then centrifuged and the solution transferred into a new tube, frozen, and lyophilized.

Isotopic analyses

Scale and bone collagen stable carbon and nitrogen isotope analyses were performed on 0.5 mg subsamples using an elemental analyser (EA) coupled via continuous flow to an isotope ratio mass spectrometer (CF-IRMS) at the Department of Anthropology at The University of British Columbia (UBC; Vancouver, BC, Canada; n = 266) and at the Water Quality Research Centre at Trent University (TU; Peterborough, ON, Canada; n = 48). Duplicate or triplicate analyses were performed on 17% of samples (S1 Text). Collagen samples were combusted in tin capsules in a Vario MICRO cube EA coupled to an Isoprime IRMS (Elementar, Hanover, Germany) at UBC and an EA 300 (Eurovector, Pavia, Italy) coupled to a Horizon IRMS (Nu Instruments, Wrexham, UK) at TU. Carbon and nitrogen isotopic compositions were calibrated relative to VPDB and AIR, respectively, using USGS-40 and USGS-41 or USGS-41a [86, 87]. Instrumental accuracy and precision were monitored using internal collagen standards (S1 Text). Following Szpak and colleagues [88], analytical uncertainty was calculated to be ±0.13‰ and ±0.21‰ for δ13C and δ15N, respectively (S1 Text).

Collagen quality

There has been some suggestion in the literature that fish bone collagen is particularly susceptible to diagenesis [89] and it may therefore be useful to outline factors relevant to assessing collagen quality (QC) for specimens of different ages. Because the amino acid composition of collagen is well understood [89], the percent elemental concentrations (%C and %N) and carbon-to-nitrogen ratio (C:Natomic) of collagen provides a robust proxy for the presence of exogenous carbon and nitrogen, which could skew isotopic measurements. These indicators can therefore be used to screen for collagen diagenesis and assess overall collagen preservation.

It is worth highlighting that, while archaeological fish bone collagen is often found to have higher C:Natomic than bone from other taxa, we can still expect that the isotopic compositions of archaeological and historical specimens will be directly comparable with data from modern specimens, provided they have C:Natomic falling within the acceptable range [89]. Owing to their higher proline and hydroxyproline content, fish collagens have a lower melting point and less stability relative to those of mammals [90] and may therefore be more prone to the effects of leaching and selective amino acid loss [89]. However, because this kind of diagenetic alteration would leave collagen enriched in amino acids like glycine and alanine, which have lower C:Natomic [91], diagenetic processes such as selective amino acid loss are not likely to be behind the trend in higher C:Natomic observed in archaeological fish. Instead, factors contributing to higher C:Natomic likely involve the greater susceptibility of fish bone (being porous, with loosely packed collagen fibrils) to exogenous contaminants [89], a factor that would also apply to historical specimen fixed with formalin.

To assess collagen quality, we used established ranges for carbon (>13.8%) and nitrogen (>4.0%) as well as C:Natomic (2.9–3.6) [92, 93]. These collagen quality indicators are routinely used to screen for collagen degradation and diagenesis in archaeological studies and can also serve as an indicator for issues related to residual museum preservatives (i.e., formalin).

Carbon isotope corrections

Comparing the stable carbon isotope compositions of O. nerka from a range of time-periods as well as with varying curatorial histories requires additional isotopic corrections.

Suess effect correction.

Globally, the δ13C of CO2, which supplies dissolved inorganic carbon (DIC) for primary production in both freshwater and marine environments, has been declining since industrial processes began releasing additional 13C-depleted carbon into the atmosphere through fossil fuel combustion [94]. To compare data from different time periods, we calculated the correction for the Suess effect [94] on δ13C values for tissues from modern and historical samples (Δ13CSuess) as outlined below. Because the residence time for DIC differs between water bodies based on several factors, different Δ13CSuess are applied for tissues from freshwater and marine organisms [95].

For δ13C from anadromous fish, we applied a correction for the Suess effect following Hilton and colleagues [96, 97] as follows:where “a”is the annual rate of decrease in δ13C for the water body, “yrs” is the number of years since 1850 CE, and “b” is the shape of the exponential curve (0.027) defined by the decrease in oceanic δ13C observed by Gruber and colleagues [98] between 1945 and 1997 CE. Sockeye analyzed as part of this study span a wide geographical area across the North Pacific Ocean. Reflecting this broad distribution, and following others working in this region [e.g., 97, 99], we use Quay and colleagues’ [100] estimate of –0.014 for the annual rate of decrease in δ13C in the North Pacific Ocean.

Following others [e.g., 101] working in freshwater environments, for δ13C from potamodromous fish, we applied a correction for the Suess effect following Verberg [102] as follows:where “yrs” is the number of years since 1850 CE. It is important to note that, unlike the correction used for anadromous fish, the equation used for potamodromous fish does not account for temporal offsets related to the complexities of freshwater aquatic carbon cycling. An increasing number of 14C studies highlight the wide range of variation that can occur in the age and reservoir period for DIC used by primary producers at the base of freshwater environments, which can extend to hundreds and even thousands of years [e.g., 103–110]. This offset means that there may be considerable temporal lag between the impact of the Suess effect on the isotopic composition of atmospheric and freshwater DIC pools. Unfortunately, data are not available to calculate the offsets for each freshwater system in this study. It is therefore likely that the correction applied for the Suess effect on freshwater fish overestimates Δ13CSuess, which would serve to decrease the magnitude of potential difference between anadromous and potamodromous O. nerka.

Muscle correction.

Several studies have observed a systematic offset [e.g., 48, 81, 111–113] between fish scale collagen and muscle δ13C, which has previously been characterized for Pacific salmonid species (⁓ +3.69 ‰) [81]. This correction has been applied to data for muscle that we have included in our baseline from the literature.

Formalin fixation correction.

Samples from museum archived specimens have often been preserved using formalin and ethanol. Typically, this is achieved via a multiday soaking in a 10% formalin solution followed by long-term storage in ethanol. This process is known to introduce carbon from formalin to the preserved tissues [114, 115]. Numerous studies have investigated the effects of formalin fixation on fish muscle tissues [e.g., 116–121] and typically show offsets for δ13C (⁓0 to –2 ‰) and δ15N (⁓0 to +1‰). To date, we are unaware of any studies that have investigated the impact of formalin fixation or other preservation techniques on bone or scale collagen. However, based on recent experimental results from analyses of salmonid tissues (Salmo salar and Oncorhynchus tshawytscha) formalin fixation appears to have little effect on the δ13C (–0.4‰) and δ15N (0.0‰) composition of bone and scale collagen. In this context, we have used a conservative δ13C correction of +0.5‰ for museums specimens that have been formalin fixed. In comparison with the range for δ13C that we are considering in this study (>25‰), these effects should be inconsequential for distinguishing between ecotypes. Moreover, any specimens with excessive carbon contamination resulting from formalin fixation should be identifiable based on an elevated C:Natomic ratio and excluded from the study (see Section Collagen quality).

Statistical analyses.

Statistical comparisons of sockeye and kokanee stable isotope compositions were performed using PAST version 3.22 [122]. To establish whether δ13C and δ15N differed significantly between our baseline sockeye and kokanee groups, we used Mann-Whitney U tests with Bonferroni corrected p values. The Shapiro-Wilk’s test showed that δ13C and δ15N values for both sockeye (p <0.005) and kokanee (p <0.003) were not normally distributed. Results of the Mann-Whitney U test showed that sockeye and kokanee differed significantly from one another in both δ13C and δ15N values (p <0.0001).

In order to assess whether individual specimens of unknown ecotype could be assigned to either sockeye or kokanee ecotypes according to their δ13C values, we used the ‘single-case’ test, which can test whether a single value comes from the same population as a given sample. With this test we evaluated whether values from individual specimens in the unknown group were statistically more similar to values in sockeye or kokanee baseline groups using a modified t-test. While this test is imperfect because it assumes that the individual value given is an estimate, with variance and standard deviation similar to the sample with which it is being compared, it is a useful way to assess relationships between individual and sample values [123, 124].

Ancient DNA analysis

Our isotopic analysis included 25 archaeological salmonid vertebrae from two archaeological sites located within the Interior Plateau of British Columbia, Canada, in Xeni Gwet’in territory dating to between ca. 2400 and 50 years before present [125]. This sample consisted of 8 vertebrae from the Plateau Pithouse Tradition-associated Shields site (EkSa-13), and 17 vertebrae from the ancestral Tsilhqot’in Bear Lake site (EkSa-36) (Fig 1). These vertebrae were preliminarily identified as O. nerka (n = 24) or Oncorhynchus sp. (n = 1; much larger) based on their size. However, as speciating Pacific salmonid vertebrae through non-biomolecular methods is difficult (See Section Species identification of archaeological Pacific salmonid remains), we used aDNA analysis to confirm species identifications and obtain sex identifications [70, 126] for a subset of the archaeological vertebrae (n = 10). Detailed descriptions of the methods employed in the aDNA analysis are provided in the supplementary materials (S2 Text).

All pre-PCR procedures (decontamination, DNA extraction, and PCR setup) were conducted at the Department of Archaeology at Simon Fraser University (Burnaby, BC, Canada), in a dedicated aDNA laboratory and followed strict contamination control protocols [127]. All of the analyzed samples were decontaminated prior to DNA extraction using established protocols [128]. DNA was extracted from the decontaminated samples using a silica-spin column method [129] as modified by Yang and colleagues [130]. Sex identities were assigned to the samples using two PCR sex identification assays (termed clock1a/sdY and D-loop/sdY) designed by Royle and colleagues [126]. Following Royle and colleagues [126], we sought to assign species identifications to the remains by sequencing a 249 bp fragment of the mitochondrial D-loop co-amplified as an internal positive control in one of the sex identification assays (D-loop/sdY assay). To confirm the species identifications assigned to the samples, we also sequenced a 168 bp fragment of cytochrome b (cytb), which was amplified in a singleplex PCR [70, 131]. D-loop was also amplified from a single sample (IUBC 5226) through a singleplex PCR in order to improve sequencing quality. To monitor for contamination, blank extraction controls were processed alongside the samples and negative PCR controls were included in each PCR run.

The obtained sequences were visually edited, truncated to remove the primer sequences, and compiled in ChromasPro v 2.1.8 (http://technelysium.com.au). Following Yang and colleagues [70], the obtained sequences were compared with Pacific salmonid reference sequences through BLASTn searches [132] against GenBank [133], and neighbour-joining trees constructed in MEGA X [134]. Species-level identifications were assigned to samples if the obtained cytb and D-loop sequences matched or closely resembled sequences from a single species and differed from closely related species [70].

Results

Isotopic analyses

The results of the isotopic analyses are presented in Fig 3 and Table 5 in S1 Text. We analyzed bone or scale collagen samples from 342 specimens, 92% (n = 314) of which produced acceptable collagen quality criteria. Our baseline, including data sourced from the literature, is comprised of 239 kokanee and 417 sockeye collected at 87 sampling localities (Fig 1). These sampling localities included sites in the North Pacific Ocean as well as lakes and rivers in Canada, Russia, and the United States of America (Fig 1). As expected based on their behavioural ecology and use of freshwater environments, kokanee (δ13C range = 11.8‰, between –29.9 and –18.1‰, mean = –24.3‰; δ15N range = 12.1‰, between +4.1 and +16.2‰, mean = +9.0‰) have a much wider range of isotopic compositions relative to sockeye (δ13C range = 4.8‰, between –18.4 and –13.6 ‰, mean = –16.3‰; δ15N range = 6.9‰, between +7.1 and +14.0‰, mean = +11.0‰). While the food webs that kokanee inhabit are subject to a highly diverse set of carbon sources and carbon cycling processes (for review see [31]), those of marine sockeye δ13C appear to be much less variable.

[Figure omitted. See PDF.]

Fig 3. Stable carbon and nitrogen isotope compositions of O. nerka samples.

https://doi.org/10.1371/journal.pone.0232180.g003

Notably, the ranges for δ13C of sockeye and kokanee collagen overlap by only 0.3‰. Kokanee samples with δ13C values falling within the range observed for sockeye (i.e., > –18.4‰) are all derived from a population that inhabited a single system, the Alouette Reservoir, British Columbia, Canada (see Table 5 in S1 Text for details), in which 13C-enriched carbon sources or environmental variables have contributed to high δ13C baseline values. The Alouette Reservoir ecosystem is part of a nutrient addition program, aimed at supporting its kokanee population [135, 136], and the increased productivity resulting from the addition of agricultural fertilizers could be responsible for the higher δ13C values observed in kokanee from this location. It is important to note, however, that the freshwater Suess effect corrections (see Section Suess effect correction) applied to this population (+2.1‰; for comparison, marine Suess correction is only +1.1‰ for the same year) are also large. In not accounting for the possibility of old carbon sources (i.e., from non-atmospheric DIC and reservoir offsets), δ13C for the Alouette Reservoir kokanee has potentially been overcorrected, thereby pushing them into the baseline range for sockeye. Nonetheless, the Alouette Reservoir specimens demonstrate that, while atypical, it is possible for kokanee and sockeye to have similar δ13C values.

Despite slight overlap between ecotypes, kokanee and sockeye δ13C values are significantly different (Mann Whitney U, p <0.0001) at a broad temporal and geographical scale, suggesting that these data can provide a robust baseline for distinguishing between these O. nerka ecotypes on the basis of collagen δ13C values. Using δ13C values, the single-case test found that all values from the unknown ecotype group that were equal to or lower than –19‰ were statistically similar to the kokanee baseline group, while values of –18‰ or greater were statistically similar to the sockeye baseline group. Therefore, the test confirmed visual inspection of the data suggesting that O. nerka with collagen δ13C values either side of –18 to –19‰ are broadly distinguishable as either the sockeye or kokanee ecotypes. This test also confirmed that collagen δ13C values between –18 to –19‰ cannot be reliably assigned to either ecotype.

Based on O. nerka behavioural ecology as well as the ways in which nitrogen is sourced and cycled through aquatic ecosystems [31], we had not anticipated that δ15N would provide a viable basis for distinguishing between freshwater and marine behavioural types. Due to a stepwise enrichment of 3–4‰ between trophic levels [37], δ15N is typically associated with trophic position in ecological and archaeological studies [38, 39]. Adult size for kokanee can range widely between populations and size variation could offer a possible partial explanation for the wide range observed here in kokanee δ15N, with larger individuals feeding at a higher trophic level. However, the δ15N range observed in kokanee, which would be consistent with a span of at least two to three trophic levels, is not consistent with what is known about the dietary behaviour of this species, which are typically planktivores (although sockeye have also been noted to sometimes eat small fish and squid [137]). Instead, a more parsimonious explanation would be relative differences in baseline δ15N for food webs in the freshwater environments inhabited by each kokanee population. The δ15N compositions of dissolved inorganic nitrogen that is assimilated by primary producers at the base of freshwater food webs is subject to a much wider range of processes (relative to DIC) that alter nitrogen isotopic compositions (for reviews see [31, 41, 138]). In this context, given the more recent collection date and location (interior regions that may have been logged or subjected to substantial urban development), it seems likely that, rather than feeding at a much higher trophic level, elevated δ15N for certain kokanee populations is the result of higher local δ15N baselines that have been altered through twentieth and twenty-first century anthropogenic activities [139, 140]. Therefore, while these δ15N data are not considered as a means of distinguishing between ecotypes for the purposes of our study, we include them here as they may be useful for future research on anthropogenic impacts such as urbanization and deforestation.

Ancient DNA analysis

Table 1 summarizes the results of the aDNA analysis. Both cytb and D-loop were successfully amplified from each of the 10 samples included in the analysis. The results of the BLASTn searches and phylogenetic analyses indicate the cytb and D-loop sequences obtained from 9 of the 10 samples matched or closely resembled O. nerka references sequences (Figs 1 and 2 in S2 Text). Consequently, these samples were confidently identified as O. nerka (Table 1). The remaining sample, IUBC 5207 (ELS8), was identified as coho salmon (O. kisutch) as it yielded cytb and D-loop sequences that matched reference sequences from that species (Table 1; Figs 1 and 2 in S2 Text). This finding is consistent with the larger size of this vertebral specimen, relative to the others analyzed here, as well as its unusual isotopic composition, falling outside of the observed range for O. nerka δ13C (–13.6‰) and δ15N (+14.8‰) (Fig 1 in S1 Text). As sample IUBC 5207 was not identified as O. nerka, it was excluded from further analysis and discussion. DNA was amplified with both sex identification assays from 9 of the 10 salmonid remains. In total, 4 samples were identified as male and 4 were identified as female (Table 1). Two samples could not be assigned to a sex due to the assays generating conflicting results or one of the assays failing to amplify DNA (Table 1). No DNA was amplified from any of the blank extraction or negative PCR controls, indicating a lack of systematic contamination. The cytb and D-loop sequences obtained from each of the samples have been deposited in GenBank under accession numbers MN598987 to MN599006.

[Figure omitted. See PDF.]

Table 1. Genetic species and sex identification results for the archaeological samples included in the ancient DNA analysis.

https://doi.org/10.1371/journal.pone.0232180.t001

Discussion

To demonstrate the utility of this approach, we undertook three case studies illustrating how this O. nerka δ13C baseline can be used in a range of research fields including archaeology, natural history, and conservation biology.

Archaeology

Aquatic resources, and fish in particular, have been of tremendous importance both dietarily and culturally throughout much of the course of humanity’s evolution [141]. Consequently, understanding past fishing strategies—what, where, and how particular species were exploited—has long been a prominent archaeological research theme [142–144]. This is particularly true in North America’s Pacific Northwest where salmon and other fish have played a vital role in human cultural and dietary practices since the Late Pleistocene [12, 53, 55]. While numerous studies have demonstrated the importance of O. nerka to Indigenous fisheries in the Pacific Northwest [e.g., 59, 64, 66, 68], clear faunal evidence for the relative importance of sockeye versus kokanee has remained elusive.

The lack of information concerning the relative importance of sockeye and kokanee to Indigenous fisheries in the Pacific Northwest partially reflects methodological difficulties surrounding the differentiation of these ecotypes. As noted above (see Section Species identification of archaeological Pacific salmonid remains), Pacific salmonid vertebrae and most other elements cannot be confidently assigned to a species, let alone ecotype, through conventional morphology-based zooarchaeological approaches or a variety of other approaches. While useful for species identification, aDNA analysis and ZooMS likewise have limitations that hamper their ability to differentiate archaeological sockeye and kokanee specimens. In modern contexts, microsatellite and nuclear SNP loci are often used to assign O. nerka to populations [e.g., 145–147]. However, due to DNA degradation, the PCR amplification of these markers from archaeological specimens is prone to high rates of drop-out or is often not possible [e.g., 128, 148]. Although high-throughput sequencing methods are improving capacity for recovering genome-wide data useful for population assignment from ancient fish remains [149], these methods are expensive relative to isotopic approaches. Moreover, even when genome-wide data is available, genetically assigning bones from extinct populations to the correct ecotype may be difficult due to their polyphyly [150, 151]. Since they do not represent distinct lineages, genetically assigning individuals to the correct ecotype requires genetic reference data from local populations of the different ecotypes, which may be unavailable for extinct populations. In the case of ZooMS, type 1 collagen’s relatively slow rate of evolution amongst salmonids (ca. 6–12 amino acid substitutions per million years in Oncorhynchus) [71, 72] means recently diverged ecotypes, such as kokanee and sockeye (which diverged post-glacially [151]), cannot be differentiated by peptide mass fingerprinting techniques [e.g., 139, 152].

The applicability of our δ13C approach for assessing the ecotype of archaeological specimens requires a caveat. At the extreme end of their growth range, sockeye smolts can overlap in size with adult kokanee [153] and, because both have freshwater life histories, should therefore be indistinguishable based on δ13C. As adults and smolts can typically be differentiated morphologically, this issue would not present a problem for archival or modern ecological research. However, in cases where archaeological O. nerka bones are particularly small, this δ13C approach may not always be able to distinguish between adult kokanee and juvenile sockeye. Nevertheless, this approach would still provide evidence for the relative importance of freshwater versus marine resource use.

Based on the δ13C composition of archaeological bone collagen (Fig 3), we were able to assign all of the O. nerka specimens from the Shields and Bear Lake sites to the anadromous (sockeye) ecotype. Previous interpretations of whether these bones represent evidence of fishing strategies targeting kokanee, sockeye, both ecotypes, or other salmonids were inconclusive, in part because of the morphological and size overlap between vertebrae from different salmonids [154]. However, due to these sites’ proximity to Choelquoit Lake (known locally as Eagle or Big Eagle Lake), a known kokanee habitat, it was hypothesized that these bones might be derived from a kokanee fishery, assuming the ecotype also historically inhabited the lake [125, 154]. Our isotopic data not only provide novel evidence for the targeting of anadromous fish by Plateau Pithouse Tradition and Tsilhqot’in fisheries but also have broader implications for our understanding of salmon fisheries in the British Columbia Interior Plateau. In the case of the Bear Lake samples, our data contribute to the long-standing debate regarding the extent to which Tsilhqot’in fisheries were oriented towards resident lacustrine taxa as opposed to anadromous salmon [125, 155, 156]. Although our sample size is small, the identification of only sea-run sockeye at both sites do not support Matson and Magne’s [125] hypothesis that Tsilhqot’in fisheries may have been more freshwater-lake oriented than those of Plateau Pithouse Tradition peoples. At least in the case of O. nerka, our data indicate Tsilhqot’in and Plateau Pithouse Tradition fisheries may have targeted anadromous O. nerka to a similar degree. These data support Tyhurst’s [156] hypothesis that anadromous salmon were an important food resource for the Tsilhqot’in. At the same time, our data provide new support for early isotopic analyses of human remains that suggested people in the Chilcotin region had consumed large amounts of marine protein during the Late Holocene [157].

While interpreting the relative use of anadromous and potamodromous fish in the archaeological past, it is also important to remain cognizant of the broader range of economic, social, and sensorial factors that may have been taken into consideration when choosing between available fish stocks. For instance, there are significant differences in body size [153] and flavour profile [158, 159] between sockeye and kokanee. Moreover, these data could have further implications for reconstructing the seasonality of past fisheries because kokanee and sockeye migrate to spawn at slightly different times [137].

Our data also highlight the utility of conducting combined isotopic and aDNA analysis on single ancient fish bones. By applying both our δ13C approach and aDNA analysis to single vertebrae, we were able to generate more information about a single specimen than is possible with either method in isolation. While δ13C data provided information about archaeological specimens’ migratory behaviour, we were able to assign specimens to both a species and sex through aDNA analysis; information that contributes to a more nuanced understanding of past human fishing strategies. Moreover, through additional genetic analyses of ancient specimens isotopically identified as sockeye or kokanee, tracking natural and anthropogenic changes in the genetic diversity [e.g., 160] and phenotypic variation [e.g, 161] of these ecotypes is possible. Ultimately, by conducting multiple analyses on a single sample, this combined aDNA-isotopic approach will reduce the need for destructive sampling of different specimens for individual biomolecular analyses.

Natural history collections

Globally, fish and other biological specimens archived in natural history collections represent an invaluable resource for reconstructing past species’ behaviours and environmental conditions [162–164]. The earliest specimens can be especially valuable for providing baseline evidence for species’ genetics, behaviour, and environmental conditions prior to the onset of major impacts from industrialization. However, early specimens may have lost their associated provenance information, or may have been collected at a time when fewer contextual details were typically recorded. In this context, methods for reconstructing aspects of a specimen’s life history may provide a useful tool for re-establishing key provenance or contextual details.

While collecting scales from sockeye and kokanee at museums, we observed a number of O. nerka specimens that did not have associated information regarding their ecotype or had limited provenance details. Some of these specimens were large enough to have been either kokanee or sockeye and many were collected in regions where these ecotypes are sympatric. We used scale δ13C from these behaviourally ambiguous O. nerka specimens to provide improved life history information for museum records. In total, we analyzed scales from 20 fish collected between 1926 and 1960 from 9 locations along a large latitudinal transect (Figs 1 and 3). Through our analyses, we were able to assign all specimens to either kokanee (n = 17) or sockeye (n = 3) behavioural types.

A number of specimens produced notable results. For instance, two fish from Osoyoos Lake collected in 1956 (IUBC 4380 and 4381) had museum labels suggesting that they were kokanee but were categorized more cautiously here as “unknowns” due to their large size (ca. 40 cm) for the drainage as well as a handwritten postscript in associated field notes that led to questions about certainty of their initial identification. At the time that these specimens were collected it was thought that human impacts on water flow in rivers in this region limited sockeye returns, a fact that may have influenced the biologists’ choice in identifying these specimens. Stable carbon isotope values indicate that these fish were, in fact, anadromous sockeye and provide a more complete record for these historical specimens, enhancing their potential value for future research. Application of this technique to additional museum specimens may identify extinct populations of sockeye or kokanee, allowing for the historic distribution of these ecotypes to be reconstructed.

Conservation

Worldwide, freshwater environments are a key conservation focus due to their high biodiversity as well as their vulnerability to overfishing [165] and other human impacts [166]. In coastal regions of the North Pacific Ocean, kokanee and sockeye are a major focus of fisheries conservation efforts. In many places, monitoring programs are used to track the health and progress of rehabilitating O. nerka populations at a variety of scales and using a wide range of techniques. In areas where sockeye and kokanee populations are sympatric, having an independent and non-invasive (in the case of scales) means of distinguishing between these ecotypes may be important. In the past, genetic methods have been successfully used to differentiate kokanee and sockeye [147]. However, the genetic differentiation of kokanee from sockeye is confounded by the fact that these ecotypes not only interbreed [147] but can also undergo intergenerational shifts between anadromy and potamodromy [22]. In this context, stable carbon isotope analyses of collagen from O. nerka can provide a simple alternative method for establishing whether a fish was anadromous or potamodromous.

We measured the δ13C of bone and scale collagen from modern specimens collected during the 2018 spawning run on the Okanagan River between Skaha and Okanagan Lakes in British Columbia (Fig 1). Skaha Lake’s resident kokanee population has received long-term monitoring from fisheries management agencies and in 2003 was stocked with sockeye as part of conservation efforts aimed at reintroducing the anadromous ecotype to the region [167]. In this context, our δ13C approach could provide independent validation of other methods and would be particularly important if other methods result in uncertainty or conflicting estimates as may be the case when anadromous or potamodromous populations mix. Based on bone collagen δ13C, all 13 O. nerka specimens analyzed were identified as either potamodromous (n = 5) or anadromous (n = 8). A comparison of these isotopic ecotype identifications with forthcoming genetic assignments will provide a cost-effective (and therefore easily scalable) means for monitoring the extent to which ecotype information from genetic analyses reflects O. nerka migratory behaviour.

Conclusion

Using the collagen δ13C baseline presented here, each case study has demonstrated some of the potential for using this approach for differentiating sockeye (anadromous) and kokanee (potamodromous) populations across the North Pacific Ocean and its coastal regions. Not only is this approach potentially less invasive (in using scales rather than otoliths) it relies on δ13C analyses, a well-established technique that is widely available at many commercial laboratories at a fraction of the cost of other methods. For this reason, our approach could provide a widely applicable alternative to conventional methods for ecologists and conservationists interested in identifying ecotypes for large numbers of O. nerka. Moreover, it could be possible to adapt this approach to other taxa as well provided they meet behavioural and environmental criteria (see Section Stable car bon isotopes in aquatic environments). For instance, other species for which the presence and relative importance of freshwater versus marine ecotypes could be discerned include cherry salmon (Oncorhynchus masou), cutthroat trout (Oncorhynchus clarkii), and rainbow trout (Oncorhynchus mykiss) in the Pacific Ocean; Atlantic salmon and brown trout (Salmo trutta) in the Atlantic Ocean; and Arctic char (Salvelinus alpinus) in the Arctic Ocean.

Given the considerable variability in freshwater food web δ13C, as this study shows, it is critical that the approach is ‘ground truthed’ with a large and geographically expansive baseline of δ13C data from individuals of known anadromous and potamodromous ecotypes. This is essential in order to confirm that δ13C of fish scale and bone collagen from freshwater environments within a given study region do not significantly overlap with their marine conspecifics. For retrospective applications, it is also important to remain cognizant of the potential for temporal shifts in aquatic δ13C baselines, particularly in freshwater ecosystems. For instance, as shown by samples from the Alouette Reservoir it is possible for environmental conditions (particularly those impacted by cultural eutrophication) to push δ13C baselines for freshwater specimens into the range observed for marine fish. Due to the context dependent nature of the factors that control aquatic δ13C baselines, it may be difficult or impossible to fully assess these kinds of issues a priori for past environments. Nonetheless, considering some of the key axes along which δ13C is known to vary in modern environments (e.g., carbon sources and cycling; for review see [31]) in the context of other relevant paleoenvironmental indicators, such as diatom, elemental, and isotopic records from sediment cores, may help to constrain interpretations.

This collagen δ13C approach to the differentiation of fish anadromy and potamodromy should also be particularly useful for archaeologists interested in understanding the relative importance for freshwater and marine resources, both for exploring ancient fisheries as well as generating more realistic isotopic baselines for paleodietary reconstructions using human isotopic compositions. While ZooMS and aDNA analysis offer a means of establishing from which taxon a bone originates, they are often unable to assign archaeological fish remains to an ecotype or do so in a cost-effective manner. While it is possible to establish this information using elemental concentration and isotopic (e.g., strontium and sulfur; see Section Introduction) analyses of otoliths, these elements are archaeologically rare and, relative to measuring the δ13C composition of archaeological bone collagen, require a more costly and laborious analytical process. Considering the broad significance of salmon and other anadromy-flexible species to human subsistence and cultural practices across the globe, wider adaptation of this approach to include other species could provide novel details about the seasonality of archaeological site use, clues about past fishing technologies, and possibly even reveal new evidence for trade in fish products.

Supporting information

[Figure omitted. See PDF.]

S1 Text. Supporting information for the stable isotope analysis.

https://doi.org/10.1371/journal.pone.0232180.s001

(PDF)

S2 Text. Supporting information for the ancient DNA analysis.

https://doi.org/10.1371/journal.pone.0232180.s002

(PDF)

Acknowledgments

Sampling permissions and logistical assistance: Gavin Hanke and the Royal British Columbia Museum, Erling Holm and the Royal Ontario Museum, Eric Taylor and the Beaty Biodiversity Museum, Chief Jimmy Lulua and the Xeni Gwet’in First Nation, Dylan Hilliis (University of Victoria). Data: Many scholars provided access to their previously published raw data: Joachim Molkentin (Max Rubner-Institute), David Beauchamp (United States Geological Survey Western Fisheries Research Center). We also thank the following peer reviewers: Rachel C. Johnson (National Oceanic and Atmospheric Administration and University of California, Davis) and Malte Willmes (National Oceanic and Atmospheric Administration and University of California, Santa Cruz), as well as Adam Boethius (Lund University).

Citation: Guiry E, Royle TCA, Matson RG, Ward H, Weir T, Waber N, et al. (2020) Differentiating salmonid migratory ecotypes through stable isotope analysis of collagen: Archaeological and ecological applications. PLoS ONE 15(4): e0232180. https://doi.org/10.1371/journal.pone.0232180

References

1. Hendry AP. To sea or not to sea: Anadromy versus non-anadromy in salmonids. In: Stearns SC, Stearns SC, Hendry AP, editors. Evolution illuminated: Salmon and their relatives. Oxford: Oxford University Press; 2003. pp. 92–125.

2. Dodson JJ, Aubin‐Horth N, Thériault V, Páez DJ. The evolutionary ecology of alternative migratory tactics in salmonid fishes. Biol Rev Camb Philos Soc. 2013;88(3):602–625. pmid:23347290

3. Northcote T. Migration and residency in stream salmonids-Some ecological considerations and evolutionary consequences. Nord J Freshw Res. 1992;67:5–17.

4. Acolas M-L, Lambert P. Life histories of anadromous fishes. In: Morais P, Daverat F, editors. An introduction to fish migration. Boca Raton: CRC Press; 2016. pp. 55–77.

5. Thurow RF. Life histories of potamodromous fishes. In: Morais P, Daverat F, editors. An introduction to fish migration. Boca Raton: CRC Press; 2016. pp. 29–54.

6. Chapman BB, Brönmark C, Nilsson J-Å, Hansson L-A. The ecology and evolution of partial migration. Oikos. 2011;120(12):1764–1775.

7. Hesslein R, Capel M, Fox D, Hallard K. Stable isotopes of sulfur, carbon, and nitrogen as indicators of trophic level and fish migration in the lower Mackenzie River basin, Canada. Can J Fish Aquat Sci. 1991;48(11):2258–65.

8. Elsdon TS, Wells BK, Campana SE, Gillanders BM, Jones CM, Limburg KE, et al. Otolith chemistry to describe movements and life-history parameters of fishes: Hypotheses, assumptions, limitations and inferences. Oceanogr Mar Biol. 2008;46:297–330.

9. Blair AA. Scales of Lake Ontario salmon indicate a land-locked form. Copeia. 1938;4:206.

10. Guiry E, Karavanić I, Klindžić RŠ, Talamo S, Radović S, Richards MP. Stable isotope palaeodietary and radiocarbon evidence from the Early Neolithic site of Zemunica, Dalmatia, Croatia. Eur J Archaeol. 2017;20(2):235–256.

11. Robson HK, Andersen SH, Clarke L, Craig OE, Gron KJ, Jones AKG, et al. Carbon and nitrogen stable isotope values in freshwater, brackish and marine fish bone collagen from Mesolithic and Neolithic sites in central and northern Europe. Environ Archaeol. 2015; 21(2):105–118.

12. Halffman CM, Potter BA, McKinney HJ, Finney BP, Rodrigues AT, Yang DY, et al. Early human use of anadromous salmon in North America at 11,500 y ago. Proc Natl Acad Sci U S A. 2015;112(40):12344–12348. pmid:26392548

13. Miller JA. Effects of water temperature and barium concentration on otolith composition along a salinity gradient: Implications for migratory reconstructions. J Exp Mar Biol Ecol. 2011;405(1–2):42–52.

14. Kennedy BP, Folt CL, Blum JD, Chamberlain CP. Natural isotope markers in salmon. Nature. 1997;387(6635):766.

15. Ingram BL, Weber PK. Salmon origin in California’s Sacramento–San Joaquin river system as determined by otolith strontium isotopic composition. Geology. 1999;27(9):851–854.

16. Godbout L, Trudel M, Irvine JR, Wood CC, Grove MJ, Schmitt AK, et al. Sulfur isotopes in otoliths allow discrimination of anadromous and non-anadromous ecotypes of sockeye salmon (Oncorhynchus nerka). Environ Biol Fishes. 2010;89(3–4):521–532.

17. Zazzo A, Smith GR, Patterson WP, Dufour E. Life history reconstruction of modern and fossil sockeye salmon (Oncorhynchus nerka) by oxygen isotopic analysis of otoliths, vertebrae, and teeth: Implication for paleoenvironmental reconstructions. Earth Planet Sci Lett. 2006;249(3–4):200–215.

18. Koch PL, Halliday AN, Walter LM, Stearley RF, Huston TJ, Smith GR. Sr isotopic composition of hydroxyapatite from recent and fossil salmon: The record of lifetime migration and diagenesis. Earth Planet Sci Lett. 1992;108(4):277–287.

19. Robinson BS, Jacobson GL, Yates MG, Spiess AE, Cowie ER. Atlantic salmon, archaeology and climate change in New England. J Archaeol Sci. 2009;36(10):2184–2191.

20. Dufour E, Holmden C, Van Neer W, Zazzo A, Patterson WP, Degryse P, et al. Oxygen and strontium isotopes as provenance indicators of fish at archaeological sites: The case study of Sagalassos, SW Turkey. J Archaeol Sci. 2007;34(8):1226–1239.

21. Disspain MC, Ulm S, Gillanders BM. Otoliths in archaeology: Methods, applications and future prospects. Journal of Archaeol Sci Rep. 2016;6:623–632.

22. Godbout L, Wood CC, Withler RE, Latham S, Nelson RJ, Wetzel L, et al. 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(9):1590–1602.

23. Cook PK, Dufour E, Languille M-A, Mocuta C, Réguer S, Bertrand L. Strontium speciation in archaeological otoliths. J Anal At Spectrom. 2016;31(3):700–711.

24. Hoppe KA, Koch PL, Furutani TT. Assessing the preservation of biogenic strontium in fossil bones and tooth enamel. Int J Osteoarchaeol. 2003;13(1‐2):20–28.

25. Baker RF, Blanchfield PJ, Paterson MJ, Flett RJ, Wesson L. Evaluation of nonlethal methods for the analysis of mercury in fish tissue. Trans Am Fish Soc. 2004;133(3):568–576.

26. Sanderson BL, Tran CD, Coe HJ, Pelekis V, Steel EA, Reichert WL. Nonlethal sampling of fish caudal fins yields valuable stable isotope data for threatened and endangered fishes. Trans Am Fish Soc. 2009;138(5):1166–1177.

27. Kelly MH, Hagar WG, Jardine TD, Cunjak RA. Nonlethal sampling of sunfish and slimy sculpin for stable isotope analysis: How scale and fin tissue compare with muscle tissue. N Am J Fish Manag. 2006;26(4):921–925.

28. Carlson CC, Klein K. Late Pleistocene salmon of Kamloops Lake. In: Ludvigsen R, editor. Life in stone: A natural history of British Columbia’s fossils. Vancouver: The Univiersity of British Columbia Press; 1996. pp. 274–80.

29. Doucett RR, Power M, Power G, Caron F, Reist JD. Evidence for anadromy in a southern relict population of Arctic charr from North America. J Fish Biol. 1999;55(1):84–93.

30. Doucett RR, Hooper W, Power G. Identification of anadromous and nonanadromous adult brook trout and their progeny in the Tabusintac River, New Brunswick, by means of multiple-stable-isotope analysis. Trans Am Fish Soc. 1999;128(2):278–288

31. Guiry EJ. Complexities of stable carbon and nitrogen isotope biogeochemistry in ancient freshwater ecosystems: Implications for the study of past subsistence and environmental change. Front Ecol Evol. 2019;7: 313

32. Finlay JC. Patterns and controls of lotic algal stable carbon isotope ratios. Limnol Oceanogr. 2004;49(3):850–861.

33. Finlay JC. Controls of streamwater dissolved inorganic carbon dynamics in a forested watershed. Biogeochemistry. 2003;62(3):231–252

34. Finlay JC, Kendall C. Stable isotope tracing of temporal and spatial variability in organic matter sources to freshwater ecosystems. In: Michener RH, Lajtha K, editors. Stable isotopes in ecology and environmental science. 2nd ed. New York: Blackwell; 2007. pp. 283–333.

35. Ishikawa NF, Doi H, Finlay JC. Global meta-analysis for controlling factors on carbon stable isotope ratios of lotic periphyton. Oecologia. 2012;170(2):541–549. pmid:22466861

36. Rau GH, Takahashi T, Des Marais DJ. Latitudinal variations in plankton δ13C: Implications for CO2 and productivity in past oceans. Nature. 1989;341(6242):516–518. pmid:11536614

37. DeNiro MJ, Epstein S. Influence of diet on the distribution of nitrogen isotopes in animals. Geochim Cosmochim Acta. 1981;45(3):341–351.

38. Post DM. Using stable isotopes to estimate trophic position: Models, methods, and assumptions. Ecology. 2002;83(3):703–718.

39. Caut S, Angulo E, Courchamp F. Variation in discrimination factors (Δ15N and Δ13C): The effect of diet isotopic values and applications for diet reconstruction. J Appl Ecol. 2009;46(2):443–453.

40. Sigman DM, Karsh KL, Casciotti KL. Nitrogen isotopes in the ocean. In: Steele JH, editor. Encyclopedia of ocean science. 2nd ed. San Diego: Academic Press; 2009. pp. 40–54.

41. Kendall C, Elliott EM, Wankel SD. Tracing anthropogenic inputs of nitrogen to ecosystems. In: Michener RH, Lajtha K, editors. Stable isotopes in ecology and environmental science. 2nd ed. New York: Blackwell; 2007. pp. 375–449.

42. Selbie DT, Finney BP, Barto D, Bunting L, Chen G, Leavitt PR, et al. Ecological, landscape, and climatic regulation of sediment geochemistry in North American sockeye salmon nursery lakes: Insights for paleoecological salmon investigations. Limnol Oceanogr. 2009;54(5):1733–1745.

43. Vander Zanden MJ, Clayton MK, Moody EK, Solomon CT, Weidel BC. Stable isotope turnover and half-life in animal tissues: A literature synthesis. PLoS ONE. 2015;10(1):e0116182. pmid:25635686

44. Hobson KA, Clark RG. Assessing avian diets using stable isotopes I: Turnover of 13C in tissues. Condor. 1992;94(1):181–188.

45. Ambrose SH, Norr L. Experimental evidence for the relationship of the carbon isotope ratios of whole diet and dietary protein to those of bone collagen and carbonate. In: Lambert JB, Grupe G, editors. Prehistoric human bone. Berlin: Springer; 1993. pp. 1–37.

46. Hedges REM, Clement JG, Thomas CDL, O’Connell TC. Collagen turnover in the adult femoral mid-shaft: Modeled from anthropogenic radiocarbon tracer measurements. Am J Phys Anthropol. 2007;133(2):808–816. pmid:17405135

47. Matsubayashi J, Tayasu I. Collagen turnover and isotopic records in cortical bone. J Archaeol Sci. 2019;106:37–44.

48. Guiry EJ, Hunt BPV. Integrating fish scale and bone isotopic compositions for ‘deep time’ retrospective studies. Mar Environ Res. 2020:104982.

49. Trueman CN, Moore A. Use of the stable isotope composition of fish scales for monitoring aquatic ecosystems. Terr Ecol. 2007;1:145–161.

50. Wainright SC, Fogarty MJ, Greenfield RC, Fry B. Long-term changes in the Georges Bank food web: Trends in stable isotopic compositions of fish scales. Mar Biol. 1993;115(3):481–493.

51. Dixon HJ, Power M, Dempson JB, Sheehan TF, Chaput G. Characterizing the trophic position and shift in Atlantic salmon (Salmo salar) from freshwater to marine life-cycle phases using stable isotopes. ICES J Mar Sci. 2012;69(9):1646–1655.

52. Bump JK, Fox-Dobbs K, Bada JL, Koch PL, Peterson RO, Vucetich JA. Stable isotopes, ecological integration and environmental change: Wolves record atmospheric carbon isotope trend better than tree rings. Proc Biol Sci. 2007;274(1624):2471–2480. pmid:17686730

53. Butler VL, Campbell SK. Resource intensification and resource depression in the Pacific Northwest of North America: A zooarchaeological review. J World Prehist. 2004;18(4):327–405.

54. Matsui A. Archaeological investigations of anadromous salmonid fishing in Japan. World Archaeol. 1996;27(3):444–460.

55. McKechnie I, Moss ML. Meta-analysis in zooarchaeology expands perspectives on Indigenous fisheries of the Northwest Coast of North America. J Archaeol Sci Rep. 2016;8:470–485.

56. Butler VL, Chatters JC. The role of bone density in structuring prehistoric salmon bone assemblages. J Archaeol Sci. 1994;21(3):413–424.

57. Cannon DY. Marine fish osteology: A manual for archaeologists. Burnaby: Archaeology Press, Simon Fraser University; 1987.

58. Butler VL, Bowers NJ. Ancient DNA from salmon bone: A preliminary study. Anc Biomol. 1998;2(1):17–26.

59. Cannon A, Yang DY. Early storage and sedentism on the Pacific Northwest Coast: Ancient DNA analysis of salmon remains from Namu, British Columbia. Am Antiq. 2006;71(1):123–140.

60. Huber HR, Jorgensen JC, Butler VL, Baker G, Stevens R. Can salmonids (Oncorhynchus spp.) be identified to species using vertebral morphometrics?. J Archaeol Sci. 2011;38(1):136–146.

61. Cannon A. Radiographic age determination of Pacific salmon: Species and seasonal inferences. J Field Archaeol. 1988;15(1):103–108.

62. Orchard TJ, Szpak P. Identification of salmon species from archaeological remains on the Northwest Coast. In: Moss ML, Cannon A, editors. The archaeology of North Pacific fisheries. Fairbanks: University of Alaska Press; 2011. pp. 17–29.

63. Orchard TJ. Late Holocene fisheries in Gwaii Haanas: Species composition, trends in abundance, and environmental or cultural explanations. In: Moss ML, Cannon A, editors. The archaeology of North Pacific fisheries. Chicago: University of Chicago Press; 2011. pp. 111–128.

64. Speller CF, Yang DY, Hayden B. Ancient DNA investigation of prehistoric salmon resource utilization at Keatley Creek, British Columbia, Canada. J Archaeol Sci. 2005;32(9):1378–1389.

65. Moss ML, Judd KG, Kemp BM. Can salmonids (Oncorhynchus spp.) be identified to species using vertebral morphometrics? A test using ancient DNA from Coffman Cove, Alaska. J Archaeol Sci. 2014;41:879–889.

66. Grier C, Flanigan K, Winters M, Jordan LG, Lukowski S, Kemp BM. Using ancient DNA identification and osteometric measures of archaeological Pacific salmon vertebrae for reconstructing salmon fisheries and site seasonality at Dionisio Point, British Columbia. J Archaeol Sci. 2013;40(1):544–555.

67. Hofkamp AR, Butler VL. On the validity of the radiographic method for determining age of ancient salmon. J Archaeol Sci Rep. 2017;12:449–456.

68. Cannon A, Yang D, Speller C. Site-specific salmon fisheries on the central coast of British Columbia. In: Moss ML, Cannon A, editors. The archaeology of North Pacific fisheries. Fairbanks: University of Alaska Press; 2011. pp. 57–74.

69. Ewonus PA, Cannon A, Yang DY. Addressing seasonal site use through ancient DNA species identification of Pacific salmon at Dionisio Point, Galiano Island, British Columbia. J Archaeol Sci. 2011;38(10):2536–2546.

70. Yang DY, Cannon A, Saunders SR. DNA species identification of archaeological salmon bone from the Pacific Northwest Coast of North America. J Archaeol Sci. 2004;31(5):619–631.

71. Korzow Richter K, McGrath K, Masson-MacLean E, Hickinbotham S, Tedder A, Britton K, et al. What’s the catch? Archaeological application of rapid collagen-based species identification for Pacific salmon. J Archaeol Sci. 2020;116:105116.

72. Buckley M. Zooarchaeology by mass spectrometry (ZooMS) collagen fingerprinting for the species identification of archaeological bone fragments. In: Giovas CM, LeFebvre MJ, editors. Zooarchaeology in practice. Cham: Springer; 2018. pp. 227–247.

73. Meeuwig MH, Peacock MM. Food web interactions associated with a Lahontan cutthroat trout reintroduction effort in an alpine lake. J Fish Wildl Manag. 2017;8(2):449–464.

74. Overman NC, Beauchamp DA, Berge HB, Mazur MM, McIntyre JK. Differing forage fish assemblages influence trophic structure in neighboring urban lakes. Trans Am Fish Soc. 2009;138(4):741–755.

75. Welch DW, Parsons TR. δ13C‐δ15N values as indicators of trophic position and competitive overlap for Pacific salmon (Oncorhynchus spp.). Fish Oceanogr. 1993;2(1):11–23.

76. Qin Y, Kaeriyama M. Feeding habits and trophic levels of Pacific salmon (Oncorhynchus spp.) in the North Pacific Ocean. N Pac Anadromous Fish Com Bul. 2016;6:469–481.

77. Wang YV, Wan AHL, Lock E-J, Andersen N, Winter-Schuh C, Larsen T. Know your fish: A novel compound-specific isotope approach for tracing wild and farmed salmon. Food Chem. 2018;256:380–389. pmid:29606463

78. Kaeriyama M, Nakamura M, Edpalina R, Bower JR, Yamaguchi H, Walker RV, et al. Change in feeding ecology and trophic dynamics of Pacific salmon (Oncorhynchus spp.) in the central Gulf of Alaska in relation to climate events. Fish Oceanogr. 2004;13(3):197–207.

79. Molkentin J, Lehmann I, Ostermeyer U, Rehbein H. Traceability of organic fish–Authenticating the production origin of salmonids by chemical and isotopic analyses. Food Control. 2015;53:55–66.

80. Guiry EJ, Szpak P, Richards MP. Effects of lipid extraction and ultrafiltration on stable carbon and nitrogen isotopic compositions of fish bone collagen. Rapid Commun Mass Spectrom. 2016;30(13):1591–1600. pmid:27321847

81. Satterfield FR IV, Finney BP. Stable isotope analysis of Pacific salmon: Insight into trophic status and oceanographic conditions over the last 30 years. Prog Oceanogr. 2002;53(2–4):231–246.

82. Horii S, Takahashi K, Furuya K. Effects of ethanol-preservation on stable carbon and nitrogen isotopic signatures in marine predators. Plankton Benthos Res. 2015;10(2):91–97.

83. Guiry EJ, Needs-Howarth S, Friedland KD, Hawkins AL, Szpak P, Macdonald R, et al. Lake Ontario salmon (Salmo salar) were not migratory: A long-standing historical debate solved through stable isotope analysis. Sci Rep. 2016;6: 36249. pmid:27824097

84. Longin R. New method of collagen extraction for radiocarbon dating. Nature. 1971;230(5291):241–242. pmid:4926713

85. Folch J, Lees M, Sloane-Stanley G. A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem. 1957;226(1):497–509. pmid:13428781

86. Qi H, Coplen TB, Mroczkowski SJ, Brand WA, Brandes L, Geilmann H, et al. A new organic reference material, l‐glutamic acid, USGS41a, for δ13C and δ15N measurements − A replacement for USGS41. Rapid Commun Mass Spectrom. 2016;30(7):859–866. pmid:26969927

87. Qi H, Coplen TB, Geilmann H, Brand WA, Böhlke JK. Two new organic reference materials for δ13C and δ15N measurements and a new value for the δ13C of NBS 22 oil. Rapid Commun Mass Spectrom. 2003;17(22):2483–2487. pmid:14608617

88. Szpak P, Metcalfe JZ, Macdonald RA. Best practices for calibrating and reporting stable isotope measurements in archaeology. J Archaeol Sci Rep. 2017;13:609–616.

89. Szpak P. Fish bone chemistry and ultrastructure: Implications for taphonomy and stable isotope analysis. J Archaeol Sci. 2011;38(12):3358–3372.

90. Regenstein JM, Zhou P. Collagen and gelatin from marine by-products. In: Shahidi F, editor. Maximising the value of marine by-products. Cambridge: Woodhead Publishing; 2007. pp. 279–303.

91. Turban-Just S, Schramm S. Stable carbon and nitrogen isotope ratios of individual amino acids give new insights into bone collagen degradation. Bull Soc Géol Fr. 1998;169(1):109–114.

92. DeNiro MJ. Postmortem preservation and alteration of in vivo bone collagen isotope ratios in relation to palaeodietary reconstruction. Nature. 1985;317(6040):806–809.

93. Van Klinken GJ. Bone collagen quality indicators for palaeodietary and radiocarbon measurements. J Archaeol Sci. 1999;26(6):687–695.

94. Keeling CD. The Suess effect: 13Carbon-14Carbon interrelations. Environ Int. 1979;2(4–6):229–300.

95. Eide M, Olsen A, Ninnemann US, Eldevik T. A global estimate of the full oceanic 13C Suess effect since the preindustrial. Global Biogeochem Cycles. 2017;31(3):492–514.

96. Hilton GM, Thompson DR, Sagar PM, Cuthbert RJ, Cherel Y, Bury SJ. A stable isotopic investigation into the causes of decline in a sub-Antarctic predator, the rockhopper penguin Eudyptes chrysocome. Glob Chang Biol. 2006;12(4):611–625.

97. Misarti N, Finney B, Maschner H, Wooller MJ. Changes in northeast Pacific marine ecosystems over the last 4500 years: Evidence from stable isotope analysis of bone collagen from archeological middens. Holocene. 2009;19(8):1139–1151.

98. Gruber N, Keeling CD, Bacastow RB, Guenther PR, Lueker TJ, Wahlen M, et al. Spatiotemporal patterns of carbon-13 in the global surface oceans and the oceanic Suess effect. Global Biogeochem Cycles. 1999;13(2):307–335.

99. Szpak P, Buckley M, Darwent CM, Richards MP. Long-term ecological changes in marine mammals driven by recent warming in northwestern Alaska. Glob Change Biol. 2018;24(1):490–503.

100. Quay PD, Tilbrook B, Wong CS. Oceanic uptake of fossil fuel CO2: Carbon-13 evidence. Science 1992;256(5053):74–79. pmid:17802595

101. Fera SA, Rennie MD, Dunlop ES. Broad shifts in the resource use of a commercially harvested fish following the invasion of dreissenid mussels. Ecology. 2017;98(6):1681–1692. pmid:28369860

102. Verburg P. The need to correct for the Suess effect in the application of δ13C in sediment of autotrophic Lake Tanganyika, as a productivity proxy in the Anthropocene. J Paleolimnol. 2007;37(4):591–602.

103. Cook GT, Bonsall C, Hedges REM, McSweeney K, Boronean V, Pettitt PB. A freshwater diet-derived 14C reservoir effect at the Stone Age sites in the Iron Gates gorge. Radiocarbon. 2001;43(2A):453–460.

104. Philippsen B. The freshwater reservoir effect in radiocarbon dating. Herit Sci. 2013;1(1):24.

105. Svyatko SV, Mertz IV, Reimer PJ. Freshwater reservoir effect on redating of Eurasian steppe cultures: First results for Eneolithic and Early Bronze Age northeast Kazakhstan. Radiocarbon. 2015;57(4):625–644.

106. Philippsen B, Heinemeier J. Freshwater reservoir effect variability in northern Germany. Radiocarbon. 2013;55(3):1085–1101.

107. Marchenko ZV, Orlova LA, Panov VS, Zubova AV, Molodin VI, Pozdnyakova OA, et al. Paleodiet, radiocarbon chronology, and the possibility of freshwater reservoir effect for Preobrazhenka 6 burial ground, western Siberia: Preliminary results. Radiocarbon. 2015;57(4):595–610.

108. Ascough PL, Cook GT, Church MJ, Dunbar E, Einarsson Á, McGovern TH, et al. Temporal and spatial variations in freshwater 14C reservoir effects: Lake Mývatn, northern Iceland. Radiocarbon. 2010;52(3):1098–1112

109. Keaveney EM, Reimer PJ. Understanding the variability in freshwater radiocarbon reservoir offsets: A cautionary tale. J Archaeol Sci. 2012;39(5):1306–1316.

110. Hart JP, Feranec RS, Abel TJ, Vavrasek JL. Freshwater reservoir offsets on radiocarbon-dated dog bone from the headwaters of the St. Lawrence River, USA. PeerJ. 2019;7:e7174. pmid:31275759

111. Johnson SP, Schindler DE. Four decades of foraging history: Stock-specific variation in the carbon and nitrogen stable isotope signatures of Alaskan sockeye salmon. Mar Ecol Prog Ser. 2012;460:155–167.

112. Coll YD. Stable isotope analysis of Rivers Inlet sockeye salmon (Oncorhynchus nerka): Investigating the contribution of environmental conditions in the high seas to British Columbia population declines. M.Sc. Thesis, University of British Columbia. 2015.

113. Espinasse B, Hunt BPV, Coll YD, Pakhomov EA. Investigating high seas foraging conditions for salmon in the North Pacific: Insights from a 100-year scale archive for Rivers Inlet sockeye salmon. Can J Fish Aquat Sci. 2019;76(6):918–927.

114. Bosley KL, Wainright SC. Effects of preservatives and acidification on the stable isotope ratios (15N:14N, 13C:12C) of two species of marine animals. Can J Fish Aquat Sci. 1999;56(11):2181–2185.

115. Sarakinos HC, Johnson ML, Vander Zanden MJ. A synthesis of tissue-preservation effects on carbon and nitrogen stable isotope signatures. Can J Zool. 2002;80(2):381–387.

116. Kelly B, Dempson JB, Power M. The effects of preservation on fish tissue stable isotope signatures. J Fish Biol. 2006;69(6):1595–1611.

117. Sweeting CJ, Polunin NV, Jennings S. Tissue and fixative dependent shifts of δ13C and δ15N in preserved ecological material. Rapid Commun Mass Spectrom. 2004;18(21):2587–2592. pmid:15468144

118. Akın Ş, Şahin C, Turan D, Gözler AM, Verep B, Bozkurt A, et al. Effects of preservation methods on the δ13C and δ15N signatures is muscle of two freshwater fish species. Fresenius Environ Bull. 2011;20(9a):2419–2426.

119. Edwards MS, Turner TF, Sharp ZD. Short-and long-term effects of fixation and preservation on stable isotope values (δ13C, δ15N, δ34S) of fluid-preserved museum specimens. Copeia. 2002;4:1106–1112.

120. Fanelli E, Cartes JE, Papiol V, Rumolo P, Sprovieri M. Effects of preservation on the δ13C and δ15N values of deep sea macrofauna. J Exp Mar Biol Ecol. 2010;395(1–2):93–97.

121. Kaehler S, Pakhomov EA. Effects of storage and preservation on the δ13C and δ15N signatures of selected marine organisms. Mar Ecol Prog Ser. 2001;219:299–304.

122. Hammer Ø, Harper DA, Ryan PD. PAST: Paleontological statistics software package for education and data analysis. Palaeontol Electron. 2001;4(1):9.

123. Crawford JR, Howell DC. Comparing an individual’s test score against norms derived from small samples. Clin Neuropsychol. 1998;12(4):482–486.

124. Sokal R, Rohlf F. Biometry. New York: W. H. H Freeman and Company; 1995.

125. Matson RG, Magne MPR. Athapaskan migrations: The archaeology of Eagle Lake, British Columbia: University of Arizona Press; 2007.

126. Royle TCA, Sakhrani D, Speller CF, Butler VL, Devlin RH, Cannon A, et al. An efficient and reliable DNA-based sex identification method for archaeological Pacific salmonid (Oncorhynchus spp.) remains. PLoS ONE. 2018;13(3):e0193212. pmid:29538397

127. Yang DY, Watt K. Contamination controls when preparing archaeological remains for ancient DNA analysis. J Archaeol Sci. 2005;32(3):331–336.

128. Speller CF, Hauser L, Lepofsky D, Moore J, Rodrigues AT, Moss ML, et al. High potential for using DNA from ancient herring bones to inform modern fisheries management and conservation. PLoS ONE. 2012;7(11):e51122. pmid:23226474

129. Yang DY, Eng B, Waye JS, Dudar JC, Saunders SR. Improved DNA extraction from ancient bones using silica-based spin columns. Am J Phys Anthropol. 1998;105(4):539–543. pmid:9584894

130. Yang DY, Liu L, Chen X, Speller CF. Wild or domesticated: DNA analysis of ancient water buffalo remains from north China. J Archaeol Sci. 2008;35(10):2778–2785.

131. Royle TCA, Zhang H, Guiry EJ, Orchard TJ, Needs-Howarth S, Yang DY. Investigating the sex-selectivity of a Middle Ontario Iroquoian Atlantic salmon (Salmo salar) and lake trout (Salvelinus namaycush) fishery through ancient DNA analysis. J Archaeol Sci Rep. 2020;31:102301.

132. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–410. pmid:2231712

133. Sayers EW, Cavanaugh M, Clark K, Ostell J, Pruitt KD, Karsch-Mizrachi I. GenBank. Nucleic Acids Res. 2019;47(D1):D94–D99. pmid:30365038

134. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: Molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018;35(6):1547–1549. pmid:29722887

135. Scott DC, Harris SL, Hebert AS, van Poorten BT. Nutrient dynamics in a highly managed reservoir system: Considering anadromous sockeye salmon (Oncorhynchus nerka) and nutrient restoration. Lake Reserv Manag. 2017;33(1):14–22.

136. Harris SL, Cruickshank A, Andrusak G, Andrusak H, Sebastian D, Weir T, et al. The Alouette Reservoir nutrient restoration program, 2009–2010. Victoria: Province of British Columbia, Ministry of Environment, Aquatic Ecosystem Protection Branch; 2011. Fisheries Project Report No.: RD 129.

137. McPhail JD. The freshwater fishes of British Columbia. Edmonton: University of Alberta Press; 2007.

138. Kendall C. Tracing nitrogen sources and cycling in catchments. In: Kendall C, McDonnell JJ, editors. Isotope tracers in catchment hydrology. Amsterdam: Elsevier; 1998. pp. 519–576.

139. Guiry EJ, Buckley M, Orchard TJ, Hawkins AL, Needs-Howarth S, Holm E, et al. Deforestation caused abrupt shift in Great Lakes nitrogen cycle. Limnol Oceanogr. 2020.

140. Brugam RB, Little K, Kohn L, Brunkow P, Vogel G, Martin T. Tracking change in the Illinois River using stable isotopes in modern and ancient fishes. River Res Appl. 2017;33(3):341–352.

141. Fagan BM. Fishing: How the sea fed civilization. New Haven: Yale University Press; 2017.

142. Colley SM. The analysis and interpretation of archaeological fish remains. Archaeol Method Theory. 1990;2:207–253.

143. Lambrides AB, Weisler MI. Pacific islands ichthyoarchaeology: Implications for the development of prehistoric fishing studies and global sustainability. J Archaeol Res. 2016;24(3):275–324.

144. Casteel RW. Fish remains in archaeology and paleo-environmental studies. New York: Academic Press; 1976.

145. Beacham TD, McIntosh B, Wallace C. A comparison of stock and individual identification for sockeye salmon (Oncorhynchus nerka) in British Columbia provided by microsatellites and single nucleotide polymorphisms. Can J Fish Aquat Sci. 2010;67(8):1274–1290.

146. Beacham TD, Candy JR, McIntosh B, MacConnachie C, Tabata A, Kaukinen K, et al. Estimation of stock composition and individual identification of sockeye salmon on a Pacific Rim basis using microsatellite and major histocompatibility complex variation. Trans Am Fish Soc. 2005;134(5):1124–1146.

147. Veale AJ, Russello MA. Sockeye salmon repatriation leads to population re-establishment and rapid introgression with native kokanee. Ecol App. 2016;9(10):1301–1311.

148. Hutchinson WF, Culling M, Orton DC, Hänfling B, Lawson Handley L, Hamilton-Dyer S, et al. The globalization of naval provisioning: Ancient DNA and stable isotope analyses of stored cod from the wreck of the Mary Rose, AD 1545. R Soc Open Sci. 2015;2(9):150199. pmid:26473047

149. Oosting T, Star B, Barrett JH, Wellenreuther M, Ritchie PA, Rawlence NJ. Unlocking the potential of ancient fish DNA in the genomic era. Ecol App. 2019;12(8):1513–1522.

150. Beacham TD, Withler RE. Population structure of sea-type and lake-type sockeye salmon and kokanee in the Fraser River and Columbia River drainages. PLoS ONE. 2017;12(9):e0183713. pmid:28886033

151. 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(1):401–416. pmid:28568856

152. Harvey VL, Daugnora L, Buckley M. Species identification of ancient Lithuanian fish remains using collagen fingerprinting. J Archaeol Sci. 2018;98:102–111.

153. Gustafson RG, Wainwright TC, Winans GA, Waknitz FW, Parker LT, Waples RS. Status review of sockeye salmon from Washington and Oregon. Seattle: United States of America Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service; 1997. NOAA Technical Memorandum: NMFS-NWFSC-33.

154. Roberts L, Magne MPR. Faunal analysis. In: Matson RG, Magne MPR, editors. Appendices for “Athapaskan migrations: The archaeology of Eagle Lake, British Columbia”. Vancouver: Vancouver: Laboratory of Archaeology, Department of Anthropology and Sociology, University of British Columbia; 2007. pp. 51–67.

155. Lane RB. Cultural relations of the Chilcotin Indians of west central British Columbia. Ph.D. Dissertation, University of Washington. 1953.

156. Tyhurst RJS. An ethnographic history of the Chilcotin. Draft Ph.D. Dissertation, University of British Columbia. 1984.

157. Lovell NC, Chisholm BS, Nelson DE, Schwarcz HP. Prehistoric salmon consumption in interior British Columbia. Can J Archaeol. 1986;10:99–106.

158. Josephson DB, Lindsay RC, Stuiber DA. Variations in the occurrences of enzymically derived volatile aroma compounds in salt-and freshwater fish. J Agric Food Chem. 1984;32(6):1344–1347.

159. Lindsay RC. Flavour of fish. In: Shahidi F, Botta JR, editors. Seafoods: Chemistry, processing technology and quality. Boston: Springer; 1994. pp. 75–84.

160. Johnson BM, Kemp BM, Thorgaard GH. Increased mitochondrial DNA diversity in ancient Columbia River basin Chinook salmon Oncorhynchus tshawytscha. PLoS ONE. 2018;13(1):e0190059. pmid:29320518

161. Thompson TQ, Bellinger MR, O’Rourke SM, Prince DJ, Stevenson AE, Rodrigues AT, et al. Anthropogenic habitat alteration leads to rapid loss of adaptive variation and restoration potential in wild salmon populations. Proc Natl Acad Sci U S A. 2019;116(1):177–186. pmid:30514813

162. Holmes MW, Hammond TT, Wogan GO, Walsh RE, LaBarbera K, Wommack EA, et al. Natural history collections as windows on evolutionary processes. Mol Ecol. 2016;25(4):864–881. pmid:26757135

163. Wandeler P, Hoeck PE, Keller LF. Back to the future: Museum specimens in population genetics. Trends Ecol Evol. 2007;22(12):634–642. pmid:17988758

164. Rick TC, Lockwood R. Integrating paleobiology, archeology, and history to inform biological conservation. Conserv Biol. 2013;27(1):45–54. pmid:22979917

165. Allan JD, Abell R, Hogan Z, Revenga C, Taylor BW, Welcomme RL, et al. Overfishing of inland waters. Bioscience. 2005;55(12):1041–1051.

166. Dudgeon D, Arthington AH, Gessner MO, Kawabata Z-I, Knowler DJ, Lévêque C, et al. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol Rev Camb Philos Soc. 2006;81(2):163–182. pmid:16336747

167. Elliott LD. There and back again: An investigation of the biological and genetic consequences of a sockeye stocking program. M.Sc. Thesis, University of British Columbia. 2019.