Biomagnification and Fate of Persistent Organic Pollutants (Pops) in Marine Mammal Food Webs in the Northeastern Pacific Ocean

Home , BX (sternwheeler), Cowichan (steamship)

Biomagnification and Fate of Persistent Organic Pollutants (POPs) in Marine Mammal Food Webs in the Northeastern Pacific Ocean

Donna Lynn Cullon

B.Sc., University of Victoria, 1993 M.Sc., Royal Roads University, 2001

A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

in the School of Earth and Ocean Sciences

Biomagnification and Fate of Persistent Organic Pollutants (POPs) in Marine Mammal Food Webs in the Northeastern Pacific Ocean

Donna Lynn Cullon

B.Sc., University of Victoria, 1993 M.Sc., Royal Roads University, 2001

Supervisory Committee

Dr. Michael J. Whiticar, CoSupervisor (School of Earth and Ocean Sciences)

Dr. Peter S. Ross, CoSupervisor (Fisheries and Oceans Canada; School of Earth and Ocean Sciences)

Dr. Kevin Telmer, Departmental Member (School of Earth and Ocean Sciences)

Dr. Robie W. Macdonald, Departmental Member (Fisheries and Oceans Canada; School of Earth and Ocean Sciences)

Dr. John F. Dower, Outside Member (Department of Biology)

iii

Supervisory Committee

Dr. Michael J. Whiticar, CoSupervisor (School of Earth and Ocean Sciences)

Dr. Peter S. Ross, CoSupervisor (Fisheries and Oceans Canada; School of Earth and Ocean Sciences)

Dr. Kevin Telmer, Departmental Member (School of Earth and Ocean Sciences)

Dr. Robie W. Macdonald, Departmental Member (Fisheries and Oceans Canada; School of Earth and Ocean Sciences)

Dr. John F. Dower, Outside Member (Department of Biology)

Abstract

Elevated polychlorinated biphenyl (PCB) concentrations have been detected in marine mammals inhabiting the Strait of Georgia, British Columbia (Canada) and Puget

Sound, Washington State (USA). This raises concerns about adverse health effects and underscores the importance of documenting source, transport, and fate of contaminants.

This marine mammaloriented study (1) examines dietary exposure to complex mixtures of persistent organic pollutants (POPs); (2) characterizes POP accumulations using congenerspecific contaminant analyses, stable isotope ratios, and multivariate statistical methods; and (3) explores some of the influencing factors for POP bioaccumulation in marine mammals.

A first application of a food basket approach to assessing realworld dietary exposure to mixtures of chemicals in marine mammals has revealed Puget Sound as a regional “hotspot” for PCB contamination. The consistency between PCB concentrations in Puget Sound and the Strait of Georgia harbour seals ( Phoca vitulina ) and their food iv baskets validates the use of this method as a basis for exploring dietary exposure, metabolism, biomagnification, and health risks in marine mammals. Concentration rankings of POPs and estimated daily intakes based on our food baskets suggests that both legacy (e.g., PCB, dichlorodiphenyltrichloroethane [DDT]) and new

(polybrominated diphenyl ethers [PBDEs]) POPs may pose potential health risks to seals.

Accumulations of PCBs in the Strait of Georgia seal food web demonstrate the bioaccumulative nature and persistence of PCBs. Correlations of PCB concentrations with physicochemical properties and trophic level revealed the important role that metabolism plays in biomagnification in seals, alongside trophic level and log K ow . We estimate a PCB load of 77 kg within the Strait of Georgia biomass, with the largest proportion (36 %) detected in marine mammals.

Dietary exposure of POPs to resident killer whales ( Orcinus orca ) was assessed by measuring POPs in four stocks of chinook salmon ( Oncorhynchus tshawytscha ), their primary prey. Differences in POP concentrations between chinook smolts and returning adults suggest that the majority of POPs are acquired at sea during the major growth period in their life cycle. Higher POP concentrations and low lipid content were observed among the more southerly stocks suggesting a migrationassociated metabolism and loss of lighter congeners, thereby exposing southern residents to more highly contaminated chinook salmon. Consumption on a lipidweight basis, (higher consumption on a wet weight basis), as well as consuming prey from a more contaminated region, likely increases killer whale exposure to POPs, offering an explanation for higher contaminant burdens in southern residents. v

While previous research has examined species inhabiting different trophic levels or food chains in other regions, this study has provided an assessment of POP dietary exposure, biomagnification, and influencing factors on trophic accumulations in a North eastern Pacific marine mammal food web. These results have provided further insight into the influence of such factors as age, sex, lipid content, diet, migrationrelated metabolism, physicochemical properties (degree of chlorination, log K ow ), and chemical structure on POP accumulation in marine mammals. We have identified the largely unregulated PBDEs as posing potential health risks to marine mammals and offered a means to update existing tissue residue guidelines for the protection of wildlife.

Table of Contents

Supervisory Committee ………………………………………………………………… ii Abstract …………………………………………………………………………………. iii Table of Contents …………………………………………………………………….…..vi List of Tables ………………………………………………………………………..….viii List of Figures ………………………………………………………………………….…x Glossary ……………………………………………………………………………..…..xii Acknowledgements ……………………………………………………………………..xiv Dedication …………………………………………………………………………...….xvi

Chapter 1 Introduction ……………………………………………………………………. 1 1.1 Statement of problem …………………………………………… 1 1.2 Background …………………………………………………...… 1 1.3 Objectives …………………………………………………….... 12

Chapter 2 Persistent Organic Pollutants in the Diet of Harbour Seals ( Phoca vitulina ) Inhabiting Puget Sound, Washington (USA), and the Strait of Georgia, British Columbia (Canada): A Food Basket Approach …………………… 15 2.1 Abstract ……………………………………………………...… 15 2.2 Introduction ……………………………………………………. 16 2.3 Materials and Methods ……………………………………….... 21 2.4 Results and Discussion ………………………………………… 29 2.5 Conclusion ………………………………………………….….. 45

Chapter 3 Biomagnification and Trophic Transfer of Polychorinated Biphenyls within British Columbia (Canada) Harbour Seal ( Phoca vitulina ) Food Webs ……46 3.1 Abstract …………………………………………………………46 3.2 Introduction ………………………………………………….… 47 3.3 Materials and Methods ………………………………………… 49 3.4 Results and Discussion ………………………………………… 58 3.5 Conclusion …………………………………………………...… 78

Chapter 4 Characterizing Dietary Exposure of Persistent Organic Pollutants in Resident Killer Whales ( Orcinus orca ) of British Columbia and Adjacent Waters …………………………………………………………………………. 80 4.1 Abstract …………………………………………………………80 4.2 Introduction …………………………………………………..…81 4.3 Materials and Methods ………………………………………… 83 4.4 Results and Discussion ……………………………………....… 92 4.5 Conclusion ……………………………………………………..113

vii

Chapter 5 Conclusions ………………………………………………………………...... 115 5.1 Introduction ………………………………………………….... 115 5.2 Is 15 N: 14 N a good prediction of PCB accumulation in marine mammal food webs? ………………………………………….. 115 5.3 Can we predict PCB source (within vs outside of the Strait of Georgia) using deviations from δ15 N vs PCB regression lines in marine mammal food webs? ………………………………..… 117 5.4 Does lipid content in diet explain PCB contamination in Puget Sound marine mammals Does eating on a lipidweight basis explain PCB contamination in Puget Sound marine mammals? ………………………………………………………………...... 118 5.5 Does the phrase “You are what you eat” provide a relatively accurate descriptor for marine food web contamination? ……. 119 5.6 What amount of PCBs are in the Strait of Georgia biota? ……. 120 5.7 How do PBDEs compare with PCBs in the Strait of Georgia? ……………………………………………………………… … 122 5.8 Can we put our estimate of PCBs in the Strait of Georgia into a worldwide context? …………………………………………….126 5.9 How will climate change affect the Strait of Georgia marine mammal food webs? ………………………………………...…130 5.10 How does this research contribute to existing Strait of Georgia bioaccumulation models? ………………………………………131 5.11 Do the Strait of Georgia marine mammals fall within the current approaches to risk assessment? ……………………………...…132 5.12 What questions were answered in this research? ………………134 5.13 Recommendations for future research …………………………135

Literature Cited …………………………………………...... 137

Appendix I Principal components analysis (PCA) variables………………………. 158 Appendix II Prey preferences for harbour seal food web species ……………...……161 Appendix III Stable carbon and nitrogen isotope ratios for the Strait of Georgia harbour seal food web species …………………………………………………..162 Appendix IV Biomagnification factors for harbour seal/food basket increase with log Kow in the Strait of Georgia harbour seal ( Phoca vitulina ) food web, with the highest magnification observed between log K ow 7.27 and 7.7 …... 163 Appendix V List of food web magnification factors (FWMFs) of polychlorinated biphenyl (PCB) congeners for the Strait of Georgia harbour seal ( Phoca vitulina ) food web ...... 164

viii

List of Tables

Table 2.1 Annual prey consumption estimates for Puget Sound (WA, USA) harbour seals ………………………………………………………………………...…19

Table 2.2 Annual prey consumption estimates for Strait of Georgia (BC, Canada) harbour seals ……………………………………………………………….... 20

Table 2.3 Concentrations of persistent organic pollutants in harbour seal food baskets from the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) expressed on both a lipid and wet weight basis …………………………...… 31

Table 2.4 Toxic equivalents (TEQs) to 2,3,7,8tetrachlorodibenzopdioxin (TCDD) for polychlorinated biphenyls (PCBs), polychlorinated dibenzopdioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs) in the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets ……………………………………………………………………….. 33

Table 2.5 Nutritional content of the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets ...... 42

Table 2.6 Estimated daily intake and estimated annual intake of POPs, on a wet weight basis, by adult harbour seals for Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) based on contaminant analyses of our food baskets ………………………………………………………………………...44

Table 3.1 Capture location, sex, estimated age, and mass data for harbour seals ( Phoca vitulina ) collected from the Strait of Georgia, B.C., Canada in 2001 ………………………………………………………………………...…52

Table 3.2 Lipid weight concentrations of total polychlorinated biphenyls (PCBs), lipid percentage, stable nitrogen and carbon isotope ratios ( δ15 N, δ13 C) and trophic level estimations are provided for the Strait of Georgia (BC, Canada) harbour seal food web ………………………………………………61.

Table 3.4 Estimated polychlorinated biphenyl (PCB) loadings for Strait of Georgia biota using biomass estimates from a massbalance model reconstruction of the Strait of Georgia present day and updated biomass estimates with recent PCB concentrations .…………………………………………………………..76

Table 4.1 Morphometric and related data for chinook salmon ( Oncorhynchus tshawytscha ) collected from Johnstone Strait and Lower Fraser River (British Columbia, Canada), Duwamish River and Deschutes River (Washington State, USA), and chinook smolts from the Strait of Georgia (British Columbia, Canada) and Puget Sound (Washington State, USA) ……………………………………………………………………...…..95

Table 4.2 Wet weightbased concentrations of persistent organic pollutants and toxic equivalents (TEQs) to 2,3,7,8tetrachlorodibenzopdioxin (TCDD) for polychlorinatedbiphenyls(PCBs),polychlorinateddibenzopdioxins(PCDDs), and polychlorinated dibenzofurans (PCDFs) in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait and Lower Fraser River (British Columbia, Canada); Duwamish River and Deschutes River (Washington State, USA); and chinook smolts from the Strait of Georgia (British Columbia, Canada) and Puget (Washington State, USA) …………………………………………………………………………………97

Table 4.3 Wet weightbased concentrations of organochlorine pesticides in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait and Lower Fraser River (British Columbia, Canada); Duwamish River and chinook smolts from the Strait of Georgia (British Columbia, Canada) and Puget Sound (Washington State, USA) ………………………………………99

Table 4.4 Estimated body burdens of persistent organic pollutants in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait and Lower Fraser River (British Columbia, Canada); Duwamish River and Deschutes River (Washington State, USA); and chinook smolts from the Strait of Georgia (British Columbia, Canada) and Puget Sound (Washington State, USA) ………………………………………………………………….104

Table 4.5 Estimated daily intake (EDI) of persistent organic pollutants (POPs) by northern and southern resident killer whales ………………………………..112

Table 5.1 Ratio of ΣPCB to ΣPBDE concentrations for 8 of the 15 species analyzed for PCBs in the Strait of Georgia (BC, Canada) ………………………………..123

Table 5.2 Food web magnification factors (FWMFs) and top predator ΣPCB and PCB 153 concentrations differ between freshwater and marine food webs and regions in the world …………………………………………………………128

List of Figures

Figure 1.1 Molecular structures for several of the major compounds analyzed in this study ………………………………………………………………………… 4

Figure 1.2 Strait of Georgia harbour seals are omnivorous, however have a high preference for hake and herring species …………………………………..…11

Figure 2.1 Sampling locations for the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food basket prey items …………………………... 23

Figure 2.2 Polychlorinated biphenyl (PCB), polychlorinated dibenzopdioxin (PCDD), and polychlorinated dibenzofuran (PCDF), polybrominated diphenyl ether (PBDE), polychlorinated naphthalene (PCN), and polybrominated biphenyl (PBB) homolog group patterns in the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets ……………………….. 35

Figure 2.3 Ratio of polychlorinated biphenyl (PCB) patterns in the Puget Sound (WA, USA) to the Strait of Georgia (BC, Canada) harbour seal food baskets …… 38

Figure 3.1 Collection sites for the Strait of Georgia, British Columbia (Canada) harbour seals and their prey ………………………………………………………….. 51

Figure 3.2 Polychlorinated biphenyl congener (PCB) patterns change from lower trophic level prey to harbour seal ……………………………………………….……64

Figure 3.3 Biomagnification is clearly observed in the regression of log ΣPCB concentrations with stable nitrogen isotope ratios in the Strait of Georgia harbour seal ( Phoca vitulina ) food web …………………………………….. 65

Figure 3.4 Food web magnification factors (FWMFs) increase with log K ow in the Strait of Georgia harbour seal ( Phoca vitulina ) food web, with the highest magnification observed between log K ow 7.27 and 7.7 …………………….. 69

Figure 3.5 Principle components analysis (PCA) provides insight into physicochemical and trophic level influences on biomagnification of PCBs in the Strait of Georgia harbor seal food web ……………………………………………….70

Figure 3.6 Biomagnification regression slopes may offer insight into the contribution of local sources to the global background ……………………………………... 74

Figure 4.1 Migratory routes and collection sites for British Columbia (Canada) and Washington State (USA) adult chinook salmon (Oncorhynchus tshawytscha ) ………………………………………………………………………………. 84

Figure 4.2 Contaminant patterns in chinook salmon are relevant to assessing the influence of feeding ecology in chinook and dietary exposure of persistent organic pollutants in resident killer whales …………………………………………109

Figure 5.1 We estimate that high trophic level (TL) marine mammals (TL 45) carry almost 14 kg of PCBs, representing approximately 35 % of the PCBs in the Strait of Georgia food web …………………………………………………121

Figure 5.2 In the regression of the ratio of recalcitrant congeners PCB153 to PBDE47 concentrations with trophic level for 8 of the 15 species analyzed for PCBs in the Strait of Georgia (BC, Canada) we observe higher PCB concentrations in harbour seals, spot prawns, and ghost shrimp ………………………………124

Figure 5.3 Total PCB and ΣPBDE concentrations detected in the Strait of Georgia biota among assigned trophic levels. The majority of both ΣPCBs and ΣPBDEs reside in trophic levels 2 and 3 ……………………………………………..125

xii

Glossary benthic species plants, algae, and animals (nonfloaters and nonswimmers) that inhabit the ocean floor (Thurman et al. 2001). bioaccumulation process by which chemical concentration in an aquatic organism exceeds that of its surrounding water, through all possible routes of chemical exposure (dietary exposure, transport across respiratory surfaces, dermal absorption, inhalation) (Gobas and Morrison 2000). bioavailability “the fraction of chemical in a medium that is in a state which can be absorbed by the organism” (Gobas and Morrison 2000). bioconcentration process by which chemical concentration in an aquatic organism exceeds that of its surrounding water through exposure to waterborne chemical (Gobas and Morrison 2000). biomagnification process by which chemical concentration in an organism exceeds that of its diet, through dietary absorption (Gobas and Morrison 2000). biotransformation “process by which chemicals undergo chemical or biochemical reactions in organism” (Gobas and Morrison 2000). body burden mass of contaminant in an individual (Newman 1998). contamination concentration of a toxic substance exceeds that of normal ambient conditions in an environment (Freedman 1995).

δδδ15 stable isotope ratio of 15 N: 14 N that can provide information about diet of organism (Hobson et al. 1996).

δδδ13 C stable isotope ratio of 13 C: 12 C that can provide information about sources of carbon and feeding habitat of organism (France 1995). ecosystem a community of organisms functioning together and interacting with their physical environment through a flow of energy and a cycling of materials (Starr and Taggart 1981). equilibrium chemical equilibrium occurs when a chemical is distributed among environmental media according to its physicochemical partitioning properties and remains constant over time. “This would be the end result of a physicochemical partitioning process.” (Gobas and Morrison 2000) fugacity a thermodynamic quantity describing the “escaping tendency of a chemical substance from a phase”. (Gobas and Morrison 2000)

xiii hydrophobic substances having high solubility in water, usually low solubility in lipid (Gobas and Morrison 2000).

Kow octanol/water partition coefficient, usually reported as its common logarithm (log Kow ). A chemical having a large log K ow value indicates affinity for the noctanol phase and is more lipophilic (hydrophobic) (Gobas and Morrison 2000). lipophilic substances having high solubility in lipid, usually low solubility in water (Gobas and Morrison 2000). pelagic species organisms (floaters and swimmers) that inhabit the ocean water, ie. live above the ocean floor (Thurman et al. 2001). persistent Chemicals that resist the natural processes of degradation, having long half lives in soils, sediments, air or biota, and therefore remain in the environment a long time (Jones and De Voogt 1999; US Environmental Protection Agency [USEPA] 2002; World Wildlife Fund [WWF] 2005). physicochemical physical and chemical properties of a chemical such as molecular structure and vapour pressure (Gobas and Morrison 2000). sentinel species that serve as indicators of their environment; where selection of sentinel species depends on defining the hypothesis and questions to be answered. RM# “Based on their life history and physiological attributes, selected species can provide insight about environmental changes at various spatial, temporal, and trophic scales”(Tabor and Aguirre 2004). steady state occurs when the total amount of chemical entering an organism equals the amount exiting the organism without change in chemical mass or concentration. It is the result of transport and transformation processes on the chemical (Gobas and Morrison 2000). trophic level an energy transfer step within a food chain or food web. Various trophic levels include primary producers, primary consumers, secondary consumers, tertiary consumers, and top consumers/predators (Starr and Taggart 1981).

xiv

Acknowledgements

I would like to sincerely thank my research advisors Peter Ross and Michael

Whiticar for their continued support, encouragement, and expertise throughout this study.

I would also like to thank my committee members, Rob Macdonald, Kevin Telmer, and

John Dower for their support and constructive feedback which served to strengthen many aspects of this project. A very special thank you to my good friend and colleague Neil

Dangerfield for his encouragement and assistance with all aspects of this project, aside from almost drowning me a couple of times! I also wish to thank Paul Eby for his teaching and assistance with the stable isotope analyses.

I want to gratefully acknowledge the assistance of the many people who participated in this research B. Andrews, J. Beam, D. Bowen, D. Bright, P. Browne, J.

Christensen, N. Crewe, R. Dhillon, C. Dubetz, G. Ellis, M. Fischer, J. Ford, T. Fraser, M.

Gembala, B. Hickie, M. Ikonomou, J. Irvine, S. Iverson, S. Jeffries, S. Johannessen, P.

Kimber, D. Lambourn, M. Lance, G. Lichota, L. Loseto, L. MacDougall, S. MacLellan,

K. Miller, L. Mos, P. Olesiuk, S. O’Neill, S. Quinnell, S. Redman, M. Saunders, J.

Schweiggert, P. Shaw, K. Shew, D. Swanston (Seacology), M. Tabuchi, M. Trudel, R.

Veefkind, J. West, R. Withler, and the crews of the C.C.G.S. Vector , C.C.G.S. Ricker , and F.V. Chasina .

Financial support was provided by Environment Canada; the Environmental

Sciences Strategic Research Fund (Fisheries and Oceans Canada); the Puget Sound

Action Team; the Washington Department of Fish and Wildlife; the SeaDoc Society xv through the Wildlife Health Center, School of Veterinary Medicine, University of

California, Davis; and the Toxic Substances Research Initiative.

Thank you to my family, especially my parents and inlaws, who helped out with our three children to allow me to complete this degree not an easy task! Finally, to my husband Richard for his encouragement, commitment, and support for a project that he says he never really understood! His dedication and efforts have exceeded what most partners could sustain. …And, my children, Jessica, Russell, and Nicole (“the twins”), who have brought me to a new level of exhaustion, given me a broader perspective, and above all, forced me to keep my sense of humour throughout all of this!

xvi

……for my family

Chapter 1: Introduction

1.1 Statement of problem

Toxic contaminants are important stressors to coastal ecosystems in the world today. Marine mammals are useful indicators for marine ecosystem contamination since they are exposed to a mixture of chemicals and their accumulations reflect uptake, biotransformation and/or excretion from all trophic levels in the food web (Ross 2000).

High trophic level, top predators such as killer whales ( Orcinus orca ) and harbour seals

(Phoca vitulina ) are especially vulnerable to high accumulations of a group of contaminants known as persistent organic pollutants (POPs). They often accumulate very high concentrations of POPs as a result of trophic level, long life span, and limited ability to metabolize contaminants (Hickie et al. 2007; Ross et al. 2004; Muir et al. 1988).

Characterizing the movement and fate of these compounds through food webs can provide information for prioritizing chemicals and identifying priority species for risk assessments. This project characterized the accumulation of complex mixtures of POPs in marine mammal food webs in the Strait of Georgia, British Columbia (Canada) and

Puget Sound, Washington State, (USA).

1.2 Background

Persistent Organic Pollutants

The Stockholm Convention evaluates Persistent Organic Pollutants (POPs) on the basis of toxicity, persistence, bioaccumulation, and potential for longrange transport

(World Wildlife Fund [WWF] 2005). Persistent Organic Pollutants are lipophilic or “fat soluble” compounds not easily metabolized that can bioaccumulate within organisms and 2 biomagnify to high trophic levels within food webs (Jones and De Voogt 1999). Their bioaccumulation and biomagnification potential poses particular elevated risk to high trophic level wildlife and subsistenceoriented human populations. This group of compounds includes polychlorinated biphenyls (PCBs), polychlorinated dibenzop dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), organochlorine pesticides

[DDT (dichlorodiphenyltrichloroethane), HCH (hexachlorocyclohexane), toxaphene, chlordane, dieldrin, and heptachlor] and the fungicide HCB (hexachlorobenzene) ( Figure

1.1 ). In 2009, nine new POPs were added to the Convention including HCH ( α, β), lindane (primarily γHCH), and PBDEs (tetra, penta, hexa, hepta) (United Nations

Environment Programme 2009). Remaining POP candidates to be evaluated include

PBDEs (octa, deca), PCNs, and PBBs (hexa) (World Wildlife Fund [WWF] 2005).

Of the chemicals examined in this study, the majority were phased out under the

Stockholm Convention in 2001, including the legacy DDT and PCBs which were banned earlier in Canada and the U.S. Polychlorinated biphenyls were an industrial chemical manufactured from 1929 until banned in most of the industrial world in the late 1970s.

There are 209 theoretically possible congeners which vary by degree of chlorination, location of chlorine atoms, as well physicochemical properties and toxicology (Shiu and

Mackay 1986; Safe 1984). Commercial PCB mixtures (Arochlors) were produced by catalytic chlorination of biphenyl, the last two digits of Arochlor numbers representing the target weight percentage of chlorine (Frame 1997). Desired for its flame retardant and insulatory properties, PCBs were used in transformers and capacitors, ink additives, oils, and plastics (Borlakoglu and Dils 1990). A figure of 1.3 million tonnes for global production has been estimated with almost 97 % usage occurring in the Northern 3

Hemisphere (Breivik et al. 2002). Although banned, PCBs remain a global problem due to their persistence, recycling between environmental compartments and bioavailability for uptake in organisms.

Unintentional byproducts including PCDDs and PCDFs were released into the environment as a result of combustion/incineration of municipal and chemical wastes

(Hites 1990) (Figure 1.1). Of the 75 theoretically possible congeners for PCDDs and 135 for PCDFs, the most toxic synthetic compound is known to be 2,3,7,8 tetrachlorodibenzopdioxin (Hites 1990). Elevated levels of PCDDs and PCDFs detected in sediments (Macdonald et al. 1992) and biota (Yunker and Cretney 2000;

Addison et al. 2005) in the Strait of Georgia have been linked to effluent discharges from pulp mills located along its coastlines. Although listed among the “dirty dozen” POPs under the Stockholm Convention (World Wildlife Fund [WWF] 2005), knowledge gaps still exist in estimating environmental inventories and identifying sources (historical or recent) (Jones and De Voogt 1999). Their global recycling, bioavailability, and uptake into food webs highlight the need for continued monitoring of food web accumulations as well as help to establish temporal trend datasets worldwide.

Figure 1.1 Molecular structures for individual congeners of each of the major compounds analyzed in this study. All of the compound classes here, including PCBs, PCDDs, PCDFs, DDT, HCH, HCB, and tetraBDEs are termed Persistent Organic Pollutants under the Stockholm Convention (United ations Environment Programme 2009).

Organochlorine pesticides, including the wellknown DDT that has been associated with eggshell thinning in birds (Carson 1962), were produced for their insecticide properties. These compounds, evidenced for their “grasshopper” effect

(Gouin et al. 2004), pose particular risk to northern latitude food webs (Wania and

Mackay 2001). The movement of semivolatile POPs through revolatilization and deposition at their condensation temperature has been termed the “grasshopper effect”

(Wania and Mackay 1996). Although, 10 out of the 13 compound classes analyzed in this study are listed under the Stockholm Convention (United Nations Environment

Programme 2009), further understanding of their environmental distribution is needed.

Their persistence and bioaccumulative nature remain a concern for high trophic level organisms and human populations with heavy reliance on marine foods.

The more recently studied classes of POPs, including PBDEs, PBBs, and PCNs, have been detected in resident killer whales (Rayne et al. 2004) and harbour seal prey

(Cullon et al. 2005) known to frequent and forage in the Strait of Georgia (Ford and Ellis

2006; Olesiuk 1993; Cottrell et al. 2002). The similarities in physicochemical properties between PBDEs and PCBs give cause for health concern, especially when PBDE concentrations are projected to exceed those of PCBs in marine mammals within 10 years

(Ross 2006; Ross et al. 2009). While there is growing data on emerging POPs concentrations in marine biota in this region (Ikonomou et al. 2002; Christensen et al.

2005; Cullon et al. 2005; Rayne et al. 2004; Cullon et al. 2009), further research in comparative food web accumulations of these compounds (e.g., PCB vs PBDE) is needed to fully understand their fate and risk to higher trophic organisms. 6

Determining source, transport, and fate functions for many of these POPs is difficult. The multitude of congeners and isomers in conjunction with physicochemical differences complicate understanding and predicting behaviour in environmental matrices and biota. Many of these compounds are subject to longrange global transport through either oceanic currents or atmospheric processes. Where the more heavily chlorinated

PCBs may be deposited near their source release, the lighter chlorinated PCBs may be transported to distant regions entering local food webs. Food web studies provide a central basis for the study of source, transport, and fate mechanisms for POPs in the environment (Hoekstra et al. 2003; Kidd et al. 1998; Borgå et al. 2004; Fisk et al. 2001;

Hop et al. 2002; Ruus et al. 2002; Muir et al. 2000; Borgå et al. 2001; Moisey et al.

2001). Increased understanding of how physiological processes (uptake, biotransformation, and excretion), feeding ecology, migration and physicochemical properties influence POP accumulations is needed to assess POP sources.

Persistent organic pollutants in Marine Mammal Food Webs

Marine mammals have been used as sentinels of marine ecosystem contamination

(Ross 2000). Their fish diet and elevated trophic position make them excellent species to examine bioaccumulation of fatsoluble compounds between prey to predator and trophic levels. The “real world” complex mixtures of chemicals that marine mammals are exposed to would be similar to other high trophic level species, including marine birds

(bald eagles [ Haliaeetus leucocephalus ]), terrestrial mammals (grizzly bears [ Ursus arctos horribilis ]), and subsistenceoriented human groups. It is in this way that marine 7 mammal food web studies play an important role in ecological risk assessments, complementing species specific laboratory and field studies.

Trophic transfer and POP accumulation in food webs has been well studied in marine and freshwater systems (Hoekstra et al. 2003; Kidd et al. 1998; Ruus et al. 2002;

Hop et al. 2002; Morrison et al. 2002; Kiriluk et al. 1995; Law et al. 2006). In the past, regions of study have been concentrated in the Arctic, Baltic, and Great Lakes. The northern regions have been of greatest interest for several reasons. It is an area of minimal POP usage, historically considered one of the more “pristine” regions in the world. However, longrange transport processes have led to polar regions becoming a

“global sink” for POPs (Muir et al. 1999). Bioaccumulation of these compounds within food webs has resulted in elevated concentrations in high trophic level and high lipid organisms such as marine mammals (seals (Ross et al. 2004; Shaw et al. 2005), whales

(Ross et al. 2000; Rayne et al. 2004), and terrestrial mammals (grizzly bears (Christensen et al. 2005). Entry of POPs into the base of the food web can begin diffusion through cellular membranes of phytoplankton (adsorbed in cellular carbon or adsorbed to cell surfaces) (Kujawinski et al. 2000; Skoglund et al. 1996). Other mechanisms include uptake through bacterial cell walls, grazing of POPs adsorbed to organic particles, and gill exchange by invertebrates and vertebrates (Kujawinski et al. 2000; Magnusson and

Tiselius 2010; Gobas and Morrison 2000).

Several tools can be used to help characterize contaminant accumulation patterns.

Multivariate statistical applications such as principal component analysis (PCA) are used to group samples according to similarities, and also to gain information about relationships among samples (GrahlNielsen 1999). Biomagnification factors (BMFs) 8 describe the chemical concentrations in organisms relative to that of their diet (Gobas and

Morrison 2000) and food web magnification factors (FWMFs) provide an overall magnification for a food web (Fisk et al. 2001; Hoekstra et al. 2003). Both BAFs and

BMFs can provide information about contaminant accumulations within organisms and between organisms occupying different trophic levels. Examining congenerspecific contaminants and homolog groups with age, sex, and physical condition of organisms can further provide information about metabolism and accumulation patterns.

Stable nitrogen and carbon isotope ratios have been used in food web studies to characterize feeding ecology (Kurle and Worthy 2001; Hooker et al. 2001; Hirons et al.

2001; Walker and Macko 2000; Lesage et al. 2002; Hobson et al. 1997; Burton and Koch

1999; Lesage et al. 2001) and trophic accumulations of chemicals (Hoekstra et al. 2003;

Kidd et al. 1998; Borgå et al. 2004; Ruus et al. 2002; Hop et al. 2002; Fisk et al. 2001;

Hobson et al. 2002). The enrichment of heavy isotopes of nitrogen and carbon relative to the lighter isotopes or the fractionation of these isotopes through food webs has been well studied (Tieszen et al. 1983; McConnaughey and McRoy 1979; Minagawa and Wada

1984; Deniro and Epstein 1978; Deniro and Epstein 1981). Ratios of 13 C: 12 C can provide information about feeding habitat, freshwater vs marine, benthic vs pelagic, and inshore vs offshore sources of carbon (France 1995; France and Peters 1997) dependent upon tissues used (Tieszen et al. 1983). Ratios of 15 N: 14 N can provide information about diet, trophic structure of food webs, integration of 15 N in animal tissues, dependent upon tissue used (Hobson et al. 1996), turnover rate. Accepted levels of stepwise enrichments are 3

5 ‰ for δ15 N (Kelly 2000; Kurle and Worthy 2001; Hoekstra et al. 2002), with an average of 3.4 ± 1.1 ‰ independent of habitat (Minagawa and Wada 1984), and 02 ‰ 9 for δ13 C (France and Peters 1997; Kelly 2000; Rau et al. 1983). However, stable isotope ratios can be affected by confounding factors such as tissue type analyzed, metabolic turnover rate, nutritional status of organisms (fasting), migratory movement, lipid content

(δ13 depleted) (Borgå et al. 2004; Fisk et al. 2001), and many studies using stable isotopes as measures of feeding ecology/trophic level have failed to adequately explain chemical bioaccumulation in food webs.

There is increasing concern for the ecological and human health risks of POPs since elevated levels can affect humans and wildlife (Ross and Birnbaum 2003). As high trophic level organisms, marine mammals have the ability to accumulate high concentrations of POPs in their tissues; including killer whales (Ross et al. 2000; Rayne et al. 2004), northern fur seals ( Callorhinus ursinus ) (Mossner and Ballschmiter 1997), harbour seals (Ross et al. 2004; Mossner and Ballschmiter 1997), ringed seals ( Phoca hispida ) (Kucklick et al. 2006), bottlenose dolphins ( Tursiops truncatus ) (Pulster et al.

2009), and Stellar sea lions ( Eumetopias jubatus ) (Myers et al. 2008). Studies have shown exposure to POPs can result in adverse effects including immunotoxicity (Ross et al. 1996), vitamin A disruption (Simms et al. 2000), developmental abnormalities, and thyroid disruption in harbour seals (Brouwer et al. 1989; Tabuchi et al. 2006). Human populations at increased risk for adverse effects include those groups (e.g. indigenous people) that rely heavily on fish and marine mammal tissues for both nutritional and cultural purposes (van Oostdam et al. 2005; Kuhnlein and Chan 2000). While human health risk assessments have been carried out using market basket analyses (Bolles et al.

1999; Newsome et al. 2000) and total diet studies (Yess et al. 1993; Gunderson 1995), 10 and until this study, no such approach had been carried out in wildlife, in part due to the difficulty in carrying out feeding preference assessments.

Strait of Georgia marine mammals

The Strait of Georgia is a semienclosed marine region approximately 6900 km 2, with a mean depth of 156 metres, and relatively high sedimentation rates (Johannessen et al. 2008). It opens to the Pacific Ocean in the north at Queen Charlotte Strait and in the south at Juan de Fuca Strait. It is subject to strong and variable tidal influences as well as freshwater input, mostly from the Fraser River (LeBlond 1983). Changes associated with climate variability and contaminant inputs (point source, freshwater, atmospheric) have drawn attention to the area and nurtured interest on the part of the scientific and resource management communities. It is a site for coastal industries (pulp and paper mills, fish farms, fish hatcheries, shipping, commercial fishing), transport vessels (ferries), and recreational activities (sport fishing). The Strait of Georgia also provides habitat for a diverse group of species including such top predators as harbour seals and resident killer whales.

There are an estimated 105,000 harbour seals (Fisheries and Oceans Canada

2010) and 285 resident (both northern and southern) killer whales (Hickie et al. 2007) inhabiting coastal areas of British Columbia. Harbour seals are largely nonmigratory in the NE Pacific Ocean (Cottrell et al. 2002; Cottrell 1996; Bigg 1981), whereas resident killer whales have large foraging ranges (Ross et al. 2000; Ford and Ellis 2006).

Reasonably complete information exists on foraging behaviour and prey consumption for both harbour seals ( Figure 1.2) (Olesiuk 1993; Olesiuk et al. 1990; Cottrell et al. 11

1995; Bigg et al. 1990) and resident killer whales (Ford and Ellis 2006), providing the foundation for this food web study.

Longlived, marine mammals occupying high trophic levels can provide an overview of aquatic food web contamination by chemicals that are persistent, bioaccumulative and toxic. In this way, marine mammals can serve as sentinels in identifying those chemicals that may be problemmatic to high trophic level consumers, including humans (Ross 2000; Ross and Birnbaum 2003). This can inform posthoc or proactive human health risk assessments, wildlife risk assessments, chemical risk assessments, and chemical regulations.

Strait of Georgia harbour seal food web

Harbour seal Trophic level 4 Rockfish Salmon Tomcod Lingcod Hake 3 Herring Sandlance Midshipman Shrimp/Prawns 2 Zooplankton 1

Figure 1.2 Strait of Georgia harbour seals are omnivorous, but have a high preference for Pacific hake ( Merluccius productus ) and Pacific herring ( Clupea pallasi ). Figure adapted from J. West (West 1997).

1.3 Objectives

The primary objective of this research was to characterize food chain accumulations of POPs in marine mammal food webs. This included the exploration of dietary exposure to real world complex mixtures of POPs; influences of lipid, feeding ecology, and physicochemical properties on POP concentrations and patterns in marine mammal food webs. Specific research questions that have guided and been addressed throughout the chapters include:

1 Can we characterize POP accumulations in marine mammal food webs using

tools such as congenerspecific contaminant analyses, stable isotope ratios, and

multivariate statistical methods?

2 Are there differences in POP accumulations among prey species and trophic

levels in Northeastern Pacific Ocean marine mammal food webs? If so, could

these observations help to explain differences in contaminant concentrations

between northern and southern resident killer whales and between Strait of

Georgia and Puget Sound harbour seals?

3 Can we use a food webbased approach to POP biomagnification to better

understand health risks associated with dietary exposure to POPs?

In Chapter Two, a food basket approach documents dietary exposure to “real world” mixtures of POPs in British Columbia and Washington State harbour seals to 13 better understand contaminants in our regional environment. “Legacy” POPs, including

PCBs, PCDDs, and DDT, as well as the more recently studied PBDEs, PBBs, and PCNs detected in seal food baskets from the Strait of Georgia, BC and Puget Sound, WA are compared and characterized. Estimated daily and annual intakes of POPs for harbour seals are calculated for dietary risk and potential health risk assessments. In addition, the seal food baskets are used to assess biomagnification (prey to predator), metabolism, and the use of harbour seals as marine ecosystem sentinels.

Chapter Three provides a food webbased study of PCB biomagnification in the

Strait of Georgia harbour seals to improve our understanding of POP accumulation and health risks for marine mammals in our region. The harbour seal prey species that comprised the food basket in chapter two are analyzed individually in order to characterize PCB accumulations in the harbour seal food web. Prey to predator biomagnification, overall food web magnification, feeding ecology (trophic level), metabolism, and physicochemical properties of individual congeners are explored. An analysis of regression slopes is suggested as a means of estimating PCB source contribution for the Strait of Georgia harbour seal food web. Finally, an application of our marine mammal food web data to an existing Strait of Georgia mass balance ecosystem model is presented in an attempt to estimate mass of PCBs in the Strait of

Georgia biota.

Chapter Four characterizes dietary exposure of POPs to Pacific resident killer whales by examining their primary prey, chinook salmon to better understand both the sources of POPs to salmon and the extent to which they deliver POPs to resident killer whales. Persistent organic pollutant concentrations from four different salmon stocks, 14 two from British Columbia and two from Puget Sound are compared in an attempt to explain the higher POP levels previously detected in southern resident killer whales.

Confounding factors, including migrationassociated lipid changes and metabolism, feeding ecology, and POP sources are explored. Finally, estimated daily intakes are calculated for resident killer whales by adjusting prey consumption to lipid content, providing a possible explanation of the differences in contaminant burdens observed previously in northern and southern resident killer whales.

In Chapter Five, the major findings and conclusions of this research are discussed within the context of the following questions: Is 15 N: 14 N a good predictor of PCB accumulation in marine mammal food webs? Can we predict PCB source (within vs outside of the Strait of Georgia) using deviations from δ15 N vs PCB regression lines in marine mammal food webs? Does lipid content in diet or eating on a lipidweight basis explain PCB contamination in Puget Sound marine mammals? Does the phrase “You are what you eat” provide a relatively accurate descriptor for marine food web contamination?

The latter half of this chapter compares PCB loading estimates for the Strait of

Georgia biota (Chapter 3) with those for PBDEs. The contributions this research has made towards further understanding of how climate change may affect Strait of Georgia marine mammals and to existing bioaccumulation models are discussed. As well, an evaluation of whether current approaches to risk assessment for wildlife are adequate.

Finally, a global perspective of POPs is discussed with recommendations for future research.

Chapter 2: Persistent Organic Pollutants in the Diet of Harbour Seals (Phoca vitulina ) Inhabiting Puget Sound, Washington (USA), and the Strait of Georgia, British Columbia (Canada): A Food Basket Approach †

2.1 Abstract

Harbour seal pups (Phoca vitulina ) inhabiting Puget Sound (WA, USA) were recently found to be seven times more contaminated with polychlorinated biphenyls

(PCBs) than those inhabiting the adjacent Strait of Georgia (BC, Canada). We carried out a food basket approach to approximate realistic dietary exposures of both new (e.g., polybrominated diphenyl ethers [PBDEs]) and legacy (e.g., dichlorodiphenyltrichloroethane [DDT]) persistent organic pollutants (POPs) for these harbour seals. Food basket homogenates, each consisting of over 200 individual prey items, were constructed using documented dietary preferences for harbour seals in these basins, and analyzed for organochlorine pesticides, flame retardants, and other persistent contaminants. Concentration rankings for the major contaminant classes in the Puget

Sound food basket were ΣPCBs> ΣPBDEs> ΣDDT, and for the Strait of Georgia food basket were Σ PCBs> ΣDDT> ΣPBDEs, highlighting the emergence of PBDEs as a significant concern in the regional environment. Consistent with observations in harbour seals, PCB concentrations in the Puget Sound food basket were seven times higher than in its Strait of Georgia counterpart. Based on our food basket results, the estimated daily intake (EDI) of ΣPCB toxic equivalents (TEQs) to dioxin by Puget Sound harbour seals exceeds some wildlife consumption guidelines for PCBs. Our results indicate that both legacy and new POPs present a health risk to these marine mammals.

† Adapted from Cullon DL, Jeffries SJ, Ross PS. 2005. Persistent organic pollutants in the diet of harbour seals (Phoca vitulina) inhabiting Puget Sound, Washington (USA) and the Strait of Georgia, British Columbia (Canada): A food basket approach. Environ Toxicol Chem 24: 25622572.

2.2 Introduction

Persistent organic pollutants (POPs) include a wide array of compounds characterized by low water solubility, high lipid solubility, and resistance to metabolism and degradation in the environment. Since these fatsoluble compounds are not easily metabolized, they can reach high concentrations in organisms through bioaccumulation and in food webs through biomagnification. Persistent organic pollutants include industrial compounds and flameretardants such as polychlorinated biphenyls (PCBs), polychlorinated naphthalenes (PCNs), polybrominated diphenyl ethers (PBDEs), polybrominated biphenyls (PBBs), industrial byproducts such as polychlorinated dibenzopdioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs), as well as the organochlorine (OC) pesticides such as dichlorodiphenyltrichloroethane (DDT), hexachlorobenzene (HCB), and hexachlorocyclohexane (HCH). Exposure to POPs has been associated with immunotoxicity, endocrine disruption, reproductive impairment, and developmental abnormalities in humans and wildlife (Ross 2000; Ross and Birnbaum

2003). High trophic level marine mammals appear particularly vulnerable; POP mixtures

(dominated by PCBs) have been associated with impaired reproduction and reduced immune function, and the disruption of vitamin A and thyroid hormones in captive and freeranging harbour seals (De Swart et al. 1996; Ross et al. 1996; Reijnders 1986;

Simms et al. 2000).

Freeranging (livecaptured) harbour seal pups inhabiting Puget Sound,

Washington State (USA) were recently found to be seven times more contaminated with

PCBs (18.1 ± 3.1 mg/kg lipid in blubber) than those inhabiting the adjacent Strait of

Georgia (BC, Canada) (2.5 ± 0.2 mg/kg lipid) (Ross et al. 2004). Elevated POP 17 concentrations have been detected in other biota inhabiting Puget Sound and the Strait of

Georgia, with both regions being subjected to a combination of a regional introduction of contaminants along with those introduced from other regions by other processes including longrange atmospheric transport. While historical differences in regulations, sources, usage, and spills partly explain differences in POP levels between the two basins, differences in contaminant fate processes (e.g., as influenced by geological hydrological, topographical, and oceanographic features) are also at play. Relative to the

Strait of Georgia, Puget Sound is smaller, semienclosed, and has limited sedimentation rates (Macdonald and Crecelius 1994). The Strait of Georgia encompasses a larger area, is subject to greater tidal influences, and has relatively high sedimentation rates.

As with terrestrial mammals, including humans, the dominant pathway for bioaccumulation of POPs in marine mammals is through dietary intake (CACAR 1997).

Two basic components are needed when assessing contaminant exposure: The identification and quantification of food items consumed, and the measurement of contaminants in these food items. In the case of humans, surveys (questionnaires) generally provide the information required when quantifying food items consumed.

Market basket or food basket studies are then used in human health risk assessments to approximate realistic exposures to contaminants (including PCBs, organochlorine pesticides, metals, and radionuclides) through dietary intake (Gunderson 1995; Bolles et al. 1999; Newsome et al. 2000). In the case of wildlife, generating such information represents a major challenge, reflecting the oftenincomplete knowledge of their consumption patterns (e.g., prey selection, age, quality, and spatial/temporal abundance).

A range of techniques has been used to quantify or estimate feeding preferences for 18 different species including visual observations, stomach content analyses, scat (fecal) analyses, and fatty acid signatures (Olesiuk et al. 1990; Bowen 2000; Cottrell et al. 1995;

Smith et al. 1997).

With their omnivorous nature, high trophic position, and nonmigratory nature in many regions, harbour seals can serve as local sentinels of food web contamination (Ross

2000). However, seasonal and regional variation in prey availability must be considered when describing feeding preferences within and among harbour seal populations (Olesiuk

1993), something that represents a complication for ecotoxicological studies. Indeed, differences in dietary intakes have been observed between seals of Puget Sound and seals of Strait of Georgia, likely reflecting differences in relative prey abundance. Puget

Sound harbour seals consume a wide variety of prey species, whereas seals in the Strait of Georgia rely largely on two species: Pacific hake ( Merluccius productus ) and Pacific herring ( Clupea harengus pallasi ) ( Tables 2.1 and 2.2 ).

Given the availability of information on the dietary preferences of harbour seals inhabiting both Puget Sound and the Strait of Georgia, an opportunity existed for an assessment of dietary exposure to contaminants in these basins. We hypothesize that two mechanisms could explain the increased PCB concentrations in Puget Sound harbour seals (Ross et al. 2004): Regional variation in contaminant inputs, or differences in prey selection between basins. Our objectives in this study were to determine whether a food basket approach represents a viable risk assessment method for dietary exposure, and to improve our understanding of contaminants of current concern in regional harbour seal habitat.

Table 2.1. Annual prey consumption estimates for Puget Sound (WA, USA) harbour seals (estimates from S. Jeffries, unpublished data). A wide variety of prey species are reflected in the construction of a weighted, preyspecific food basket for contaminant analysis. C = not collected

Annual Our harbour seal Species estimated food basket prey Final Common name Latin name consumption g in 250g composition in (%) pool food basket (%) Pacific tomcod Microgadus proximus 35.5 94.4 37.7 Pacific herring Clupea pallasi 18.0 47.8 19.1 English sole Parophrys vetulus 8.9 23.7 9.5 Plainfin midshipman Porichthys notatus 8.7 23.1 9.2 Pacific hake Merluccius productus 5.4 14.4 5.7 Shiner surfperch Cymatogaster 4.8 12.8 5.1 aggregata Market Squid Loligo opalescens 3.9 10.4 4.2 Dover sole Microstomus pacificus 3.8 10.1 4.0 Pacific cod Gadus macrocephalus 2.2 NC NC Wattled eelpout Lycodes palearis 2.0 NC NC Blackbelly eelpout Lycodopsis pacifica 0.1 0.3 0.1 Rockfish spp. Family Scorpaenidae 1.4 3.7 1.5 Salmonid spp. Oncorhynchus spp. 1.3 3.5 1.4 Slender sole Lyopsetta exilis 0.8 2.1 0.8 Sculpin spp. Family Cottidae 0.7 1.9 0.7 Hexagrammids Family 0.7 NC NC Hexagrammidae Starry flounder Platichthys stellatus 0.3 0.8 0.3 Pacific lamprey Lampetra tridentatus 0.3 NC NC Sculpins Icelinus spp. 0.2 NC NC Sand sole Psettichthys 0.2 NC NC melanostictus Rex sole Glyptocephalus 0.2 0.5 0.2 zachirus Rock sole Lepidopsetta bilineata 0.1 NC NC Octopus Octopus spp. 0.1 NC NC Pacific sandaab Citharichthys sordidus 0.1 0.3 0.1 Rainbow smelt Osmerus mordax 0.1 NC NC dentex

Totals 100.0 250.0 100.0 20

Table 2.2. Annual prey consumption estimates for Strait of Georgia (BC, Canada) harbour seals (Olesiuk 1993). Pacific hake and Pacific herring dominate the diet of Strait of Georgia seals (~75% of annual estimated prey) in our weighted, prey specific food basket for contaminant analysis. C = not collected

Annual Our harbour seal Species estimated food basket prey Final Common name Latin name consumption g in 250g composition in (%) pool food basket (%) Pacific hake Merluccius 42.6 110.5 44.2 productus Pacific herring Clupea pallasi 32.4 84.0 33.6 Salmonid spp. Oncorhynchus spp. 4.0 10.4 4.2 Plainfin midshipman Porichthys notatus 3.4 8.8 3.5 Lingcod Ophiodon elongatus 3.0 7.8 3.1 Surfperches Family 2.3 6.0 2.4 Embiotocidae Cephalopods (squid) Class Cephalopoda 2.1 5.4 2.2 Sculpins Family Cottidae 1.2 3.1 1.2 Flatfish spp. Order 1.2 3.1 1.2 Pleuronectiformes Rockfish spp. Family Scorpaenidae 1.1 2.9 1.1 Pacific tomcod Microgadus 1.0 2.6 1.0 proximus Walleye pollock Theragra 1.0 2.6 1.0 chalcogramma Pacific sandlance Ammodytes 0.8 2.1 0.8 hexapterus Pacific cod Gadus 0.5 NC NC macrocephalus Smelts (eulachon) Family Osmeridae 0.4 NC NC Other invertebrates 0.2 0.5 0.2 Unidentified prey 2.7 NC NC

Totals 100.0 250.0 100.0

2.3 Materials and Methods

Harbour seal diets

To construct harbour seal food baskets, we used dietary preferences documented for Strait of Georgia harbour seals (Olesiuk 1993). Since no published account of the dietary preferences of Puget Sound harbour seals exists, one of the authors (S. Jeffries) conducted a pilot study of harbour seal feeding preferences in this basin. Briefly, harbour seal diet was estimated from scat samples collected from Gertrude Island (southern Puget

Sound, WA, USA) during the period June 1994 to October 1995. The hard parts of prey species (e.g. bones and otoliths) isolated from scat samples were identified to taxon in

207 samples at the National Marine Mammal Laboratory (P. Browne, unpublished data) using methods described elsewhere (Olesiuk 1993). Percentage of diet by weight was estimated using frequency of occurrence, minimum number of individuals identified using all prey hard parts, mean mass of prey species calculated from otolith length, and total mass of prey consumed. Harbour seals diet preferences are summarized for Puget

Sound (Table 2.1) and the Strait of Georgia (Table 2.2).

Sample collection

Harbour seal prey items (Tables 2.1 and 2.2) were collected from Puget Sound,

Washington State (USA) and the Strait of Georgia, British Columbia (Canada) between

August 2000 and August 2001 using trawls (gang, otter, beam), rods, and beachseine nets. Puget Sound prey collections were carried out on board the FV Chasina between

47º12'N, 122º38'W and 47º13'N, to 122º39'W at depths of 25 to 150 meters ( Figure 2.1 ).

The Strait of Georgia prey collections were largely carried out on board the CCGS Vector and CCGS Ricker between 49º26'N, 124º38'W and 48º53'N, 123º27'W at depths of 20 to 22

50 meters (Figure 2.1). We collected seal preysized individual fish of each species (i.e.,

< 25 cm) for homogenization and harbour seal food basket construction. Gross measurement data including weight, fork or mantle length, and sex were recorded (results not shown). Samples were individually wrapped in acetone and hexane rinsed aluminium foil, bagged, and frozen at –20 °C for subsequent analysis.

Food basket design

Whole fish were homogenized individually ( n=1012 per species consisting of 56 male and 56 female) and subsequently homogenized into 100g species pools for each basin using a Sorvall Omni Mixer (Newton, CT, USA), a Hobart grinder (Hobart, Troy,

OH, USA), and a Polytron PT 10/35 Brinkmann homogenizer (Luzern, Switzerland). All equipment was rinsed with 4% Extran 300 (EM Science, Darmstadt, Germany), followed with double MilliQ® deionized water (Millipore, Nepean, ON, Canada) between fish of the same species and finally rinsed twice with acetone (Omnisolve gas chromatography

[GC] grade; EM Science, Gibbstown, NJ, USA) and twice with hexane (Omnisolve GC grade) between different species. A 250g harbour seal food basket was constructed from individual species pools for both Puget Sound and the Strait of Georgia. Since some harbour seal prey items were not caught during trawls, the final diet compositions were slightly adjusted from the original diet composition estimates. The Puget Sound food basket consisted of approximately 94% of the annual harbour seal intake and the Strait of

Georgia food basket consisted of approximately 96% of the annual harbour seal intake

(Olesiuk 1993).

Figure 2.1 Sampling locations for the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food basket prey items. The majority of prey items for the Strait of Georgia food basket were collected from locations 1 to 4 and 7. Chinook salmon were collected from location 5, as well as from Johnstone Strait (outside of northern boundary of map), and Pacific sandlance were collected from location 6. Most prey items for Puget Sound food basket were collected from locations 10 and 11. Chinook salmon were caught at locations 8 and 12, and Chinook smolts were collected from location 9.

Two additional food baskets were constructed for a diet switching exercise aimed at assessing whether local contamination or dietary differences explained the degree of contamination in Puget Sound: A food basket for Puget Sound seals if they were to adopt the Strait of Georgia seal dietary preferences, i.e., Strait of Georgia diet using Puget

Sound prey samples; and a food basket for Strait of Georgia seals if they were to adopt the Puget Sound seal diet, i.e., Puget Sound diet using Strait of Georgia prey samples. If differences in feeding preferences between the two seal populations explained the increased contamination of Puget Sound seals, then we would expect the PCB concentrations to decrease if they were to adopt the Strait of Georgia harbour seal diet.

Contaminant analyses

Four food basket samples (10g) were analyzed for congenerspecific PCBs,

PCDDs, PCDFs, PBDEs, PBBs, and PCNs (reported as individual or coeluting congeners) using highresolution gas chromatography/highresolution mass spectrometry

(HRGC/HRMS). Organochlorine pesticides [including total DDT ( o,p ' DDT, DDD,

DDE and p,p ' DDT, DDD, DDE), HCB, HCH ( αHCH, βHCH, γHCH), heptachlor, aldrin, chlordane (oxy, γ, α), nonachlor (trans, cis), and mirex] were measured using lowresolution gas chromatography/mass spectrometry and gas chromatography with electron capture detection.

Extraction and cleanup procedures, instrumental analysis and conditions, and quality assurance/quality control criteria used for PCBs, PCDDs, and PCDFs for the

Regional Contaminants Laboratory (Fisheries and Oceans Canada, Sidney, BC) are described elsewhere (Ross et al. 2000; Ikonomou et al. 2001). The sample batch for

PCBs, PCDDs, and PCDFs included a procedural blank, a replicate, and a certified 25 reference material (herring) sample (Ikonomou et al. 2001). Based upon 45 individual, congenerspecific PCB analyses by this laboratory, we calculated the reproducibility of this certified reference material as 94.5 ± 15.4 µg/kg wet weight (mean ± standard deviation) as assessed by homolog group. The calculated reproducibility for 2,3,7,8 tetrachlorodibenzofuran (TCDF) during the same exercise was 2.46 ± 0.42 µg/kg wet weight.

Polybrominated diphenyl ethers, PBBs, PCNs, and organochlorine pesticides were analyzed by AXYS Analytical Services (Sidney, BC, Canada) according to their laboratory procedures and criteria. The PBDE method is based on the U.S.

Environmental Protection Agency (U.S. EPA) draft analytical methods and procedures number 1614 (USEPA Office of Science and Technology 2003). The gas chromatographymass spectrometry method for organochlorine pesticide determination is based on U.S. EPA 8270 (USEPA Solid Waste and Emergency Response 1998) modified to include isotope dilution quantification, and the gas chromatography with electron capture detection method is based on a modified U.S. EPA 8081 method (USEPA Solid

Waste and Emergency Response 1998). The PCN and PBB analyses were carried out using an inhouse GC/HRMS method with isotope dilution or internal standard quantification.

Samples for PBDEs, PBBs, PCNs, and organochlorine pesticides were spiked with 13 Clabelled surrogate standards and then ground with anhydrous sodium sulphate.

Samples were transferred to a soxhlet thimble, surrogate standard was added, and samples were refluxed for 16 h with dichloromethane (DCM). The extract was eluted through a gel permeation column with 1:1 DCM:hexane. The extract was applied to a 26 partially deactivated Florisil column and eluted with hexane followed by 15:85

DCM:hexane. Eluates were then combined and eluted with 1:1 DCM:hexane and each fraction concentrated. All solvents used were pesticide grade (J.T. Baker, Phillipsburg,

NJ, USA).

Instrumental analysis by lowresolution gas chromatography/mass spectrometry was carried out using a Finnigan INCOS 50 mass spectrometer (Finnigan, Sunnyvale,

CA, USA) equipped with a Varian 3400 GC (Varian, Palo Alto, CA, USA), a CTC a

Prolab/Enviroquant data system. Chromatographic autosampler (LEAP Technologies,

Carrboro, NC, USA) and separation was achieved through a DB5 capillary chromatography column (60 m x 0.25 mm, 0.1 µm film). Instrumental analysis by gas chromatography with electron capture detection was carried out using a HewlettPackard

5890 GC (Agilent Technologies, Wilmington, DE, USA) with a 63 Ni electron capture detector and a DB5 Durabond fused silica capillary column (60 m x 0.25 mm, 0.10 µm film). Instrumental analysis by highresolution GC/MS was carried out using an Ultima high resolution MS (Micromass, Manchester, UK) equipped with a HewlettPackard

5890 GC (Agilent Technologies) and a DB5 capillary column (60 m x 0.25 mm, 0.10

µm film).

utrient analyses

Nutrient analyses including total calories (energy), carbohydrate, lipid, and protein were determined through carbohydrate analysis by M.B. Laboratories (Sidney,

BC, Canada). 27

Stable isotopes

Subsamples of whole prey homogenates for stable isotope analyses were freeze dried for 48 to 72 h and then ground to a fine powder using a mortar and pestle. Seal skin biopsies were collected from freeranging animals inhabiting Puget Sound ( n=12 pups; n= 8 adults; Gertrude Island) and the Strait of Georgia ( n=6 pups; n=15 adults; Fraser

River Estuary). Seals were captured and released following sampling as described elsewhere (Ross et al. 2004) under the auspices of animal care and scientific permits issued by Fisheries and Oceans Canada, and in accordance with the Canadian Council of

Animal Care. Skin tissues were freezedried for 48 to 72 h and cut into finite piece fragments using scalpel blades. Bulk stable carbon and nitrogen isotope ratio ( 15 N: 14 N and 13 C: 12 C) measurements were made using a Fisons NA 1500 elemental analyzer

(Milano, Italy) interfaced to a Finnigan 252 isotope ratio mass spectrometer (Bremen,

Germany). Isotopic composition is expressed in δ notation as the proportional deviation in parts per thousand (‰) of the isotope ratio in a sample from that of a standard:

δX = ( Rsample /Rstandard – 1) X 1,000 (1)

13 15 13 12 15 14 where X is C or N, and Rsample and Rstandard are the ratios of C: C or N: N for the sample and standard (Hobson et al. 1997). The standards used for carbon and nitrogen isotope ratio analyses included PeeDee belemnite (PDB), atmospheric nitrogen (air), and acetanilide (Baker), an inhouse standard.

Data reporting

The numbers of PCB congeners detected either as individual or coeluting congeners were 183 and 176 out of 209 in the Puget Sound and the Strait of Georgia food 28 baskets, respectively. The numbers of PCDD congeners detected were 5 and 5 out of 75 congeners for the two food baskets, and the numbers of PCDFs were 3 and 2 out of 135 congeners, respectively. The numbers of PBDE congeners detected either as individual or coeluting congeners were 24 and 19 out of 209, numbers of PBB congeners were 10 and 8 out of 209, and numbers of PCN congeners were 47 and 40 out of 75, for the Puget

Sound and Strait of Georgia food baskets, respectively.

Concentrations were reported for individual PCB congeners when 70% or more of samples (within a sample batch of ten) had detectable concentrations. Congener concentrations detected in less than 70% of samples were not reported nor included in total concentrations. Where 70 to 100% of samples had detectable concentrations, detection limit values, described elsewhere (Ikonomou et al. 2001), were substituted for nondetected values. All results have been reported on both a wet weight basis and lipid weight basis. Toxic equivalent concentrations (TEQs) to 2,3,7,8tetrachlordibenzop dioxin (TCDD) were calculated for PCBs, PCDDs, and PCDFs using World Health

Organization International toxic equivalent factors (TEFs) for humans and wildlife (Van den Berg et al. 1998).

For the purposes of this paper, we compare lipidadjusted contaminant concentrations for the two food baskets, while we use wetweight contaminant concentrations that are equalized to lipid across basins to explore dietary exposure scenarios. In other words, we assume for the purposes of our exposure estimates that seals of both Strait of Georgia and Puget Sound consume similar amounts of lipid, and hence, higher total mass of prey for Puget Sound seals.

2.4 Results and Discussion

To help explain why Puget Sound seals are more contaminated with PCBs than the Strait of Georgia seals (Ross et al. 2004), we carried out a food basket approach to assessing dietary exposure by seals of both basins to POPs. Our food basket approach generated integrated dietary signals for two adjacent and nonmigratory seal populations that consume different diets, something that enables the exploration of such features as dietary exposure and exposureassociated health risks.

Food basket contaminant concentrations

Highest total concentrations of POPs in our two food baskets (Puget Sound and

Strait of Georgia) included those that have been regulated in North America (i.e., PCBs and DDT), as well as surprisingly high levels of the largely unregulated PBDEs ( Table

2.3). Concentration rankings for the top three POP classes were ΣPCBs > ΣPBDEs >

ΣDDT for the Puget Sound food basket, and ΣPCBs > ΣDDT > ΣPBDEs for the Strait of

Georgia food basket. The ratio of PBDE: PCB concentrations (0.2 for Puget Sound; 0.3 for Strait of Georgia) underscore the emerging risk that PBDEs present to marine wildlife in this coastal environment. Although PBDEs are largely unregulated in North America and toxicological data is limited, their chemical similarity to PCBs, laboratory rodent studies (Hallgren et al. 2001) and associative evidence in grey seal field studies (Hall et al. 2003) suggest that they possess endocrinedisrupting potential. Moves in Canada and the USA are currently being made to regulate or voluntarily withdraw penta and octa

PBDE formulations.

The Puget Sound food basket was seven times more contaminated with PCBs

(2.90 mg/kg lipid) than the Strait of Georgia seal food basket (0.41 mg/kg lipid) (Table

3). This is consistent with previous observations that Puget Sound harbour seal pups are seven times more contaminated (18.1 ± 3.1 mg/kg lipid) than those inhabiting the Strait of Georgia (2.5 ± 0.2 mg/kg lipid) (Ross et al. 2004).

The ΣTEQs of the Puget Sound food basket were 4.7 times higher (lipid wt) than the Strait of Georgia food basket ( Table 2.4 ). The ΣPCB TEQs largely accounted for the

ΣTEQs in Puget Sound food basket, whereas the ΣPCDD and ΣPCDF TEQs accounted for a greater proportion of ΣTEQs in the Strait of Georgia food basket. While higher

PCDD and PCDF concentrations were previously documented for Strait of Georgia harbour seals relative to Puget Sound harbour seals (Ross et al. 2004), the higher PCDD and PCDF concentrations in the Puget Sound food basket compared to the Strait of

Georgia food basket may indicate a contaminantrelated induction of detoxifying enzymes in the more contaminated seals, which could preferentially eliminate planar

PCDDs and PCDFs. Another study noted the relative ease with which harbour seals can eliminate planar POPs, such as 2,3,7,8TCDD (De Swart et al. 1995), compared to nonplanar or globular POPs, such as the monoortho PCBs. Alternatively, these results may indicate that the sampled seals were feeding on prey items or in areas other than those captured by our food basket collections.

Table 2.3. Concentrations of persistent organic pollutants in harbour seal food baskets from the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) expressed on both a lipid and wet weight basis. Although concentrations of most classes of contaminants were higher in the Puget Sound food basket than in the Strait of Georgia food basket, larger differences were observed with the flameretardants and industrial byproducts than with the organochlorine pesticides. Subtotals are listed in italics. PS = Puget Sound; SG = Strait of Georgia

µg/kg Wet weight µg/kg Lipid weight Sum congeners/isomers a (except for ΣΣΣPCDD and ΣΣΣPCDF) Strait of Puget x factor Strait of Puget x factor Georgia Sound difference Georgia Sound difference food food (PS/SG) food food (PS/SG) basket basket basket 1 basket 2 Flame retardants ΣPCB 16.91 54.03 3.2 407.51 2904.93 7.1 ΣPBDE 5.53 11.81 2.1 120.15 590.32 4.9 ΣPBB 0.02 0.09 4.5 0.52 4.41 8.5 ΣPCN 0.06 0.14 2.3 1.21 7.11 5.9

Industrial byproducts (ng/kg) ΣPCDD 1.42 2.23 1.6 0.03 0.12 4.0 ΣPCDF 0.86 1.06 1.2 0.02 0.06 3.0

aPCB = polychlorinated biphenyl; PBDE = polybrominated diphenyl ether; PBB = polybrominated biphenyl; PCN = polychlorinated naphthalene; PCDD = polychlorinated dibenzopdioxin; PCDF = polychlorinated dibenzofuran; DDT = dichlorodiphenyltrichloroethane; DDD = dichlorodiphenyldichloroethane; DDE = dichlorodiphenyldichloroethylene; HCB = hexachlorobenzene; HCH = hexachlorocyclohexane. b Lipid content for the Strait of Georgia food basket was 4.15 % for PCBs, PCDDs, PCDFs, and 4.6% for all other contaminants. c Lipid content for the Puget Sound food basket was 1.86 % for PCBs, PCDDs, PCDFs, and 2.0 % for all other contaminants.

Organochlorine pesticides ΣDDT (DDT, DDD, DDE) 7.21 5.06 0.7 156.80 253.20 1.6 o,p' −DDT, p,p'DDT 0.42 0.24 0.6 9.09 11.85 1.3 o,p' −DDD, p,p'DDD 0.95 0.59 0.6 20.80 29.75 1.4 o,p' −DDE, p,p'DDE 5.84 4.23 0.7 126.91 211.60 1.7 ΗCB 1.00 0.39 0.4 21.74 19.50 0.9 ΣHCH ( α, β, γ) 1.45 0.97 0.7 31.52 48.65 1.5 alpha (α) 0.74 0.27 0.4 16.09 13.50 0.8 beta (β) 0.55 0.61 1.1 11.95 30.50 2.6 gamma (γ) 0.16 0.09 0.6 3.48 4.65 1.3 Heptachlor 0.05 0.08 1.6 1.02 4.25 4.2 Chlordane (oxy, γ, α) 0.82 0.89 1.1 17.76 44.50 2.5 Nonachlor (trans, cis) 1.20 1.12 0.9 26.09 58.00 2.2 Mirex 0.02 0.03 1.5 0.39 1.55 4.0 33

Table 2.4. Toxic equivalents (TEQs) to 2,3,7,8tetrachlorodibenzopdioxin (TCDD) for polychlorinated biphenyls (PCBs), polychlorinated dibenzop dioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs) in Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets. The PCBs represent the largest contributor to the ΣΣΣTEQs in the Puget Sound food basket (77%), whereas ΣΣΣPCDD/Fs are the largest contributors to the ΣΣΣTEQs in the Strait of Georgia food basket (55%). Subtotals for PCB TEQs are listed in italics.

Strait of Georgia Puget Sound food basket food basket TEQ % ΣTEQ TEQ % ΣTEQ ng/kg lipid ng/kg lipid

ΣPCBs 6.63 44.4 54.46 77.4 Σ planar PCBs 3.57 23.9 28.66 40.8 Σ monoortho PCBs 3.06 20.5 25.81 36.7

ΣPCDDs 4.99 33.4 7.74 11.0 ΣPCDFs 3.33 22.3 8.12 11.5

ΣTEQs 14.94 100 70.32 100

The concentration rankings of the major organochlorine pesticides in the Puget

Sound food basket were ΣDDT > Nonachlor > Σ HCH > Chlordane > HCB > Heptachlor

> Mirex, and in the Strait of Georgia food basket were ΣDDT > Σ HCH > Nonachlor >

HCB > Chlordane > Heptachlor > Mirex (Table 3). The high relative concentrations of

DDT breakdown products, o,p 'DDD + p,p 'DDD (i.e., 12% and 13% of ΣDDT in Puget

Sound and Strait of Georgia, respectively), and o,p 'DDE + p,p 'DDE (i.e., 84% and 81% of ΣDDT), compared with unmetabolized o,p 'DDT + p,p 'DDT (i.e., 5% and 6% of

ΣDDT), suggests that DDT contamination of these basins is old, with relatively little in the way of fresh inputs. 34

Although the concentrations of PCBs, PCDDs, PCDFs, and flameretardants were considerably higher in the Puget Sound food basket, many of OC pesticides, notably those with high volatilities (e.g., HCB, HCH), were more equally represented in the two food baskets. Hexachlorocyclohexane and HCB are often touted as pesticides which exemplify the grasshopper effect by readily volatilizing from source regions and moving great distances prior to deposition and incorporation into remote aquatic food webs

(Wania and Mackay 2001). Differences in regulations, historical applications/disposal, physicochemical properties, transport and fate processes in the regional environment, as well as differences in uptake, bioaccumulation, and metabolism within the many species present in the two food baskets, contribute in varying ways to the observed differences in contaminant concentrations and contaminant patterns.

Food basket contaminant patterns

Contaminant homolog patterns reveal some basic differences between the Puget

Sound and the Strait of Georgia food baskets (Figure 2.2). The PCB homolog pattern appears heavier in the Puget Sound food basket than in the Strait of Georgia food basket

(Figure 2.2), consistent with previous observations in harbour seals (Ross et al. 2004).

Puget Sound is a highly industrialized area where physical and oceanographic features favor the retention of more heavily chlorinated PCBs. Less chlorinated PCB congeners are more volatile, leading to their relative depletion in source regions and their subsequent atmospheric transport into more remote areas (Ross et al. 2004; Wania and

Mackay 2001; Ross et al. 2004).

Figure 2.2. Polychlorinated biphenyl (PCB), polychlorinated dibenzopdioxin (PCDD), and polychlorinated dibenzofuran (PCDF), polybrominated diphenyl ether (PBDE), polychlorinated naphthalene (PC), and polybrominated biphenyl (PBB) homolog group (i.e., degree of halogenation of congeners) patterns in the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets. (White: Strait of Georgia; Pattern: Puget Sound.) 36

Polychlorinated dibenzopdioxin/F patterns in the Strait of Georgia food basket

(Figure 2.2) appear consistent with an extensive history of pulp and paper and wood treatment facilities (pentachlorodibenzopdioxins [PnCDDs], hexachlorodibenzop dioxins [HxCDDs], tetrachlorodibenzofurans [TCDFs], pentachlorodibenzofurans

[PnCDFs]), coupled with industrial/municipal combustion (octachlorodibenzopdioxins

[OCDDs]), while the Puget Sound food basket appears to reflect a more pronounced combustion signal (Yunker and Cretney 2000). Temporal trend analysis of environmental samples suggest that regulatory controls have contributed to greatly reduced inputs of dioxins and furans into the Strait of Georgia over the past 15 years (Hagen et al. 1997).

The PBDE patterns were similar in the two food baskets, consisting predominantly of the tetra and penta brominated compounds observed in other biota in the region (Rayne et al. 2003; Rayne et al. 2004) (Figure 2.2). Polychlorinated naphthalenes patterns show a predominance of tetra and penta chlorinated compounds in both food baskets. Tetra, penta, and hexa chlorinated naphthalenes were the predominant homolog patterns observed in Baltic Sea benthic food chain species

(Lundgren et al. 2002). Polybrominated biphenyl patterns in both food baskets were characterized by tri through hexa brominated biphenyls, with tetra through hexa brominated biphenyls being most prominent. Our food basket PBB patterns may, in part, reflect hexaBB commercial mixtures and/or debromination products of octa, and deca

BB commercial mixtures (De Boer et al. 2000).

The PCB pattern differences between food baskets of Puget Sound and Strait of

Georgia are more clearly captured when individual congener profiles are corrected to

PCB153 and plotted as ratios between the two basins ( Figure 2.3). These PCB153 37 corrected profiles clearly indicate that the Puget Sound food basket is characterized by a heavier (more chlorinated) PCB mixture, consistent with its role as a source or PCB hotspot in the regional environment (Ross et al. 2004). These observations in both seals and seal food baskets also provide further evidence of relative dispersion of lighter (i.e., lower log H and Kow ) PCB congeners in food chains away from sources and into more remote regions and food chains.

Stable isotope ratios in seals and their prey from different regions should provide an integrated measure of trophic level (Hobson et al. 1997) as well as generate information on feeding ecology over extended time periods (Kurle and Worthy 2001).

Although limited sample size precluded us from determining whether significant differences existed between our two food baskets, nitrogen ratios indicate that the Puget

Sound food basket ( δ15 N= 14.4 ± 0.1) was slightly higher in trophic position than the

Strait of Georgia food basket ( δ15 N= 12.9 ± 0.4). In addition, slight differences in δ15 N ratios were observed between Puget Sound adult harbour seals and young harbour seals, suggesting that Puget Sound seals feed slightly higher trophically than the Strait of

Georgia seals (results not shown). While this may partly explain the observed differences, a trophic levelassociated bioaccumulation of contaminants is very unlikely to explain a sevenfold interbasin difference in the contamination of seals. Although pups are not yet directly feeding on prey, they will reflect the integrated intakes of the transplacental and lactational transfer from their mothers (Ross et al. 1993; Pomeroy et al. 1996).

Figure 2.3. Ratio of polychlorinated biphenyl (PCB) patterns in the Puget Sound (WA, USA) to the Strait of Georgia (BC, Canada) harbour seal food baskets. The Strait of Georgia food basket is dominated by the lowerchlorinated (lighter) PCBs, while the Puget Sound food basket is dominated by the higherchlorinated (heavier) PCBs, similar to patterns observed in seals inhabiting these two basins (Ross et al. 2004). We lognormalized the food basket congener data from both basins to PCB 153 to eliminate concentration differences from congener patterns: log {(Puget Sound Food Basket [PCB congener] / [PCB153]) / (Strait of Georgia Food Basket [PCB congener] / [PCB153])}.

Contamination from feeding differences or source differences?

The contaminant concentrations in the Puget Sound and Strait of Georgia food baskets explain the sevenfold difference in PCB levels in harbour seals inhabiting the two basins, yet beg the question of whether the Puget Sound harbour seals are more contaminated due to differences in feeding (prey selection) or whether Puget Sound is a generally more contaminated ecosystem. Two scenarios may underlie the observed differences between the two basins: Seals and the food baskets may be more contaminated in Puget Sound because the Puget Sound ecosystem is seven times more contaminated than the Strait of Georgia; or seals select more contaminated prey species in Puget Sound (e.g., longer lived, higher trophic level, more contaminated prey). If the same prey species are equally contaminated in both basins, we would predict that if Puget

Sound prey samples were used to construct a Strait of Georgia food basket (i.e., the menu consumed by Strait of Georgia seals), this would decrease contaminant concentrations to the level observed for the Strait of Georgia food basket. Alternatively, if the same prey species were more contaminated in Puget Sound as a consequence of local (i.e., Puget

Soundspecific) environmental contamination, we would not expect such a decrease when Puget Sound prey samples were used to construct a Strait of Georgia food basket.

Our two additionally constructed food baskets found Puget Sound food basket PCB levels to increase with the Strait of Georgia seal diet (from 2904 µg/kg lipid to 4363 µg/kg lipid) and Strait of Georgia food basket PCB levels to decrease with the Puget Sound seal diet (from 407 µg/kg lipid to 354 µg/kg lipid). These results clearly indicate that prey selection does not explain the observed differences, and that the degree of contamination of the original Puget Sound food basket (and harbour seals) relates to an effect of local 40 contamination within Puget Sound.

Risk characterization using prey tissue concentrations

Food basket ΣPCB (Table 2.3) and ΣPCB TEQ (Table 2.4) concentrations were compared with published estimated dietary threshold concentrations (based on biomagnification factors and measured as the mean of the noobservedadverse effect level and lowestobservedadverse effect level) as a means of assessing potential health risks associated with dietary exposure in seals. Both Puget Sound and Strait of Georgia food basket concentrations were below a PCB dietary threshold concentration of 140

µg/kg diet wet weight for immune and reproductive impairment in seals (Kannan et al.

2000) based on semifield feeding studies (De Swart et al. 1994; Ross et al. 1995; Boon et al. 1987; Brouwer et al. 1989; Reijnders 1986). Both were also below the 250 µg/kg wet weight threshold for reproductive toxicity in mink ( Mustela vison ) (Kannan et al.

2000) established in laboratory feeding studies (Tillitt et al. 1996; Heaton et al. 1995).

However, the Strait of Georgia food basket is approaching, and the Puget Sound food basket exceeded, the dietary threshold of 20 µg/kg wet weight for Vitamin A disruption in European otter ( Lutra lutra ) (Kannan et al. 2000) established in semifield feeding studies (Murk et al. 1998; Smit et al. 1996; Leonards et al. 1997).

The Strait of Georgia food basket ΣPCB TEQs (0.28 ng TEQ/kg diet wet wt) fell below the Canadian PCB tissue residue guidelines for the protection of mammalian wildlife consumers of aquatic biota (0.79 ng TEQ/kg diet wet wt). However, the Puget

Sound food basket ΣPCB TEQs (1.01 ng TEQ/kg diet wet wt) exceeded these guidelines

(Canadian Council of Ministers of the Environment 1999; Canadian Council of Ministers 41 of the Environment 2002). Both Puget Sound and the Strait of Georgia food baskets fell below the U.S. National Academy of Sciences/National Academy of Engineering

(NAS/NAE) ΣPCB guideline (500 µg/kg diet wet wt) and the New York State

Department of Environmental Conservation (NYSDEC) ΣPCB guideline (110 µg/kg diet wet wt) for fisheating wildlife (Canadian Council of Ministers of the Environment

2002). The Canadian guidelines are more conservative than the NAS/NAE and

NYSDEC guidelines because they consider chronic toxicity, potential carcinogenicity, and reproductive effects (Canadian Council of Ministers of the Environment 1999).

Given that we have observed adverse health effects (Vitamin A disruption) in Puget

Sound harbour seals (Simms et al. 2000), our food basket results may provide a means to generate updated guidelines for the protection of wildlife.

Risk characterization using estimated daily intakes

A comparison of tissue residue concentrations in the two seal food baskets to consumption guidelines ignores an important consideration. Qualitative features such as composition and energy content (related to lipid content) of the seal diet influence quantitative aspects of dietary intakes (Markussen et al. 1989). Harbour seals likely prefer highenergy prey such as herring whenever available (Olesiuk 1993). Studies of pinniped energetics suggest that seals consuming a low energy (low lipid content) diet need to compensate by consuming more mass of food than seals consuming a high energy

(high lipid content) diet (Markussen et al. 1994). Energy content of prey, therefore, needs to be considered when estimating diet consumption (Markussen et al. 1989;

Markussen et al. 1990; Markussen et al. 1994). Nutritional content (including energy, 42 protein, carbohydrate, and lipid) differed slightly between the Puget Sound and the Strait of Georgia food baskets ( Table 2.5), leading us to normalize contaminant exposure estimates on the basis of lipid content. The consistency in lipidadjusted PCB concentrations across basins for both seals (7.2x) and seal food baskets (7.1x) may, in fact, provide support for the possibility that seals forage on a lipidweight basis

(Markussen et al. 1989; Markussen et al. 1994; Markussen et al. 1990).

Table 2.5. utritional content of the Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) harbour seal food baskets. These analyses suggest that Puget Sound harbour seals may have to consume more prey in order to achieve the same energy intake as Strait of Georgia seals. Analyses conducted by M.B. Laboratories (Sidney, BC, Canada), with the exception of lipid content, which represents mean ± standard deviation values generated by AXYS Analytical Services (Sidney, BC), and the Regional Contaminant Laboratory (Fisheries and Oceans Canada, Sidney, BC). PS = Puget Sound; SG = Strait of Georgia

utrient Units Strait of Puget Sound x factor Georgia food basket difference food basket /100g (PS/SG) /100g

Calories (Energy) kj 398 285 0.7 Protein % 15.1 14.7 1.0 Total lipid (Fat) % 4.4 ± 0.3 1.9 ± 0.1 0.4 Carbohydrate % 2.8 5.4 2.0 Ash % 3.0 2.9 1.0 Moisture % 74.8 75.8 1.0

Characterizing risks associated with dietary contaminant concentrations may in this manner be better captured in the calculation of the estimated daily intake. Tissue residue guidelines used alone do not provide enough information to fully estimate dietary exposure to contaminants. Since prey consumption differs between harbour seals 43 inhabiting Puget Sound and the Strait of Georgia (Tables 2.1 and 2.2), estimated dietary exposure using calculated EDIs based on amount of prey (quantity) with consideration of lipid content (quality) better reflects contaminant intake. As expected, the estimated daily intakes and estimated annual intakes of POPs for harbour seals ( Table 2.6) indicate that Puget Sound seals are exposed to much higher dietary concentrations of PCBs than

Strait of Georgia seals.

In captive feeding studies where harbour seals were fed similar prey (herring) which differed in lipid content, intake was adjusted accordingly to compensate for energy requirements and toxicological exposure assessments; body weights in the two groups remained similar during this 30month study (De Swart et al. 1994). Impaired immune function and diminished serum vitamin A were observed in juvenile harbour seals having an EDI of 1460 µg ΣPCB /day and 288 ng ΣTEQ/day (De Swart et al. 1994). Our EDI for adult seals based on both Puget Sound and Strait of Georgia food baskets for ΣPCB and ΣTEQ concentrations fell below these EDIs, but higher metabolic requirements in younger seals would lead to higher contaminant intakes per kg body weight. We estimate that a 25 kg seal would be exposed to 3.2x higher concentrations of PCBs on a body weight basis than adults, and might therefore be at increased risk for adverse health effects.

Table 2.6. Estimated daily intake and estimated annual intake of POPs, on a wet weight basis, by adult harbour seals for Strait of Georgia (BC, Canada) and Puget Sound (WA, USA) based on contaminant analyses of our food baskets. Energetic studies suggest that pinnipeds may consume on a lipid content (energy) basis (Markussen et al. 1990); in this case, Puget Sound seals would consume more prey on a wet weight basis to compensate for lower lipid content (Strait of Georgia = 4.4 ±±± 0.3% vs Puget Sound = 1.9 ±±± 0.1%). Based on energetic values derived from captive studies (Markussen et al. 1990), Olesiuk estimated the daily food requirement for a Strait of Georgia adult harbour seal (60 kg female; 75 kg male) to be 2.5 kg of prey (Olesiuk 1993). If Puget Sound seals consumed an energetically equivalent amount as Strait of Georgia seals, they would have to consume approximately 5.0 kg of prey.

Estimated daily intake Estimated annual intake µg/day mg/year (except for ΣΣΣPCDD, (except for ΣΣΣPCDD and Sum congeners/isomers a ΣΣΣPCDF and ΣΣΣTEQ ΣΣΣPCDF µg/year) ng/day) Strait of Puget Sound Strait of Puget Sound Georgia Georgia

Flame retardants ΣPCB 42.28 270.15 15.43 98.60 ΣPBDE 13.82 59.05 5.05 21.55 ΣPBB 0.06 0.44 0.02 0.16 ΣPCN 0.14 0.70 0.05 0.26

Industrial byproducts ΣPCDD 3.50 11.00 1.28 4.02 ΣPCDF 2.15 5.50 0.78 2.01

ΣTEQ (PCB/PCDD/PCDF) 1.55 6.54 0.57 2.39

OC pesticides ΣDDT (DDT, DDD, DDE) 18.02 25.20 6.58 9.20 ΗCB 2.50 1.95 0.91 0.71 HCH ( α, β, γ) 3.62 4.85 1.32 1.77 Heptachlor 0.12 0.40 0.04 0.15 Chlordane (oxy, γ, α) 2.05 4.45 0.75 1.62 Nonachlor (trans, cis) 3.00 5.60 1.10 2.04 Mirex 0.05 0.15 0.02 0.05 aPCB = polychlorinated biphenyl; PBDE = polybrominated diphenyl ether; PBB = polybrominated biphenyl; PCN = polychlorinated naphthalene; PCDD = polychlorinated dibenzopdioxin; PCDF = polychlorinated dibenzofuran; DDT = dichlorodiphenyltrichloroethane; DDD = dichlorodiphenyldichloroethane; DDE = dichlorodiphenyldichloroethylene; HCB = hexachlorobenzene; HCH = hexachlorocyclohexane. 45

2.5 Conclusion

To our knowledge, this is the first time that a food basket approach has been applied to estimating POP intakes in a marine mammal. Sample size in our study is limited and does not incorporate all of the temporal, spatial, and prey size variations that are at play in the feeding ecology of freeranging seals. However, our food baskets consisted of more than 200 individual prey items sampled from each basin, and therefore provide an integrated dietary intake signal for harbour seals in these areas. We conclude that a food basket approach represents: A relatively straight forward method of assessing dietary exposure to real world mixtures of POPs if prey consumption patterns are known; a realistic and integrated dietary signal as a basis for exploring issues of dietary exposure, biomagnification, metabolism, and health risks; and information that is complementary to other studies which rely on the harbour seal as a sentinel of marine ecosystem contamination. Further research into characterizing accumulation patterns of POPs in northeastern Pacific Ocean food webs is needed to better understand sources, pathways, and fate of new and legacy chemicals. Our study demonstrates that important differences in food web contamination exist between Puget Sound and the Strait of Georgia, and also highlights the emerging concern of the largely unregulated PBDEs in this region. The recent observations of PCBs and flame retardants in the region’s resident and transient killer whales ( Orcinus orca ) (Ross et al. 2000; Rayne et al. 2004), and in marine mammals from other regions of the world (Rayne et al. 2004; Ikonomou et al. 2002), further underscore the importance of food webbased contaminant studies.

Chapter 3: Biomagnification of polychlorinated biphenyls in a harbour seal

(Phoca vitulina ) food web from British Columbia (Canada) †

3.1 Abstract

The accumulation of the environmentally persistent polychlorinated biphenyls

(PCBs) has been characterized for a harbour seal food web in the Strait of Georgia,

British Columbia, Canada. Concentrations of the recalcitrant PCB congener 153 and

ΣPCBs increased with δ15 N ratios in the seal food web ( r2 = 0.38 and r2 = 0.34, respectively; p < 0.0001 for both) illustrating the bioaccumulative nature of PCBs. The preypredator trophic leveladjusted biomagnification factor (BMF TLC ) for our seal food basket to harbour seals is 13.4 and the overall seal food web magnification factor

(FWMF) is 3.6 for ΣPCBs. The highest BMFs TLC and FWMFs were observed among octa and heptachlorinated PCBs, and values increased with log K ow . This provides evidence that chemical structure and physicochemical properties influence biomagnification and trophic transfer of PCBs in the seal food web. Principal components analyses (PCA) further suggested that although both trophic level and log

Kow influence PCB biomagnification in seals, metabolic capacity of seals, (based on dominance of structural activity groups 1 and 2), plays an important role in food web bioaccumulation. We estimate there to be approximately 77 kg of PCBs within the Strait of Georgia biota, of which 73 % are found in resident species, and 27 % in migratory species. Our results provide insight into the biomagnification of the legacy PCBs in a

† Adapted from Cullon DL, Yunker MB, Christensen JR, Whiticar MJ, Macdonald RW, Ross PS. 2010. Biomagnification of polychlorinated biphenyls in a harbour seal ( Phoca vitulina ) food web from British Columbia (Canada), manuscript in prep . 47 moderately contaminated coastal basin, and also highlight the continued biological importation of these contaminants from the open Pacific Ocean.

3.2 Introduction

Persistent organic pollutants (POPs), such as polychlorinated biphenyls (PCBs), are lipophilic, not easily metabolized, and can reach high concentrations in top predator marine mammals. Polychlorinated biphenyls were mass produced between 1929 and

1977 for use in transformers and capacitors, plastics, ink, and as flame retardants, due to their high chemical stability and low aqueous solubility (Borlakoglu and Dils 1990).

Although banned by most industrialized nations in the 1970s, chemical persistence, volatility, and continuous cycling among environmental compartments (air, water, soil, biota) has resulted in ongoing ecological health concern.

Biomagnification is the process where a chemical concentration in an organism

(predator) exceeds that of its diet (prey) due to dietary absorption (Mackay and Fraser

2000; Gobas and Morrison 2000), whereas trophic accumulation of contaminants in aquatic food webs (increasing chemical concentration in organisms with increasing trophic level) may include biomagnification and bioconcentration (Borgå et al. 2004).

Stable nitrogen isotope values ( δ15 N) are useful for determining trophic level or feeding position within a food web as assimilated over time (Vander Zanden and Rasmussen

1999; Cabana and Rasmussen 1996), and thus can be used in characterizing trophic transfer and biomagnification of contaminants. Trophic transfer of contaminants along with energy (food) transfer from prey to predator is a complex process, influenced by factors such as age, sex, body size, life cycle, seasonality, reproduction, migration, 48 biotransformation, lipids, and feeding ecology (Borgå et al. 2004). Using trophic level adjusted biomagnification factors helps to determine biomagnification independent of trophic level lipid increases in animals (Fisk et al. 2001). Food web magnification is a further means of determining contaminant increases within the entire food web, offering an overall mean characterization of trophic transfer within food webs (Hop et al. 2002).

These tools have been used in many aquatic food web studies (Fisk et al. 2001; Ruus et al. 2002; Borgå et al. 2001; Hoekstra et al. 2003); however, many knowledge gaps exist in Pacific northwest marine mammal food webs.

The Strait of Georgia, situated between the British Columbia mainland and

Vancouver Island, is a semienclosed marine basin covering approximately 6900 km 2 that has a mean depth of 156 metres. The basin has large fresh water input from many rivers

(principally the Fraser River), and relatively high sedimentation rates. These waters support a plethora of human activities, including commercial and recreational fishing, shipping and passenger ferries, pulp and paper mills, urban centers, and fish farms. The

Strait of Georgia has a rich ecosystem providing habitat for many species, including more approximately 39,000 harbour seals (Fisheries and Oceans Canada 2010).

Harbour seals occupy high trophic levels, are nonmigratory, and have a relatively long lifespan (2030 years) (Ross 2000). With these characteristics, harbour seals become highly useful indicators of marine ecosystem contamination and can provide an integrated overview of PCB contamination in the Strait of Georgia food web. The demonstrated effects of PCBs on the health of harbour seals (reproductive impairment, immunotoxicity, vitamin A and thyroid hormone disruption (De Swart et al. 1996; Ross et al. 1996; Reijnders 1986; Simms et al. 2000)) attests to the importance of continued 49 research into the sourcetransport and fate features of this legacy contaminant. In an effort to better understand dietary exposure in Strait of Georgia harbour seals, we previously constructed a harbour seal food basket which consisted of their preferred prey species (Cullon et al. 2005; Olesiuk 1993). In the present study, we conduct an individual speciesbased analysis of harbour seal prey and individual harbour seals in an effort to characterize biomagnification and trophic transfer of PCBs in the Strait of

Georgia harbour seal food web.

3.3 Materials and Methods

Sample collection

Blubber biopsies from 10 livecaptured freeranging adult harbour seals (5 male:

5 female) were collected from the Fraser River (n = 3) and Boundary Bay (n = 7) in the

Strait of Georgia, BC, Canada in 2001 ( Figure 3.1). Blubber biopsies from 10 live captured harbour seal pups were collected from Hornby Island, BC, Canada in 2001.

Details of capture and biopsy sampling are described elsewhere (Simms et al. 2000). For seal pups age estimates were determined by umbilicus physiology and for adult seals

(except for PV0133 and 34) by teeth. All handling and biopsy procedures were carried out according to Canadian Council on Animal care guidelines by research personnel with scientific research permits. Meristic data for sampled harbour seals are provided in

Table 3.1.

Seal ‘preysized’ individual fish (i.e. < 25 cm), including Pacific lingcod

(Ophiodon elongatus ), rockfish (Family Scorpaenidae ), Pacific tomcod ( Microgadus proximus ), Pacific midshipman ( Porichthys notatus ), English sole ( Parophrys vetulus ), 50 and Pacific hake ( Merluccius productus ) were collected on board the CCGS Vector and the CCGS Ricker between 49º26΄N, 124º38΄W and 48º53΄N, 123º27΄W in the Strait of

Georgia at depths of 2050 metres between August and September 2001. Methods and criteria for construction of the Strait of Georgia harbour seal food basket, which consisted of more than 200 individual prey items, are described in detail elsewhere (Cullon et al.

2005). The composition of our seal food basket sample represents approximately 96 % of the seal’s annual consumption.

Six Pacific herring ( Clupea pallasi ) were collected from central Strait of Georgia in January 2001. Sockeye salmon ( Oncorhynchus nerka ) were collected from the Fraser

River in October 2000. Twelve adult chinook salmon ( Oncorhynchus tshawytscha ) were collected, six from Johnstone Strait and six from the mouth of the Fraser River in October

2000. In addition, twelve chinook smolts were collected from central Strait of Georgia in August 2000. The above mentioned herring and salmonids were netcaptured. Spot prawns ( Pandalus platyceros ) were collected from prawn traps set in Patricia Bay, BC,

Canada (48º38΄N, 123º28΄). Pacific sandlance ( Ammodytes hexapterus ) and ghost shrimp

(Callianassa californiensis ) were collected at Crescent Beach, BC, Canada, (49º03΄N,

122º53΄W) using beach seine nets to a maximum depth of 1.5 m in October 2001.

Four samples of zooplankton were collected from Trincomali channel in the Strait of Georgia using a 450 µm cod end following lateral and oblique tows, at 1276 m depths, between November 2001 and October 2002 at seasonal intervals. Copepods

(Pseudocalanus spp. ) were the predominant species among all samples collected with seasonal variations seen in abundance of species rather than variety of species (data not shown). Twentyfour butter clams ( Saxidomus gigantea ) were collected in total from two 51 areas in Patricia Bay, BC, 12 in 2002, and 12 in 2003. These bivalves were used as our baseline organism for calculating trophic level estimations derived from stable nitrogen isotope ratios.

Figure 3.1 Collection sites for the Strait of Georgia, British Columbia (Canada) harbour seals and their prey. Adult chinook salmon were collected from Johnstone Strait (1) and the mouth of the Fraser River (7), harbour seal pups were collected from Hornby Island (2), and chinook smolts were collected from central Strait of Georgia (3). The majority of seal prey items were collected from central and southern Strait of Georgia (4, 5, 6, and 9). Adult harbour seals, Pacific sandlance and ghost shrimp were collected from Crescent Beach/Boundary Bay (8), zooplankton was collected from Trincomali channel (9), and spot prawns and butter clams were collected from Patricia Bay (10). WA = Washington; OR = Oregon; CA = California. 52

Table 3.1 Capture location, sex, estimated age, and mass data for harbour seals ( Phoca vitulina ) collected from the Strait of Georgia, B.C., Canada in 2001.

Seal Adult/Pup Area Sex Estimated Age Mass (kg) Identification Captured (F = female (wks or yrs) M = male) PV0101 Pup Hornby Island F 3.55 wks 21.0 PV0102 Pup Hornby Island M 3.55 wks 39.0 PV0103 Pup Hornby Island F 3.55 wks 22.8 PV0104 Pup Hornby Island M 3.55 wks 19.2 PV0105 Pup Hornby Island F 3.55 wks 22.5 PV0106 Pup Hornby Island M 3.55 wks 23.4 PV0107 Pup Hornby Island M 3.55 wks 19.0 PV0108 Pup Hornby Island F 3.55 wks 24.0 PV0109 Pup Hornby Island M 3.55 wks 25.0 PV0110 Pup Hornby Island M 3.55 wks 23.0 PV0118 Adult Fraser River F >4 yr 78.5 PV0120 Adult Fraser River M >5 yr 111.0 PV0123 Adult Fraser River F >4 yr 62.5 PV0125 Adult Boundary Bay F 89 yr 67.0 PV0130 Adult Boundary Bay F >6yr 79.0 PV0131 Adult Boundary Bay M 1215 yr 104.0 PV0132 Adult Boundary Bay M 10 yr 103.0 PV0133 Adult Boundary Bay F ND 61.5 PV0134 Adult Boundary Bay M ND 89.5 PV0138 Adult Boundary Bay M 69 yr 77.0 53

Fish morphometrics and sample preparation

Gross measurement data including body weight, fork or mantle length, and sex were recorded for all samples (data not shown). Middorsal 1cm 3 muscle sections were immediately removed for subsequent stable isotope analysis. Samples and subsamples were individually wrapped in acetone and hexane rinsed aluminium foil, bagged, and frozen at –20 °C for subsequent analysis.

Whole fish samples were homogenized using either a Sorvall Omni Mixer

(Newton, CT, USA) or a Hobart Brinkmann homogenizer (Luzern, Switzerland) following procedures described in detail elsewhere (Cullon et al. 2005).

Stable isotope analyses

Stable isotope analyses on seal skin biopsies (n = 10 adults; n = 8 pups) were done after samples were freezedried for 4872 hours and cut into finite (< 0.5 mm 2) fragments using scalpel blades. Subsamples of individual whole fish homogenates were also freezedried for 4872 hours and then ground to a fine powder using a mortar and pestle. Bulk stable carbon and nitrogen isotope ratio ( 15 N: 14 N and 13 C: 12 C) measurements were carried out at the University of Victoria Biogeochemistry Laboratory

(Victoria, BC, Canada), with analytical equipment and standards described elsewhere

(Cullon et al. 2005). Isotopic composition is expressed in δ notation as the proportional deviation in parts per thousand (‰) of the stable isotope ratio in a sample from that of a standard:

δX = ( Rsample /Rstandard – 1) • 1000 (1)

13 15 13 12 15 14 where X is C or N, and Rsample and Rstandard are the ratios of C: C or N: N for the sample and standard (Hobson et al. 1997). 54

Trophic level estimations

The δ15 N ratios were used to calculate trophic level (TL) position for individual organisms in the harbour seal food web using the following formula (Cabana and

Rasmussen 1996):

TL = [mean fish δ15 N – mean clam δ15 N / 3.4] + 2 (2)

We used clams as our baseline species for trophic level estimations as they are a larger primary consumer with a longer life span than zooplankton species. They therefore offer a more stable δ15 N tissue signature (slower nitrogen turnover) with less sensitivity to seasonal variations in δ15 N ratios than smaller primary consumers (zooplankton) and primary producers (phytoplankton) (Cabana and Rasmussen 1996; Vander Zanden and

Rasmussen 1999).

PCB analyses and lipid determination

Harbour seal biopsies, whole individual fish and fillet samples, whole fish pools, and zooplankton pool underwent PCB analyses using highresolution gas chromatography/high resolution mass spectrometry (HRGC/HRMS). Congenerspecific

PCB analyses and lipid determinations were done at the Regional Contaminant

Laboratory (Fisheries and Oceans Canada), methods described elsewhere (Ikonomou et al. 2001; Cullon et al. 2005). The whole bodied clam pool, used for the biomass estimate only, was analyzed for PCBs and lipid content at AXYS Analytical Services, (Sidney, 55

BC, Canada) according to their methods and criteria, described elsewhere (Christensen et al. 2005).

The numbers of PCB congeners detected as either individual or coeluting congeners out of a possible 209 congeners were 141 in adult harbour seal biopsies and ranged between 135 and 176 in prey samples. Detection limit substitutions were made when more than 70 % of samples had detectable PCB concentrations. Where 70 % or more samples did not have detectable PCB concentrations, that congener was not reported nor considered further.

Biomagnification factors and food web magnification factors

Biomagnification factors (BMFs) were trophiclevel adjusted (BMF TLC ) to account for increases in contaminant concentrations due to lipid increases in organisms with increasing trophic levels (Christensen et al. 2005).

BMF TLC = (Predator [PCB LW ]) / (Prey [PCB LW ]) (3)

(TL Predator / TL Prey )

Food web magnification factors (FWMFs) were calculated using the linear regression line slope of the relationship of trophic level and log 10 transformed PCB concentrations

(lipid normalized) of the harbour seal food web (Borgå et al. 2004; Hoekstra et al. 2003).

FWMF = 10 b (4)

For coeluting congeners, the concentration value was assigned to the congener having the highest weight percentage distribution found in Aroclors 1221, 1016, 1242, 1254,

1260, and 1262, described elsewhere (Frame 1997).

Statistical analyses

For comparisons of PCB concentrations, single factor analysis of variance

(ANOVA) was conducted. Total PCB concentration data was lognormalised to meet assumptions of normality and homogeneity of variance. Where significant differences existed, Tukey posthoc tests were conducted to assess specific group differences among adult male and female seals, and pups. All reported logged data has been done using a log (base 10) transformation.

Principal Components Analysis

Principal Components Analysis (PCA) was used to characterize PCB patterns in harbour seals (20 samples) and their prey species (58 samples). Each PCB congener was evaluated for potential interferences, closeness to the limit of detection and the percentage of undetectable (random value estimated) values. Borderline variables were tested in preliminary PCA models before inclusion in the final PCA data set, which included 109 PCBs. Undetectable values (75 instances, or 0.88 % of the data set; maximum of 21 out of 78 undetectable values for PCB 80 and 102/93) were replaced by a random number between zero and the limit of detection.

Samples were normalised to the concentration total to remove artifacts related to concentration differences between samples. The centred logratio transformation 57

(division by the geometric mean of the concentrationnormalised sample followed by log transformation) was then applied to this compositional data to produce a data set that was unaffected by negative bias or closure (Bonn 1998; Ross et al. 2004). Data were then autoscaled (congeners were scaled by subtracting the variable mean and dividing by the variable standard deviation) to give every variable equal weight. Finally, a Varimax rotation was applied to the first three principal components (PCs) to simplify the physical interpretation of the PCA projections (Ross et al. 2004; Ross et al. 2004); this rotation maximised or minimised the loading of each variable on each PC while preserving trends.

Linear relationships involving the PCA results were quantified using geometric mean (GM) linear regression (Ricker 1984; Cretney and Yunker 2000). The GM slope was calculated by dividing the Y on X slope by the correlation coefficient for the regression, r (Ricker 1984); the mean values for the X and Y variables were then used to calculate the intercept for the GM regression equation. To estimate the relative shift in contaminant distribution for each sample we used the linear distance along the GM linear regression line between adult harbour seals samples and prey species, with the intersection point between the regression line and a perpendicular between the line and the sample position calculated using standard trigonometry mensuration formulae (Ricker

1984).

PCB loadings in Strait of Georgia biota

We estimated the total mass of PCB residing in Strait of Georgia biota by combining PCB concentration data with biomass estimates for this 6900 km 2 inland sea.

Polychlorinated biphenyl estimates for zooplankton and the majority of fish species in the

Strait of Georgia were derived from our study samples. Additional PCB data were 58 obtained for other species for which we did not have samples, including some salmon species, seabirds, and harbour porpoises ( Table 3.4 ). Marine mammal PCB loadings were estimated using data for the relatively abundant harbour seals, harbour porpoises, and killer whales. These estimates were derived from wet weight, total body burdens by adjusting reported lipid weight PCB concentrations in blubber biopsies on the basis of average estimated lipid content in harbour seals (Muelbert and Bowen 1993) and harbour porpoises (McLellan et al. 2002). For killer whales, we adapted an existing loadings estimate of 4.65 kg for southern resident killer whales (Ross et al. 2000) to marine mammaleating transient killer whales using published PCB concentrations (Ross et al.

2000) and updated census information (220 individuals; Ford pers comm.) as follows:

PCB ratio = (251.2/146.3) ppm x (220/89) x 4.65 kg = 19.76 kg (5)

3.4 Results and Discussion

This study of predatorprey biomagnification and biomagnification of PCBs in the

Strait of Georgia harbour seal food web is founded on earlier studies of the Strait of

Georgia harbour seal diet (Olesiuk 1993) and our seal food basket study (Cullon et al.

2005). As an integrated dietary signal for assessing realistic mixtures of POPs, the food basket not only revealed consistencies in PCB concentrations detected in harbour seals and their diet across two basins, but also provided insight into the influences of prey preference, lipid content in diet, and regional contamination may have on PCB accumulation.

Trophic Ecology

Mean δ15 N and δ13 C ratios varied among our harbour seal food web species, with

δ15 N ratios ranging from 8.6 ‰ for zooplankton to 18.4 ‰ for adult male seals and δ13 C ratios ranging from 24.1 ‰ for sockeye salmon to 12.9 ‰ for ghost shrimp ( Table 3.2).

Ranges of δ15 N and δ13 C ratios reflect varied prey preferences and habitat (carbon sources) among species. Estimated 15 N enrichment between primary consumers

(zooplankton) and secondary consumers (fish/invertebrates) was 4.7 ± 0.3 ‰ (mean ±

SEM) and between secondary consumers and tertiary consumers (seals) was 5.0 ± 0.3 ‰

(mean ± SEM), both being within accepted tissue enrichment ranges of 35 ‰ for δ15 N ratios (Minagawa and Wada 1984). Trophic level estimates from our δ15 N ratios derived from a formula using clams as our baseline organism (TL = 2) (Cabana and Rasmussen

1996) indicate a range of three trophic levels among the species collected (1.6 for zooplankton to 4.4 for adult harbour seals) (Table 3.2), with seals trophically higher than their prey, consistent with studies on similar groups of species (Hobson and Welch 1992).

Our harbour seals appear to have more direct coupling with the pelagic food web, evidenced by a stepwise increase trophic enrichment in δ15 N and δ13 C (r2 = 0.20, y =

0.51x + 23.62, p < 0.0001, v = 156), and their consumption of a largely pelagic prey diet

(Olesiuk 1993). Our lower than predicted adult harbour seal δ13 C values (16.7 ‰) may indicate a partial reliance on benthic as well as pelagic carbon sources (Table 3.2)

(France 1995). This has been attributed to an increase in omnivory (reliance upon several carbon sources) with increasing trophic positioning through the food web (Hobson et al.

2002; Kidd et al. 1998). Similar links to both benthic and pelagic food webs have been observed in ringed seals (Kidd et al. 1998), bearded seals (Hoekstra et al. 2003; Hobson 60 et al. 2002), and arctic char (Kidd et al. 1998). As this study was designed to be harbour seal centric and seals rely heavily on vertebrates and to lesser extent invertebrates, we did not collect a full range of species to fully analyze both benthic and pelagic food web components.

PCB concentrations and patterns

The highest PCB concentrations were observed in adult male seals (between the ages of approximately 5 and 15) (Table 3.2), where lognormalized lipidcorrected concentrations increased with estimated age ( r2 = 0.90; p = 0.05). Significant differences in lognormalized lipidcorrected PCB concentrations were found between adult males and females ( p = 0.01; ν = 19) and between adult males and pups (p = 0.00; ν = 19).

Adult male seal concentrations (lipid wt) were more than twice that of adult females, reflecting both the bioaccumulative nature of PCBs and potential lactational transfer of contaminants from mother to pup. Total toxic equivalents ( ΣTEQs) to 2,3,7,8 tetrachlorodibenzopdioxin (TCDD) for adult seals was 29.7 ± 4.4 ng/kg lw (lipid weight) with males higher than females (males 33.7 ± 6.4 ng/kg lw; females 25.6 ± 6.1 ng/kg lw, respectively). Although these ΣTEQ levels are well below concentrations found to immunotoxic and endocrine disrupting in harbour seals (209 ng TEQ/kg) (Ross

2000; Ross et al. 1995; De Swart et al. 1994), the ΣPCB TEQs accounted for 84.5 ± 2.7

% of the ΣTEQ suggesting PCBs continue to pose the largest dioxinlike toxic risk to the

Strait of Georgia harbour seals, consistent with previous data on seals (Ross et al. 2004) and their food basket (Cullon et al. 2005). 61

Table 3.2 Lipid weight (lw) concentrations of total polychlorinated biphenyls ( ΣΣΣPCBs), lipid percentage, stable nitrogen and carbon isotope ratios ( δδδ15 , δδδ13 C) and trophic level estimations are provided for the Strait of Georgia (BC, Canada) harbour seal food web. Values represent mean ±±± standard error of the mean (SEM).

Species n % Lipid a δδδ15 (‰) δδδ13 C (‰) Trophic level ΣΣΣPCB (ug/kg lw) Harbour seals Male adults 5 36.0 ± 6.9 18.4 ± 0.4 16.7 ± 0.3 4.4 ± 0.1 8652 ± 1888 Female adults 5 36.0 ± 1.6 18.3 ± 0.3 16.7 ± 0.3 4.4 ± 0.1 3248 ± 978 Pups 10 51.4 ± 3.4 17.9 ± 0.4 17.6 ± 0.6 4.3 ± 0.1 1645 ± 235 Chinook salmon Adults 12 7.8 ± 1.1 15.2 ± 0.1 19.6 ± 0.4 3.5 ± 0.0 581 ± 212 Smolts 6 0.9 ± 0.3 13.7 ± 0.2 18.2 ± 0.2 3.0 ± 0.0 1924 ± 474 Rockfish 30 b 3.7 15.2 ± 0.1 18.8 ± 0.1 3.5 ± 0.0 690 Pacific lingcod 6 1.0 ± 0.2 14.3 ± 0.2 17.2 ± 0.1 3.2 ± 0.1 1768 ± 561 English sole 18 0.3 ± 0.1 14.3 ± 0.1 16.7 ± 0.3 3.2 ± 0.0 6862 ± 2320 Pacific tomcod 11 b 1.4 13.9 ± 0.2 16.3 ± 0.1 3.1 ± 0.0 453 Pacific hake 6 1.1 ± 0.2 13.2 ± 0.3 18.1 ± 0.1 2.9 ± 0.1 1549 ± 284 Pacific herring 6 9.6 ± 0.8 13.1 ± 0.2 19.9 ± 0.4 2.9 ± 0.0 270 ± 74 Food basket 231 b 4.2 12.9 ± 0.2 19.2 ± 0.0 2.8 408 Pacific midshipman 12 b 2.7 13.0 ± 0.2 18.9 ± 0.2 2.8 ± 0.0 243 Pacific sandlance 18 b 2.7 12.9 ± 0.1 17.8 ± 0.1 2.8 ± 0.0 226 Ghost shrimp 40 b 0.5 12.4 ± 0.2 12.9 ± 0.1 2.7 ± 0.1 150 Spot prawns 8b 0.8 11.8 ± 0.1 15.8 ± 0.1 2.5 ± 0.0 712 Sockeye salmon 6 7.4 ± 0.4 11.1 ± 0.2 24.1 ± 0.3 2.3 ± 0.1 93 ± 11 Butter clams 24 0.7 ± 0.1 10.2 ± 0.0 16.6 ± 0.0 2.0 ± 0.0 105 ± 40 Zooplankton 4b 6.1 8.6 ± 0.5 20.5 ± 0.6 1.6 ± 0.2 128 a Blubber biopsy percentage lipid for harbour seals, muscle/fillet percentage lipid for chinook adults, sockeye, English sole, Pacific herring, and whole fish percentage lipid for all other species. b Pooled homogenate analyzed.

Rankings for the six top PCB congeners in adult male seals were 153 > 138 > 180

> 187 > 99 > 170, accounting for 69 % of ΣPCBs ( Table 3.3; Figure 3.2). Dominance of recalcitrant PCB congeners (CB153 and CB138) in species throughout the seal food web demonstrates their persistence in marine food webs. Differences in dominance of other congeners between seals and their prey has been attributed, in part, to metabolic biotransformation capacity for PCB congeners with specific structural characteristics

(Boon et al. 1997).

Congener pattern differences were observed between harbour seals and their prey, with an overall decrease in lesschlorinated PCB congeners (Figure 3.2). Top consumers

(seals) are dominated by the more heavily chlorinated PCBs, suggesting a heavier PCB source, depuration (e.g. biotransformation, excretion) of the lowerchlorinated congeners, or a combination thereof. Secondary and tertiary consumers (herring, hake) (comprising approximately 75 % of seal diet (Olesiuk 1993)), had similar congener patterns to our harbour seal food basket in an earlier study (Cullon et al. 2005). Many of our fish species

(rockfish, lingcod, hake, herring, and midshipman) were depleted in the lesschlorinated

PCBs relative to lower trophic prey (sandlance, sockeye, and zooplankton) possibly are reflecting their capacity to biotransform or excrete the lighter congeners. Although midshipman and sandlance occupy similar trophic levels, differences in prey preferences may explain their dissimilar congener patterns. Midshipman prefer fishes and crustaceans, whereas sandlance prefer copepods (Hart J.L. 1973). The lesschlorinated

PCB dominant pattern of the primary consumer, zooplankton, is likely due to exposure to atmospherically transported PCBs and baseline PCB patterns in water and suspended particulate. 63

Table 3.3 Food web magnification factors (FWMFs) and trophiclevel adjusted biomagnification factors (BMFTLC ) for the six most dominant polychlorinated biphenyl (PCB) congeners (69 % of ΣΣΣPCB) detected in Strait of Georgia (BC, Canada) male harbour seals. We have used male seals for our BMF calculations because female concentrations may be affected by lactation and pups may not yet be fully consuming an adult seal diet. FWMF = 10 b, where b is the slope of the linear regression trophic level (TL) vs log 10 transformed lipid weight (LW) PCB concentrations; and BMF TLC = (Predator [PCB LW]) / (Prey [PCB LW])

(TL Predator / TL Prey )

% of ΣΣΣPCB BMF TLC PCB FWMF 2 Congener (r ) Seals Food Hake Herring Seal/FB Seal/Hake Seal/Herring basket

153 6.5 (0.61) 25.8 18.4 16.6 14.7 18.8 5.7 36.4 138 5.9 (0.62) 17.5 12.0 12.6 9.7 19.6 5.1 37.4 180 7.0 (0.60) 10.5 2.6 4.3 2.2 53.5 9.0 101.0 187 7.1 (0.68) 6.8 3.0 4.5 3.5 30.4 5.5 40.3 99 5.3 (0.64) 4.6 3.2 3.9 3.2 19.1 4.3 30.2 170 7.1 (0.58) 3.7 0.7 1.5 0.6 69.9 9.4 134.6 ΣΣΣPCB 3.6 (0.53) 69.0 39.9 43.3 33.8 13.4 3.7 20.8

2.0 Harbor seal Rockfish Lingcod 1.5 TL = 4.4 TL = 3.5 TL = 3.2

153 153 153 1.0 138 138 138 180 118 0.5 180 101 187 118 149 99 170 101 187 99 0.0 2.0 Hake Herring Midshipman 1.5 TL = 2.9 TL = 2.9 TL = 2.8

153 153 1.0 153 138 138 138 149 0.5 101 149 101 101 149 118 187 118 187 118 187 0.0

2.0 11 Sandlance Sockeye Zooplankton 1.5 TL = 2.8 TL = 2.3 TL = 1.6 138 138 153 153 1.0 101 153 28 31 118

Concentration ratios (relative toPCB153) (relative ratios Concentration 118 110 101 138 99 52 0.5 118 149

0.0 1 209 1 209 1 209 PCB congener

Figure 3.2 Polychlorinated biphenyl congener (PCB) patterns change from lower trophic level prey to harbour seal. The ratio of PCB congener to the recalcitrant PCB153 shows similar predominance of PCB congeners among fish and harbour seals, whereas zooplankton show a predominance of the lesschlorinated PCBs. Harbour seals further show a simpler pattern that is more depleted in the less chlorinated congeners than the lower trophic level fish species revealing a potential metabolic capacity for the lighter PCB congeners. TL = Trophic level

Biomagnification

Log ΣPCB concentrations exhibited a linear increase with δ15 N ratios ( r2 = 0.34, y

= 0.16x + 0.56, p < 0.0001; v = 72) in the seal food web ( Figure 3.3). This suggests that in our dataset PCB concentration increases by 8fold between trophic levels two and 65 three, threefold between three and four, and 1.5fold between four and 4.5. Logged congener CB52, 101, and 153 concentrations also increased linearly with δ15 N ratios

(r2 = 0.42, y = 0.13x + (0.67), p < 0.0001, v = 71; r 2 = 0.31, y = 0.14x + (0.46), p <

0.0001, v = 71; r2 = 0.38, y = 0.19x + (0.75), p < 0.0001, v = 72, respectively) (Fig. 3).

The regression slopes of these light to heavy congeners increased with increasing number of chlorine atoms and log K ow illustrating not only the biomagnification potential of these chemicals, but the influence of physicochemical properties on biomagnification in harbour seal food webs (Figure 3.3).

4 r2 = 0.34 HSM 4 153 (m = 0.19) p < 0.0001 ES 101 (m = 0.14) HSF m = 0.16 CS 52 (m = 0.13) HK LC HSP 3 3 153 SP RF FB TC HE CH MS SL 101 ZO 2 SHP 2 SK 52 PCB g/kg g/kg lipid PCB PCBx g/kg lipid g/kg PCBx Σ Σ 1 1 log log log log

0 0

08 10 12 14 16 18 20 08 10 12 14 16 18 20 δ15 N δ15 N

Figure 3.3 Biomagnification is clearly observed in the regression of log ΣΣΣPCB concentrations with stable nitrogen isotope ratios in the Strait of Georgia harbour seal ( Phoca vitulina ) food web. Regression slope differences between individual PCB congeners 52, 101, and 153 likely reflect physicochemical influences on PCB accumulations, with the higher accumulation slopes observed in the more heavily chlorinated congeners having higher K ow values. HSM: male harbour seals; HSF: female harbour seals; HSP: harbour seal pups; CH: chinook salmon; RF: rockfish; ES: English sole; LC: lingcod; TC: tomcod; CS: chinook smolts; HK: hake; SL: sandlance; HE: herring; MS: midshipman; FB: seal food basket; SK: sockeye salmon; SHP: ghost shrimp; SP: spot prawns; ZO: zooplankton. 66

Biology and feeding ecology differences, as suggested from our δ15 N and δ13 C ratios (Table 3.2), exist among the seal food web species. Factors including baseline organism signature, age, partitioning among tissues, lipid, migration, nutritional status, can greatly affect δ15 N and δ13 C ratios and subsequent chemical biomagnification interpretation (France and Peters 1997; Robbins et al. 2005; Doucett et al. 1999;

Minagawa and Wada 1984; Lesage et al. 2002), and these factors have proven useful in explaining contaminant concentrations in organisms (Jardine et al. 2006).

Biomagnification has not been observed in the transfer of PCBs from phytoplankton to zooplankton (Berglund et al. 2000; Magnusson et al. 2007).

BMFs and FWMFs

Trophic leveladjusted biomagnification factors (BMFs) between predator (seals) and prey (food basket) reveal a biomagnification for ΣPCBs of more than 13 times (Table

3.3). Biomagnification is said to have occurred when BMF > 1. Ranking of highest

BMFs among PCB congeners were 205 > 194 > 170 > 180 > 191 > 206, octa (CB205,

194), hepta (CB170, 180, 191), and nonachlorinated (CB206) groups. The highly persistent congener CB153 ranked 27 for BMFs, illustrating physicochemical and pharmacokinetic process differences among congeners between seals and their diet.

Differences in BMFs among seal/hake and seal/herring, both prey occupying similar trophic positions and comprising similar percentages of seal diet (43 % and 32 %, respectively (Olesiuk 1993)), likely reflect the differences in PCB concentrations, lipid content, and feeding ecology that can influence BMFs. Our adult hake collected are a resident stock of the Strait of Georgia and consume a diet of primarily zooplankton and 67 sandlance, whereas herring are a semiresident stock that largely consumes zooplankton and crustaceans (McFarlane et al. 2001).

Another method of examining biomagnification potential of contaminant is through the calculation of FWMFs. The FWMFs provide a value representing an average rate of increase in lipidnormalized POP concentrations per trophic level for an entire food web (Hop et al. 2002; Nfon et al. 2008). A significant increase in log ΣPCBs measured in biota with increasing trophic level ( r2 = 0.53, y = 0.55x + 1.08, p = 0.0006, v

= 17) (Figure not shown) was observed (FWMF = 3.6). Our calculated FWMF values for

ΣPCB (3.6) and PCB153 (6.5) (Table 3.3) were most comparable with coastal marine food web studies from BeaufortChukchi Seas ( ΣPCB = 3.26; PCB153 = 6.69) (Hoekstra et al. 2003b), North Baffin Bay ( ΣPCB = 4.6; PCB153 = 9.7) (Fisk et al. 2001), Iceland

(ΣPCB = 3.11; PCB153 = 5.04) (Skarphedinsdottir et al. 2009), and a freshwater lake from China ( ΣPCB = 3.68; PCB153 = 4.73) (Wu et al. 2009). Food web magnification factors can, however, underestimate biomagnification for homeotherms and overestimate for poikilotherms because of differences in energy requirements and biotransformation capacities (Borgå et al. 2004), and should therefore be used in conjunction with predator/prey BMFs (Fisk et al. 2001).

The top ten highest FWMFs among PCB congeners were 205 > 194 > 195 > 203

> 170 > 187 > 180 > 206 > 201 > 109 and ranged between 6.7 and 9.4 for reported congeners. Seven of the ten highest FWMFs were observed for those PCB congeners that belong to structural activity group (SAG) 1 (congeners without any vicinal hydrogen

(H)atoms (Boon et al. 1997)). Interestingly, CB153 is generally considered the most persistent congener, but it ranked 11 for FWMFs (FWMF = 6.5), suggesting other 68 congeners may be more persistent, at least in a harbour seal food web. Work in salmon eating grizzly bears also found that CB153 was not the most persistent congener in predatory mammals (Christensen et al. 2009).

Physicochemical properties of contaminants (e.g. log K ow ) can play an important role in determining and/or explaining bioaccumulation within food webs.

Biomagnification of PCB congeners between seals and their diet (food basket) reveal a positive linear relationship with log K ow , with the greatest degree of biomagnification

2 (log BMF > 1.5) occurring with those congeners possessing the highest log K ow ( r =

0.48, y = 0.64x – 3.57, p < 0.0001, v = 108) (Figure not shown). Biomagnification is observed in the majority of detected congeners having log K ow values greater than 6.8.

Depuration (loss) of congeners with log K ow values less than 6.8 are likely due to pharmacokinetic processes, such as metabolic biotransformation, excretion, and/or lack of uptake.

Food web magnification factors of individual PCB congeners were also related to

2 log K ow ( r = 0.56, y = 2.12x – 10.36, p < 0.0001, v = 119); Figure 3.4). Despite a wide range of log K ow values among congeners (from 4.76 to 8.18) (Hawker and Connell

1988), almost all biomagnified to some degree, with only minor exceptions (PCB 15,

19, 22, 35, and 37). The steady increase in FWMFs with log K ow in the harbour seal food web may be a result of the combination of bioconcentration from both water respiring fish/invertebrates and biomagnification from airbreathing mammals (Kelly et al. 2007), or may also be a result of the “averaging” effect of using FWMFs (all species in a regression, rather than specific predatorprey). This relationship has been evidenced in other marine food webs (Fisk et al. 2001; Kelly et al. 2008) however, a decrease at 69

higher log K ow values has also been observed in an Arctic marine food web (>10.75)

(Kelly et al. 2008) and a freshwater food web in South China (>7.0) (Wu et al. 2009).

Our dataset does not provide a clear interpretation of a decrease at higher log K ow values

(Figure 3.4).

10 205

2 8 r = 0.56 195 194 p < 0.0001 203 m = 2.12 153 6

4 FWMF

2 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5

log K ow

PCB patterns in the seal food web

Principal component analysis (PCA) provides a means of characterizing PCB patterns in harbour seals and their prey species, and evaluating the influence of δ15 N, metabolism, and log K ow on these PCB patterns. Principal components analysis 70 differentiates harbour seals and their prey according to the amount of difference in PCB congener proportions.

In the PCA samples plot, prey species (fish, invertebrates, and the food basket samples) project in the upper left quadrant, while adult seals project on the lower right

(Figure 3.5a). The variables plot indicates that the prey/predator separation is a reflection of high proportions of tetra and pentachlorinated PCBs in the fish and invertebrates and high proportions of hepta to decachlorinated PCBs in the adult seals

(Figure 3.5b ). Seal pups project in the lower centre indicating higher proportions of tri and tetrachlorinated PCBs than in the adult seals (Figures 3.5a and 3.5b). This composition difference is likely a reflection of a transfer barrier and/or metabolic removal of the more heavily chlorinated PCBs through motherpup lactational transfer (Ross et al.

2004; Addison and Brodie 1987).

Figure 3.5a 71

Figure 3.5b

Figure 3.5c 72

Figure 3.5d

Figures 3.5ad Principal components analysis (PCA) provides insight into physicochemical and trophic level influences on biomagnification of PCBs in the Strait of Georgia harbour seal ( Phoca vitulina ) food web. (a) Score plot separating harbour seals and prey species, (b) PCB congener variable loadings (p1 and p2) by congener number, (c) trophic level (TL) is plotted against the sample positions of the GM regression line between seals and prey, and (d) PCB variable separations in PC1 are strongly correlated to both log K ow and FWMFs.

The sample separations in PC1 are strongly related to trophic level, as demonstrated by a significant correlation between δ15 N and t1 ( r2 = 0.61, p = 7.6E15, v =

65; Figure 3.5c ). The correlation between δ15 N and the sample positions on the GM regression line between adult seals and the prey samples is only a little greater ( r2 = 0.65), indicating that most of the trophic level relationship is encompassed by PC1. 73

Similarly, the PCB variable separations in PC1 are strongly correlated to both log

2 2 Kow ( r = 0.65, p = 3.4E26) and the FWMF ( r = 0.72, p = 2.7E31, v = 107 for both;

Figure 3.5d). These correlations are higher than the correlations between the two

2 parameters ( r = 0.52 for log K ow vs. FWMF). The dominance of PCB congeners belonging to structural activity groups (SAGs) 1 and 2 in adult seals (PCBs in Figure 3.5d are shown by the SAG) clearly illustrates their inability to metabolize congeners without vicinal hydrogen (H) atoms and congeners with vicinal H atoms in ortho and meta positions in combination with ≥ 2 ortho Cl substituents (Boon et al. 1997). This suggests that although both trophic level and log K ow influence PCB biomagnification in harbour seals, the seals’ metabolic capacity for some congeners plays an important role.

PCB loadings in Strait of Georgia biota

In this study, some of the deviations from biomagnification regression slopes may in fact show some species are spending different amounts of time in the Strait of Georgia.

In an attempt to characterize sources of PCB contamination, we examined the regression slope for δ15 N vs log PCB153 resident species only (r2 = 0.34, y = 0.14x + 0.32, p <

0.0001; v = 48) (Figure 3.6). Distance from mean resident slope for migratory pelagic species may offer an approach for estimating the relative contribution of local (Strait of

Georgia) vs North Pacific Ocean sources for the harbour seal food web. For the purposes of this exercise, migratory (chinook and sockeye salmon) (Healey 1991; Burgner 1991) or semiresident (herring) (Burgner 1991) species are defined as spending onethird or less time within the Strait of Georgia waters.

4.0

3.5 Resident spp. HSM r2 = 0.34 3.0 p < 0.0001 ES m = 0.14 HSP HSF 2.5 CS HK LC 2.0 SP TC RF FB MS 1.5 HE CH

log PCB153 ug/kg lipid PCB153ug/kg log SL 1.0 SK

ZO SHP 0.5 6 8 10 12 14 16 18 20

15 δ N (‰)

Figure 3.6 Biomagnification regression slopes may offer insight into the contribution of local sources to the global background. Calculating the distance from the mean regression slope of resident species for migratory species (chinook, sockeye, herring) estimates increases in PCB153 concentrations of 47 %, 40 %, and 23 %, respectively. This would suggest that sources within the Strait of Georgia have a greater contribution rather than those from the orth Pacific Ocean to the harbour seal ( Phoca vitulina ) PCB accumulation slope. HSM: male harbour seals; HSF: female harbour seals; HSP: harbour seal pups; CH: chinook salmon; RF: rockfish; ES: English sole; LC: lingcod; TC: tomcod; CS: chinook smolts; HK: hake; SL: sandlance; HE: herring; MS: midshipman; FB: seal food basket; SK: sockeye salmon; SHP: ghost shrimp; SP: spot prawns; ZO: zooplankton.

Assuming δ15 N is an accurate predictor of trophic feeding level, and no change in trophic level feeding within or outside of the Strait of Georgia, calculating the difference between predicted PCB153 concentrations using the residency regression line and actual observed concentrations, we predict log PCB153 concentrations in chinook, sockeye, and herring to increase by 47%, 40%, and 23%, respectively if they were to spend 100 % 75 of their time within the Strait of Georgia. This suggests sources within the Strait of

Georgia to be the greater source, while open ocean sources contribute to a lesser extent to the seal PCB accumulation slope.

Global production of PCBs has been estimated to be approximately 1.3 million tonnes, with almost 97 % of historical use in the Northern Hemisphere, and 3 % in

Canada (Breivik et al. 2002). We estimated PCB burdens in Strait of Georgia biota by updating a food web biomass model (Dalsgaard et al. 1998) with available PCB concentration data (Table 3.4). We estimate that 77.4 kg of PCBs can be found in the biota of the Strait of Georgia, of which approximately 0.6 kg are removed through the fisheries sector annually. Our estimates suggest that the largest PCB loads, on a per tonne (t) basis, reside in invertebrates and high trophic level marine mammals.

This PCB loadings estimate in Strait of Georgia marine biota includes both resident and migratory species. If we consider only those species that reside in the Strait of Georgia 86 100% of the time, our loadings estimate declines to 56 kg of PCBs, or 73

% of our total estimate. For the purposes of this exercise, migratory species include those that reside in the Strait of Georgia less than 86 % of the year (herring, salmonids, dogfish, killer whales (Therriault et al. 2009; Ford and Ellis 2006; Osborne 1999; Freedman 1995;

McFarlane and King 2003; Fisheries and Oceans Canada 2008). Despite what appears to be an important “regional” signal, our results add to a growing body of literature that document the biological importation of PCBs into coastal British Columbia (Cullon et al.

2005; Christensen et al. 2005).

Table 3.4. Estimated polychlorinated biphenyl (PCB) loadings for Strait of Georgia biota using biomass estimates from a massbalance model reconstruction of the Strait of Georgia present day (Dalsgaard et al. 1998) and updated biomass estimates with recent PCB concentrations. ww = wet weight; lw = lipid weight

Biota Category Species used for Biomass ΣPCB ΣPCB in Biota Reference (s) a calculation estimates (kg ww) (mg/kg lw for marine (kg) Biomass mammals; ug/kg ww ΣPCBs for all other species) b Zooplankton Zooplankton 3.35 x 10 8 7.85 2.63 (Dalsgaard This study et al. 1998) Shellfish invertebrates Pacific oysters, Butter clams, Blue 1.52 x 10 9 16.48 25.07 (Dalsgaard This study, mussels et al. 1998) P. Ross unpubl. data Grazing invertebrates Spot prawns, Ghost shrimp 2.76 x 10 9 3.34 9.20 (Dalsgaard This study et al. 1998) Predatory invertebrates Dungeness crab 6.28 x 10 7 2.38 0.15 (Dalsgaard Grant et al., et al. 1998) in prep Small pelagic fish Pacific sandlance 9.98 x 10 7 6.08 0.61 (Dalsgaard This study et al. 1998) Pacific herring Pacific herring 5.90 x 10 7 26.00 1.53 (Fisheries This study and Oceans Canada 2009) Pacific hake Pacific hake 6.00 x 10 7 14.52 0.87 (Fisheries This study and Oceans Canada 1999) Dermersal fishes Plainfin midshipman, Pacific 8.69 x 10 7 13.23 1.15 (Dalsgaard This study tomcod, English sole, Rockfish c et al. 1998) Lingcod Lingcod 4.00 x 10 6 17.05 0.07 (Fisheries This study and Oceans Canada 2005) a Biomass estimates have been updated where data was available for species (herring, hake, lingcod, dogfish, and marine mammals) b PCB concentrations (based on lipid fraction for marine mammals and whole body for all other species) averaged from concentrations detected in species used for estimates c Rockfish was a pooled sample of three species: greenstriped rockfish ( Sebastes elongates ), quillback rockfish ( Sebastes maliger ), and splitnose rockfish ( Sebastes diploproa ) 77

Salmon Sockeye salmon, Chum salmon, 4.84 x 10 7 8.57 0.42 (Dalsgaard This study, Pink salmon, Coho salmon, et al. 1998) (Carlson Chinook salmon D.L. and Hites 2005) Dogfish Spiny dogfish 6.00 x 10 7 130.31 d 7.82 (Saunders (Mackintosh 1988) et al. 2004) Marine birds Doublecrested cormorant eggs, (Dalsgaard (Harris et al. Great blue heron eggs, Osprey eggs 1.38 x 10 5 638.67 0.09 et al. 1998) 2005; Elliot et al. 1989; Elliott et al. 2000) Marine mammals Harbour seals 1.72 x 10 6 males 8.65 (Fisheries This study females 3.25 3.31 e and Oceans Canada 2010) Harbour porpoises 2.76 x 10 4 males 9.27 (Baird 2003) (Lucas females 4.27 0.05 e 1998) Southern resident killer whales 2.07 x 10 5 males 146.30 (Ross et al. (Ross et al. females 55.40 4.65 e 2000) 2000) Transient killer whales 6.21 x 10 5 males 251.20 J. Ford, pers. (Ross et al. females 58.80 19.76 e comm 2000) TOTAL 77.38

d Sum of six PCB congeners (18, 99, 118, 180, 194, 209) e Estimated PCB load (kg) for population (harbour seals n = 39,000; harbour porpoises n = 911; southern resident killer whales n = 89; transient killer whales n = 220) 78

3.5 Conclusion

Concentrations of PCBs and dominance of congeners 153 and 138 throughout the harbour seal food web demonstrate the continued persistence of these largely banned chemicals. Concentration increases with δ15 N ratios along with BMFs and FWMFs clearly illustrate biomagnification and trophic transfer of PCBs in our harbour seal food web. Regression slope differences among CB congeners 52, 101, and 153 and different compositional patterns between trophic levels, with more heavily chlorinated congeners dominating adult seals, and lesschlorinated congeners dominating fish, invertebrates, zooplankton and seal pups were clearly observed. These pattern differences throughout the seal food web and their close correlations with trophic level,

SAGs, log Kow , and degree of chlorination provide evidence of feeding ecology, biotransformation and/or excretion, and physicochemical influences on PCB accumulation. An approach of using distance from mean resident slope for migratory species suggests local (Strait of Georgia) sources to be the greater contributor than North

Pacific Ocean sources to the seal food web PCB accumulation slope. We further estimate there to be approximately 77 kg of PCBs within the Strait of Georgia biota with approximately 0.6 kg removed annually for human consumption.

Although PCBs are considered to be more stable than other contaminants of concern, it remains difficult to assess or predict their environmental fate in food webs.

Concentration levels today may, in fact, be changed from our data collected in 200001 due to changes in climate, contaminant and organic loadings, sedimentation shifts as well as species biomass. Therefore, further food web and mass balance research to establish temporal trends on a regional level would be invaluable for the Strait of Georgia as it 79 continues to be a thriving area for industry, recreation, transportation, and a rising human population.

Chapter 4: Characterizing Dietary Exposure of Persistent Organic

Pollutants in Resident Killer Whales ( Orcinus orca ) of British Columbia and

Adjacent Waters †

4.1 Abstract

We measured persistent organic pollutant (POP) concentrations in chinook salmon (Oncorhynchus tshawytscha ) in order to characterize dietary exposure in the highly contaminated, salmoneating northeastern Pacific resident killer whales. We estimate that 97 to 99 % of polychlorinated biphenyls (PCBs), polychlorinated dibenzop dioxins PCDDs), polychlorinated dibenzofurans (PCDFs), dichlorodiphenyl trichloroethane (DDT), and hexachlorocyclohexane (HCH) in returning adult chinook were acquired during their time at sea. Highest POP concentrations (including PCBs,

PCDDs, PCDFs, and DDT) and lowest lipids were observed in the more southerly chinook sampled. While feeding by salmon as they enter some more POPcontaminated nearshore environments inevitably contribute to their contamination, relationships observed between POP patterns and both lipid content and δ13 C also suggest a migration related metabolism and loss of the less chlorinated PCB congeners. This has implications for killer whales, with the more PCBcontaminated salmon stocks in the south partly explaining the 4.0 to 6.6 times higher estimated daily intake for ΣPCBs in southern resident killer whales compared to northern residents. We hypothesize that the lower lipid content of southerly chinook stocks may cause southern resident killer whales to

† Adapted from Cullon DL, Yunker MB, Alleyne C, Dangerfield NJ, O’Neill S, Whiticar MJ, Ross PS. 2009. Persistent organic pollutants in chinook salmon ( Oncorhynchus tshawytscha ): Implications for resident killer whales of British Columbia and adjacent waters. Environ Toxicol Chem 28: 148161. 81 increase their salmon consumption by as much as 50%, which would further increase their exposure to POPs.

4.2 Introduction

Two populations of resident killer whales ( Orcinus orca ) frequent the coastal waters of British Columbia (BC), Canada, and Washington State (WA), USA. The

Canadian Species at Risk Act (SARA) has designated northern resident killer whales as threatened, while SARA and the U.S. Endangered Species Act (ESA) have designated southern residents as endangered. Although both are fisheating, polychlorinated biphenyl (PCB) concentrations in the southern residents (males: 146 mg/kg lipid wt) are almost four times that of northern residents (males: 37 mg/kg lipid wt), placing them among the most PCBcontaminated marine mammals in the world (Ross et al. 2000).

Both populations elicit a strong preference for chinook salmon (Oncorhynchus tshawytscha ), which comprises 70% of their estimated diet (Ross et al. 2000), underscoring the need to characterize persistent organic pollutant (POP) concentrations in this salmonid.

Anadromous chinook, the largest of the Pacific salmon, spend the majority of their life in the pelagic marine environment, where they undergo the majority of their growth before returning to freshwater natal streams for spawning (Healey 1991). Fish accumulate persistent organic pollutants (POPs) through gill uptake (bioconcentration) and dietary uptake (biomagnification) (Qiao et al. 2000). Exposure to POPs in freshwater, estuarine, and coastal environments may explain the relative contamination of some salmon stocks (Missildine et al. 2005), especially in the relatively PCB contaminated Puget Sound (Missildine et al. 2005; Ross et al. 2004). However, it is 82 apparent that global sources acquired by salmonids during their time in the North Pacific

Ocean also contribute substantially to their contamination (Ewald et al. 1998; O'Neill and

West 2009). This has implications for wildlife, as POPs are delivered by salmon to coastal, freshwater and terrestrial ecosystems (O'Neill and West 2009; DeBruyn et al.

2004; Ewald et al. 1998). Salmon partly explain the POPs found in British Columbia wildlife, including resident killer whales (Ross et al. 2000) and grizzly bears (Christensen et al. 2005), though questions linger about their importance relative to other prey items.

As adult salmon enter nearshore marine waters en route toward their natal streams, they undergo dramatic changes in body weight, and lipid, protein, and water content (Brett 1995). Chinook salmon can lose more than 80% of their lipid during their return migration (Brett 1995), which has profound ramifications for lipidsoluble contaminant concentrations.

The extent to which chinook salmon deliver POPs to resident killer whales is unclear, (Ross et al. 2000), as are the sources of POPs to salmon. In the present study we measured POPs including flameretardants, industrial byproducts, and organochlorine

(OC) pesticides in oceanmigrating smolts and in returning adults from four stocks of chinook salmon from British Columbia (Canada) and Washington State (USA). Our objectives were to characterize in chinook salmon the POPs acquired locally as juveniles

(i.e. prior to ocean migration) and the POPs acquired during time at sea, and to estimate the contribution of chinook to POP exposure in resident killer whales. 83

4.3 Materials and Methods

Chinook salmon collections

Chinook salmon smolts ( n = 18) and adults ( n = 24) were collected in southern

British Columbia and Washington (Figure 4.1 ). British Columbia adult chinook salmon were collected from Johnstone Strait and near the mouth of the Fraser River in October

2000. Chinook smolts were collected from central Strait of Georgia in August 2000.

Washington State adult chinook salmon were collected near the mouth of the Duwamish

River and from the Tumwater Falls Hatchery on the Deschutes River in September 2001.

Puget Sound chinook smolts were collected from the Green/Duwamish River and the

Deschutes River during the period May through June 2001. Samples were individually wrapped in aluminium foil, bagged, and frozen at –20 °C for transport and subsequent analysis.

Morphometrics and stock identification

Body weight and fork length were recorded for all chinook salmon, while sex was recorded for adult chinook only ( Table 4.1). Dorsal muscle samples (1 cm 2) were collected for the adult chinook collected in Johnstone Strait and the Lower Fraser River and smolts from the Strait of Georgia and placed in 95 % ethanol for deoxyribonucleic acid (DNA) analyses. Stock identification was carried out by the Molecular Genetics Lab at the Pacific Biological Station (Fisheries and Oceans Canada, Nanaimo, BC). Thirteen microsatellite loci were amplified and DNA fragments were sized and sequenced on an automated ABI 377 DNA sequencer (Applied Biosystems, Foster City, CA, USA). Data analyses and classification procedures for species and stock identification are described 84

Figure 4.1. Migratory routes [adapted from Fisheries and Oceans Canada, 2006 (http://www.pac.dfompo.gc.ca/species/salmon/salmon_facts/chinook_e.htm)] and collection sites (inset map) for British Columbia (Canada) and Washington State (USA) adult chinook salmon ( Oncorhynchus tshawytscha ) ( n = 24). Johnstone Strait adult chinook were collected from site (1), Strait of Georgia chinook smolts were 85 collected from site (2), Lower Fraser River adults from site (3), Duwamish River adults from site (4), and Deschutes River adults from site (5), Puget Sound smolts were collected upstream from both sites (4) and (5). Although ocean distribution for BC and Washington chinook encompasses the orth Pacific Ocean and Bering Sea, greatest abundance has been observed along the orth American coastal shelf waters (Healey 1991). BC = British Columbia; WA = Washington State; OR = Oregon State; CA = California State.

elsewhere (Beacham et al. 2006). Fish scales from the left side posterior of the dorsal fin above the lateral line were removed for age determination, described in detail elsewhere

(MacLellan 1999). Age determination was carried out by the Aging Lab at the Pacific

Biological Station in accordance with their procedures and criteria, also described in detail elsewhere (Chilton and Beamish 1982).

Sample preparation

Johnstone Strait and Lower Fraser River chinook fillet (muscle) tissue homogenates were prepared for analyses for both this study and a human health hazard assessment, whereas for Duwamish River, Deschutes River, and all chinook smolts whole fish tissue homogenates were prepared. Strait of Georgia chinook smolts were prepared as individual samples. However, Puget Sound smolts were pooled due to their small body size. Additionally, pooled fillet homogenates were prepared for Lower Fraser River and Duwamish River adult chinook. For Johnstone Strait and Lower Fraser chinook, rest of fish (ROF) homogenates, which included all fish tissues except for fillet, were constructed for lipid determination in order to calculate POP body burdens, described in detail later. Fillet, ROF, and whole fish tissues were homogenized according to procedures described in detail elsewhere (Cullon et al. 2005). 86

Stable isotope analysis

Whole salmon homogenates (20 g) were freezedried for 48 to 72 h and then ground to a fine powder using a mortar and pestle. Bulk stable carbon and nitrogen isotope ratio ( 15 N: 14 N and 13 C: 12 C) measurements were carried out at the

Biogeochemistry Laboratory at the University of Victoria (Victoria, BC, Canada), equipment and standards described elsewhere (Cullon et al. 2005). Isotopic composition is expressed in δ notation as the proportional deviation in parts per thousand (‰) of the isotope ratio in a sample from that of a standard as in the following equation:

δX = ( Rsample /Rstandard – 1) X 1000 (1)

13 15 13 12 15 14 where X is C or N, and Rsample and Rstandard are the ratios of C: C or N: N for the sample and standard (Hobson et al. 1997).

Contaminant analysis and lipid determination

Whole body ( n = 12) and fillet ( n = 12) adult chinook, whole body chinook smolts

(n = 6), and one chinook smolt pool of twelve individuals (10 g) were analyzed for congenerspecific polychlorinated biphenyls (PCBs), polychlorinated dibenzopdioxins

(PCDDs), polychlorinated dibenzofurans (PCDFs), and two pooled fillet samples for polybrominated diphenyl ethers (PBDEs), polychlorinated diphenyl ethers (PCDEs), and polybrominated biphenyls (PBBs) (reported as either individual or coeluting congeners) using highresolution gas chromatography/highresolution mass spectrometry

(HRGC/HRMS). For Duwamish River adults and all chinook smolts, organochlorine pesticides, including the dichlorodiphenyltrichloroethane (DDT) group [ o,p’ DDT, dichlorodiphenyldichloroethane (DDD), dichlorodiphenyldichloroethylene (DDE) and 87 p,p’ DDT, DDD, DDE], hexachlorobenzene (HCB), hexachlorocyclohexane (HCH) [α

HCH, βHCH, γHCH], heptachlor, aldrin, chlordane [oxy, γ, α], nonachlor [ trans, cis ], and mirex were measured using lowresolution gas chromatography/mass spectrometry (LRGC/MS) and gas chromatography with electron capture detection

(GC/ECD).

Extraction and cleanup procedures, instrumental analysis and conditions, lipid determination, and quality assurance/quality control criteria used for PCBs, PCDDs, and

PCDFs by the Regional Contaminant Laboratory (Fisheries and Oceans Canada) are described elsewhere (Ross et al. 2000; Ikonomou et al. 2001). Polybrominated diphenylethers, PCDEs, PBBs, and OC pesticides were analyzed by AXYS Analytical

Services, (Sidney, BC, Canada) according to their laboratory procedures and criteria and are described in detail elsewhere (Cullon et al. 2005), PCDE method publication in process. Lipid values were also determined by AXYS for samples analyzed for PBDEs,

PCDEs, PBBs, and OC pesticides. Where whole fish lipid percentage data were compared, Regional Contaminant Laboratory values were reported.

Organochlorine pesticide analyses and lipid determinations for Johnstone Strait and Lower Fraser River fillet tissues were carried out by the Western Regional

Laboratory, Health Canada (Burnaby, BC) using an inhouse validated analytical method.

The sample batch for OC pesticides included 10 samples, a reagent blank, and a replicate.

Samples were spiked with 13 Clabeled surrogate standards and extracted with acetone/hexane (2:1 v/v) using a Polytron homogenizer (Luzern, Switzerland). The extract was centrifuged and the organic layer was further reextracted with hexane and saturated sodium chloride. An aliquot of the organic layer was taken to dryness under 88 vacuum with a rotary evaporator for lipid determination. The sample residue was redissolved in dichloromethane and the lipids removed by preparative gel permeation chromatography using a Waters highpressure liquid chromatograph (Milford, MA,

USA). The solvent of the gel permeation chromatography eluate was exchanged to hexane and the sample purified by eluting through a Florisil® (U.S. Silica, Berkeley

Springs, WV, USA) column (2% deactivated) with dichloromethane/ hexane (60/40 v/v).

The purified eluate as concentrated to near dryness, dissolved quantitatively into iso octane and spiked with a 13 Clabeled internal standard for instrumental analysis.

Instrumental analysis by HRGC/MS was carried out using a VG AutospecQ magnetic sector mass spectrometer (Manchester, UK) coupled with a HewlettPackard

5890 Series II gas chromatograph (Palo Alto, CA, USA), a CTC A200S autosampler

(Canboro, NC, USA), and Micromass OPUS data system. Chromatographic separation was achieved through an Agilent (Santa Clara, CA, USA) J&W DB5 capillary chromatography column (60m x 0.25 mm internal diameter x 0.25 µm film thickness).

The mass spectrometer was operated at a resolution of 5,000 in selected ion monitoring mode using two intense ions for each analyte.

Number of congeners detected either as individual or coeluting congeners out of a theoretical possible number in 31 salmon samples were 135 out of 209 for PCBs, three out of 75 for PCDDs, two out of 135 for PCDFs, and in two muscle pools, 26 and 28 out of 44 for PBDEs, 34 and 34 out of 44 for PCDEs, and 12 and 12 out of 21 for PBBs.

Detection limit substitutions were made for PCB, PCDD, PCDF, and OC pesticide analytes that were not detected when at least 70 % of samples had detectable concentrations. Where more than 70 % of samples did not have detectable concentrations 89 of analytes, concentrations were not reported. For the two pooled samples analyzed for

PBDE, PCDE, and PBB analytes, detection limit substitutions were made when one sample had a detectable concentration. Toxic equivalent quotients (TEQs) were calculated for PCBs, PCDDs, and PCDFs using World Health Organization (WHO)

International Toxic Equivalent Factors (TEFs) for humans and wildlife (Van den Berg et al. 1998).

Statistical analyses

For comparisons among adult chinook salmon groups single factor analysis of variance (ANOVA) was done. The degrees of freedom (ν) for ANOVA test were 23

(including numerator and denominator) for all analyses except for OC pesticides where ν was 17. Data met the assumptions of normality and homogeneity of variance, or were lognormalized. If significant differences (α = 0.05) existed among the adult chinook groups, Tukey post hoc tests were done to determine which groups were significantly different from each other. For comparison between fillet and ROF lipid percentages student’s t tests were used (equal variances). Since chinook smolt groups consisted of one group of six samples and one pooled sample a statistical comparison was not possible.

Body burden calculations

Estimates of POP body burdens for chinook salmon adults and smolts were determined from either concentrations from whole fish homogenates, or, in the case of

Johnstone Strait and Lower Fraser River salmon, from fillet concentrations. We assumed that lipidnormalized POP concentrations would be equally partitioned between whole 90 fish and fillet, as previously documented in salmon (Isosaari et al. 2004). Lipid content was determined for Johnstone Strait and Lower Fraser River salmon using a weighted combination of fillet lipid values and ROF lipid values as follows:

mass lipid (whole fish) = (mass fillet/mass whole fish) • % lipid (fillet)

+ (mass ROF/mass whole fish) • % lipid (ROF) (2)

Body burden estimates were subsequently calculated using whole fish lipid percentages as follows:

body burden (POPs) = [POP] lipid wt • mass lipid (3)

Principal components analysis

Principal components analysis (PCA) was used to characterize POP patterns among salmon and generate insight into the factors affecting them. Each PCB, dibenzo pdioxin, and dibenzofuran congener was evaluated for potential interferences, closeness to the limit of detection and the percentage of undetectable (random value estimated) values. Borderline variables were tested in preliminary PCA models before inclusion in the final PCA data set, which included two dioxins (1,2,3,6,7,8HxCDD and OCDD), two furans (2,3,7,8TCDF and 2,3,4,7,8PnCDF) and 130 PCBs ( Appendix I ). Undetectable values (42 instances, or 1.31 % of the data set; maximum of six undetectable values for

2,3,4,7,8PnCDF and PCB188) were replaced by a random number between zero and the limit of detection, while the stated concentration was used for two values reported by the 91 laboratory as not detected due to incorrect isotope ratio or NDR (peak detected but confirmingion ratios outside of the specified range).

Samples were normalized to the concentration total to remove artifacts related to concentration differences between samples. The centered logratio transformation

(division by the geometric mean of the concentrationnormalized sample followed by log transformation) was then applied to this compositional data to produce a data set that was unaffected by negative bias or closure (Bonn 1998; Ross et al. 2004). Data were then autoscaled (congeners were scaled by subtracting the variable mean and dividing by the variable standard deviation) to give every variable equal weight. Finally, a Varimax rotation was applied to the first three principal components (PCs) to simplify the physical interpretation of the PCA projections (Ross et al. 2004; Ross et al. 2004). This rotation maximised or minimised the loading of each variable on each PC while preserving trends.

With n = 24 adult chinook and p = 134 contaminants, the PCA model provided a case where n < p and the PCA model would be limited to n 1 = 23 statistically valid eigenvectors (Legendre and Legendre 1998). The first few eigenvectors are little affected when the PCA data matrix is not of full rank and having n < p does not lead to incorrect interpretations.

Linear relationships involving the PCA results were quantified using geometric mean (GM) linear regression (Ricker 1984; Cretney and Yunker 2000). The GM slope was calculated by dividing the y on x slope by the correlation coefficient for the regression, r (Ricker 1984); the mean values for the x and y variables were then used to calculate the intercept for the GM regression equation. To estimate the relative shift in contaminant distribution for each sample we used the linear distance along the GM linear 92 regression line for the fish samples, with the intersection point between the regression line and a perpendicular between the line and the sample position calculated using standard trigonometry mensuration formulae (Ricker 1984).

Dietary exposure calculations

As a means of characterizing health risks associated with dietary exposure we calculated estimated daily intakes (EDIs) of POPs by resident killer whales. Based on food consumption studies of captive killer whales, estimated intake as a function of body weight was calculated where food intake = 0.277 mass 0.663 (Kriete 1995). We used this relationship to estimate food intake by a 2500kg adult killer whale and calculated EDIs for POPs with an assumption of 71.5 % chinook consumption of a 96 % salmonid diet

(Kriete 1995). Given the limited information on nonsalmon prey items in the diet of resident killer whales, we restrict our exercise here to their dominant prey item (chinook).

Daily food intake for a 2500 kg killer whale = 0.277 mass 0.663 (4)

Salmonid portion of diet = 96% • 50 kg/d (5)

Chinook portion of diet = 71.5% • 48 kg/d (6)

EDI (µg/d) = [POP] wet wt • 34 kg/d (7)

4.4 Results and Discussion

As their primary prey item, chinook salmon provide both a source of nutrition and contaminants to northern and southern resident killer whales. The highly contaminated southern resident killer whales frequent the nearurban waters of the Strait of Georgia and

Puget Sound, while northern resident killer whales ply the more remote waters of central and northern British Columbia. While logistical and ethical challenges preclude an 93 accurate evaluation of dietary exposure to POPs by killer whales, we can estimate dietary

POP exposure in these killer whales using data from chinook salmon.

Life history and feeding ecology of chinook

Since chinook salmon is the primary prey of killer whales, an understanding of their life history and feeding ecology is important to exploring issues related to exposure and bioaccumulation in the killer whale food web. Stock identification assigned at least

67% of the adults collected from Johnstone Strait to the Thompson River region and 83% of the adults collected from the mouth of the Fraser River to the Lower Fraser River region (with Harrison River being the most probable population). Harrison River stock is known to be predominantly coastal in its marine distribution, being found on the West coast of Vancouver Island, the Strait of Georgia and Washington waters, whereas

Johnstone Strait (Thompson River stock) are known to migrate into the northern waters of British Columbia and the Gulf of Alaska ((Ford and Ellis 2006); http://www comm.pac.dfompo.gc.ca/publications/speciesbook/Salmon/chinook.fraser.html). All six chinook smolts sampled in the Strait of Georgia originated from eastern Vancouver

Island, with five from Big Qualicum River, and one from Little Qualicum River. Stock identification, δ13 C, δ15 N, morphometric, and meristic data provide an overview of the biology and ecology of sampled chinook salmon (Table 4.1).

Scale data from our chinook samples indicated that all adults migrated to marine waters during their first year of life with the exception of three British Columbia adults

(two Johnstone Strait and one Lower Fraser River) that spent one year in freshwater before going to sea. Although scale data indicated that Duwamish and Deschutes River stocks migrated to marine waters during their first year of life, some fish from these 94 populations are known to be resident stock that remain in Puget Sound waters yearround without migrating into open ocean (O'Neill and West 2009).

A significant difference in δ13 C ratios was observed among adult stocks (oneway

ANOVA, ν = 23; p = 0.001) and lipidnormalizing the δ13 C ratios (McConnaughey and

McRoy 1979) did not statistically affect δ13 C ratios among adults ( r2 = 0.24; ν = 23; p =

0.0123). However, no significant difference was apparent in δ15 N ratios among adults, suggesting similarities in trophic level (Table 4.1). Our chinook δ15 N ratios and δ13 C ratios decreased with lipid % in whole fish ( r2 = 0.18; ν = 23; p = 0.0404; r2 = 0.59; ν =

23; p < 0.0001, respectively) (data not shown). Previous studies have demonstrated that

δ13 C and δ15 N ratios can be affected by the nutritional status of organisms (Doucett et al.

1999; Hobson et al. 1993; Fuller et al. 2005) and that a range of at least 4 to 6 ‰ should be expected in both δ13 C and δ15 N values during the migration of salmon that is due to changes in lipid and protein concentrations (Doucett et al. 1999). Therefore, enrichment in δ13 C and δ15 N ratios with decreasing lipid content in our chinook salmon likely reflects the migrationassociated influence of declining lipid stores on δ13 C and δ15 N values, rather than trophic level or feeding ecology. This is consistent with changes in physiological condition in chinook salmon as they near their natal streams.

Contaminant concentrations in chinook salmon

Significant differences in PCB, PCDD, and PCDF concentrations, on a wet weight basis, were observed among adult salmon (ν = 23; p = 0.001; p = 0.011; p =

0.000, respectively), with Johnstone Strait salmon having the lowest concentrations and 95

Table 4.1. Morphometric and related data for chinook salmon ( Oncorhynchus tshawytscha ) collected from Johnstone Strait and Lower Fraser River (British Columbia, Canada), Duwamish River and Deschutes River (Washington State, USA), and chinook smolts from the Strait of Georgia (British Columbia, Canada) and Puget Sound (Washington State, USA). Values represent mean ± standard error of the mean . One way analysis of variance (AOVA) tests were used to assess significant differences (α = 0.05) between the four adult salmon groups ( ν = 23) and Tukey post hoc tests to assess which groups differed (results in italics). Student’s t test was used to assess significant differences between fillet and rest of fish lipid percentages. S = no significant difference; A = not analyzed.

Lower Fraser Strait of Johnstone Strait River Puget Sound Duwamish River Deschutes River AOVA test Georgia smolts adults adults smolts adults adults (Tukey test) Group 1 2 3 4

Number of fish ( n =) 6 6 6 12 6 6 NS

Fork length (cm) 17.98 ± 1.08 88.70 ± 2.47 73.88 ± 3.13 6.90 ± 0.19 77.28 ± 2.39 78.23 ± 2.28 0.004 (12; 13; 14)

Body weight (kg) 0.08 ± 0.02 10.86 ± 1.07 6.08 ± 0.57 0.003 ± 0.0001 6.32 ± 0.57 5.95 ± 0.44 0.001 (12; 13; 14)

Age (years) NA 3.50 ± 0.22 2.50 ± 0.34 NA 2.33 ± 0.21 2.33 ± 0.21 0.010 (12; 13; 14) Sex (male:female) 3:3 3:3 5:1 NA 3:3 3:3 0.607

Whole fish lipid (%) 0.87 ± 0.26 14.06 ± 1.37 a 9.48 ± 0.62 a 1.35 b 6.38 ± 0.61 4.29 ± 0.82 0.000 (12; 13; 14; 24) Fillet lipid (%) NA 10.03 ± 1.42 5.37 ± 0.92 NA NA NA 0.016 Rest of fish lipid (%) 19.45 ± 1.44 12.84 ± 0.77 NA 0.002

Moisture (%) 74.93 ± 0.85 62.55 ± 1.31 66.54 ± 0.17 79.41 69.70 ± 0.65 71.66 ± 1.62 0.000 (13; 14; 24)

δ15 N 13.68 ± 0.17 14.97 ± 0.23 15.51 ± 0.05 10.26 ± 0.42 b 15.59 ± 0.24 15.54 ± 0.37 0.282

δ13 C 18.23 ± 0.24 20.51 ± 0.35 18.59 ± 0.27 22.84 ± 0.26 b 18.94 ± 0.36 18.11 ± 0.50 0.001 (12; 13; 14)

a Whole fish lipid percentage calculated using fillet lipid percentage and rest of fish lipid percentage. b Data generated from a pooled sample of 12 individuals.

Deschutes River salmon having the highest concentrations ( Table 4.2). Polychlorinated biphenyls were the dominant POP detected in all salmon sampled, including smolts. Two of the six adult chinook sampled from each of the Lower Fraser, Duwamish, and

Deschutes Rivers exceeded mean PCB concentrations found in a previous study where

Puget Sound returning chinook, collected from either nearshore estuaries or river locations had a detected mean value of 49.1 µg/kg wet wt (O'Neill et al. 1998).

Significant differences in OC pesticides were observed among the adult salmon stocks ( Table 4.3). Duwamish River salmon had the highest concentrations of all OC pesticides with the exception of HCH compounds. Total DDT dominated the OC pesticide rankings among both British Columbia and Washington smolts and three out of the four returning adult groups (Table 3), while total HCH dominated the OCs in

Johnstone Strait adults. Although DDT and HCH appear to be the dominant OC pesticides detected in both British Columbia and Washington salmon, their isomeric compositions may reflect differences in distance from source/use regions. The contribution of the DDT degradation products (ΣDDE and ΣDDD) in all chinook samples was 88 to 97 % of the ΣDDT, suggesting fresh input to be minimal. The high concentrations of the predominant parent αHCH isomer, the most bioaccumulative isomer βHCH, and lower concentrations of the insecticide γHCH are apparent as one moves away from source/use regions, reflecting partitioning properties which favor colder, more northerly waters (Li et al. 2002). The volatility of HCH ensures its ready atmospheric transport from Asia to the northeastern Pacific Ocean via prevailing winds

(Wania and Mackay 2001). 97

Table 4.2. Wet weightbased concentrations of persistent organic pollutants and toxic equivalents (TEQs) to 2,3,7,8 tetrachlorodibenzopdioxin (TCDD) for polychlorinated biphenyls (PCBs), polychlorinated dibenzopdioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs) in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait (n = 6) and Lower Fraser River ( n = 6) (British Columbia, Canada); Duwamish River ( n = 6) and Deschutes River ( n = 6) (Washington State, USA); and chinook smolts ( n = 6) from the Strait of Georgia (British Columbia, Canada) and Puget Sound (n = 1 pool of 12) (Washington State, USA). Values represent mean ± standard error of the mean. Oneway analysis of variance (AOVA) tests were used to assess significant differences (α = 0.05) between the four adult salmon groups ( ν = 23) and Tukey post hoc tests to assess which groups differed (results in italics). SG = Strait of Georgia; PS = Puget Sound; A = not analyzed; D = statistical comparison not possible.

Sum congeners/ µg/kg wet weight isomers a (except for ΣΣΣPCDD and ΣΣΣPCDF)

SG Smolts Johnstone Strait Lower Fraser PS Smolts Duwamish River Deschutes River ANOVA test River (Tukey test) Group 1 2 3 4

Lipid (%) b 0.87 ± 0.26 10.03 ± 1.42 5.37 ± 0.92 1.35 6.38 ± 0.61 4.29 ± 0.82 (refer to Table 1) ΣPCB c 12.03 ± 1.46 9.07 ± 1.49 46.97 ± 8.06 9.63 34.61 ± 8.09 56.09 ± 17.97 0.001 (12; 13; 14) Σ PCB TEQ 0.30 ± 0.04 0.17 ± 0.03 0.74 ± 0.11 0.28 0.55 ± 0.12 1.09 ± 0.35 0.007 (12; 14) ΣPCDD c (ng/kg) 1.39 ± 0.32 0.58 ± 0.05 0.81 ± 0.14 1.57 0.83 ± 0.15 1.74 ± 0.63 0.011 (14) Σ PCDD TEQ 0.35 ± 0.10 0.03 ± 0.02 0.27 ± 0.08 0.00 0.12 ± 0.04 0.31 ± 0.05 0.00 (12; 13; 14) ΣPCDF c (ng/kg) 2.03 ± 0.48 0.50 ± 0.12 1.90 ± 0.38 0.26 1.30 ± 0.24 1.92 ± 0.31 0.000 (12; 13; 14) ΣPCDF TEQ 0.24 ± 0.06 0.06 ± 0.02 0.28 ± 0.05 0.00 0.11 ± 0.04 0.25 ± 0.06 0.007 (12; 14)

a PCB = polychlorinated biphenyl; PBDE = polybrominated diphenyl ether; PCDE = polychlorinated diphenyl ether; PBB = polybrominated biphenyl; PCDD = polychlorinated dibenzopdioxin; PCDF = polychlorinated dibenzofuran. b Whole fish percentage lipid for SG smolts, PS smolts, Duwamish and Deschutes River adults; fillet percentage lipid for Johnstone Strait and Lower Fraser River adults. c Whole fish analyzed for SG smolts, PS smolts, Duwamish and Deschutes River adults; fillet analyzed for Johnstone Strait and Lower Fraser River adults.

ΣPBDE d NA NA 17.71 NA 6.43 NA ND ΣPCDE d NA NA 0.53 NA 0.24 NA ND ΣPBB d NA NA 0.10 NA 0.04 NA ND ΣTEQs 0.89 ± 0.20 0.26 ± 0.06 1.30 ± 0.19 0.28 0.78 ± 0.18 1.65 ± 0.44 0.006 (12; 14)

d Pooled fillet ( n = 1 pool of 6 fish) analyzed

Table 4.3. Wet weightbased concentrations of organochlorine pesticides in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait ( n = 6) and Lower Fraser River ( n = 6) (British Columbia, Canada); Duwamish River ( n = 6) and chinook smolts ( n = 6) from the Strait of Georgia (British Columbia, Canada) and Puget Sound ( n = 1 pool of 12) (Washington State, USA). Values represent mean ± standard error of the mean. Oneway analysis of variance (AOVA) tests were used to assess significant differences (α = 0.05) between the three adult salmon groups (ν = 17) and Tukey post hoc tests to assess which groups differed (results in italics). SG = Strait of Georgia; PS = Puget Sound; A = not analyzed;

Sum congeners/isomers a µg/kg wet weight

SG Smolts b Johnstone Strait c Lower Fraser PS Smolts b Duwamish Deschutes ANOVA test River c River b River b (Tukey test) Group 1 2 3 (refer to Table Lipid (%) 0.87 ± 0.26 10.03 ± 1.42 5.37 ± 0.92 1.35 6.38 ± 0.61 4.29 ± 0.82 1) ΣDDT 4.38 ± 0.55 1.46 ± 0.27 4.29 ± 0.50 2.68 18.31 ± 3.94 NA (0.000) (12; 13; 23) o,p' −DDT 0.05 ± 0.01 0.07 ± 0.02 0.04 ± 0.00 0.02 0.12 ± 0.01 NA (0.014) (23) p,p'DDT 0.27 ± 0.04 0.10 ± 0.02 0.22 ± 0.03 0.10 0.40 ± 0.08 NA (0.000) (12; 13) o,p' −DDD 0.06 ± 0.01 0.07 ± 0.01 0.06 ± 0.01 0.01 0.21 ± 0.02 NA (0.001) (13; 23) p,p'DDD 0.40 ± 0.08 0.25 ± 0.03 0.60 ± 0.05 0.13 2.88 ± 0.50 NA (0.000) (12; 13; 23) o,p' −DDE 0.08 ± 0.02 0.07 ± 0.01 0.04 ± 0.00 0.02 0.12 ± 0.02 NA (0.000) (12; 13; 23) p,p'DDE 3.52 ± 0.43 0.90 ± 0.21 3.34 ± 0.42 2.40 14.58 ± 3.39 NA (0.000) (12; 13; 23) ΣDDE + Σ DDD 0.93 ± 0.00 0.88 ± 0.01 0.94 ± 0.00 0.95 0.97 ± 0.01 NA (0.002) ΣDDT (12; 13) ΗCB 0.36 ± 0.06 1.50 ± 0.16 0.85 ± 0.09 0.29 2.15 ± 0.12 NA 0.000 (12; 13; 23) ΣHCH (α, β, γ) 1.09 ± 0.25 2.28 ± 0.23 0.68 ± 0.14 0.27 1.60 ± 0.20 NA 0.000 (12; 13; 23) 100

alpha (α) 0.32 ± 0.08 0.98 ± 0.10 0.25 ± 0.05 0.08 0.84 ± 0.10 NA 0.000 (12; 23) beta (β) 0.34 ±0.08 1.09 ± 0.11 0.37 ± 0.08 0.10 0.63 ± 0.08 NA 0.000 (12; 13) gamma (γ) 0.43 ± 0.10 0.21 ± 0.02 0.06 ± 0.01 0.08 0.12 ± 0.03 NA 0.005 (12; 13) Heptachlor

a DDT = dichlorodiphenyltrichloroethane; DDD = dichlorodiphenyldichloroethane; DDE = dichlorodiphenyldichloroethylene; HCB = hexachlorobenzene; HCH = hexachlorocyclohexane. b Whole fish analyzed c Fillet analyzed d ΣChlordanes = Heptachlor + Heptachlor epoxide + Oxychlordane + cis and trans Chlordane + cis and tran s Nonachlor 102

Of the two adult chinook pools analyzed for PBDEs, the most predominant congeners detected were BDE47 and BDE99, respectively. The PBDE profile for

Lower Fraser River chinook was 47 > 99 > 100 > 49 > 209, and for Duwamish River chinook was 47 > 99 > 100 > 49 > 120. Similar congener profiles have been observed in chinook from Oregon (BDE47 > 99 > 100> 49 > 154) (Stone 2006) and in Lake

Michigan salmonids (BDE47 > 99 > 100 > 154 > 153) (ManchesterNeesvig et al.

2001). The ratio of PBDE: PCB concentrations were 0.4:1 for the Lower Fraser River adults and 0.2:1 for the Duwamish River adults, highlighting the emergence of PBDEs as a significant environmental contaminant.

Significant differences in ΣPCB TEQs, ΣPCDD TEQs, ΣPCDF TEQs, and

ΣTEQs were observed among the four adult salmon stocks (Table 2). Although PCBs explained the majority of ΣTEQs in the adult salmon groups, Lower Fraser River adults had a significantly lower Σplanar PCB TEQ contribution to the ΣTEQ compared with the

Johnstone Strait and Duwamish River adults (ν = 23; p = 0.007; p = 0.011, respectively).

This in part is due to the higher ΣPCDD contribution to the ΣTEQ in the latter samples.

While this may in part reflect differences in dietary exposure for the different stocks, metabolic removal or preferential elimination of the planar PCBs may also explain this observation. Polychlorinated biphenyls made up 100 % of the ΣTEQs in the Puget Sound smolts, whereas in the Strait of Georgia the ΣPCDD and ΣPCDF TEQs made up a greater proportion of the ΣTEQs. These results are consistent with those found in our Puget

Sound and Strait of Georgia harbour seal food baskets, likely reflecting differences in regional source inputs between pulp mills in the Strait of Georgia and more industries in

Puget Sound (Cullon et al. 2005). 103

POP origin in returning adult salmon

By comparing body burdens of POPs in returning adult chinook to outmigrating smolts and juveniles, we estimate that 97 to 99 % of the body burden of PCBs, PCDDs,

PCDFs, DDT, and HCH in all stocks originated during their time at sea (Table 4.4).

Field sampling provided us with salmon that could be identified only after genetic analysis. As a result of differences in stock identification between some smolts and adults, stockspecific assignment of the POPs in adults was not directly possible. Our estimation that the majority of POPs in chinook salmon can be ascribed to their growth stage in coastal and marine waters is consistent with other studies. A study of chinook from Washington State ascribed 99 % of PCBs in returning Duwamish River adults to the waters of Puget Sound and the Pacific Ocean (O'Neill et al. 1998). A later study which included returning adult chinook sampled from five Puget Sound River locations also found that > 96 % of PCBs were accumulated in marine waters (O'Neill and West 2009).

The concentrations of POPs detected in our smolts are comparable to values previously reported in outmigrating chinook salmon smolts from a number of stocks in Washington and Oregon (Johnson et al. 2007), further underscoring the limited contribution of locally acquired contaminants during the juvenile stage. It is increasingly clear that salmon acquire the majority POPs during their growth period at sea and that more research is needed on the extent of Pacific Ocean food web contamination.

104

Table 4.4. Estimated body burdens of persistent organic pollutants in returning adult chinook salmon ( Oncorhynchus tshawytscha ) from Johnstone Strait ( n = 6) and Lower Fraser River ( n = 6) (British Columbia, Canada); Duwamish River ( n = 6) and Deschutes River ( n = 6) (Washington State, USA); and chinook smolts ( n = 6) from the Strait of Georgia (British Columbia, Canada) and Puget Sound ( n = 1 pool of 12) (Washington State, USA). Body burden estimates calculated using POP lipid weight concentrations and whole fish lipid mass. Values represent mean ± standard error of the mean. Oneway analysis of variance (AOVA) tests were used to assess significant differences (α = 0.05) between the four adult salmon groups (ν = 23) and three adult groups for total dichlorodiphenyltrichloroethane ( ΣΣΣDDT) and total hexachlorocyclohexane ( ΣΣΣHCH) (ν = 17). Tukey post hoc tests were used to assess which groups differed ( results in italics ). A = not analyzed; D = statistical comparison not possible.

Sum congeners/ Body Burden (µg) isomers a (except for ΣΣΣPCDD and ΣΣΣPCDF)

Strait of Georgia Johnstone Strait Lower Fraser Puget Sound Duwamish Deschutes River ANOVA test Smolts Adults River Smolts River Adults (Tukey test) Adults Adults

Group 1 2 3 4

ΣPCBs 1.12 ± 0.40 141.54 ± 30.48 537.58 ± 99.01 0.03 216.32 ± 56.88 339.62 ± 108.82 0.017 as % returning burden 99.21 % 99.79 % 99.99 % 99.99 % (12)

ΣPCDDs (ng) 0.13 ± 0.04 8.99 ± 1.44 10.14 ± 3.18 0.005 5.50 ± 1.34 9.76 ± 3.07 0.194 as % returning burden 98.56 % 98.72 % 99.91 % 99.95 %

ΣPCDFs (ng) 0.21 ± 0.10 8.07 ± 2.30 23.94 ± 8.05 0.001 8.21 ± 1.87 11.48 ± 2.01 0.018 as % returning burden 97.40 % 99.12 % 99.99 % 99.99 % (12; 23)

ΣPBDEs NA NA 165.75 NA 54.34 NA ND

a PCB = polychlorinated biphenyl; PCDD = polychlorinated dibenzopdioxin; PCDF = polychlorinated dibenzofuran; PBDE = polybrominated diphenyl ether; DDT = dichlorodiphenyltrichloroethane; DDD = dichlorodiphenyldichloroethane; DDE = dichlorodiphenyldichloroethylene; HCH = hexachlorocyclohexane. b ΣDDT includes DDT ( o,p' −DDT, p,p'DDT ), DDD ( o,p' −DDD, p,p'DDD ), and DDE ( o,p' −DDE, p,p'DDE ) 105

ΣDDT b 0.15 ± 0.08 24.11 ± 6.30 44.46 ± 8.00 0.01 107.96 ± 28.08 NA 0.003 as % returning burden 99.38 % 99.66 % 99.99 % (13)

ΣHCH c 0.04 ± 0.02 35.01 ± 3.74 6.68 ± 1.21 0.001 9.32 ± 1.55 NA 0.000 as % returning burden 99.88 % 99.40 % 99.99 % (12; 13)

c ΣHCH includes (α, β, γ) HCH

106

Lipidnormalized ΣPCB and ΣDDT concentrations increased with δ15 N ratios among adult chinook ( r2 = 0.31; ν = 23; p = 0.0046; r2 = 0.46; ν = 17; p = 0.0020, respectively), as did ΣPCB and ΣDDT body burdens ( r2 = 0.25; ν = 23; p = 0.0306; r2 =

0.34; ν = 17; p = 0.155; respectively) (results not shown). While our observed relationship between these POPs and δ15 N could be interpreted as reflecting an influence of trophic level, it may also signal an effect of migrationassociated lipid changes.

Changes in tissue concentrations of lipid and protein in migrating salmon complicate this interpretation of stable isotopedefined trophic level assignment (Doucett et al. 1999).

Contaminant patterns in adult chinook

In the present study the primary purpose of PCA modeling is to quantitatively compare the contaminant distributions between different adult chinook populations.

Because the PCA algorithm uses the variable magnitudes when decomposing the data set into a series of orthonormal rank 1 matrices or PCs, the substantial concentration differences between populations (Table 4.2) have to be removed by normalizing each sample before PCA. The difficulty is that this normalisation step introduces closure

(spurious negative correlations in the highest variables and negative correlations in the smallest). Centered log ratio transformation removes this closure and produces a data set where the average concentration and concentration total are identical for every sample

(Bonn 1998; Bonn 1998). In the PCA model the first two PCs account for the largest percentage of total variance in the data set and, particularly when data are normalised, reflect the most discriminating compositional features. The contaminants with variable 107 loadings near axis centre have essentially no contribution to a PC, while the contribution to a PC increases as the absolute magnitude of the variable loading.

Principal component analyses differentiated adult chinook on the basis of variation in PCB, PCDD, and PCDF congener proportions ( Figure 4.2a). In the Varimax rotated PCA model, chinook salmon samples project along a line from the upper left to the lower right of the samples plot (Figure 4.2a). Because both variables (the PCs) in this relationship between chinook samples are affected by natural variability, rather than just measurement error, the appropriate regression line to use quantify the relationship is GM linear regression (Yunker et al. 2005; Ricker 1984; Cretney and Yunker 2000).

Geometric mean linear regression for the sample projections of chinook samples indicates that this linear relationship in Figure 4.2a is highly significant ( r2 = 0.840, v =

22, p = 3.1E10). In the corresponding variables plot, most PCDD, PCDF, and lower chlorine number PCB congeners project in the upper left quadrant while the higher chlorine number PCB congeners project in the lower right quadrant ( Figure 4.2b).

Geometric mean regression for the variables also indicates that this linear relationship is highly significant ( r2 = 0.377, v = 132, p = 3.1E15), despite the greater amount of scatter in the variables plot. Comparison of samples and variables indicates that salmon samples projecting towards the upper left of the samples plot have higher proportions of the

PCDD, PCDF and lower chlorine number and non and monoortho PCB congeners while samples projecting on the lower right have higher proportions of the higher chlorine number, diortho PCB congeners.

The differences in contaminant composition are not obviously related to either sex or sampling location (urban vs remote, or BC vs Puget Sound) for the salmon samples 108

(Figure 4.2a). The shifts in contaminant composition along the GM regression line for the samples ( Figure 4.2c) correlated with lipid content ( r2 = 0.328; p = 0.0034), δ13 C composition ( r2 = 0.605; p = 7.7E6), and body weight (r2 = 0.214; p = 0.0227), but are not significantly related to δ15 N (r2 = 0.152; p = 0.0596; ν = 22 in all cases).

Accordingly, the change in contaminant composition for the salmon appears to reflect metabolism or solubilization of the PCDD, PCDF and lower chlorine number and non and monoorthoPCB congeners as the salmon lose lipid during migration. This suggests that the migrating salmon PCB burdens will be increasingly dominated by the more heavilychlorinated congeners. Similar observations in sockeye salmon were thought to reflect a greater metabolic capacity by salmonids for PCDDs and PCDFs as compared to

PCBs (DeBruyn et al. 2004). While our results support the notion of compositional loss associated with depleting lipid reserves during migrating salmon, a contribution of local

POP sources from more contaminated areas, such as Puget Sound, cannot be ruled out

(DeBruyn et al. 2004; Missildine et al. 2005). Indeed, feeding in such an area during outward and inwardbound migrations likely does lead to increased POP concentrations in certain salmon individuals and stocks.

109

110

Figure 4.2. Contaminant patterns in chinook salmon are relevant to assessing the influence of feeding ecology in chinook and dietary exposure of persistent organic pollutants in resident killer whales. Varimax rotated projections of the first two principal components (PCs) for a principal components analysis model based on normalised concentrations (see text) showing (a) chinook salmon scores (t1 and t2) by sampling location and sex (M = male; F = female), and (b) dibenzopdioxin (PCDD), dibenzofuran (PCDF) and polychlorinated biphenyl (PCB) congener variable loadings (p1 and p2) by chlorine number. In (b) PCDDs and PCDFs have a “D” preceding the number, the dioxinlike nonortho and monoortho PCBs are in italics and diorthoPCBs use a regular font. In (c) the lipid, δδδ13 C and δ 15 content is plotted by sampling location and sex against the relative distance along the GM linear regression line in the first PC for the salmon samples (a). ( PCA plots prepared by M. Yunker )

While our results suggest that salmon accumulate the majority of POPs during their growth period at sea, lipid depletion and metabolism in salmon associated with migration may have profound consequences for dietary exposure to POPs in resident killer whales. While both northern and southern resident killer whales preferentially consume chinook salmon, southern residents likely intercept more chinook in relatively contaminated, nearurban areas, and at points closer to their natal streams. Southern residents may therefore be consuming chinook salmon that is both more contaminated and less lipidrich.

Health risks for killer whales

Dietary POP concentrations and patterns have profound implications for killer whale POP accumulation and consequent related health risks. High trophic level marine mammals have shown susceptibility to adverse health effects such as immunotoxicity, endocrine disruption, reproductive impairment, and developmental abnormalities with elevated exposure to POPs (Ross et al. 2000). To characterize health risks in killer 111 whales associated with dietary exposure to POPs, chinook ΣPCB, ΣPCDD/PCDF TEQ and ΣDDT concentrations were compared with Canadian Council of Ministers of the

Environment (CCME) tissue residue guidelines for the protection of mammalian wildlife consumers of aquatic biota ((Canadian Council of Ministers of the Environment 2001); http://www.ccme.ca/assets/pdf/trg_summary_table.pdf). The Deschutes River salmon exceeded, and the Lower Fraser River salmon were approaching (Table 4.2), the CCME

PCB tissue residue guidelines (0.79 ng TEQ/kg diet wet wt)(Canadian Council of

Ministers of the Environment 2001). The Duwamish River salmon (Table 4.2) exceeded the ΣDDT tissue residue guidelines (14.0 µg/kg diet wet wt)(Canadian Council of

Ministers of the Environment 2001).

The ΣPCB and ΣDDT concentrations in all salmon groups were below the less conservative U.S. guidelines (New York State Department of Environmental

Conservation) for protection of fisheating wildlife (Newell et al. 1987). All of the chinook analyzed in the present study exceeded the 8 µg/kg dietary PCBs that was estimated to protect 95 % of a killer whale population, based on a 17 mg/kg PCBs adverse effects threshold for marine mammals (Hickie et al. 2007).

Another means of characterizing health risks associated with dietary exposure is through the calculation of EDI. Based on food intake estimates derived from studies of captive killer whales (Kriete 1995), we have estimated the food intake of a 2500kg resident killer whale to be approximately 50 kg per day. We have further estimated the chinook portion, 71.5 % of a 96 % salmonid diet (Ford and Ellis 2006), to be 34 kg per day. Taking into account observed ranges for resident killer whales (Ford et al. 2000), 112

POP concentrations (wet wt) for Johnstone Strait chinook were used to calculate EDIs for northern residents and all four chinook stocks for southern residents ( Table 4.5).

Table 4.5. Estimated daily intake (EDI) of persistent organic pollutants (POPs) by northern and southern resident killer whales. Johnstone Strait (British Columbia, Canada) chinook POP concentrations have been used to calculate EDIs for northern residents and all four chinook stocks for southern residents. We have further calculated the EDI for southern residents if they were to consume chinook of equivalent lipid content to that of northern residents, i.e., if northern residents were to consume 34 kg chinook per day (8.5% lipid); southern residents may consume up to 85 kg chinook per day (3.4% lipid).

Estimated daily intake µg/d (except for ΣΣΣPCDD, ΣΣΣPCDF and ΣΣΣTEQ ng/d)

Northern residents Southern residents Southern residents Sum congeners/isomers a (lipidequivalent)

ΣPCB 308.49 1248.00 2051.38 ΣPBDE No Data 410.62 674.95 ΣPCDD 19.75 33.67 55.34 ΣPCDF 17.17 47.82 78.60 ΣDDT (DDT, DDD, DDE) 49.72 272.88 448.55 HCH ( α-, β-, γ-) 77.76 51.74 85.05

ΣTEQ (PCB, PCDD, PCDF) 9.00 36.85 60.58 a PCB = polychlorinated biphenyl; PBDE = polybrominated diphenyl ether; PCDE = polychlorinated biphenyl ether; PBB = polybrominated biphenyl; PCDD = polychlorinated dibenzopdioxin; PCDF = polychlorinated dibenzofuran; TEQ = toxic equivalents; DDT = dichlorodiphenyltrichloroethane; DDD = dichlorodiphenyldichloroethane; DDE = dichlorodiphenyldichloroethylene; HCB = hexachlorobenzene; HCH = hexachlorocyclohexane.

113

Our EDIs suggest that southern residents may be consuming, on a body weight basis, 4.0 times more PCBs than their northern counterparts, consistent with the differences in PCB concentrations measured in biopsies collected from freeranging northern and southern resident killer whales (Ross et al. 2000). However, since studies of marine mammal energetics suggest that calorimetric content is an integral component of food needs (Kriete 1995; Kastelein and Vaughan 1989), we have also adjusted consumption to reflect equivalent lipid content. Because of the lower lipid content of our more southerly chinook salmon, there may be a compensatory increase in consumption by southern resident killer whales. This nutritionally adjusted scenario would predict that southern residents would consume 6.6 times more PCBs than northern residents.

Similarly, we previously speculated that Puget Sound harbour seals consume nearly twice as much prey as Strait of Georgia seals in order to compensate for the lower lipid content in their prey, with results explaining a neardoubling of their contaminant burden (Cullon et al. 2005). Additional studies on killer whale feeding ecology and on the behavior of

POPs in salmon during different life history stages will shed more insight into the sources and fate of contaminants in killer whale food webs.

4.5 Conclusion

The present study underscores the global nature of contaminant dispersion with chinook salmon acquiring the majority of their POPs during their time at sea. As the two resident killer whale populations in British Columbia intercept these returning salmon, they are exposed to different dietary POP concentrations. We conclude that the endangered southern resident killer whales are exposed to much higher concentrations of 114

POPs than their northern counterparts through the consumption of more POP contaminated chinook salmon, and may increase their consumption of salmon in order to compensate for the reduced lipid content observed in southerly chinook. In this regard, increasing climaterelated stresses on salmon abundance and lipid content raise the specter of increased contaminant exposures for resident killer whales in the future.

115

Chapter 5: Conclusions

5.1 Introduction

The information drawn from field research such as this provide the most conclusive evidence of biomagnification of “real world” mixtures of chemicals in food webs (Gobas et al. 2009). This thesis has provided increased understanding of biomagnification and fate of persistent organic pollutants (POPs) in Northeastern Pacific

Ocean marine mammal food webs. The major conclusions and contributions to the existing science of food web bioaccumulation of POPs are discussed within the context of the following questions.

5.2 Is 15 : 14 a good predictor of PCB accumulation in marine mammal food webs?

Many studies have provided evidence of PCB accumulation with increasing stable nitrogen isotope ratios ( δ15 N) ratios in aquatic food webs (Ruus et al. 2002; Hoekstra et al. 2003; Fisk et al. 2001; Hobson et al. 2002; Nfon et al. 2008). Although this demonstrates value for characterizing POP accumulations in the harbour seal food web, the utility of δ15 N ratios remains limited when used on its own. Stable nitrogen isotope ratios are greatly affected by the physiological state of an organism (Doucett et al. 1999), as well by diet and tissue types sampled for analysis (Deniro and Epstein 1981). The strength of this tool increases when used in conjunction with other methods, such as congener specific POP analyses, lipid content analysis, and principal components analysis (PCA). It may also be best applied to entire food web analyses instead of single species comparisons. 116

In the food basket study (Chapter 2), we found that δ15 N ratios in the Puget Sound harbour seal food basket were slightly higher than the Strait of Georgia seal food basket.

While this suggests that Puget Sound harbour seals feed at a slightly higher trophic level

(TL), these differences were unlikely to explain a sevenfold difference in PCB concentrations in the two food baskets. Our diet switching exercise supported this, as

PCB concentrations in our inverted food baskets did not increase with higher δ15 N ratios nor decrease with lower δ15 N ratios. Therefore, although δ15 N ratios provided insight into the feeding ecology of harbour seals inhabiting the two regions, on their own they could not explain the magnitude of the differences in food web contamination observed between Strait of Georgia and Puget Sound.

The harbour seal food web study (Chapter 3) demonstrated trophic level accumulations of POPs, as illustrated in the more recalcitrant congener CB153.

However, this study also showed that other factors such as log K ow and metabolic capacity in seals were important in influencing many of the less recalcitrant POPs. In the salmon study (Chapter 4), lipidnormalized ΣPCBs and ΣDDT concentrations increased with δ15 N ratios, however compositional patterns in POPs did not significantly correlate with δ15 N ratios ( p = 0.06). Therefore, although POP accumulations may be influenced by trophic level, they may also be affected by migrationassociated lipid and protein changes in the salmon.

This research has shown that δ15 N ratios can be a strong predictor of PCB accumulation in food webs, but also indicates that high trophic level (high δ15 N ratios) does not necessarily equate with highest concentrations in a food web. An example of this is with our ΣPCB levels detected in spot prawns ( δ15 N 11.8 ± 0.1 ‰), which 117 exceeded concentrations in both our rockfish species ( δ15 N 15.2 ± 0.1 ‰) and adult chinook salmon ( δ15 N 15.2 ± 0.1 ‰) (Table 3.2). Stable nitrogen isotope ratios provide understanding to food web structure which includes feeding ecology and habitat as well as a means of overall contaminant accumulations within food webs. It cannot provide a clear interpretation of contaminant accumulations in the Strait of Georgia marine mammal food webs, as it does not capture the full influence of feeding ecology, habitat use, physiology and other factors.

5.3 Can we predict the importance of local (within the Georgia Basin) vs global POP sources using deviations from δδδ15 vs PCB regression lines in marine mammal food webs?

The Strait of Georgia is subject to both local and global sources of POPs (Ross et al. 2004; Noël et al. 2009). Predicting sources of PCBs in regions such as this would be beneficial to many stakeholders. Information as to local, regional or other sources of contaminants is used in environmental assessments, human health assessments, species management, industry, and research. Analyses of PCB compositional patterns in this study has shown that patterns may reflect differences in source (local vs global), region

(Strait of Georgia vs Puget Sound), or feeding ecology (resident vs migratory species), as illustrated in our food baskets, food webs, and chinook salmon.

Other approaches to determining POP sources have included stable carbon isotope ratios (France and Peters 1997; Kurle and Worthy 2001; Tieszen et al. 1983), POP patterns in migratory species (Kelly et al. 2007; Krümmel et al. 2003) and nonmigratory low trophic level species (Ikonomou et al. 2002). While these methods offer insight into 118 feeding ecology and sources of POPs, they are often hindered and/or restricted by seasonal and individual variation (Deniro and Epstein 1978). Our collective datasets provide three approaches for characterizing the real importance of local source differences the comparison of Strait of Georgia and Puget Sound food baskets, regression slope differences in migratory and resident species in a food web, and the acquisition of POPs by adult Chinook salmon during their time at sea. The cumulative lines of evidence from our research provides greater confidence in interpretation and further understanding of POP sources to marine mammals in the Strait of Georgia and

Puget Sound. Biomagnification regression slopes may offer a means to guide further analyses, such as selection of specific species for indepth analyses. A good understanding of feeding ecology and migratory behaviour of species is necessary to avoid misinterpretation. Used in conjunction with stable isotope ratios and congener specific analyses, it serves as an additional means of interpretation of data that provides insight into biomagnification of POPs in the harbour seal food web.

5.4 Does lipid content in diet or eating on a lipidweight basis explain PCB contamination in marine mammals?

Although studies have shown that POPs reside in the lipid fraction of tissue without preference for specific tissues (Isosaari et al. 2004), lipid may be an under emphasized factor which influences contaminant accumulation in marine mammals.

Throughout this study the data has provided evidence that lipid is an important factor in

POP accumulations. Both the food basket study and the killer whale exposure studies have suggested that if a species consumes on a lipid weight basis, they may be exposed to 119 much higher POP concentrations when they are forced to consume a low lipid diet. The further correlations between PCB compositional patterns and lipid in chinook salmon have shown a migrationassociated metabolism and loss of lesschlorinated PCBs. This may result in resident killer whales consuming salmon enriched in the more heavily chlorinated PCBs. Studies of returning Pacific sockeye salmon ( Oncorhynchus nerka ) have also observed changes in lipid reserves associated with magnification of POP concentrations (Kelly et al. 2007; DeBruyn et al. 2004). With the existing knowledge base of feeding ecology for harbour seals (Olesiuk 1993; Olesiuk et al. 1990) and resident killer whales (Ford and Ellis 2006), this study has provided insight into health risks associated with dietary exposure for marine mammals.

5.5 Does the phrase “You are what you eat” provide a relatively accurate descriptor for marine food web contamination?

For high trophic level marine mammals such as killer whales and harbour seals, the phrase “ You are what you eat ” is indeed a relatively accurate descriptor for POPs.

This study provides the first ever “food basket” approach to characterizing POPs for a wildlife species, and offers a more robust alternative to singleitem or species approach carried out elsewhere (StorrHansen et al. 1995). Our harbour seal food baskets were representative of the diets of both the Strait of Georgia and Puget Sound seals (Cullon et al. 2005) and results were consistent with observations of burdens in the seals themselves from the two regions (Ross et al. 2004). Again, in the killer whale diet, the high PCB concentrations detected in the more southerly chinook stocks were reflective of the higher concentrations in the Southern resident killer whales (Ross et al. 2000). 120

Influencing factors within organisms such as sex, age, lipid, reproduction, metabolism, excretion as well as the physicochemical factors such as log K ow , chemical structure, etc. make it very difficult to interpret contaminant patterns (Borgå et al. 2004).

Through the use of stable isotope ratios, congenerspecific contaminant analyses, and principal components analysis, we have provided further evidence of the influences of metabolism and physicochemical properties of chemicals on POP accumulations in marine mammals. Perhaps a better descriptor for POPs might be “You are what you eat and can’t excrete”.

5.6 What amount of PCBs are in the Strait of Georgia biota?

An opportunity existed to explore the mass of PCBs in the Strait of Georgia biota using data from our food web studies, and data obtained from the literature (see Chapter

3). A massbalance ecosystem model was developed for present day Strait of Georgia marine biomass by trophic level using the Ecopath modelling software (Dalsgaard et al.

1998). Using PCB data from this study and other recent Strait of Georgia studies suggest that more than 77 kg of PCBs resides within the Strait of Georgia biota. Of this, fisheries resources containing an estimated 0.6 kg of PCBs would be removed annually for human consumption. Detritus was previously estimated to be 7.0 t/km 2 (Dalsgaard et al. 1998) in an earlier model, however, has been excluded in this model because the linkage between ecosystem maturity with retaining and recycling detritus for the Strait of Georgia is not fully understood (Dalsgaard et al. 1998).

Trophic level breakdown of this mass balance in the Strait of Georgia indicates that 16 % of ΣPCBs reside in biota occupying trophic levels 23 and 36% reside in high 121 trophic level marine mammals ( Figure 5.1 ). Approximately 20 % of ΣPCBs are found in shortlived species (<10 years), with the remainder found in longlived species (>=10 years) (Figure 5.1). Very longlived species (>30 years) are estimated to carry approximately 43 % of ΣPCBs in the Strait of Georgia food web, highlighting the important role that bioaccumulation plays over a lifetime, and that biomagnification plays in the food web (Figure 5.1).

PCB Biomass (kg) Biomass PCB 10 Σ

0 12 23 34 45 01 110 1030 30+

Trophic level Age (yrs)

Figure 5.1 We estimate that high trophic level (TL) marine mammals (TL 45) carry almost 28 kg of PCBs, representing approximately 36 % of the PCBs in the Strait of Georgia food web.

Within the Strait of Georgia biota, there appears to be a portion of PCBs with a short turnover time, thus become bioavailable to other compartments (water, sediment, and biota) quickly, and a portion remaining within the biotic compartment for several decades. Recent modeled PCB concentrations for male resident killer whales predict 40

% higher past history dietary exposure to PCBs than a simulation run to steady state, 122 reflecting the long lasting contribution of contaminated prey to high trophic level, long lifespan marine mammals (Hickie et al. 2007).

5.7 How do polybrominated diphenyl ethers (PBDEs) compare with PCBs in the Strait of Georgia? †

Although PBDEs are similar in chemical structure to PCBs, they have distinct physical and chemical properties, therefore different toxicological and environmental properties (de Wit 2002) . Disruption of neurobehavioural development, alterations in thyroid hormone and vitamin A levels, and reproductive impairment has been linked with

PBDE exposure in mammal laboratory studies (Hallgren et al. 2001; Kuriyama et al.

2005; Darnerud 2003), as well bioaccumulative behaviour in marine food webs (Burreau et al. 2006; Kelly et al. 2008; Stapleton and Baker 2003). The three PBDE commercial formulations available (penta, octa, and deca) are currently being considered under the

Stockholm Convention with their screening criteria (persistent, bioaccumulative, toxic, and potential for longrange transport) (World Wildlife Fund [WWF] 2005).

Since many of our samples were analyzed for both PCBs and PBDEs, we were able to provide a comparison of concentrations, bioaccumulation and contaminant load for Strait of Georgia marine biota ( Table 5.1; Figure 5.2; Figure 5.3 ). Ratios of

ΣPCB: Σ10 PBDE concentrations indicate that PBDEs are nearing PCB levels in some species, however the range of ratios among species may be reflecting speciesspecific differences in uptake, biotransformation, and excretion, and/or continued loading of

† Adapted from Christensen JR, Cullon DL, Shaw, P, Ross PS. 2010. Polybrominated diphenyl ether rivals polychlorinated biphenyls in a harbour seal ( Phoca vitulina ) food web, British Columbia Canada, manuscript in prep .

123

PBDEs into the system (Table 5.1). The high ΣPCB: Σ10 PBDE concentration ratio in male harbour seals (Table 5.1) and higher bioaccumulation slope observed for PCB153 in the harbour seal food web (slope = 0.19) than for PBDE47 (slope = 0.14) (Figure 5.2) both suggest that PBDEs have not yet reached peak concentrations in high trophic level marine mammals. Interestingly, is the projection of our food basket sample, very near to the regression line, further validating our food basket as representative of the harbour seal diet (Figure 5.2).

Table 5.1 Ratio of total polychlorinated biphenyl (ΣΣΣPCB) to total polybrominated diphenyl ether ( ΣΣΣPBDE) concentrations for 8 of the 15 species analyzed for PCBs in the Strait of Georgia (BC, Canada). ΣΣΣ10 PBDE: 10 congeners accounting for > 90% of ΣΣΣPBDEs in food web species, with the exception of ghost shrimp (58 %).

Species ΣΣΣPCB: ΣΣΣ10 PBDE

Male harbour seals 13.8 Female harbour seals 4.3 Chinook salmon 2.9 Pacific hake 2.6 Pacific herring 4.4 Food basket 3.4 Pacific sandlance 0.2 Ghost shrimp 34.7 Spot prawns 13.1 Zooplankton 3.9

Biomagnification was clearly observed in the harbour seal food web, evidenced

2 by TL vs Σ10 PBDEs ( r = 0.39, y = 0.41x + 3.95, p = 0.0004), FWMF = 2.6, and TL vs

PBDE153 ( r2 = 0.69, y = 0.63x + 1.38, p < 0.0001), FWMF = 4.3. Biomagnification factors for seals and their prey (food basket) were 13.4 for ΣPCBs and 3.2 for Σ10 PBDEs likely reflecting differences in physicochemical properties between the two compounds. 124

Small number of species precludes an analysis of the benthic food web; however, these species may have higher ratios due to life history differences such as life span and close association with the benthic environment and sediments.

5 HSM 4 SHP

SP 3 Pelagic spp. 2 2 r = 0.63 SL p = 0.03 m = 1.33 HE 1 FB CH HK PCB153 : : PBDE47 PCB153 ZO 0

1 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Trophic level

We conducted a similar exercise to estimate PBDE loadings in Strait of Georgia biota. We estimate there to be 9 kg of Σ10 PBDEs within the Strait of Georgia biota

(Figure 5.3), approximately 12 % of ΣPCBs (77 kg), consistent with ranges of

ΣPCBs/ ΣPBDEs reported in previous studies on Strait of Georgia fish and invertebrates

(Ikonomou et al. 2002; Ikonomou et al. 2006). The higher trophic level biota carried the 125 most PBDEs, similar to observations with ΣPCBs (Figure 5.3). Global production of

PBDEs has been estimated, as of the late 1990s, to be 67,400 metric tons/year with approximately 50 % of use occurring in North and South America (Birnbaum and Staskal

2004). Given the current limited global regulation, these compounds likely have not reached their maximum levels in high trophic organisms. In addition to contributing to the Strait of Georgia PBDE database for biota (Ikonomou et al. 2002; Ikonomou et al.

2006; Rayne et al. 2004; Cullon et al. 2005), air (Noël et al. 2009), water (Dangerfield et al. 2007), and sediment (Johannessen et al. 2008), these results highlight the accumulation similarities between PBDEs and PCBs and need for further food web research.

100

ΣPCB 80 Σ10 PBDE

20 Mass of contaminant (kg) of contaminant Mass

0 12 23 34 45 All Biota

Trophic level

Figure 5.3 Mass of ΣΣΣPCBs and ΣΣΣPBDEs (kg) detected in the Strait of Georgia biota among assigned trophic levels. The majority of both ΣΣΣPCBs and ΣΣΣPBDEs reside in 126

higher trophic level biota, although an almost equal portion of ΣΣΣPCBs reside in lower trophic level invertebrates.

5.8 Can we put our estimate of PCBs in the Strait of Georgia into a worldwide context?

If the global production of PCBs is estimated to be 1.3 million tonnes (Breivik et al. 2002), our biomass estimate for the Strait of Georgia biota would account for 3.1 × 10

6 %. The Marine Research Institute in Svalbard has estimated PCB biomass to be 6 kg in all zooplankton and 14 kg in all cod for the Barents and Norwegian Sea (Governor of

Svalbard 2008). By comparison, our estimated PCB burden was 2.6 kg for zooplankton and 12.5 kg for all fish species (Table 3.4). An estimated 8.07 kg of PCBs resides in

Zebra mussels in Saginaw Bay, Lake Huron (USEPA 2010), however the overabundance of this species would distort any regional comparison.

A study of freshwater lakes in Sweden calculated a mean ΣPCB biomass of 91 ng/m 2 in fish and 22 ng/m 2 in zooplankton in oligotrophic lakes (n = 6) and 440 ng/m 2 in fish and 48 ng/m 2 in zooplankton in eutrophic lakes (n = 5) (Berglund et al. 2001). Our estimates for the Strait of Georgia zooplankton and fish species are thousandfold higher than the eutrophic lakes and ten thousandfold higher than the oligotrophic lake zooplankton and fish values. Along with food web ecology differences between freshwater and marine ecosystems, sources (global and/or local) and transport processes

(atmospheric and/or oceanic) differ between lakes and coastal basins. These values offer a gross comparison between regions, reflecting larger PCB biomasses observed in the northern marine seas and less biomass in northern freshwater lakes. 127

Another means for regional comparisons is using food web magnification factors with top predator contaminant concentrations ( Table 5.2 ). Comparable FWMFs to results in this study were in the Southern BeaufortChukchi Seas, Iceland, South China, and North Baffin Bay (with avian data excluded) (Table 5.2). The calculated FWMFs reflect concentrations detected in high trophic level top predators as well as food chain length, homeotherm and poikilotherm differences, longrange transport, and inter and intraspecies differences. Our food web study provides additional data for regional comparisons of FWMFs, and further evidence of POP biomagnification in aquatic food webs.

128

Table 5.2 Food web magnification factors (FWMFs) and top predator total polychlorinated biphenyl ( ΣΣΣPCB) and PCB congener 153 concentrations differ between freshwater and marine food webs and regions in the world.

Region Marine/ FWMFs Top predator Reference Freshwater (Mean ±±± SE) PCB concentrations (Mean ±±± SE) Strait of Georgia, Marine ΣPCB: 3.6 Harbour seals This study Pacific Ocean PCB153: 6.5 Males (8652 ± 1888 ng/g lw) Females (3248 ± 978 ng/g lw) Southern Beaufort Marine ΣPCB: 3.26 ± 0.12 Ringed seals (Hoekstra et al. 2003) Chukchi Seas PCB153: 6.69 ± 0.14 (700 ± 118 ng/g lw) Barents Sea Marine PCB153: 18.8 ± 1.44 Glaucous gulls (Borgå et al. 2001; Hoekstra et (130442 ± 33936 ng/g lw) al. 2003) White Sea, Russia Marine ΣPCB: 1.52 ± 0.23 Harp seals (Muir et al. 2003) (Pelagic) PCB 153: 2.93 ± 0.59 (1070 ± 504 ng/g lw) Ringed seals Males (995 ± 385 ng/g lw) Females (999 ± 304 ng/g lw) Northwater Polynya Marine ΣPCB: 4.6 Ringed seals and seabirds b (Fisk et al. 2001) (North Baffin Bay) PCB 153: 9.7

ΣPCB: 2.90 ± 0.09 a (Fisk et al. 2001; Hoekstra et PCB 153: 5.87 ± 0.11 a al. 2003) Norway coastal region Marine ΣPCB: 6.2 Harbour seals (Ruus et al. 2002) (North Sea) (21,348 ng/g lw) Polar front to the Marine PCB 153: 4.1(poikilotherms) Harp seals (Hop et al. 2002) Marginal Ice Zone in PCB 153: 26.3(homeotherms) (3394 ± 891 ng/g lw), the Barents Sea Ringed seals a Excludes avian data in Fisk et al., 2001. b Concentration data not provided. 129

(523 ± 134 ng/g lw) Glaucous gulls (35927 ± 10789 ng/g lw) Baltic Sea Marine PCB 153: 1.45 (pelagic) SculpinsBenthic (Nfon et al. 2008) PCB 153: 0.88 (benthic) (0.5301 ng/g lw) Herring Pelagic (12220 ng/g lw) Hudson Bay region Marine PCB 153: 11 Ringed seals (Kelly et al. 2008) (Northeastern Canada) (602 ng/g lw) Beluga whales Males (3690 ng/g lw Females (661 ng/g lw) Iroise Sea and Seine Marine PCB 153: 15.3 and 4.1 (Iroise Spider crab (Bodin et al. 2008) Bay (Eastern English Sea) Seine Bay Channel) PCB 153: 9.0 (Seine Bay) (655.5 ± 342.4 ng/g lw) Iroise Sea (127.8 ± 49.6 ng/g lw) Iceland (subArctic) Marine ΣPCB: 3.11 Black guillemot (Skarphedinsdottir et al. 2009) PCB 153: 5.04 (2352 ± 1077 ng/g lw) Lake Ontario Freshwater ΣPCB: 5.7 c Lake trout (Tomy et al. 2004) (4.177 ± 0.297 ug/g ww) c (Kiriluk et al. 1995) Lake Superior Freshwater ΣPCB: 1.6 Lake trout (Wong et al. 2004) (5440 ± 3870 ng/g lw) Freshwater lake (South Freshwater ΣPCB: 3.68 Watersnake and northern (Wu et al. 2009) China) PCB 153: 4.73 snakehead b South Carolina stream Freshwater ΣPCB: 1.6 Fishes (Walters et al. 2008) (~ 40 ng/g ww; n=3 > 2000 ng/g ww )

c FWMF calculated from data in Kiriluk et al., 1995

130

5.9 How will climate change affect the Strait of Georgia marine mammal food webs?

Understanding the influence and impacts of climate change on coastal ecosystems such as the Strait of Georgia is equally important as understanding other stressors such as urbanization, loss of natural habitat, biodiversity changes, commercial and sport fishing, increased vessel traffic, and toxic contaminants. Recorded temperatures indicate an ocean warming trend of 0.5 to 1.0 °C along the southern British Columbia coast has occurred during the last 50 years (Environment Canada 2006). Some of the projections for the Strait of Georgia such as alterations in geochemical cycling (inorganic/organic carbon, nutrient), decreased pH and dissolved oxygen concentrations have been described elsewhere (Johannessen and Macdonald 2009).

For the Strait of Georgia marine mammals, specific climate change effects could result in loss of physical habitat for harbour seals with sea level rises and changes in prey availability with water temperature changes. Decreases and earlier peaks for zooplankton species (Johannessen and Macdonald 2009), the primary consumer link to both benthic and pelagic food webs, could alter food web structures, with changes to species abundance, richness, and biodiversity. As a major food source for chinook salmon,

Pacific hake, and Pacific herring, this inevitably affects primary prey availability for resident killer whales and Strait of Georgia harbour seals. Research has also shown that inshore, estuarine, and riverine cetaceans may be at elevated risk for infectious disease because these ecosystems are often heavily impacted by human activities, biological and chemical contamination, and effects of climate change (Van Bressem et al. 2009).

A recent application of the BerkeleyTrent global multimedia mass balance model predicts increased volatilization emissions and mobility of POPs (with properties similar 131 to PCBs) with climate change scenarios (Lamon et al. 2009). Changes in POP levels observed in antarctic biota suggest global redistribution and increased transfer of POPs to antarctic waters due to southern hemisphere sources/usage and climate changes (Weber and Goerke 2003). Climaterelated stresses on salmon abundance and increased consumption of a lower lipid, more POPcontaminated salmon (Chapter 4), suggest both decreased prey availability and increased contaminant exposure for resident killer whales.

This raises particular concern for species already designated as threatened or endangered, such as resident killer whales.

5.10 How does this research contribute to existing Strait of Georgia bioaccumulation models?

Food web models provide a means to assess and/or evaluate bioaccumulation of chemicals and are useful for risk assessment and regulatory purposes (Gobas and

Morrison 2000; Thomann 1989). Our high resolution contaminant analyses include not only total concentrations, but also individual congeners and isomers increasing the database for the Strait of Georgia. The data and interpretation of processrelated variables (halflife, log K ow , metabolism, age, condition) are an inherent aspect of understanding distribution of chemicals in food webs. Providing empirical evidence and observations will enable a more thorough analysis and development of a Strait of Georgia

POP food web bioaccumulation model.

132

5.11 Do the Strait of Georgia marine mammals fall within the current approaches to risk assessment?

This research has provided an excellent opportunity to assess risk to marine mammals within Canadian and US risk assessment protocols. We have taken several approaches to assessing POP risk to harbour seals and killer whales (1) dietary threshold concentrations for adverse effects in marine mammals, (2) tissue residue guidelines for the protection of mammalian wildlife consumers of aquatic biota (diet), and (3) estimated dietary intakes with consideration of nutritional (lipid) content. Our results have shown that taking several approaches to evaluating health risks provides a more complete picture of exposure and risk. Nutritional content and subsequent increases in dietary intake can greatly affect bioaccumulation of compounds and adverse health risks, and thus should be taken into account when establishing tissue residue guidelines.

Undoubtedly, the Canadian Tissue Residue Guidelines for the Protection of

Wildlife Consumers of Aquatic Biota, published by the Canadian Council of Ministers of the Environment (CCME), are more conservative than the US National Academy of

Sciences/National Academy of Engineering and New York State Department of

Environmental Conservation guidelines for PCBs since they consider chronic toxicity, carcinogenicity, and reproductive effects (Canadian Council of Ministers of the

Environment 1999). Our Puget Sound harbour seal food basket exceeded the Canadian tissue residue guidelines for the protection of mammalian wildlife consumers of aquatic biota. One of the southern (Puget Sound) chinook salmon stocks exceeded and one of the

Strait of Georgia stocks was approaching the CCME PCB guidelines, and one southern stock exceeded DDT tissue residue guidelines. Further, all our chinook stocks sampled 133 exceeded the dietary threshold estimated to protect 95 % of a killer whale population based on a 17 mg/kg PCB adverse effects threshold for marine mammals (Hickie et al.

2007).

These results highlight the need for ongoing monitoring of chemicals in these two basins as well as the need for updated US guidelines. This study has also shown that the current risk assessment approaches for PBDEs is in need of increased regulation. The levels of PBDEs detected in our seal food baskets (Chapter 2) and adult chinook salmon

(Chapter 4) were approaching PCB levels. As well, our comparison of biomagnification and biomass estimates for PCBs and PBDEs suggest PBDEs have not yet reached environmental equilibrium. Currently, Penta, octa and decaPBDEs are among the POP candidates for consideration under the Stockholm Convention (World Wildlife Fund

[WWF] 2005).

Chemicals do not recognize nor behave with consideration to species, borders, politics, or economics. A universal or global set of guidelines for chemicals of concern would be very beneficial to scientists, regional authorities, international agencies, and policymakers alike. The time frame from release of compounds to the environment to restrictions and regulations has shown to be too long of a time span to eliminate these environmental threats, 40+ years for PCBs, 30+ years for DDT, and 30+ years for

PBDEs. With their physicochemical properties favouring bioaccumulation, toxicity, longrange transport, and persistence, reducing this time period should be a global priority.

134

5.12 What questions were answered in this research?

This project represents the most thorough analysis of marine mammal food webs in the Strait of Georgia, BC (Canada).

This research has provided

• Increased understanding of the sources, pathways, and fate of POPs within

marine mammal food webs (PCBs remain persistent in harbour seals and their

prey in the Strait of Georgia, continuing to cycle throughout this marine

ecosystem.

• Further understanding of dietary exposure by wildlife to real world complex

mixtures of POPs.

• Evidence that factors such as age, sex, lipid content, diet, feeding strategies,

migrationrelated metabolism, as well as physicochemical properties such as

degree of chlorination, log Kow, and chemical structure all influence POP

bioaccumulation (concentrations and patterns) in organisms.

• Further validation of the utility of harbour seals as sentinels of marine ecosystem

contamination.

This research could not address in detail the many influencing factors for POP accumulations within each species of the harbour seal food web. It further does not address the abiotic components (air, water, and sediment) in the system which undoubtedly should be incorporated in order to construct an “ecosystem picture” of the

Strait of Georgia marine food web.

135

5.13 Recommendations for future research

Food web based studies allow the exploration of “real world” exposures to complex mixtures of POPs. A good understanding of the biology (feeding ecology, diet, lifecycle, reproduction, and physiological processes) of organisms is crucial to interpreting contaminant data within and between different species. Study design and selection of organisms is another very important aspect to this research. In evaluating putative contaminantrelated effects in biota, it is important to consider the many additional stressors that can influence health and productivity such as species diversity and trophic imbalances (Tabor and Aguirre 2004).

Tools such as stable isotope ratios, congenerspecific contaminant analyses, principal components analysis have wellillustrated their utility in this study. The interpretation of contaminant data is strengthened when these methods and tools are used in conjunction with each other rather than individually. Finally, as important as indepth analyses are in this area of research, “big picture” analyses and interpretation

(atmospheric and oceanic transport) must be included in hopes of furthering our understanding of POP accumulations in aquatic food webs in the future.

Rachel Carson brought about a public awareness to environmental contamination that instigated change from the bottom up (Carson 1962). Achieving balance between the production of synthetic compounds and their eventual fate and effects to address the needs of a growing population remains a great challenge to scientists, economists, and policymakers. Climate change may alter food web structure leading to alterations in the way in which contaminants move, accumulate and magnify in food webs, and elicit effects in aquatic biota (Macdonald 2005). This may diminish fitness in marine 136 mammals and/or increase the risk of contaminantrelated health effects (Ross 2002;

Daszak et al. 2001).

137

Literature Cited

Addison,R.F. and Brodie,P.F. 1987. Transfer of organochlorine residues from blubber through the circulatory system to milk in the lactating gray seal Halichoerus grypus , Canadian Journal of Fisheries and Aquatic Sciences 44: 782786.

Addison,R.F., Ikonomou,M.G., and Smith,T.G. 2005. PCDD/F and PCB in harbour seals (Phoca vitulina ) from British Columbia: response to exposure from pulp mill effluents, Marine Environmental Research 59: 165176.

Baird, R. W. Update COSEWIC status report on the harbour porpoise Phocoena phocoena (Pacific Ocean population) in Canada, in COSEWIC assessment and update status report on the harbour porpoise Phocoena phocoena Pacific Ocean population) in Canada. Committee on the Status of Endangered Wildlife in Canada. 122. 2003. Ottawa, ON, Environment Canada.

Beacham,T.D., Candy,J.R., Jonsen,K.L., Supernault,J., Wetklo,M., Deng,L., Miller,K.M., and Withler,R.E. 2006. Estimation of stock composition and individual identification of Chinook salmon across the Pacific Rim using microsatellite variation, Transactions of the American Fisheries Society 135: 861888.

Berglund,O., Larsson,P., Ewald,G., and Okla,L. 2000. Bioaccumulation and differential partioning of polychlorinated biphenyls in freshwater, planktonic food webs, Canadian Journal of Fisheries and Aquatic Sciences 57: 11601168.

Berglund,O., Larsson,P., Ewald,G., and Okla,L. 2001. Influence of trophic status on PCB distribution in lake sediments and biota, Environmental Pollution 113: 199210.

Bigg,M.A. 1981. Harbour seal Phoca vitulina linnaeus , 1758 and Phoca largha pallas , 1811, pp. 127. In S.A.Ridgway and R.J.Harrison, editors, Handbook of marine mammals Volume 2: Seals. Academic Press, London.

Bigg,M.A., Olesiuk,P.F., Ellis,G.M., Ford,J.K.B., and Balcomb,K.C. 1990. Social organization and geneology of resident killer whales ( Orcinus orca ) in the coastal waters of British Columbia and Washington State, Report of the International Whaling Commission 12: 383405.

Birnbaum,L.S. and Staskal,D.F. 2004. Brominated flame retardants: cause for concern?, Environmental Health Perspectives 112: 1017.

Bodin,N., Le Loc'h,F., Caisey,X., Guellec,A.M., Abarnou,A., Loizeau,V., and Latrouite,D. 2008. Congenerspecific accumulation and trophic transfer of polychlorinated biphenyls in spider crab food webs revealed by stable isotope analysis, Environmental Pollution 151: 252261. 138

Bolles,H.G., DixonWhite,H.E., Peterson,R.K., Tomerlin,J.R., Day,E.W., Jr., and Oliver,G.R. 1999. U.S. market basket study to determine residues of the insecticide chlorpyrifos., Journal of Agricultural and Food Chemistry 47: 18171822.

Bonn,B.A. 1998. Polychlorinated dibenzopdioxin and dibenzofuran concentration profiles in sediment and fish tissue of the Willamette basin, Oregon, Environmental Science and Technology 32: 729735.

Boon,J.P., Reijnders,P.J.H., Dols,J., Wensvoort,P., and Hillebrand,M.T.J. 1987. The kinetics of individual polychlorinated biphenyl congeners in female harbour seals ( Phoca vitulina ), with evidence for structurerelated metabolism, Aquatic Toxicology 10: 307 324.

Boon,J.P., Van der Meer,J., Allchin,C.R., Law,R.J., Klungsoyr,J., Leonards,P.E.G., Spliid,H., StorrHansen,E., Mckenzie,C., and Wells,D.E. 1997. Concentrationdependent changes of PCB patterns in fisheating mammals: Structural evidence for induction of cytochrome P450, Archives of Environmental Contamination and Toxicology 33: 298 311.

Borgå,K., Fisk,A.T., Hoekstra,P.F., and Muir,D.C.G. 2004. Biological and chemical factors of importance in the bioaccumulation and trophic transfer of persistent organochlorine contaminants in arctic marine food webs, Environmental Toxicology and Chemistry 23(10): 23672385.

Borgå,K., Gabrielsen,G.W., and Skaare,J.U. 2001. Biomagnification of organochlorines along a Barents Sea food chain, Environmental Pollution 113: 187198.

Borlakoglu,J. and Dils,R. 1990. Polychlorinated biphenyls (PCBs) and marine food chains, Biologist 37(5): 145147.

Bowen,W.D. 2000. Reconstruction of pinniped diets: accounting for complete digestion of otoliths and cephalopod beaks, Canadian Journal of Fisheries and Aquatic Sciences 57: 898905.

Breivik,K., Sweetman,A., Pacyna,J.M., and Jones,K.C. 2002. Towards a global historical emission inventory for selected PCB congeners a mass balance approach. 1. Global production and consumption, Science of the Total Environment 290: 181198.

Brett,J.R. 1995. Energetics, pp. 368. In C.Groot, L.Margolis, and W.C.Clarke, editors, Physiological ecology of Pacific salmon. UBC Press, Vancouver.

Brouwer,A., Reijnders,P.J.H., and Koeman,J.H. 1989. Polychlorinated biphenyl (PCB) contaminated fish induces vitamin A and thyroid hormone deficiency in the common seal (Phoca vitulina ), Aquatic Toxicology 15: 99106.

Burgner,R.L. 1991. Life histories of sockeye salmon, Oncorhynchus nerka , pp. 1118. In C.Groot and L.Margolis, editors, Pacific salmon life histories. UBC Press, Vancouver, BC. 139

Burreau,S., Zebuhr,Y., Broman,D., and Ishaq,R. 2006. Biomagnification of PBDEs and PCBs in food webs from the Baltic Sea and the northern Atlantic Ocean, Science of the Total Environment 366: 659672.

Burton,R.K. and Koch,P.L. 1999. Isotopic tracking of foraging and longdistance migration in northeastern Pacific pinnipeds, Oecologia 119: 578585.

Cabana,G. and Rasmussen,J.B. 1996. Comparison of aquatic food chains using nitrogen isotopes, Proceedings of the ational Academy of Sciences of the United States of America 93: 1084410847.

CACAR 1997. Canadian Arctic Contaminants Assessment Report , Indian and Northern Affairs Canada, Ottawa.

Canadian Council of Ministers of the Environment 1999. Canadian tissue residue guidelines for the protection of wildlife consumers of aquatic biota: polychlorinated biphenyls (PCBs), pp. 19. Canadian environmental quality guidelines. Canadian Council of Ministers of the Environment, Winnipeg.

Canadian Council of Ministers of the Environment. Canadian Environmental Quality Guidelines. 2001.

Canadian Council of Ministers of the Environment. Canadian Environmental Quality Guidelines Update 2002. 112. 2002.

Carlson D.L. and Hites,R.A. 2005. Polychlorinated biphenyls in salmon and salmon feed: Global differences and bioaccumulation, Environmental Science and Technology 39: 73897395.

Carson,R. 1962. Silent Spring , HoughtonMifflin Company, New York, NY.

Chilton,D.E. and Beamish,R.J. 1982. Age determination methods for fishes studied by the Groundfish Program at the Pacific Biological Station, Canadian Special Publication of Fisheries and Aquatic Sciences 60: 1102.

Christensen,J.R., Letcher,R.J., and Ross,P.S. 2009. Persistent or not persistent? Polychlorinated biphenyls are readily depurated by grizzly bears ( Ursus arctos horribilis ), Environmental Toxicology and Chemistry 28: 22062215.

Christensen,J.R., MacDuffee,M., Macdonald,R.W., Whiticar,M., and Ross,P.S. 2005. Persistent organic pollutants in British Columbia grizzly bears: Consequence of divergent diets, Environmental Science and Technology 39: 69526960.

Cottrell,P.E. 1996. Diet, activity budgets, and movement patterns of harbour seals (Phoca vitulina) in Cowichan Bay and adjacent areas, University of Victoria, Victoria. 140

Cottrell,P.E., Jeffries,S.J., Beck,B., and Ross,P.S. 2002. Growth and development in free ranging harbour seal ( Phoca vitulina ) pups from southern British Columbia, Marine Mammal Science 18(3): 721733.

Cottrell,P.E., Trites,A.W., and Miller,E.H. 1995. Assessing the use of hard parts in faeces to identify harbour seal prey: results of captivefeeding trials, Canadian Journal of Zoology 74: 875880.

Cretney,W.J. and Yunker,M.B. 2000. Concentration dependency of biotasediment accumulation factors for chlorinated dibenzopdioxins and dibenzofurans in dungeness crab ( Cancer magister ) at marine pulp mill sites in British Columbia, Canada, Environmental Toxicology and Chemistry 19(12): 30123023.

Cullon,D., Jeffries,S., and Ross,P.S. 2005. Persistent organic pollutants in the diet of harbor seals ( Phoca vitulina ) inhabiting Puget Sound, Washington (USA), and the Strait of Georgia, British Columbia (Canada): A food basket approach, Environmental Toxicology and Chemistry 24: 25622572.

Cullon,D.L., Yunker,M.B., Alleyne,C., Dangerfield,N.J., O'Neil,S., Whiticar,M.J., and Ross,P.S. 2009. Persistent organic pollutants (POPs) in chinook salmon ( Oncorhyncus tshawytscha ): Implications for resident killer whales of British Columbia and adjacent waters, Environmental Toxicology and Chemistry 28: 148161.

Dalsgaard, J., Wallace, S., Salas, S, and Preikshot, D. Massbalance model reconstruction of the Strait of Georgia: the present, one hundred, and five hundred years ago. Pauly, D., Pitcher, T. J., and Preikshot, D. Volume 6 Number 5, 7290. 1998. Vancouver, BC, Canada, University of British Columbia. Back to the Future : Reconstructing the Strait of Georgia Ecosystem.

Dangerfield, N., Macdonald, R., Crewe, N., Shaw, P., and Ross, P. S. PCBs and PBDEs in the Georgia Basin water column. Georgia Basin Puget Sound Research Conference March 2629, 2007; Vancouver, BC. 2007.

Darnerud,P.O. 2003. Toxic effects of brominated flame retardants in man and in wildlife, Environment International 29: 841853.

Daszak,P., Cunningham,A.A., and Hyatt,A. 2001. Anthropogenic environmental change and the emergence of infectious diseases in wildlife, Acta Tropica 78(2): 103116.

De Boer,J., De Boer,K., and Boon,J.P. 2000. Polybrominated biphenyls and diphenylethers, pp. 6395. In J.Paasivirta, editor, Handbook of environmental chemistry Vol. 3 Part K. ew types of persistent halogenated compounds. SpringerVerlag, Heidelberg.

De Swart,R.L., Ross,P.S., Timmerman,H.H., Hijman,W.C., De Ruiter,E., Liem,A.K.D., Brouwer,A., Van Loveren,H., Reijnders,P.J.H., Vos,J.G., and Osterhaus,A.D.M.E. 1995. Shortterm fasting does not aggravate immunosuppression in harbour seals ( Phoca vitulina ) with high body burdens of organochlorines, Chemosphere 31: 42894306. 141

De Swart,R.L., Ross,P.S., Vedder,L.J., Timmerman,H.H., Heisterkamp,S.H., Van Loveren,H., Vos,J.G., Reijnders,P.J.H., and Osterhaus,A.D.M.E. 1994. Impairment of immune function in harbor seals ( Phoca vitulina ) feeding on fish from polluted waters, Ambio 23: 155159.

De Swart,R.L., Ross,P.S., Vos,J.G., and Osterhaus,A.D.M.E. 1996. Impaired immunity in harbour seals ( Phoca vitulina ) exposed to bioaccumulated environmental contaminants: review of a longterm study, Environmental Health Perspectives 104 (suppl. 4): 823828. de Wit,C.A. 2002. An overview of brominated flame retardants in the environment, Chemosphere 46: 583623.

DeBruyn,A.M.H., Ikonomou,M.G., and Gobas,F.A.P.C. 2004. Magnification and toxicity of PCBs, PCDDs, and PCDFs in uprivermigrating Pacific salmon, Environmental Science and Technology 38: 62176224.

Deniro,M.J. and Epstein,S. 1978. Influence of diet on the distribution of carbon isotopes in animals, Geochimica et Cosmochimica Acta 42: 495506.

Deniro,M.J. and Epstein,S. 1981. Influence of diet on the distribution of nitrogen isotopes in animals, Geochimica et Cosmochimica Acta 45: 341351.

Doucett,R.R., Booth,R.K., Power,G., and McKinley,R.S. 1999. Effects of the spawning migration on the nutritional status of anadromous Atlantic salmon ( Salmo salar ): insights from stableisotope analysis, Canadian Journal of Fisheries and Aquatic Sciences 56: 21722180.

Elliot,J.E., Butler,R.W., Norstrom,R.J., and Whitehead,P.E. 1989. Enviornmental eontaminants and reproductive success of great blue herons Ardea herodias in British Columbia, 198687, Environmental Pollution 59: 91114.

Elliott,J.E., Machmer,M.M., Wilson,L.K., and Henny,C.J. 2000. Contaminants in ospreys from the Pacific Northwest: II. Organochlorine pesticides, polychlorinated biphenyls, and mercury, 19911997, Archives of Environmental Contamination and Toxicology 38: 93 106.

Environment Canada. British Columbia Coast and Marine Environment Project 2006: Climate Change. 2006. Environment Canada.

Ewald,G., Larsson,P., Linge,H., Okla,L., and Szarzi,N. 1998. Biotransport of organic pollutants to an inland Alaska lake by migrating sockeye salmon ( Oncorhyncus nerka ), Arctic 51: 4047.

Fisheries and Oceans Canada. Pacific hake Strait of Georgia stock assessment for 1999 and recommended yield options for 2000. Saunders, M. W. and McFarlane, G. A. Canadian Stock Assessment Secretariat Research Document 99/200. 1999. 142

Fisheries and Oceans Canada. Management framework for Strait of Georgia lingcod. Logan, G., de la Mare, W., King, J., and Haggarty, D. Canadian Science Advisory Secretariat Research Document 2005/048. 2005.

Fisheries and Oceans Canada. Recovery Strategy for the Northern and Southern Resident Killer Whales ( Orcinus orca ) in Canada. 181. 2008. Ottawa, Fisheries and Oceans Canada. Species at Risk Act Recovery Strategy Series.

Fisheries and Oceans Canada. Stock assessment report on Pacific herring in British Columbia. DFO Canadian Science Advisory Secretariat. Science Advisory Report 2009/059. 2009.

Fisheries and Oceans Canada. Population Assessment Pacific harbour seal ( Phoca vitulina richardsi ). 2009/011, 110. 2010. Fisheries and Oceans Canada. Canadian Science Advisory Secretariat.

Fisk,A.T., Hobson,K.A., and Norstrom,R.J. 2001. Influence of chemical and biological factors on trophic transfer of persistent organic pollutants in the Northwater Polynya marine food web, Environmental Science and Technology 35: 732738.

Ford,J.K.B. and Ellis,G.M. 2006. Selective foraging by fisheating killer whales ( Orcinus orca ) in British Columbia, Marine Ecology Progress Series 316 : 185199.

Ford,J.K.B., Ellis,G.M., and Balcomb,K.C. 2000. Killer whales , UBC Press, Vancouver.

Frame, G. M. 1997. A collaborative study of 209 PCB congeners and 6 Aroclors on 20 different HRGC columns. Fresenius Journal of Analytical Chemistry 357, 714722.

France,R.L. 1995. Carbon13 enrichment in benthic compared to planktonic algae: foodweb implications, Marine Ecology Progress Series 124: 307312.

France,R.L. and Peters,R.H. 1997. Ecosystem differences in the trophic enrichment of 13 C in aquatic food webs, Canadian Journal of Fisheries and Aquatic Sciences 54: 1255 1258.

Freedman,B. 1995. Environmental ecology: the ecological effects of pollution, disturbance, and other stresses , Academic Press, Inc., San Diego, CA.

Fuller,B.T., Fuller,J.L., Sage,N.E., Harris,D.A., O'Connell,T.C., and Hedges,R.E.M. 2005. Nitrogen balance and δ15 N: why you're not what you eat during nutritional stress, Rapid Communications in Mass Spectrometry 19: 24972506.

Gobas,F.A.P.C., de Wolf,W., Burkhard,L.P., Verbruggen,E., and Plotzke,K. 2009. Revisiting bioaccumulation criteria for POPs and PBT assessments, Integrated Environmental Assessment and Management 5: 624637. 143

Gobas,F.A.P.C. and Morrison,H.A. 2000. Bioconcentration and biomagnification in the aquatic environment, pp. 189231. In R.S.Boethling and D.Mackay, editors, Handbook of Property Estimation Methods for Chemicals. CRC Press LLC.

Gouin,T., Mackay,D., Jones,K.C., Harner,T., and Meijer,S.N. 2004. Evidence for the "grasshopper" effect and fractionation during longrange atmospheric transport of organic contaminants, Environmental Pollution 128: 139148.

Governor of Svalbard. PCBs in Svalbard Status of knowledge and management. 1/2008, 138. 2008.

GrahlNielsen,O. 1999. Comment: Fatty acid signatures and classification trees: new tools for investigating the foraging ecology of seals, Canadian Journal of Fisheries and Aquatic Sciences 56: 22192223.

Gunderson,E.L. 1995. Dietary intakes of pesticides, selected elements, and other chemicals: FDA total diet study, June 1984April 1986, Journal of AOAC International 78: 910921.

Hagen,M.E., Colodey,A.G., Knapp,W.D., and Samis,S.C. 1997. Environmental response to decreased dioxin and furan loadings from British Columbia coastal pulp mills, Chemosphere 34: 12211229.

Hall,A.J., Kalantzi,O.I., Thomas,G.O. 2003. Polybrominated diphenyl ethers (PBDEs) in gray seals during their first years of life are they thyroid hormone endocrine disrupters? Environmental Pollution 126:2937.

Hallgren,S., Sinjari,T., Hakansson,H., and Darnerud,P.O. 2001. Effects of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls (PCBs) on thyroid hormone and vitamin A levels in rats and mice, Archives of Toxicology 75: 200 208.

Harris,M.L., Wilson,L.K., and Elliot,J.E. 2005. An assessment of PCBs and OC pesticides in eggs of doublecrested ( Phalacrocorax auritus ) and pelagic ( P. pelagicus ) cormorants from the west coast of Canada, 1970 to 2002, Ecotoxicology 14: 607625.

Hart J.L. 1973. Pacific fishes of Canada , Fisheries Research Board of Canada, Ottawa, Canada.

Hawker,D.W. and Connell,D.W. 1988. Octanolwater partition coefficients of polychlorinated biphenyl congeners, Environmental Science and Technology 22: 382 387.

Healey,M.C. 1991. Life history of chinook salmon ( Oncorhynchus tshawytscha ), pp. 311 394. In C.Groot and L.Margolis, editors, Pacific salmon life histories. UBC Press, Vancouver. 144

Heaton,S.N., Bursian,S.J., Giesy,J.P., Tillitt,D.E., Render,J.A., Jones,P.D., Verbrugge,D.A., Kubiak,T.J., and Aulerich,R.J. 1995. Dietary exposure of mink to carp from Saginaw Bay, Michigan. 1. Effects on reproduction and survival, and the potential risks to wild mink populations, Archives of Environmental Contamination and Toxicology 28: 334343.

Hickie,B.E., Ross,P.S., Macdonald,R.W., and Ford,J.K.B. 2007. Killer whales ( Orcinus orca) face protracted health risks associated with lifetime exposure to PCBs, Environmental Science and Technology 41: 66136619.

Hirons,A.C., Schell,D.M., and Finney,B.P. 2001. Temporal records of δ13 C and δ15 N in North Pacific pinnipeds: Inferences regarding environmental change and diet, Oecologia 129: 591601.

Hites,R.A. 1990. Environmental behavior of chlorinated dioxins and furans, Accounts of Chemical Research 23: 194201.

Hobson,K.A., Alisauskas,R.T., and Clark,R.G. 1993. Stablenitrogen isotope enrichment in avian tissues due to fasting and nutritional stress: Implications for isotopic analyses of diet, The Condor 95: 388394.

Hobson,K.A., Fisk,A., Karnovsky,N., Holst,M., Gagnon,J.M., and Fortier,M. 2002. A stable isotope ( δ13 C, δ15 N) model for the North Water food web: implications for evaluating trophodynamics and the flow of energy and contaminants, DeepSea Research 49: 51315150.

Hobson,K.A., Schell,D.M., Renouf,D., and Noseworthy,E. 1996. Stable carbon and nitrogen isotopic fractionation between diet and tissues of captive seals: implications for dietary reconstructions involving marine mammals, Canadian Journal of Fisheries and Aquatic Sciences 53: 528533.

Hobson,K.A., Sease,J.L., Merrick,R.L., and Piatt,J.F. 1997. Investigating trophic relationships of pinnipeds in Alaska and Washington using stable isotope ratios of nitrogen and carbon, Marine Mammal Science 13(1): 114132.

Hobson,K.A. and Welch,H.E. 1992. Determination of trophic relationships within a high Arctic marine food web using δ13 C and δ15 N analysis, Marine Ecology Progress Series 84: 918.

Hoekstra,P.F., Dehn,L.A., George,J.C., Soloman,K.R., Muir,D.C.G., and O'Hara,T.M. 2002. Trophic ecology of bowhead whales (Balaena mysticetus) compared with that of other arctic marine biota as interpreted from carbon, nitrogen, and sulfurisotope signatures, Canadian Journal of Zoology 80: 223231.

Hoekstra,P.F., O'Hara,T.M., Fisk,A.T., Borgå,K., Soloman,K.R., and Muir,D.C.G. 2003. Trophic transfer of persistent organochlorine contaminants (OCs) within an Arctic marine food web from the southern BeaufortChukchi Seas, Environmental Pollution 124: 509 522. 145

Hooker,S.K., Iverson,S.J., Ostrom,P., and Smith,S.C. 2001. Diet of northern bottlenose whales inferred from fattyacid and stableisotope analyses of biopsy samples, Canadian Journal of Zoology 79: 14421454.

Hop,H., Borgå,K., Gabrielsen,G.W., Kleivane,L., and Skaare,J.U. 2002. Food web magnification of persistent organic pollutants in poikilotherms and homeotherms from the Barents Sea, Environmental Science and Technology 36: 25892597.

Ikonomou,M.G., Fernandez,M.P., and Hickman,Z.L. 2006. Spatiotemporal and species specific variation in PBDE levels/patterns in British Columbia's coastal waters, Environmental Pollution 140: 355363.

Ikonomou,M.G., Fernandez,M.P., Knapp,W., and Sather,P. 2002. PCBs in Dungeness crab reflect distinct source fingerprints among harbor/industrial sites in British Columbia, Environmental Science and Technology 36: 25452551.

Ikonomou,M.G., Fraser,T., Crewe,N., Fischer,M.B., Rogers,I.H., He,T., Sather,P., and Lamb,R. 2001. A comprehensive multiresidue ultratrace analytical method, based on HRGC/HRMS, for the determination of PCDDs, PCDFs, PCBs, PBDEs, PCDEs, and organochlorine pesticides in six different environmental matrices, Canadian Technical Report of Fisheries and Aquatic Sciences 2389: 195.

Ikonomou,M.G., Rayne,S., and Addison,R.F. 2002. Exponential increases of the brominated flame retardants, polybrominated diphenyl ethers, in the Canadian Arctic from 1981 to 2000, Environmental Science and Technology 36: 18861892.

Ikonomou,M.G., Rayne,S., Fischer,M., Fernandez,M.P., and Cretney,W. 2002. Occurrence and congener profiles of polybrominated diphenyl ethers (PBDEs) in environmental samples from coastal British Columbia, Canada, Chemosphere 46: 649 663.

Isosaari,P., Kiviranta,H., Lie,O., Lundebye,A., Ritchie,G., and Vartiainen,T. 2004. Accumulation and distribution of polychlorinated dibenzopdioxin, dibenzofuran, and polychlorinated biphenyl congeners in Atlantic salmon ( Salmo salar ), Environmental Toxicology and Chemistry 23: 16721679.

Jardine,T.D., Kidd,K.A., and Fisk,A.T. 2006. Applications, considerations, and sources of uncertainty when using stable isotope analysis in ecotoxicology, Environmental Science and Technology 40: 75017511.

Johannessen,S.C. and Macdonald,R.W. 2009. Effects of local and global change on an inland sea: the Strait of Georgia, British Columbia, Canada, Climate Research 40: 121.

Johannessen,S.C., Macdonald,R.W., Wright,C.A., Burd,B., Shaw,D.P., and van Roodselaar,A. 2008. Joined by geochemistry, divided by history: PCBs and PBDEs in Strait of Georgia sediments, Marine Environmental Research 66: S112S120. 146

Johnson,L.L., Ylitalo,G.M., Arkoosh,M.R., Kagley,A.N., Stafford,C.J., Bolton,J.L., Buzitis,J., Anulacion,B.F., and Collier,T.K. 2007. Contaminant exposure in outmigrating juvenile salmon from Pacific Northwest estuaries of the United States, Environmental Monitoring and Assessment 124: 167194.

Jones,K.C. and De Voogt,P. 1999. Persistent organic pollutants (POPs): State of science, Environmental Pollution 100: 209221.

Kannan,K., Blankenship,A.L., Jones,P.D., and Giesy,J.P. 2000. Toxicity reference values for the toxic effects of polychlorinated biphenyls to aquatic mammals, Human and Ecological Risk Assessment 6(1): 181201.

Kastelein,R.A. and Vaughan,N. 1989. Food consumption, body measurements and weight changes of a female killer whale ( Orcinus orca ), Aquatic Mammals 15: 1821.

Kelly,B.C., Gray S.L., Ikonomou,M.G., MacDonald,J.S., Bandiera,S.M., and Hrycay,E.G. 2007. Lipid reserve dynamics and magnification of persistent organic pollutants in spawning sockeye salmon ( Oncorhynchus nerka ) from the Fraser River, British Columbia, Environmental Science and Technology 41: 30833089.

Kelly,B.C., Ikonomou,M.G., Blair,J., and Gobas,F.A.P.C. 2008. Bioaccumulation behaviour of polybrominated diphenyl ethers (PBDEs) in a Canadian Arctic marine food web, Science of the Total Environment 401: 6072.

Kelly,B.C., Ikonomou,M.G., Blair,J.D., and Gobas,F.A.P.C. 2008. Hydroxylated and methoxylated polybrominated diphenyl ethers in a Canadian Arctic marine food web, Environmental Science and Technology 42: 70697077.

Kelly,B.C., Ikonomou,M.G., Blair,J.D., Morin,A.E., and Gobas,F.A.P.C. 2007. Food webspecific biomagnification of persistent organic pollutants, Science 317: 236239.

Kelly,J.F. 2000. Stable isotopes of carbon and nitrogen in the study of avian and mammalian trophic ecology, Canadian Journal of Zoology 78: 127.

Kidd,K.A., Hesslein,R.H., Ross,B.J., Koczanski,K., Stephens,G.R., and Muir,D.C.G. 1998. Bioaccumulation of organochlorines through a remote freshwater food web in the Canadian Arctic, Environmental Pollution 102: 91103.

Kiriluk,R.M., Servos,M.R., Whittle,D.M., Cabana,G., and Rasmussen,J.B. 1995. Using ratios of stable nitrogen and carbon isotopes to characterize the biomagnification of DDE, Mirex, and PCB in a Lake Ontario pelagic food web, Canadian Journal of Fisheries and Aquatic Sciences 52: 26602674.

Kriete,B. 1995. Bioenergetics in the killer whale, Orcinus orca , University of British Columbia, Vancouver, B.C. 147

Krümmel,E.M., Macdonald,R.W., Kimpe,L.E., GregoryEaves,I., Demers,M.J., Smol,J.P., and Blais,J.M. 2003. Delivery of pollutants by spawning salmon, ature 425: 255256.

Kucklick,J.R., Krahn,M.M., Becker,P.R., Porter,B.J., Schantz,M.M., York,G.S., O'Hara,T.M., and Wise,S.A. 2006. Persistent organic pollutants in Alaskan ringed seal (Phoca hispida ) and walrus ( Odobenus rosmarus ) blubber, Journal of Environmental Monitoring 8: 848854.

Kuhnlein,H.V. and Chan,H.M. 2000. Environment and contaminants in traditional food systems of northern indigenous peoples, Annual Review of utrition 20: 595626.

Kujawinski,E.B., Farrington,J.W., and Moffett,J.W. 2000. Importance of passive diffusion in the uptake of polychlorinated biphenyls by phagotrophic protozoa, Applied and Environmental Microbiology 66: 19871993.

Kuriyama,S.N., Talsness,C.E., Grote,K., and Chahoud,I. 2005. Developmental exposure of LowDose PBDE99: Effects on male fertility and neurobehavior in rat offspring, Environmental Health Perspectives 113: 149154.

Kurle,C.M. and Worthy,G.A.J. 2001. Stable isotope assessment of temporal and geographic differences in feeding ecology of northern fur seals ( Callorhinus ursinus ) and their prey, Oecologia 126(2): 254265.

Lamon,L., von Waldow,H., Macleod,M., Scheringer,M., Marcomini,A., and Hungerbühler,K. 2009. Modeling the global levels and distribution of polychlorinated biphenyls in air under a climate change scenario, Environmental Science and Technology 43: 58185824.

Law,K., Halldorson,T., Danell,R., Stern,G., Gewurtz,S., Alaee,M., Marvin,C., Whittle,M., and Tomy,G. 2006. Bioaccumulation and trophic transfer of some brominated flame retardants in a Lake Winnipeg (Canada) food web, Environmental Toxicology and Chemistry 25: 21772186.

LeBlond,P.H. 1983. The Strait of Georgia: functional anatomy of a coastal sea, Canadian Journal of Fisheries and Aquatic Science 40: 10331063.

Legendre,P. and Legendre,L. 1998. umerical Ecology , Elsevier Science Publishers.

Leonards,P.E.G., Zierikzee,Y., Brinkman,U.A.Th., Cofino,W.P., Van Straalen,N.M., and Van Hattum,B. 1997. The selective dietary accumulation of planar polychlorinated biphenyls in the otter ( Lutra lutra ), Environmental Toxicology and Chemistry 16: 1807 1815.

Lesage,V., Hammill,M.O., and Kovacs,K.M. 2001. Marine mammals and the community structure of the Estuary and Gulf of St Lawrence, Canada: evidence from stable isotope analysis, Marine Ecology Progress Series 210: 203221. 148

Lesage,V., Hammill,M.O., and Kovacs,K.M. 2002. Diettissue fractionation of stable carbon and nitrogen isotopes in Phocid seals, Marine Mammal Science 18: 182193.

Li,Y.F., Macdonald,R.W., Jantunen,L.M.M., Harner,T., Bidleman,T.F., and Strachan,W.M.J. 2002. The transport of βhexachlorocyclohexane to the western Arctic Ocean: a contrast to αHCH, Science of the Total Environment 291: 229246.

Lucas, B. G. Organochlorines in harbour porpoise ( Phocoena phocoena ) of British Columbia. 1998. University of Victoria.

Lundgren,K., Tysklind,M., Ishaq,R., Broman,D., and Van Bavel,B. 2002. Polychlorinated naphthalene levels, distribution, and biomagnification in a benthic food chain in the Baltic Sea, Environmental Science and Technology 36: 50055013.

Macdonald,R.W. 2005. Climate change, risks and contaminants: a perspective from studying the Arctic, Human and Ecological Risk Assessment 11: 10991104.

Macdonald,R.W. and Crecelius,E.A. 1994. Marine sediments in the Strait of Georgia, Juan de Fuca Strait and Puget Sound: What can they tell us about contamination?, pp. 101137. In R.C.H.Wilson, R.J.Beamish, F.Aitkens, and J.Bell, editors, Review of the marine environment and biota of the Strait of Georgia, Puget Sound and Juan de Fuca Strait: Proceedings of the BC/Washington Symposium on the Marine Environment. Fisheries and Oceans Canada.

Macdonald,R.W., Cretney,W.J., Crewe,N., and Paton,D. 1992. A history of octachlorodibenzopdioxin, 2,3,7,8tetrachlorodibenzofuran, and 3,3',4,4' tetrachlorobiphenyl contamination in Howe Sound, British Columbia, Environmental Science and Technology 26: 15441550.

Mackay,D. and Fraser,A. 2000. Bioaccumulation of persistent organic chemicals: mechanisms and models, Environmental Pollution 110: 375391.

Mackintosh,C.E., Maldonaldo,J., Hongwu,J., Hoover,N., Chong,A., Ikonomou,M.G., and Gobas,F.A.P.C. 2004. Distribution of phthalate esters in a marine aquatic food web: comparison to polychlorinated biphenyls, Environmental Science and Technology 38: 20112020.

MacLellan,S.E. 1999. Guide for sampling structures used in age determination of Pacific salmon , Fisheries and Oceans Canada, Nanaimo.

Magnusson,K., Magnusson,M., Ostberg,P., Granberg,M., and Tiselius,P. 2007. Bioaccumulation of 14 CPCB 101 and 14 CPBDE 99 in the marine planktonic copepod Calanus finmarchicus under different food regimes, Marine Environmental Research 63: 6781.

Magnusson,K. and Tiselius,P. 2010. The importance of uptake from food for the bioaccumulation of PCB and PBDE in the marine planktonic copepod Acartia clausi , Aquatic Toxicology 98: 374380. 149

ManchesterNeesvig,J., Valters,K., and Sonzogni,W.C. 2001. Comparison of polybrominated diphenyl ethers (PBDEs) and polychlroinated biphenyls (PCBs) in Lake Michigan salmonids, Environmental Science and Technology 35: 10721077.

Markussen,N.H., Bjørge,A., and Øritsland,N.A. 1989. Growth in harbour seals ( Phoca vitulina ) on the Norwegian coast, Journal of Zoology 219: 433440.

Markussen,N.H., Ryg,M., and Oritsland,N.A. 1990. Energy requirements for maintenance and growth of captive harbour seals, Phoca vitulina , Canadian Journal of Zoology 68: 423426.

Markussen,N.H., Ryg,M., and Oritsland,N.A. 1994. The effect of feeding on the metabolic rate in harbour seals ( Phoca vitulina ), Journal of Comparative Physiology B 164: 8993.

McConnaughey,T. and McRoy,C.P. 1979. Foodweb structure and the fractionation of carbon isotopes in the Bering Sea, Marine Biology 53: 257262.

McFarlane,G.A., Beamish,R.J., and Schweigert,J. 2001. Pacific herring: Common factors have opposite impacts in adjacent ecosystems in Herring: Expectations for a new millennium. , Alaska Sea Grant College Program.

McFarlane,G.A. and King,J.R. 2003. Migration patterns of spiny dogfish ( Squalus acanthias ) in the North Pacific Ocean, Fishery Bulletin 101: 358367.

McLellan,W.A., Koopman,H.N., Rommel,S.A., Read,A.J., Potter,C.W., Nicolas,J.R., Westgate,A.J., and Pabst,D.A. 2002. Ontogenetic allometry and body composition of harbour porpoises ( Phocoena phocoena , L.) from the western North Atlantic, Journal of Zoology London 257: 457471.

Minagawa,M. and Wada,E. 1984. Stepwise enrichment of 15 N along food chains: Further evidence and the relation between δ15 N and animal age, Geochimica et Cosmochimica Acta 48: 11351140.

Missildine,B.R., Peters,R.J., ChinLeo,G., and Houck,D. 2005. Polychlorinated biphenyl concentrations in adult Chinook salmon ( Oncorhynchus tshawytscha ) returning to coastal and Puget Sound hatcheries of Washington State, Environmental Science and Technology 39: 69446951.

Moisey,J., Fisk,A.T., Hobson,K.A., and Norstrom,R.J. 2001. Hexachlorocyclohexane (HCH) isomers and chiral signatures of αHCH in the Arctic marine food web of the Northwater Polynya, Environmental Science and Technology 35: 19201927.

Morrison,H.A., Whittle,D.M., and Haffner,G.D. 2002. A comparison of the transport and fate of polychlorinated biphenyl congeners in three great lakes food webs, Environmental Toxicology and Chemistry 21(4): 683692. 150

Mossner,S. and Ballschmiter,K. 1997. Marine mammals as global pollution indicators for organochlorines, Chemosphere 34: 12851296.

Muelbert,M.M.C. and Bowen,W.D. 1993. Duration of lactation and postweaning changes in mass and body composition of harbour seal, Phoca vitulina , pups, Canadian Journal of Zoology 71: 14051414.

Muir,D., Braune,B., DeMarch,B., Norstrom,R., Wagemann,R., Lockhart,L., Hargrave,B., Bright,D., Addison,R., Payne,J., and Reimer,K. 1999. Spatial and temporal trends and effects of contaminants in the Canadian Arctic marine ecosystem: a review, Science of the Total Environment 230: 83144.

Muir,D., Riget,F., Cleemann,M., Skaare,J., Kleivane,L., Nakata,H., Dietz,R., Severinsen,T., and Tanabe,S. 2000. Circumpolar trends of PCBs and organochlorine pesticides in the Arctic marine environment inferred from levels in ringed seals, Environmental Science and Technology 34: 24312438.

Muir,D., Savinova,T., Savinov,V., Alexeeva,L., Potelov,V., and Svetochev,V. 2003. Bioaccumulation of PCBs and chlorinated pesticides in seals, fishes and invertebrates from the White Sea, Russia, The Science of the Total Environment 306: 111131.

Muir,D.C.G., Norstrom,R.J., and Simon,M. 1988. Organochlorine contaminants in Arctic marine food chains: accumulation of specific polychlorinated biphenyls and chlordane related compounds, Environmental Science and Technology 22: 10711079.

Murk,A.J., Leonards,P.E.G., Van Hattum,B., Luit,R., Van der Weiden,M.E.J., and Smit,M. 1998. Application of biomarkers for exposure and effect of polyhalogenated aromatic hydrocarbons in naturally exposed European otters ( Lutra lutra ), Environmental Toxicology and Pharmacology 6: 91102.

Myers,M.J., Ylitalo,G.M., Krahn,M.M., Boyd,D., Calkins,D., Burkanov,V., and Atkinson,S. 2008. Organochlorine contaminants in endangered Stellar sea lion pups (Eumetopias jubatus ) from western Alaska and the Russian Far East, Science of the Total Environment 396: 6069.

Newell, A. J., Johnson, D. W., and Allen, L. K. Niagara River Biota Contamination Project: Fish Flesh Criteria for Piscivorous Wildlife. Technical Report 873, 1136. 1987. Albany, NY, USA, New York State Department of Environmental Conservation, Division of Fish and Wildlife, Bureau of Environmental Protection.

Newman,M.C. 1998. Fundamentals of ecotoxicology , Sleeping Bear/Ann Arbor Press, Chelsea, MI.

Newsome,W.H., Doucet,J., Davies,D., and Sun,W.F. 2000. Pesticide residues in the Canadian market basket survey 1992 to 1996., Food Additives & Contaminants 17: 847 854. 151

Nfon,E., Cousins,I.T., and Broman,D. 2008. Biomagnification of organic pollutants in benthic and pelagic marine food chains from the Baltic Sea, Science of the Total Environment 397: 190204.

Noël,M., Dangerfield,N., Hourston,R.A.S., Belzer,W., Shaw,P., Yunker,M.B., and Ross,P.S. 2009. Do transPacific air masses deliver PBDEs to coastal British Columbia, Canada?, Environmental Pollution 157: 34043412.

O'Neill,S.M. and West,J.E. 2009. Marine distribution, life history traits, and the accumulation of polychlorinated biphenyls in chinook salmon from Puget Sound, Washington, Transactions of the American Fisheries Society 138: 616632.

O'Neill,S.M., West,J.E., and Hoeman,J.C. 1998. Spatial trends in the concentration of polychlorinated biphenyls (PCBs) in Chinook ( Oncorhyncus tshawytscha ) and Coho salmon ( O. kisutch ) in Puget Sound and factors affecting PCB accumulation: Results from the Puget Sound Ambient Monitoring Program, Puget Sound Research '98 312328.

Olesiuk,P.F. 1993. Annual prey consumption by harbor seals ( Phoca vitulina ) in the Strait of Georgia, British Columbia, Fishery Bulletin 91: 491515.

Olesiuk,P.F., Bigg,M.A., Ellis,G.M., Crockford,S.J., and Wigen,R.J. 1990. An assessment of the feeding habits of harbour seals (Phoca vitulina ) in the Strait of Georgia, British Columbia, based on scat analysis, Canadian Technical Report Of Fisheries and Aquatic Sciences 1730: 1135.

Osborne,R.W. 1999. A historical ecology of Salish Sea "resident" killer whales (Orcinus orca): with implications for management , University of Victoria, Victoria, B.C.

Pomeroy,P.P., Green,N., Hall,A.J., Walton,M.J., Jones,K., and Harwood,J. 1996. Congenerspecific exposure of grey seal ( Halichoerus grypus ) pups to chlorinated biphenyls during lactation, Canadian Journal of Fisheries and Aquatic Sciences 53: 15261534.

Pulster,E.L., Smalling,K.L., Zolman,E., Schwacke,L., and Maruya,K.A. 2009. Persistent organochlorine pollutants and toxaphene congener profiles in bottlenose dolphins (Tursiops truncatus ) frequenting the Turtle/Brunswick River Estuary, Georgia, USA, Environmental Toxicology and Chemistry 28: 13901399.

Qiao,P., Gobas,F.A.P.C., and Farrell,A.P. 2000. Relative contributions of aqueous and dietary uptake of hydrophobic chemicals to the body burden in juvenile rainbow trout, Archives of Environmental Contamination and Toxicology 39: 369377.

Rau,G.H., Mearns,A.J., Young,D.R., Olson,R.J., Schafer,H.A., and Kaplan,I.R. 1983. Animal 13 C/ 12 C correlates with trophic level in pelagic food webs, Ecology 64: 1314 1318. 152

Rayne,S., Ikonomou,M.G., and Antcliffe,B. 2003. Rapidly increasing polybrominated diphenyl ether concentrations in the Columbia River system from 1992 to 2000, Environmental Science and Technology 36: 28472854.

Rayne,S., Ikonomou,M.G., Ellis,G.M., BarrettLennard,L.G., and Ross,P.S. 2004. PBDEs, PBBs, and PCNs in three communities of freeranging killer whales (Orcinus orca) from the northeastern Pacific Ocean, Environmental Science and Technology 38: 42934299.

Reijnders,P.J.H. 1986. Reproductive failure in common seals feeding on fish from polluted coastal waters, ature 324: 456457.

Ricker,W.E. 1984. Computation and uses of central trend lines, Canadian Journal of Zoology 62: 18971905.

Robbins,C.T., Felicetti,L.A., and Sponheimer,M. 2005. The effect of dietary protein quality on nitrogen isotope discrimination in mammals and birds, Oecologia 144: 534 540.

Ross,P.S. 2000. Marine mammals as sentinels in ecological risk assessment, Human and Ecological Risk Assessment 6: 2946.

Ross,P.S. 2002. The role of immunotoxic environmental contaminants in facilitating the emergence of infectious diseases in marine mammals, Human and Ecological Risk Assessment 8(2): 277292.

Ross,P.S. 2006. Fireproof killer whales: Flame retardant chemicals and the conservation imperative in the charismatic icon of British Columbia, Canadian Journal of Fisheries and Aquatic Sciences 63: 224234.

Ross,P.S. and Birnbaum,L.S. 2003. Integrated human and ecological risk assessment: A case study of persistent organic pollutants (POPs) in humans and wildlife, Human and Ecological Risk Assessment 9:1: 303324.

Ross,P.S., Couillard,C.M., Ikonomou,M.G., Johannessen,S.C., Lebeuf,M., Macdonald,R.W., and Tomy,G.T. 2009. Large and growing environmental reservoirs of DecaBDE present an emerging health risk for fish and marine mammals, Marine Pollution Bulletin 58: 710.

Ross,P.S., De Swart,R.L., Addison,R.F., Van Loveren,H., Vos,J.G., and Osterhaus,A.D.M.E. 1996. Contaminantinduced immunotoxicity in harbour seals: wildlife at risk?, Toxicology 112: 157169.

Ross,P.S., De Swart,R.L., Reijnders,P.J.H., Van Loveren,H., Vos,J.G., and Osterhaus,A.D.M.E. 1995. Contaminantrelated suppression of delayedtype hypersensitivity and antibody responses in harbor seals fed herring from the Baltic Sea, Environmental Health Perspectives 103: 162167. 153

Ross,P.S., De Swart,R.L., Timmerman,H.H., Reijnders,P.J.H., Vos,J.G., Van Loveren,H., and Osterhaus,A.D.M.E. 1996. Suppression of natural killer cell activity in harbour seals (Phoca vitulina ) fed Baltic Sea herring, Aquatic Toxicology 34: 7184.

Ross,P.S., Ellis,G.M., Ikonomou,M.G., BarrettLennard,L.G., and Addison,R.F. 2000. High PCB concentrations in freeranging Pacific killer whales, Orcinus orca : effects of age, sex and dietary preference, Marine Pollution Bulletin 40: 504515.

Ross,P.S., Jeffries,S.J., Yunker,M.B., Addison,R.F., Ikonomou,M.G., and Calambokidis,J. 2004. Harbour seals ( Phoca vitulina ) in British Columbia, Canada, and Washington, USA, reveal a combination of local and global polychlorinated biphenyl, dioxin, and furan signals, Environmental Toxicology and Chemistry 23: 157165.

Ross,P.S., Pohajdak,B., Bowen,W.D., and Addison,R.F. 1993. Immune function in free ranging harbor seal ( Phoca vitulina ) mothers and their pups during lactation, Journal of Wildlife Diseases 29: 2129.

Ruus,A., Ugland,K.I., and Skaare,J.U. 2002. Influence of trophic position on organochlorine concentrations and compostitional patterns in a marine food web, Environmental Toxicology and Chemistry 21(11): 23562364.

Safe,S.H. 1984. Polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs): biochemistry, toxicology, and mechanism of action, CRC Critical Reviews in Toxicology 13: 319395.

Saunders, M. W. Dogfish in Groundfish stock assessment for the West Coast of Canada in 1987 and recommended yield options for 1988. Fargo, J., Saunders, M. W., and Tyler, A. V. Canadian Technical Report of Fisheries Aquatic Sciences 1617: 304pp, 151158. 1988.

Shaw,S.D., Brenner,D., Bourakovsky,A., Mahaffey,C.A., and Perkins,C.R. 2005. Polychlorinated biphenyls and chlorinated pesticides in harbor seals ( Phoca vitulina concolor ) from the northwestern Atlantic coast, Marine Pollution Bulletin 50: 10691084.

Shiu,W.Y. and Mackay,D. 1986. A critical review of aqueous solubilities, vapor pressures, Henry's Law constants, and octanolwater partition coefficients of the polychlorinated biphenyls, Journal of Physical and Chemical Reference Data 15: 911 929.

Simms,W., Jeffries,S.J., Ikonomou,M.G., and Ross,P.S. 2000. Contaminantrelated disruption of vitamin A dynamics in freeranging harbor seal ( Phoca vitulina ) pups from British Columbia, Canada and Washington State, USA, Environmental Toxicology and Chemistry 19: 28442849.

Skarphedinsdottir,H., Gunnarsson,K., Gudmundsson,G.A., and Nfon,E. 2009. Bioaccumulation and biomagnfication of organochlorines in a marine food web at a pristine site in Iceland, Archives of Environmental Contamination and Toxicology . 154

Skoglund,R.S., Stange,K., and Swackhamer,D.L. 1996. A kinetics model for predicting the accumulation of PCBs in phytoplankton, Environmental Science and Technology 30: 21132120.

Smit,M.D., Leonards,P.E.G., Murk,A.J., De Jongh,A.W.J.J., and Van Hattum,B. 1996. Development of otterbased quality objectives for PCBs , Institute for Environmental Studies.

Smith,S.J., Iverson,S.J., and Bowen,W.D. 1997. Fatty acid signatures and classification trees: new tools for investigating the foraging ecology of seals, Canadian Journal of Fisheries and Aquatic Sciences 54: 13771386.

Stapleton,H.M. and Baker,J.E. 2003. Comparing polybrominated diphenyl ether and polychlorinated biphenyl bioaccumulation in a food web in Grand Traverse Bay, Lake Michigan, Archives of Environmental Contamination and Toxicology 45: 227234.

Starr,C. and Taggart,R. 1981. Biology: the unity and diversity of life , Wadsworth Publishing Company, Inc., Belmont, California.

Stone,D. 2006. Polybrominated diphenyl ethers and polychlorinated biphenyls in different tissue types from Chinook salmon ( Oncorhynchus tshawytscha ), Bulletin of Environmental Contamination and Toxicology 76: 148154.

StorrHansen,E., Spliid,H., and Boon,J.P. 1995. Patterns of chlorinated biphenyl congeners in harbor seals ( Phoca vitulina ) and in their food: statistical analysis, Archives of Environmental Contamination and Toxicology 28: 4854.

Tabor,G.M. and Aguirre,A.A. 2004. Ecosystem health and sentinel species: adding an ecological element to the proverbial "canary in the mineshaft", EcoHealth 1: 226228.

Tabuchi,M., Veldhoen,N., Dangerfield,N., Jeffries,S.J., Helbing,C.C., and Ross,P.S. 2006. PCBrelated alteration of thyroid hormones and thyroid hormone receptor gene expression in freeranging harbor seals ( Phoca vitulina ), Environmental Health Perspectives 114: 10241031.

Therriault,T.W., Hay,D.E., and Schweigert,J.F. 2009. Biological overview and trends in pelagic forage fish abundance in the Salish Sea (Strait of Georgia, British Columbia), Marine Ornithology 37: 38.

Thomann,R.V. 1989. Bioaccumulation model of organic chemical distribution in aquatic food chains, Environmental Science and Technology 23: 699707.

Thurman,H.V., Burton,E.A., Tarbuck,E.J., and Lutgens,F.K. 2001. Oceanography & Earth Science , Pearson Custom Publishing, Boston, MA.

Tieszen,L.L., Boutton,T.W., Tesdahl,K.G., and Slade,N.A. 1983. Fractionation and turnover of stable carbon isotopes in animal tissues: Implications for δ13 C analysis of diet, Oecologia 57: 3237. 155

Tillitt,D.E., Gale,R.W., Meadows,J.C., Zajicek,J.L., Peterman,P.H., Heaton,S.N., Jones,P.D., Bursian,S.J., Kubiak,T.J., Giesy,J.P., and Aulerich,R.J. 1996. Dietary exposure of mink to carp from Saginaw Bay. 3. Characterization of dietary exposure to planar halogenated hydrocarbons, dioxin equivalents, and biomagnification, Environmental Science and Technology 30: 283291.

Tomy,G.T., Budakowski,W., Halldorson,T., Whittle,D.M., Keir,M.J., Marvin,C., MacInnis,G., and Alaee,M. 2004. Biomagnification of α and γ hexabromocyclododecane isomers in a Lake Ontario food web, Environmental Science and Technology 38: 22982303.

United Nations Environment Programme. The nine new POPs An introduction to the nine chemicals added to the Stockholm Convention by the Conference of the Parties at its fourth meeting. United Nations Environment Programme , 115. 2009. Secretariat of the Stockholm Convention on Persistent Organic Pollutants, United Nations Environment Programme.

US Environmental Protection Agency [USEPA]. Persistent organic pollutants: A global issuea global response. EPA 160F02001, i24. 2002. Washington, DC, USEPA.

USEPA. The effect of zebra mussels on cycling and potential bioavailability of PCBs: Case study of Saginaw Bay. Section 5. 2010. Environmental Protection Agency. Great Lakes Pollution Prevention and Toxics.

USEPA Office of Science and Technology. Brominated diphenylethers in water, soil, sediment and tissue by HRGC/HRMS: EPA Method 1614 2003 draft. 2003. Washington, DC, USA.

USEPA Solid Waste and Emergency Response. Organochlorine pesticides by gas chromatography: Method 8081A Revision 2. 1998. Washington, DC, USA.

USEPA Solid Waste and Emergency Response. Semivolatile organic compounds by Gas Chromatography / Mass Spectrometry (GCMS): Method 8270D Revision 4. 1998. Washington, DC, USA.

Van Bressem,M.F., Raga,J.A., Di Guardo,G., Jepson,P.D., Duignan,P.J., Siebert,U., Barrett,T., de Oliveira Santos,M.C., Moreno Lopez,J., Siciliano,S., Aguilar,A., and Van Waerebeek,K. 2009. Emerging infectious diseases in cetaceans worldwide and the possible role of environmental stressors, Diseases of Aquatic Organisms 86: 143157.

Van den Berg,M., Birnbaum,L., Bosveld,A.T.C., Brunstrom,B., Cook,P., Feeley,M., Giesy,J.P., Hanberg,A., Hasegawa,R., Kennedy,S.W., Kubiak,T., Larsen,J.C., Van Leeuwen,F.X.R., Liem,A.K., Nolt,C., Peterson,R.E., Poellinger,L., Safe,S.H., Schrenk,D., Tillitt,D.E., Tysklind,M., Younes,M., Waern,F., and Zacharewski,T.R. 1998. Toxic equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife, Environmental Health Perspectives 106: 775792. 156 van Oostdam,J., Donaldson,S.G., Feeley,M., Arnold,D., Ayotte,P., Bondy,G., Chan,L., Dewaily,E., Furgal,C.M., Kuhnlein,H., Loring,E., Muckle,G., Myles,E., Receveur,O., Tracy,B., Gill,U., and Kalhok,S. 2005. Human health implications of environmental contaminants in Arctic Canada: a review, Science of the Total Environment 351352: 165246.

Vander Zanden,M.J. and Rasmussen,J.B. 1999. Primary consumer δ13 C and δ15 N and the trophic position of aquatic consumers, Ecology 80: 13951404.

Walker,J.L. and Macko,S.A. 2000. Dietary studies of marine mammals using stable carbon and nitrogen isotopic ratios of teeth, Marine Mammal Science 15(2): 314334.

Walters,D.M., Fritz,K.M., Johnson,B.R., Lazorchak,J.M., and McCormick,F.H. 2008. Influence of trophic position and spatial location on polychlorinated biphenyl (PCB) bioaccumulation in a stream food web, Environmental Science and Technology 42: 2316 2322.

Wania,F. and Mackay,D. 1996. Tracking the distribution of persistent organic pollutants., Environmental Science and Technology 30: 390A396A.

Wania,F. and Mackay,D. 2001. Global fractionation and cold condensation of low volatility organochlorine compounds in polar regions, Ambio 22: 1018.

Weber,K. and Goerke,H. 2003. Persistent organic pollutants (POPs) in antarctic fish: levels, patterns, changes, Chemosphere 53: 667678.

West,J.E. 1997. Protection and restoration of marine life in the inland waters of Washington State , Puget Sound/Georgia Basin Environmental Report Series: Number 6, Olympia, WA, USA.

Wong,C.S., Mabury,S.A., Whittle,D.M., Backus,S.M., Teixeira,C., Devault,D.S., Bronte,C.R., and Muir,D.C.G. 2004. Organochlorine compounds in Lake Superior: Chiral polychlorinated biphenyls and biotransformation in the aquatic food web, Environmental Science and Technology 38: 8492.

World Wildlife Fund [WWF]. Stockholm Convention: New POPs Screening Additional POPs Candidates. 138. 2005. World Wildlife Fund.

Wu,J.P., Luo,X.J., Zhang,Y., Yu,M., Chen,S.J., Mai,B.X., and Yang,Z.Y. 2009. Biomagnification of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls in a highly contaminated freshwater food web from South China, Environmental Pollution 157: 904909.

Yess,N.J., Gunderson,E.L., and Roy,R.R. 1993. U.S. Food and Drug Administration monitoring of pesticide residues in infant foods and adult foods eaten by infants/children, Journal of AOAC International 76: 492507. 157

Yunker,M.B., Belicka,L.L., Harvey,H.R., and Macdonald,R.W. 2005. Tracing the inputs and fate of marine and terrigenous organic matter in Arctic Ocean sediments: A multivariate analysis of lipid biomarkers, DeepSea Research II 52: 34783508.

Yunker,M.B. and Cretney,W.J. 2000. Bioavailability of chlorinated dibenzopdioxins and dibenzofurans to dungeness crab ( Cancer magister ) at marine pulp mill sites in British Columbia, Canada, Environmental Toxicology and Chemistry 19(12): 29973011.

158

Appendix I: Principal components analysis (PCA) variables. (Table taken from Cullon et al., 2009)

Variable a PCB substitution Chlorine umber < Variable PCB substitution Chlorine umber < Dibenzo pdioxins and dibenzofurans o PCB94 22'356' o 5 1 1,2,3,6,7,8 6 PCB102/93 22'456'/22'356 5 OCDD 8 PCB95 22'35'6 5 2,3,7,8 TCDF 4 PCB88 22'346 5 2,3,4,7,8 PnCD F 5 6 PCB91 22'34'6 5 PCB92/84 22'355'/22'33'6 5 onPCB11ortho PCBs 33’ 2 PCB89 22'346' 5 PCB15 44’ 2 PCB101/90 22'455'/22'34'5 5 PCB35 33’4 3 2 PCB99 22'44'5 5 PCB37 344’ 3 PCB119 23'44'6 5 PCB80 33’55’ 4 PCB109/83 233'46/22'33'5 5 PCB79 33’45’ 4 PCB97/86 22'3'45/22'345 5 PCB81 344’5 4 PCB115/87 2344'6/22'345' 5 PCB77 33’44’ 4 PCB85 22'344' 5 PCB126 33’44’5 5 PCB110 233'4'6 5 PCB169 33’44’55’ 6 PCB82 22'33'4 5 PCB15 5 22'44'66' 6 2 MonoPCB26ortho PCBs 23'5 3 PCB150 22'34'66' 6 4 PCB25 23'4 3 PCB148 22'34'56' 6 2 PCB31 24'5 3 PCB136 22'33'66' 6 PCB28 244' 3 PCB154 22'44'56 6 PCB33/20 2'34/233' 3 PCB151 22'355'6 6 PCB22 234' 3 PCB135/144 22'33'56'/22'345 '6 6 PCB72 23'55' 4 PCB147 22'34'56 6 PCB68 23'45' 4 PCB149 22'34'5'6 6 PCB57 233'5 4 1 PCB139/140 22'344'6/22'344'6' 6 PCB67 23'45 4 PCB143/134 22'3456'/22'33'56 6 PCB58 233'5' 4 2 PCB142/131 22'3456/22'33'46 6 PCB63 234'5 4 PCB146/161 22'34'55'/233'45'6 6 a PCB (polychlorinated biphenyl); HxCDD (hexachlorodibenzopdioxin); OCDD (octachlorodibenzopdioxin); TCDF (tetrachlorodibenzofuran); PnCDF (pentachlorodibenzofuran). 159

PCB61/74 2345/244'5 4 PCB132/153 22'33'46'/22'44'55' 6 PCB70/76 23'4'5/2'345 4 PCB168 23'44'5'6 6 PCB66 23'44' 4 PCB141 22'3455' 6 PCB56/60 233'4'/2344' 4 PCB137 22'344'5 6 PCB111 233'55' 5 2 PC B130 22'33'45' 6 PCB120 23'455' 5 PCB160/163/164/ 233'456/233'4'56/23 6 PCB124 2'3455' 5 PCB158 233'44'6 6 PCB108/107 233'45'/233'4'5 5 PCB129 22'33'45 6 PCB123 2'344'5 5 PCB166 2344'56 6 PCB118 23'44'5 5 PCB128 22'33'44' 6 PCB114 2344'5 5 PCB188 22'34'566' 7 6 PCB122 2'33'45 5 PCB184 22'344'66' 7 PCB105 233'44' 5 PCB179 22'33'566' 7 PCB159 233'455' 5 5 PCB176 22'33'466' 7 PCB162 233'4'55' 6 PCB178 22'33'55'6 7 PCB167 23'44'55' 6 PCB175 22'33'45'6 7 PCB156 233'44'5 6 PCB187/182 22'34'55'6/22'344'5 7 PCB157 233'44'5' 6 PCB183 22'344'5'6 7 PCB189 233'44'55' 7 PCB185 22'3455'6 7 PCB174/181 22'33'456'/22'344'5 7 DiPCB10/4ortho PCBs 26/22' 2 PCB17 7 22'33'4'56 7 PCB19 22'6 3 PCB171 22'33'44'6 7 PCB18 22'5 3 PCB192/172 233'455'6/22'33'455 7 PCB17 22'4 3 PCB180 22'344'55' 7 PCB27/24 23'6/236 3 PCB193 233'4'55'6 7 PCB16/32 22'3/24'6 3 PCB191 233'44'5'6 7 1 PCB53 22'56' 4 PCB170/190 22'33'44'5/233'44'5 7 PCB51 22'46' 4 PCB202 22'33'55'66' 8 PCB45 22'36 4 PCB201 22'33'45'66' 8 PCB46 22'36' 4 PCB197 22'33'44'66' 8 PCB73/52 23'5'6/22'55' 4 PCB200 22'33'4566' 8 PCB49 22'45' 4 PCB198 22'33'455' 6 8 3 PCB47/75/48 22'44'/244'6/22'45 4 PCB199 22'33'455'6' 8 PCB44 22'35' 4 PCB203/196 22'344'55'6/22'33'4 8 PCB59/42 233'6/22'34' 4 PCB195 22'33'44'56 8 PCB71/41/64 23'4'6/22'34/234'6 4 PCB194 22'33'44'55' 8 160

PCB40 22'33' 4 PCB208 22'33'455'66' 9 PCB96 22'366' 5 4 PCB207 22'33'44'566' 9 1 PCB103 22'45'6 5 PCB206 22'33'44'55'6 9 PCB100 22'44'6 5 PCB209 22'33'44'55'66' 10

161

Appendix II: Table showing prey preferences for harbour seal ( Phoca vitulina ) food web species

Species Preferred prey Reference Common name Latin name Harbour seals Phoca vitulina Hake, herring, salmonids, plainfin midshipman, lingcod RM 2388 Seal pups Phoca vitulina Sandlance, sandshrimp RM 963 Pacific hake Merluccius productus Herring; Hake smaller than 40cm* feed almost RM 8098 exclusively on euphausiids. (only one of our hake exceeded 40 cm) Primarily euphausiids and sand lance, lesser extent RM 8106 herring, smelt, anchovy, shrimp Pacific herring Clupea pallasi Primarily copepods, also barnacle larvae, mollusc larvae, RM 8106 bryozoans, rotifers, young fishes Chinook salmon Oncorhynchus Adults herring, sandlance, euphausiids RM 7466 tshawytscha Smolts herring, pelagic amphipods, crab megalopa (70 92%) Sockeye salmon Oncorhynchus nerka Zooplankton (euphasiids, copepods, amphipods), small RM 7469 fish, squid, sandlance, herring Pacific lingcod Ophiodon elongatus Herring, sand lance; flounders, hake walleye pollock, RM 8106 cod, rockfishes, Crustacea, octopus. Juveniles feed on copepods and small crustacea Rockfish (Greenstriped, Family Scorpaenidae Sand lance, herring, small rockfish, crustaceans; RM 8107 Quillback, Splitnose) Juveniles eat primarily plankton (small crustaceans and copepods) Pacific tomcod Microgadus proximus Shrimps RM 8106 Pacific midshipman Porichthys notatus Fishes and crustaceans RM 8106 English sole Parophrys vetulus Clams, clam siphons, other small molluscs, marine RM 8106 worms, small crabs and shrimps, brittle stars Pacific sandlance Ammodytes hexapterus Copepods RM 8106 Spot prawns Pandalus platyceros Predators small crustaceans and other invertebrates RM 8108 Ghost shrimp Callianassa californiensis Suspension feed Plankton and detritus RM 8108 Butter clams Saxidomus gigantea Filter feed Plankton RM 8108 Zooplankton Pseudocalanus spp. Carnivores, herbivores, omnivores, or detritovores RM 8105 162

Appendix III: Stable carbon and nitrogen isotope ratios for the Strait of Georgia harbour seal ( Phoca vitulina ) food web species

20 HSM HSF 18 HSP Pelagic spp. 4 r2 = 0.20 levelTrophic (TL) 16 p < 0.0001 CH RF m = 0.51 ES df = 156 LC 14 CS TC HK 3 15 SL δ N (‰) HE FB MS SP 12 SHP SK 10 2

ZO 8 1 6 26 24 22 20 18 16 14 12

13 δ C (‰)

163

Appendix IV: Biomagnification factors for harbour seal (Phoca vitulina ) /food basket increase with log K ow in the Strait of Georgia seal food web, with the highest magnification observed between log K ow 7.27 and 7.7.

2.5 205 2 ) 2.0 r = 0.48 170 194 p < 0.0001 180 1.5 m = 0.64

1.0

0.5

0.0

log (BMF Seal/Food basket Seal/Food (BMF log 0.5

1.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5

log K ow

164

Appendix V: List of food web magnification factors (FWMFs) of polychlorinated biphenyl (PCB) congeners for the Strait of Georgia harbour seal ( Phoca vitulina ) food web. FWMF = 10 b, where b is the slope of the regression trophic level (TL) vs [PCB congener]; r2 = correlation coefficient; TL vs [PCB congener] regressions found not statistically significant ( p ≥ 0.05) are annotated by italicized FWMF values.

Congener FWMF r2 Congener FWMF r2 Congener FWMF r2 4 1.53 0.114 55 1.97 0.142 102 2.81 0.221 11 0.99 0.00 56 1.39 0.084 103 1.70 0.100 13 1.00 0.00 57 1.50 0.053 105 2.66 0.328 15 0.82 0.005 58 2.04 0.174 107 2.25 0.168 16 1.30 0.027 63 1.12 0.007 109 6.68 0.475 17 1.35 0.039 64 1.75 0.143 110 1.96 0.176 18 1.35 0.050 66 1.64 0.139 114 3.36 0.504 19 0.94 0.002 67 1.00 0.00 118 2.46 0.326 22 0.94 0.001 68 1.36 0.063 119 3.60 0.385 25 1.13 0.008 70 1.43 0.102 120 4.92 0.454 26 1.05 0.002 72 1.52 0.124 123 2.78 0.202 27 1.16 0.011 74 3.07 0.518 124 1.07 0.148 28 1.32 0.070 77 2.00 0.157 128 5.48 0.514 31 1.19 0.025 80 7.96 0.417 129 2.31 0.278 33 1.10 0.004 82 1.18 0.018 130 4.88 0.649 35 0.74 0.011 84 1.39 0.027 131 2.60 0.288 37 0.72 0.016 85 4.87 0.567 134 1.94 0.134 40 1.50 0.119 87 2.70 0.314 135 2.93 0.201 42 2.42 0.231 89 1.14 0.003 136 1.78 0.102 43/49 2.63 0.430 91 3.35 0.383 137 5.06 0.594 44 1.57 0.156 94 2.58 0.205 138/163 5.94 0.621 45 1.12 0.006 95 2.36 0.274 139 4.02 0.515 46 1.22 0.018 96 2.12 0.100 141 2.47 0.283 47 2.71 0.229 97 1.69 0.111 146 5.65 0.685 51 1.16 0.010 99 5.28 0.636 147 4.47 0.602 52/73 2.87 0.629 100 2.28 0.301 148 1.96 0.181 53 1.14 0.011 101 3.42 0.428 149 3.19 0.359 165

150 2.02 0.223 195 7.85 0.579 151 3.63 0.443 197 5.52 0.524 153 6.50 0.609 198 5.58 0.576 154 3.59 0.475 199 2.19 0.154 156 4.34 0.471 200 2.81 0.232 157 3.65 0.462 201 6.78 0.564 158 5.02 0.463 202 6.15 0.590 162 2.16 0.237 203 7.31 0.562 166 3.99 0.458 205 9.43 0.647 167 1.75 0.135 206 6.90 0.502 168 3.47 0.377 207 5.63 0.508 169 1.93 0.181 208 5.82 0.510 170 7.09 0.575 209 4.89 0.470 171 5.92 0.587 ΣΣΣPCB 3.58 0.529 172 6.47 0.604 174 2.70 0.274 175 4.40 0.548 176 1.93 0.115 177 5.40 0.560 178 5.70 0.588 179 1.97 0.120 180 6.99 0.596 183 6.63 0.605 184 4.04 0.619 185 2.72 0.271 187 7.06 0.677 188 2.21 0.198 189 4.94 0.485 191 6.15 0.564 193 6.69 0.5996