EPA910-R-02-006
   COLUMBIA RIVER BASIN
 FISH CONTAMINANT SURVEY
            1996-1998
U.S. Environmental Protection Agency
             Region 10
      Seattle, Washington 98101

-------

-------
                                      Table of Contents

Table of Contents	 i
List of Tables	 vi
List of Figure	 xiv
Acknowledgment	xx
Contributors	xxii
Columbia River Basin Fish Contaminant Study Workgroup	xxiii
List of Abbreviations and Acronyms	xxiv
Units	xxv
Executive Summary	E-l
1.0    Introduction	1-1
       1.1    Report Organization	1-1
       1.2    Study Background	1-1
       1.3    Study Area  	1-3
       1.4    Sampling Locations  	1-3
       1.5    Fish Species	1-7
       1.6    Sampling Methods	1-9
       1.7    Chemical Analysis  	1-10
              1.7.1  PCB analysis	1-12
              1.7.2  Mercury and Arsenic analysis	1-12
              1.7.3  Total Chlordane and Total DDT	1-12
              1.7.4. Lead Risk Characterization	1-13
              1.7.5  Data Quality Validation of Chemical Analyses	1-13
              1.7.6  Detection limits	1-14
              1.7.7  Statistical Data Summaries	1-14
       1.8    Lipid Analysis   	1-15
       1.9    Special Studies	1-15
              1.9.1  Channel Catfish and Smallmouth Bass	1-15
              1.9.2  Acid-Labile Pesticides	1-16
              1.9.3  Radionuclide analyses	1-16

2.0    Fish Tissue Chemical Concentrations 	2-18
       2.1    Percent Lipid   	2-18
       2.2    Semi-Volatile Chemicals	2-19
       2.3    Pesticides	2-21
              2.3.1  DDMU, Hexachlorobenzene, Aldrin, Pentachloroanisole, and Mirex
                      	2-21
              2.3.2  Total Chlordane	2-23
              2.3.3  Total DDT	2-24
       2.4    Aroclors  	2-29
       2.5    Dioxin-Like PCB  congeners  	2-33
       2.6    Chlorinated Dioxins and Furans  	2-35

-------
       2.7    Toxicity Equivalence Concentrations of Chlorinated Dioxins and Furans, and
              Dioxin-Like PCB congeners	2-39
       2.8    Metals	2-40
              2.8.1   Arsenic  	2-44
              2.8.2   Mercury	2-48

3.0    Human Health Risk Assessment	3-51

4.0    Exposure Assessment	4-52
       4.1    Identification of Exposed Populations	4-52
       4.2    Exposure Pathway	4-52
       4.3    Quantification Of Exposure	4-53
       4.4    Exposure Point Concentrations (Chemical Concentrations in Fish)	4-56
       4.5    Fish Ingestion Rates	4-57
              4.5.1   Fish Ingestion Rates for the General Population	4-57
              4.5.2   Fish Ingestion Rates for CRITFC's Member Tribes 	4-58
       4.6    Exposure Frequency 	4-60
       4.7    Exposure Duration	4-60
              4.7.1   Adults	4-60
              4.7.2   Children	4-61
       4.8    Body Weight	4-61
              4.8.1   Adults	4-61
              4.8.2   Children	4-61
       4.9    Averaging Time  	4-62
       4.10   Multiple-Species Diet Exposures  	4-62

5.0    Toxicity Assessment	5-65
       5.1    Summary of Toxicity Assessment for Non-Cancer Health Effects  	5-66
       5.2    Summary of Toxicity Assessment for Cancer 	5-73
       5.3    Special Assumptions and Methods Used For Selected Chemicals	5-75
              5.3.1   Non-Cancer Toxicity Values for Chlordanes, DDT/DDE/DDD, and
                     Aroclors  	5-75
              5.3.2   Cancer Toxicity for Chlorinated Dioxins/Furans, Dioxin-Like PCB
                     congeners, and PAHs	5-76
              5.3.3   Arsenic Toxicity	5-77

6.0    Risk Characterization	6-83
       6.1    Risk Characterization Methodology 	6-83
              6.1.1   Non-Cancer Health Effects 	6-83
              6.1.2   Cancer Risk Assessment  	6-84
              6.1.3   Chemicals Not Evaluated	6-86
              6.1.4   Arsenic  	6-86
              6.1.5   Sample Type	6-86
       6.2    Risk Characterization Results	6-87
              6.2.1   Non-Cancer Hazard Evaluation	6-88

-------
                     6.2.1.1 Non-Cancer Hazard Evaluation for Resident Fish  	6-88
                     6.2.1.2 Non-cancer Hazard Evaluation for Anadromous Fish	6-104
                     6.2.1.3 Comparisons Between Anadromous Fish and Resident Fish
                            Species 	6-112
              6.2.2   Cancer Risk Evaluation 	6-115
                     6.2.2.1 Cancer Risk Evaluation for Resident Fish	6-116
                     6.2.2.2 Cancer Risk Evaluation for Anadromous Fish  	6-129
                     6.2.2.3 Comparisons of Cancer Risks Between Anadromous Fish and
                            Resident Fish Species	6-137
              6.2.3   Summary of Non-Cancer Hazards and Cancer Risks for All Species
                      	 6-140
              6.2.4   Impacts of Sample Type on Risk Characterization	6-143
              6.2.5   Risk Characterization Using a Multiple-species Diet	6-144
              6.2.6   Risk Characterization Using Different Assumptions for Percent of
                     Inorganic Arsenic	6-147

7.0    Lead Risk Assessment	7-151
       7.1    Lead Concentrations in Fish	7-151
       7.2    Overview of Lead Risk Assessment Approach	7-152
       7.3    Method for Predicting Risks to Children	7-153
       7.4    Risk Characterization for Children  	7-157
       7.5    Uncertainties in risk estimates for Children  	7-158
       7.6    Method for Predicting Risks to Fetuses	7-159
       7.7    Risk Characterization for Fetuses	7-161
       7.8    Uncertainty Analysis for Risk to Fetuses	7-162
       7.9    Conclusions	7-162

8.0    Radionuclide Assessment	8-163
       8.1    Radionuclide Data Reporting and Use	8-163
       8.2    General Information on Radiation Risk	8-164
       8.3    Risk Calculations	8-165
       8.4    Composite Study site Results	8-167
              8.4.1   Potassium-40 Results	8-167
       8.5    Background	8-167
       8.6    Uncertainties  	8-170
       8.7    Discussion	8-170
       8.8    Conclusions	8-171

9.0    Comparisons of Fish Tissue Chemical Concentrations	9-172
       9.1    Comparison by Chemical Concentration	9-172
              9.1.1  Chlordane	9-172
              9.1.2   Total  DDT 	9-172
              9.1.3   PCBs	9-173
              9.1.4   Chlorinated Dioxins andFurans  	9-177
              9.1.5   Metals	9-178

                                            iii

-------
             9.1.6   Aluminum	9-178
             9.1.7   Arsenic	9-179
             9.1.8   Cadmium  	9-180
             9.1.9   Chromium	9-180
             9.1.10  Copper	9-181
             9.1.11   Lead	9-182
             9.1.12   Mercury	9-183
             9.1.13   Nickel	9-186
             9.1.14   Selenium	9-187
             9.1.15   Vanadium	9-189
             9.1.16  Zinc 	9-189
       9.2    Comparisons By Fish Species  	9-190
             9.2.1  Largescale  Sucker (Catostomus macrocheilus) and Bridgelip Sucker (C.
                    columbianus)  	9-192
             9.2.2  Mountain Whitefish (Prosopium williamsoni)  	9-195
             9.2.3  White Sturgeon (Acipenser transmontanus)	9-196
             9.2.4  Walleye (Stizostedion vitreum)	9-196
             9.2.5  Channel catfish (Ictaluruspunctatus	9-197
             9.2.6  Smallmouth Bass (Micropterus dolomieu)	9-198
             9.2.7   Rainbow and Steelhead (Oncorhynchus my kiss)	9-199
             9.2.8  Chinook Salmon (Oncorhynchus tshawytscha)  	9-200
             9.2.9  Coho Salmon (Oncorhynchus kisutch)  	9-201
             9.2.10  Pacific Lamprey (Lampetra tridentata)	9-202
             9.2.11 Eulachon (Thaleichthyspacificus) 	9-203
       9.3 Comparisons across all species	9-203
             9.3.1  ResidentFish	9-203
             9.3.2  Pacific lamprey and eulachon	9-204
             9.3.3 Salmonids 	9-205

10.0   Uncertainty Evaluation	10-208
       10.1   Fish Tissue Collection	10-208
       10.2   Chemical Analyses	10-210
             10.2.1 Lipid analyses  	10-211
       10.3   Comparing Chemical Data Across Fish Species and with Other Studies 	10-212
       10.4   Risk Assessment 	10-212
             10.4.1 Exposure Assessment	10-212
                    10.4.1.1 Contaminant Concentrations in Fish Tissue	10-212
                    10.4.1.2 Tissue Type  	10-213
                    10.4.1.3 Exposure  Duration  	10-214
                    10.4.1.4 Consumption Rate	10-214
                    10.4.1.5 Multiple-Species Consumption Patterns	10-215
                    10.4.1.6  Effects of Cooking	10-216
             10.4.2 Toxicity Assessment	10-217
                    10.4.2.1 Toxicity Values	10-217
                    10.4.2.2 Toxicity Equivalence Factors for Dioxins, Furans, and Dioxin-

                                           iv

-------
                           like PCB Congeners and Relative Potency Factors for PAHs
                             	10-218
                     10.4.2.3  Chemicals Without Quantitative Toxicity Factors	10-219
                     10.4.2.4  Risk Characterization for PCBs	10-220
                     10.4.2.5  Non-Cancer Effects from DDT, ODD, and DDE	10-222
                     10.4.2.6  Risk Characterization for Arsenic	10-223
              10.4.3  Risk Characterization	10-224
                     10.4.3.1  Cancer Risk Estimates	10-224
                     10.4.3.2  Non-Cancer Health Effects	10-225
                     10.4.3.3  Cumulative Risk from Chemical and Radionuclide Exposure
                             	10-226
       10.5   Risk Characterization for Consumption of Fish Eggs	10-226

11.0   Conclusions	11-228

12.0   References	12-230

-------
                                        List of Tables


 Table 1-1.  Description, study site, sampling location, and river mile for Columbia River Basin
       fish sampling 1996-1998	1-6

Table l-2a.  Resident fish species collected from the Columbia River Basin, 1996 -1998	1-7

Table l-2b.  Anadromous fish species collected from the Columbia River Basin, 1996 -1998.
        	1-8

Table 1-3. Recent surveys of types offish consumed by the general public in the Columbia
       River Basin	1-9

Table l-4a. 51 semi-volatile chemicals analyzed	1-11

Table l-4b. 26 pesticides analyzed	1-11

Table l-4c. 18 Metals analyzed	1-11

Table l-4d. 7  Aroclors analyzed  	1-11

Table l-4e. 13 Dioxin-like PCB congeners analyzed	1-11

Table l-4f  7 chlorinated dioxins analyzed	1-11

Table l-4g. 10 chlorinated furans analyzed	1-11

Table 1-5. Sampling study sites and numbers of replicates for survey of chemicals in tissues of
       smallmouth bass and channel catfish collected in the Columbia River Basin, 1996-1998.
        	1-16

Table 1-6. AED pesticides detected in fish tissue from the Columbia River Basin, 1996-1998.
        	1-16

Table 1-7. Radionuclide fish tissue samples including study site,  species, and number of
       replicates from the Columbia River Basin, 1996-1998.	1-17

Table 1-8. The radionuclides analyzed in fish tissue collected in the Columbia River Basin
       1996-1998	1-17

 Table 2-la.  Basin-wide composite concentrations* of semi-volatile chemicals detected in
       resident fish species from the Columbia River Basin, 1996-1998	2-20
                                             VI

-------
Table 2-lb.  Basin-wide composite concentrations* of semi-volatile chemicals detected in
       anadromous fish species from the Columbia River Basin, 1996-1998.	2-20

Table 2.2a.  Basin-wide concentrations of pesticides in resident fish tissue from the Columbia
       River Basin, 1996-1998	2-22

Table 2.2b.  Basin-wide concentrations of pesticides in anadromous fish tissue from the
       Columbia River Basin, 1996-1998.    	2-23

Table 2-3 .  Basin-wide average concentrations of total chlordane (oxy-chlordane, gamma, beta
       and alpha chlordane, cis and trans nonachlor) in fish from the Columbia River Basin,
       1996-1998.   	2-24

Table 2-4.  Basin-wide average concentrations of total DDT (DDT, DDE, ODD)  in composite
       fish tissue samples from the Columbia River Basin, 1996-1998	2-25

Table 2-5. Basin-wide average and maximum concentrations of p,p'DDE in composite samples
       offish from the Columbia River Basin, 1996-1998.  	2-26

Table 2-6. Basin-wide average concentrations of total Aroclors (1242, 1254,1260) detected* in
       composite fish tissue samples from the Columbia River Basin	2-29

Table 2-7. Basin-wide average concentrations of the sum of dioxin-like PCB congeners in
       composite fish  samples from the Columbia River Basin, 1996-1998	2-33

Table 2-8.  Basin-wide average concentrations of the sum of chlorinated dioxins and furans in
       composite fish  samples from the Columbia River Basin, 1996-1998.  	2-36

Table 2-9a. Basin-wide concentrations of 2,3,7,8-TCDF in composite samples offish tissue
       from the Columbia River Basin, 1996-1998	2-38

Table 2-9b. Basin-wide concentrations of 2,3,7,8-TCDF in composite  samples of eggs from
       anadromous fish species in the Columbia River Basin, 1996-1998.   	2-38

Table 2-10. Toxicity Equivalence Factors (TEF) for dioxin-like PCB congeners, dioxins, and
       furans (from Van den Berg et al,  1998)	2-39

Table 2-11. Basin-wide average concentrations of the toxicity equivalence concentrations for
       composite fish  samples from the Columbia River Basin, 1996-1998	2-40

Table 2-12. Basin-wide maximum concentrations * of metals in composite fish tissues measured
       in the Columbian River Basin, 1996 -1998.  	2-41

Table 2-13. Basin-wide average concentrations of metals in samples of eggs  from anadromous
       fish collected in the Columbia River Basin, 1996-1998	2-42

                                            vii

-------
Table 2-14. Basin-wide average concentrations of metals in composite samples offish from the
       Columbia River Basin, 1996-1998.  	2-43

Table 4-1. Exposure parameters used to calculate average daily dose for assessing
       noncarcinogenic health effects for potentially exposed populations	4-55

Table 4-2. Exposure parameters used to calculate average daily dose for assessing carcinogenic
       risks for potentially exposed populations	4-56

Table 4-3. Fish consumption rates expressed in alternative units	4-59

Table 4-4. Description of the methodology used  to calculate exposure for a multiple-species diet
         	4-64

Table 5-1. Chemicals without oral reference doses and cancer slope factors	5-65

Table 5-2. Chemicals contributing to non-cancer  hazard indices  	5-68

Table 5-3. Oral reference doses (RfDs) used in this assessment, including the level of
       confidence in the RfD, uncertainty factors (UF) and modifying factor (MF) used to
       develop the RfD, and the toxic effect(s) from the critical study that the RfD was based
       upon.	5-71

Table 5-4. EPA weight-of-evidence classifications for carcinogens	5-73

Table 5-5. Oral cancer slope factors with their weight of evidence classification with the type(s)
       of tumor the cancer slope factor is based  upon.	5-74

Table 5-6. Relative potency factors for PAHs  	5-76

Table 5-7a. Results of arsenic (As) analyses from Lower Columbia River Bi-State Water
       Quality Program  	5-79

Table 5-7b. Mean concentrations** of arsenic(As) in all fish species combined	5-79

Table 5-7c. Arithmetic means** of percent inorganic arsenic by species	5-79

Table 5-8. Summary of Willamette River, speciated arsenic data  	5-80

Table 6-1. Total hazard indices (HI) and endpoint specific hazard indices (at or greater than 1.0)
       for white sturgeon.  	6-89

Table 6-2. Comparison of Estimated Total Hazard Indices Among Adult Populations	6-92
                                             Vlll

-------
Table 6-3. Comparison of Estimated Total Hazard Indices Among Child Populations	6-93

Table 6-4. Chemicals having hazard quotients at or greater than 1.0 in white sturgeon	6-96

Table 6-5   Summary of ranges in endpoint specific hazard indices across study sites for adults
       who consume resident fish from the Columbia River Basin	6-98

Table 6-6. Percent contribution of contaminant groups to total non-cancer hazards for resident
       fish species. Based on Columbia River Basin-wide averages	6-100

Table 6-7   Summary of ranges in endpoint specific hazard indices across study sites for adults
       who consume anadromous fish species from the Columbia River Basin.   	6-106

Table 6-8. Percent contribution of contaminant groups to total non-cancer hazards for
       anadromous fish species. Based on Columbia River Basin-wide averages	6-108

Table 6-9. Summary of endpoint specific hazard indices and total hazard indices (by study site
       and basin-wide) for CRITIC'S tribal member adult, high fish consumption.  	6-113

Table 6-10.    Summary of total estimated cancer risks for white sturgeon	6-117

Table 6-11. Comparison of estimated total cancer risks among adult populations  	6-118

Table 6-12. Chemicals with estimated cancer risks at or greater than 1  X 10~5 for white sturgeon,
       fillet without skin	6-121

Table 6-13.  Chemicals with estimated cancer risks at or greater than 1 X 10~5 for white
       sturgeon, whole body.  	6-121

Table 6-14.   Summary of estimated total cancer risks by study site and basin-wide, resident fish
       species	6-122

 Table 6-15. Percent contribution of contaminant groups to estimated cancer risks for resident
       fish species	6-125

Table 6-16.   Summary of estimated total cancer risks by study site and basin-wide, anadromous
       fish species	6-130

Table 6-17. Percent contribution of contaminant groups to cancer risk for anadromous fish
       species	6-133

Table 6-18.  Summary of estimated total cancer risks by study site and basin-wide, all species.
        	 6-138

Table 6-19. Summary of Hazard Indices and Cancer Risks Across Study sites	6-141

                                             ix

-------
Table 6-20.  Summary of Hazard Indices and Cancer Risks Across Study sites	6-142

Table 6-21.   Summary of Hazard Indices and Cancer Risks Across Study sites	6-142

Table 6-22. Summary of Hazard Indices and Cancer Risks Across Study sites	6-143

Table 6-23. Comparison of site specific non-cancer hazard indices (for CRITFC's member tribal
       children) and cancer risks (for CRITFC's member tribal adults) from consuming whole
       body versus fillet for different fish species	6-144

Table 6-24. Estimate cancer risks and non-cancer health effects for a hypothetical multiple-
       species diet based upon CRITFC's member average adult fish consumption (CRITFC,
       1994)  	6-145

Table 6-25. Total hazard indices (His) for adults assuming that total arsenic is 1% versus 10%
       inorganic arsenic	6-148

Table 6-26. Estimated total cancer risks for adults assuming that total arsenic was 1% versus
       10% inorganic arsenic 70 years exposure	6-149

 Table 7-1.  Default  Input Parameters Used for the IEUBK Model Adapted from (USEPA, 1994b)
        	  7-155

Table 7-2. Input Parameters Used in the IEUBK Model Meat Consumption Rate by Age in the
       IEUBK model Adapted from (USEPA, 1994b)	7-155

Table 7-3. Fish Ingestion Rates (grams/day) Used to Assess Risk for Lead and other Chemicals
        	  7-156

Table 7-4. Percentages of Child Fish Consumption Rates for Consumers of Fish	7-157

Table 7-5. Input Parameters Used for the EPA Adult Lead Model  	7-160

Table 7-6. Adult Lead Model Baseline Blood Lead and Geometric Standard Deviations	7-160

Table 8-1. Composite risks for consumption offish contaminated with radionuclides from the
       Columbia River Basin for the general public and CRITFC's member Tribes	8-169

Table 9-1. Comparison of range concentrations of sum of DDE (o,p' & p.p') in whole body
       composite fish samples Columbia River Basin	9-173

Table 9-2. PCB residues in raw agricultural commodities, 1970-76	9-174

Table 9-3. The declining trends in PCBs in ready-to-eat foods collected in markets of a number

-------
       of US cities  	9-174

Table 9-4.  The 1976-80 ranges for PCB residues from 547 finfish from the Chesapeake Bay
       and its tributaries 	9-175

Table 9-5.  Total PCB concentrations in fish tissue from studies reported in the literature from
       1978-1994.	9-175

Table 9-6.  Concentrations Aroclor 1254 & 1260 in white croaker muscle tissue from California
       water bodies in the spring of 1994. (Source: Fairey	9-175

Table 9.7. Concentrations of Aroclor 1254 in lake trout from lakes in Michigan during 1978-82
        	 9-175

Table 9-8.  Aroclor concentrations in chinook salmon eggs reported for Lake Michigan,
       Michigan, compared to our study of Aroclors in the chinook salmon eggs	9-176

Table 9-9. Concentrations of Aroclors 1254 and 1260 in composite samples offish fillets from
       Lake Roosevelt, Washington compared concentrations measured in our study of the
       Columbia River Basin	9-176

Table 9-10.  Concentrations of 2,3,7,8-TCDF in composite samples offish fillets collected from
       Lake Roosevelt, Washington in 1994 compared with our 1996-1998  survey of the
       Columbia River Basin	9-178

Table 9-11. Lead concentrations in food purchased in five Canadian cities between 1986 -1988
        	 9-183

Table 9-12. British Columbia monitoring study of mercury concentrations in fish fillet tissue
        	 9-184

Table 9-13. EPA 1984  survey of total mercury concentrations  in edible fish tissue, shrimp, and
       prepared foods.  	9-185

Table 9-14. Mercury concentrations from an EPA 1990 - 1995 national survey offish fillets
        	 9-185

Table 9-15. USGS survey of mercury concentrations in fish tissue from reservoirs and streams
       in Northern California	9-186

Table 9-16. Mercury concentrations in fish fillets collected in Lake Whatcom and Lake
       Roosevelt, Washington compared to our study of the Columbia River Basin .  	9-186

Table 9-17. Selenium concentrations in US infant diet	9-188
                                            XI

-------
Table 9-18. Concentrations of selenium in fish reported in the literature	9-188

Table 9-19. Concentrations of zinc in food groups	9-190

Table 9-20a. Comparison of chemical concentrations in composites samples of whole body
largescale sucker	9-194

Table 9-20b .  Comparison of ranges of chemical concentration in composite samples of whole
       body bridgelip sucker	9-195

Table 9-21. Comparison of ranges chemical concentrations in composite samples of whole body
       mountain whitefish.  	9-196

Table 9-22. Comparison of ranges of chemical concentrations in whole body channel catfish
       tissue from our study with the USGS-NCBP database	9-198

Table 9-23. Comparison of ranges of chemical concentrations in whole body smallmouth bass.
        	 9-199

Table 9-24. Comparison of ranges of chemical concentrations in composite samples of whole
       body rainbow trout	9-200

Table 9-25. Comparison of chemical concentrations in chinook salmon fillet with skin	9-201

Table 9-26. Comparison of chemical concentrations in coho salmon fillet with skin	9-202

Table 9-27a. Range of chemical concentrations in resident fish tissue samples from our study of
       the Columbia River Basin, 1996-1998.   	9-206

Table 9-27b. Range of chemical concentrations (|ig/kg) in anadromous fish tissue samples from
       our study of the Columbia River Basin	9-207

Table 10-1 . Percent difference in field duplicate samples from the Columbia River Basin.
       Fish are listed with study site ID in parentheses	10-211

Table 10. 2. Comparison of estimated total cancer risks and hazard indices for a hypothetical
       multiple species diet using data from Table 17 and Table 18 in the CRITFC fish
       consumption report 	10-216

Table 10-3. Estimated Cancer Risks for PCBs Using Different Methods of Calculation	10-221

Table 10-4. Comparison of Hazard Indices for the Immunological Endpoint Based on
       Alternative Treatments of Aroclor Data.	10-222

 Table 10-5. Comparison of Hazard Quotients and Hazard Indices for the Hepatic Health

                                            xii

-------
Endpoint Based on Alternative Treatments of DDT, ODD, and DDE Data	10-223
                                   Xlll

-------
                                        List of Figures

Figure 1-1. Study sites in the Columbia River Basin	1-5

Figure 2-1. Basin-wide average percent lipid in fish collected  from the Columbia River Basin..  .  2-19

Figure 2-2. Basin-wide average concentrations of total pesticides in composite fish tissue
       collected from Columbia River Basin	2-21

Figure 2-3. Percent contribution of DDT structural analogs to total DDT concentration in
       whole body largescale sucker..	2-25

Figure 2-4a. Study site specific concentrations of p,p' DDE in white sturgeon individual fish
       tissue samples in the Columbia River Basin	2-27

Figure 2-4b.  Study site specific concentrations of p,p DDE in largescale sucker composite fish
       tissue samples from the Columbia River Basin.  	2-28

Figure 2-4c.  Study site specific concentrations of p,p DDE in mountain whitefish composite
       fish tissue samples from the Columbia River Basin	2-28

Figure 2-5a.  Study site concentrations of Aroclor 1254 in white sturgeon individual fish tissue
       samples from the Columbia River Basin	2-30

Figure 2-5b.  Study site specific concentrations of Aroclor 1260 in white sturgeon individual
       fish tissue samples from the Columbia River Basin.  	2-30

Figure 2-6a. Concentration of Aroclor 1254 in largescale sucker composite fish tissue samples
       from the Columbia River Basin	2-31

Figure 2-6b. Concentration of Aroclor 1260 in largescale sucker composite fish tissue samples
       from the Columbia River Basin.  	2-31

Figure 2-7a.  Concentration of Aroclor 1254 in mountain whitefish composite fish tissue
       samples from the Columbia River Basin	2-32

Figure 2-7b.  Concentration of Aroclor 1260 in mountain whitefish composite fish tissue
       samples from the Columbia River Basin	2-32

Figure 2-8a. Percent contribution of dioxin-like PCB congeners in mountain whitefish
       composite fillet samples from the Columbia River Basin.  	2-34

Figure 2-8b.  Percent contribution of dioxin-like PCB congeners in spring chinook salmon
       composite fillet samples from the Columbia River Basin	2-34

                                            xiv

-------
Figure 2-9. Study site average dioxin-like PCB congeners in white sturgeon and mountain
       whitefish samples from the Columbia River Basin.	2-35

Figure 2-10.  Correlation of basin-wide average concentrations of Aroclors  1242,1254,1260 (x
       axis) with dioxins like PCB congeners (y axis)	2-35

 Figure 2-11.   Study site average concentrations of chlorinated dioxins and furans in mountain
       whitefish, white sturgeon, and largescale sucker from study sites in the Columbia River
       Basin. Study sites are described in Table 1-1)	2-37

Figure 2-12. Percent contribution of each chlorinated dioxin and furan in largescale sucker.
       Basin-wide average of 23 composite whole body fish tissue samples.  	2-37

Figure 2-13a. Basin-wide average percent of individual metals in largescale sucker fillets.

Figure 2-13b. Basin-wide percent of individual metals in spring chinook salmon fillets	2-41

Figure 2-14a. Site specific concentrations of arsenic in white sturgeon individual fish tissue
       samples from the Columbia River Basin	2-45

Figure 2-14b.  Site specific concentration of arsenic in largescale sucker composite fish tissue
       samples from the Columbia River Basin.	2-46

Figure 2-14c. Site specific concentration of arsenic in mountain whitefish composite fish tissue
       samples from the Columbia River Basin	2-46

Figure 2-15 a. Study site concentrations of arsenic in spring chinook composite samples from the
       Columbia River Basin	2-47

Figure 2-15b.  Site specific concentrations of arsenic in steelhead composite fish tissue samples
       from the Columbia River Basin..	2-47

Figure 2-16a.  Site specific concentrations of mercury in white sturgeon fish tissue samples
       from the Columbia River Basin	2-49

Figure 2-16b.  Site specific concentrations of mercury in largescale sucker composite fish tissue
samples from the Columbia River Basin.	2-49

Figure 2-16c. Site specific concentrations of mercury in mountain whitefish composite fish
tissue samples from the Columbia River Basin	2-49

Figure 2-17a.  Site specific concentrations of mercury in spring chinook salmon composite fish
       tissue samples from the Columbia River Basin	2-49
                                              xv

-------
Figure 2-17b.  Site specific concentrations of mercury in steelhead composite fish tissue
       samples from the Columbia River Basin..   	2-49

Figure 6-1. Total hazard index versus fish consumption rate for adults	6-91

Figure 6-2a. Hazard indices for general public adults and children, average fish consumption
       rate of white sturgeon fillets.	6-95

Figure 6-2b. Hazard indices for CRITFC's member tribal adults and children, average fish
       consumption rate for white sturgeon fillets	6-95

Figure 6-2c. Hazard indices for general public adults and children, high fish consumption rate of
       white sturgeon fillets.   	6-95

Figure 6-2d. Hazard indices for CRITFC's member tribal adults and children, high fish
       consumption rate of white sturgeon fillets	6-95

Figure 6-3. Adult total non-cancer hazard indices for resident fish species* using basin-wide
       average data	6-99

Figure 6-4. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of white sturgeon fillet without skin.  	6-101

Figure 6-5. Percent contribution of basin-wide average chemical concentrations of non-cancer
       hazards from consumption of largescale sucker fillets with skin	6-101

Figure 6-6. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of whole body bridgelip sucker.	6-102

Figure 6-7.  Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of rainbow trout fillet with skin	6-102

Figure 6-8.  Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of walleye fillet with skin	6-103

Figure 6-9. Percent contribution of basin-wide chemical concentrations to non-cancer hazards
       from consumption of mountain whitefish fillet with skin.  	6-103

Figure 6.10 Adult total non-cancer indices for anadromous fish species	6-107

Figure 6-11. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of spring chinook fillet with skin	6-109

Figure 6-12. Percent contribution of basin-wide chemical concentrations to non-cancer hazards
       from consumption of coho salmon	6-109

                                             xvi

-------
Figure 6-13. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of fall chinook fillet with skin.  	6-110

Figure 6-14. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of steelhead fillet with skin	6-110

Figure 6-15. Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of Pacific lamprey fillet with skin	6-111

Figure 6-16.  Percent contribution of basin-wide average chemical concentrations to non-cancer
       hazards from consumption of whole body eulachon..	6-111

Figure 6-17. Adult total non-cancer hazard indices across all species	6-115

Figure 6-18. Comparison of estimated total cancer risks for consumption of white sturgeon
       across study sites for adults in the general public and CRITFC's member tribes at high
       consumption rates.  	6-119

Figure 6-19. Total cancer risks versus fish consumption rate for adults	6-120

Figure 6-20. Adult cancer risks for resident fish species	6-123

Figure 6-21. Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of white sturgeon fillet without skin	6-126

Figure 6-22.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of largescale  sucker fillet with skin..	6-126

Figure 6-23.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of whole body bridgelip sucker	6-127

Figure 6-24.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of rainbow trout fillet with skin	6-127

Figure 6-25.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of walleye fillet with skin.  	6-128

Figure 6-26.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of mountain whitefish fillet with skin	6-128

Figure 6-27. Adult cancer risks for anadromous fish species	6-131

Figure 6-28.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of spring chinook fillet with skin	6-134

                                            xvii

-------
Figure 6-29.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of coho salmon fillet with skin.   	6-134

Figure 6-30.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of fall chinook salmon fillet with skin	6-135

Figure 6-31.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of steelhead fillet with skin	6-135

Figure 6-32.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of Pacific lamprey fillet with skin..	6-136

Figure 6-33.  Percent contribution of basin-wide average chemical concentrations to cancer risk
       from consumption of whole body eulachon	6-136

Figure 6-34. Adult estimated total cancer risks across all fish species sampled	6-139

Figure 6-35. Adult total hazard indices for all  fish species, with multiple-species diet results.
       Basin-wide average data	6-146

Figure 6-36.  Adult cancer risks for all species, with multiple-species diet results.  Columbia
       River Basin-wide average chemical concentration data.  .  	6-146

Figure 6-37. Impact of percent inorganic arsenic on total hazard index	6-148

Figure 6-38. Impact of percent inorganic arsenic on cancer risks	6-150

Figure 7-1. Sample IEUBK Model for Lead Output Graph	7-154

Figure 7-2. Predicted blood lead levels for children who consume offish collected from the
       Columbia River Basin assuming fish is 16% of dietary meat.  	7-157

Figure 7-3. Predicted blood lead levels for children (0-72 months) who consume 101 g/day of
       fish collected from the Columbia River Basin, 1996-1998	7-158

Figure 7-4. Predicted fetal blood lead levels with maternal fish ingestion rate of 39.2 g/day with
       baseline blood lead level at 2.2 |ig/dl and GSD = 2.1 |ig/dl.  	7-161

Figure 7-5. Predicted fetal blood lead level with maternal fish ingestion rate of 39.2 g/day with
       baseline blood lead level at 1.7 |ig/dl and GSD = 1.8 |ig/dl	7-162
                                            xvui

-------
Appendix A. Study Design for Assessment of Chemical Contaminants in Fish Consumed by
Four Native American Tribes in the Columbia River Basin
Appendix B. Fish Live Histories
Appendix C. Toxicity Profiles
Appendix D. Summary Statistics by Basin, Tributary, and Site
Appendix E. Chemical Concentrations for Three Detection Limit Rules
Appendix F. Summary of Chemicals Not Detected
Appendix G. Non-cancer Hazard Quotients
Appendix H. Percent Contribution o Non-Cancer
Appendix I.  Cancer Risk Values
Appendix J.  Percent Contribution to Cancer Risk Values
Appendix K. Radionuclide Data
Appendix L. AED Pesticide Measurement Results
Appendix M. Hazard Indices Across Study Sites
Appendix N. Estimated Cancer Risks by Study Site
Appendix O. Summary of Risk Characterization Results for Resident Species
Appendix P.  Summary of Risk Characterization Results for Anadromous Species

Volume 2. Data Appendices
Volume 3. Fish Collection and Processing Forms
Volume 4. Quality Assurance Summary to the Project Final Report
Volume 5. Quality Assurance Project Plan
                                           xrx

-------
                                ACKNOWLEDGMENTS
The authors would like to acknowledge the vision of the Columbia River Intertribal Fish
Commission, the Yakama Nation, Nez Perce Tribe, Confederated Tribes of the Umatilla Indian
Reservation, Confederated Tribes of the Warm Springs Reservation, and Craig McCormack
(Washington Department of Ecology, formerly with EPA) who saw a need to establish a clear
understanding of the presence of toxic chemicals in fish consumed from the Columbia River
Basin.  Their persistence and dedication resulted in a commitment by EPA to complete this
study.

The staff and directors of the Columbia River Intertribal Fish Commission (Ted Strong,  Anne
Watanabe, Don Sampson, Paul Lumley, in., Cat Black) are acknowledged for their persistence
and commitment beginning with the Columbia River Basin Fish Consumption Survey (CRITFC,
1994) to the completion of this contaminant survey.

EPA is grateful to the many Tribal members who made the fish sampling a major success.  As a
result of the tremendous help and work of Tribal members and a dedicated EPA/Tribal  sampling
team, the overall objectives of the project  were accomplished.  We want to thank the following
for their help in the field sampling:

       Yakama Nation sampling crews: Eugene Billy, Seymore Billy, Steve Blodgett, Bill
       Bosch, Jim Dunnigan, Ernst Edwards, Lillian Eneas, Mike George, Gina George, Lee
       Hannigan, Isaiah Hogan, Joe Hoptowit, Cecil James, Jr., Joe Jay Pinkham HI, Jamie Jim,
       Mark Johnston, James Kiona, Linda Lamebull, George Lee, Beverly Logie, Bobbi
       Looney, Jr., Donella Miller, Manard Only, Steve Parker, Lee Roy Senator, Brian
       Saluskin, Vernon Smartlowit, Greg Strom, Ceilia Walsey-Begay, Earl Wesley

       Confederated Tribes of the Warm Springs Reservation: Chris Brun, Mark Fritsch, Jim
       Griggs, Mick Jennings, Terry Luther, Patty O'Toole, Stanley Simtustus.

       Confederated Tribes of the Umatilla Indian Reservation: Brian Conner, Craig Contor,
       Gary James, Mike Jones, Jerry Rowan, Vern Spencer, Brian Zimmerman,

       Nez Perce Tribe; John Gibhards, Jay Hesse, Nancy Hoefs, Paul Kucra, Ed Larsen,
       Donna Powaukee,

       EPA Sample Collectors and Field Support: Robert Athmann, Tom Davis, Andy Hess,
       Duane Kama, Andy Osterhaus, Doc Thompson, Philip Wong

       Hatcheries Field Support:
       Dworshak National Fish Hatchery  (USFWS) Bill Miller, Bob  Semple; Oxbow Hatchery
       (IDFG) Julie Hislop;    Looking Glass Hatchery (ODFW) Bob Lund; Little White Salmon

                                            xx

-------
       National Hatchery  (USFWS) Speros Doulos; Dexter Hatchery  (ODFW) Gary Jaeger,
       Tim Wright; Leavenworth National Fish Hatchery  (USFWS) Corky Broaddus, Dan
       Davis, Greg Pratschner; Carson National Fish Hatchery  (USFWS) Bruce McCloud;
       Klickitat River Hatchery (WDFW)Ed Anderson; Priest Rapids Hatchery (WDFW) Dan
       Bozorth, Paul Peterson;

We want to especially thank Lynn Hatcher of the Yakama Nation for his encouragement of tribal
staff to participate in the field sampling.

The radionuclide analyses could not have been completed without the help of EPA's National
Air and Radiation Laboratory (John Griggs, Tonya Hudson, David Saunders).
EPA Region 10's work on this project was facilitated by the project Sample Control Manager
Melody Walker as well as the Administrative staff (Mary Moore, Lorraine Swierkos, Sharon
Buza, Ofelia Erickson).

EPA Office of Water (Jeff Bigler, Elizabeth Southerland, LeAnn Stahl, Tom Armitage) were
responsible for the initial study design and finding the funds for the study. Bill Telliard is
recognized for his development of PCB congener Method 1668.

The assistance of the following Region 10 staff was invaluable. Mary Lou Soscia helped move
the project into action through her facilitation of the completion of the Memoradum of
Agreement with the tribes and EPA. Kellie Kubena is thanked for her work on the background
material on chemicals in fish species.  Ray Peterson, Matt Gubitosa, Don Matheny, Peter
Leinenbach helped  to prepare the background maps for the project.  Carol Harrison is especially
appreciated for the work she did in putting the document together and for her patience with our
frequent last minute modifications.  Thanks to Rob Pedersen for his help in preparing the data
appendices. Thanks to Lon Kissinger for his graphic analysis of the chemical data. We are
deeply indebted to Ravi Sanga for all the work he did in completing this report. We  also thank
Tom Lewandoski for his help on the toxicological profiles. Duane Kama, Michael Watson, and
Roseanne Lorenzana are especially appreciated for their peer review of this report.

The Multi-Agency Task Force is acknowledged for their help in designing the study.
                                            xxi

-------
EPA Region 10
       Patricia Cirone
       Dana Davoli
       Duane Kama
       Robert Melton
       Rick Poeton
       Marc Stifelman
       Dave Terpening
       Michael Watson
       Peggy Knight
       Katherine Adams
       Roy Araki
       Isa Chamberlain
       Randy  Cummings
       Gerald Dodo
       Stephanie Le
       Kathy  Parker
       Steve Reimer
       Bob Rieck
       Tony Morris
       Tobi Braverman
CONTRIBUTORS
            Steve Ellis
            MCS Environmental, Inc

            EVS Consultants
            Seattle, Washington

            TetraTech Consultants
            Bellevue, Washington
                                          xxu

-------
COLUMBIA RIVER BASIN FISH TISSUE CONTAMINANT STUDY WORKGROUP
Anne Watanabe
Columbia River Intertribal Fish Commission
(formerly)

Paul Lumley HI
Columbia River Intertribal Fish Commission

Catriona (Cat) Black
Columbia River Intertribal Fish Commission

Barbara Harper
Yakama Nation (formerly)

Lynn Hatcher
Yakama Nation

Chris Walsh
Yakama Nation Health Center
Nancy Hoefs
Nez Perce Fisheries

Parti Howard
Nez Perce Tribe (formerly)

Gary James
Confederated Tribes of the Umatilla Indian
Reservation

Stuart Harris
Confederated Tribes of the Umatilla Indian
Reservation

Patty O'Toole
Confederated Tribes of the Warm Springs
Reservation
Silas Whitman
Nez Perce Tribe (formerly)

Rick Eichsteadt
Nez Perce Tribe
                                         XXlll

-------
                     LIST OF ABBREVIATIONS AND ACRONYMS

ADD               average daily dose of a specific chemical (mg/kg-day)
AFC                average fish consumption
AIM               EPA Adult Lead Model
AT                 averaging time for exposure duration (days)
ATSDR             Agency for Toxic Substances and Disease Registry
AVE                average
BCF                bioconcentration factor
BEIR               Biological Effects of Ionizing Radiation
BEST               Biomonitoring of Environmental Status and Trends
BKSF         biokinetic slope factor
BW                 body weight
C                   chemical concentration in fish tissue
CDC                Centers for Disease Control
CF                 conversion factor
CSFII               Continuing Survey of Food Intake by Individuals
CSFs                cancer slope factors
CRITFC            Columbia River Intertribal Fish Commission
DDE                1,1 -dichloro-2,2- to(p-chlorophenyl)ethylene
DDT                1,1,1 -trichloro-2,2-te(p-cMorophenyl)ethane
ODD               1,1 -dichloro-2,2- to(p-chlorophenyl)ethane
DDMU             1,1- to(p-chlorophenyl)2 chloro-ethylene
DF                 detection frequency
DMA               dimethyarsenic
EF                 exposure frequency (days/year)
ED                 Exposure duration (years)
ECRnew             Excess cancer risk for the new exposure duration
ECRvo               Excess cancer risk estimate for a lifetime exposure duration of 70 years
ED new               Individual exposure duration in years
EDvo                Default lifetime exposure duration of 70 years
EPA                Environmental Protection Agency
FS                 fillet with skin
FW                 fillet without skin
GC/AED            Gas Chromatograph/Atomic Emission Detector
GSD                Geometric Standard Deviation
GPS                global positioning system
HEAST             Health Effects Assessment Summary Tables
HFC                high fish consumption
HI                 hazard index
HQ                 hazard quotient
IEUBK             EPA integrated exposure uptake biokinetic model
IR                 ingestion rate
LLD                lower limit of detection
LOAEL             lowest observed adverse effect level
                                           xxiv

-------
MAX
MDC
MF
MIN
MMA
NA
NAERL
NCEA
NCBP
NCRP
ND
NOAEL
NS
OCDD
OERR
 PAHs
PCBs
PSAMP
RfD
RPFs
2,3,7,8-TCDD
2,3,7,8 TCDF
TEC
TEF
TRW
UF
WB
USEPA
USGS
                     maximum
                     minimum detectable concentration
                     modifying factor
                     minimum
                     monomethylarsenic
                     not applicable
                     National Air and Radiation Environmental Laboratory
                     National Center for Environmental Assessment
                     National Contaminant Biomonitoring Program
                     National Council on Radiation Protection and Measurements
                     not detected
                     no observable adverse effect level
                     not sampled
                     Octachlorodibenzo-p-dioxin
                     EPA Office of Emergency and Remedial Response
                     polycyclic aromatic hydrocarbons
                     polychlorinated biphenyls
                     Puget Sound Ambient Monitoring Program
                     reference dose
                     relative potency factors
                     2,3,7,8 -tetrachlorodibenzo-^-dioxin
                     2,3,7,8 tetrachlorodibenzo-^-furan
                     toxicity equivalence concentration
                     toxicity equivalence factor
                     EPA Technical Review Workgroup for Lead
                     uncertainty factors
                     whole body
                     U.S. Environmental Protection Agency
                     United States Geological Survey
Units
ng/kg
Hg/dl
jig/kg

g/day
mg/kg
kg
kg/g
                                                 Bq
              nanograms per kilogram (ppt)
              micrograms per deciliter
              micrograms per kilogram
one radioactive disintegration
per second
              grams per day
              milligram per kilogram (ppm)
              kilogram
              kilogram per gram
mg/kg-day    milligram per kilogram-day
                                           xxv

-------
XXVI

-------
                                EXECUTIVE SUMMARY
Introduction

This report presents the results of an assessment of chemical pollutants in fish and the potential
risks from consuming these fish.  The fish were collected throughout the Columbia River Basin in
Washington, Oregon, and Idaho.

After reviewing the results of the U. S. Environmental Protection Agency        Map of Columbia River
(USEPA. 1992a) 1989 national survey of pollutants in fish in the United        Basin
States, EPA became concerned about the potential health threat to Native
Americans who consume fish from the Columbia River Basin. The
Columbia River Intertribal Fish Commission (CRITFC) and its member
tribes  (Warm Springs Tribe, Yakama Nation, Umatilla Confederated
Tribes, Nez Perce Tribe) were also concerned for tribal members who
consume more fish than non-Indians.

In order to evaluate the likelihood that tribal people may be exposed to high levels of
contaminants in fish tissue EPA, CRITFC and its member tribes, designed a study in two phases.
The first phase was a fish consumption survey which was conducted by the staff of CRITFC and
its member tribes. The fish consumption survey was completed in 1994 (CRITFC 1994). The
conclusions of the tribal survey were:

       "The rates of tribal members' consumption across gender, age groups,
       persons who live on- vs. off-reservation, fish consumers only, seasons,
       nursing mothers, fishers,  and non-fishers range from 6 to 11 times higher
       than  the national estimate used by USEPA."(quote from CRITFC, 1994,
       Page 59)

The results of the fish consumption survey accentuated the need to complete an assessment of
chemicals in the fish being consumed by CRITFC's member tribes.

In 1994, EPA and CRITFC's member tribes initiated the second phase of the study  which was a
survey of contaminants in fish tissue in the Columbia River Basin and the subject of this report.
The contaminant survey was designed by a multi-agency group including CRITFC, Washington
Departments of Ecology and  Health, Oregon Departments of Environmental Quality and Health,
the Confederated Tribes of Warm  Springs, the Yakama Nation, the Umatilla Confederated Tribes,
the Nez  Perce Tribe, U.S. Geological Survey, and U.S. Fish and Wildlife Service. Sample
collection took place between 1996 and 1998 with the help of CRITFC's member tribes and staff
of federal and state agencies.  Chemical analyses were completed in 1999. The analyses were
done by EPA and commercial laboratories.

While the study was initiated because of concern for Native American tribes, the results are

                                            E-l

-------
important to all people who consume fish from the Columbia River Basin.
This study provided EPA with information to determine:

               1)     if fish were contaminated with toxic chemicals,

               2)     the difference in chemical concentrations among fish species and study
                     sites, and

               3)     the potential human health risks due to consumption offish from the
                     Columbia River Basin.

The results of this survey provided information on those chemicals which were most likely to be
accumulated in  fish tissue and therefore posed the greatest potential risks to people. These are the
chemicals for which regulatory strategies need to be defined to reduce these chemicals in our
environment.

This study was not designed to evaluate:

       1)      health of past or future generations of people who consume fish from the
               Columbia River Basin,

       2)      rates of disease in tribal communities,

       3)      specific sources of chemicals,

       4)      multiple exposures to chemicals from air, water, and soil,

       5)      food other than fish, and

       6)      risks for a specific tribe or individual.

It is our hope that the results of this survey will be used by CRITFC's member tribes as well as
others to  more completely evaluate and protect the quality of the fishery resource.

Study Design

This study was designed to estimate risks for a specific group of people (CRITFC's member
tribes). Therefore, the sample location, fish species, tissue type, and chemicals were not
randomly selected.  Collection sites were selected because they were important to  characterizing
risks to CRITFC's member tribes. Chemicals were chosen because they were identified in other
fish tissue surveys of the Columbia River Basin as well as being found throughout the
environment.

This type of sampling is biased with unequal sample sizes and predetermined sample locations
rather random.  This bias is to be expected when attempting to provide information for

                                             E-2

-------
individuals or groups based on their  preferences.  The results of this survey should not be
extrapolated to any other fish or fish  from other locations.

A total of 281 samples offish and fish eggs were collected from the Columbia River Basin. The
fish species included five anadromous species (Pacific lamprey, smelt, coho salmon, fall and
spring chinook salmon, steelhead) and six resident species (largescale sucker, bridgelip sucker,
mountain whitefish, rainbow trout, white sturgeon, walleye). Four types of samples were
collected: whole-body with scales, fillet with skin and scales, fillet without skin (white sturgeon
only), and eggs. The fillets were all with skin except for the white sturgeon. The armor-like skin
of the white sturgeon is considered too tough for ingestion. All the samples were composites of
individual fish, except white sturgeon. The white sturgeon were analyzed as single fish instead of
composites because of their large size. The number offish in a composite varied with species,
location, and tissue type. Eleven  samples of eggs were collected from steelhead and salmon.  Due
to availability offish, limitation in time and funds,  certain species were not sampled as frequently
as others. In particular, the bridgelip sucker, coho salmon, and eulachon were collected at only
one location. Pacific lamprey and walleye were collected at only two locations.  The type of
tissue tested (whole body, fillet, egg) varied with species and sample location.

Three replicate samples for each fish type were collected from a total of 24 study sites.  These
sites were located on 16 rivers and creeks, including, Hood River, Little White Salmon River,
Wind River, Fifteen Mile Creek, Wenatchee River, Willamette River, Deschutes River, Umatilla
River, Thomas Creek, Meacham Creek, Klickitat River, Yakima River, Snake River, Clearwater
River, Looking Glass Creek, and  the mainstream Columbia River. Different species were
collected from each site depending upon the fishing practices of CRITFC's member tribes.
Despite these many variables, general trends in the monitoring  of pollutants in these various
species  and tissues were evident.

The fish tissues were analyzed for 132 chemicals including 26 pesticides, 18 metals, 7 PCB
Aroclors, 13 dioxin-like PCBs, 7  dioxin congeners, 10 furan congeners, and 51 miscellaneous
organic chemicals.  Of these 132 chemicals, 92 were detected.  The most frequently detected
chemicals in fish tissue were 14 metals, DDT and its structural analogs (ODD, DDE), chlordane
and related compounds (c/s-chlordane, trans-chlordane, c/'s-nonachlor, ^raws-nonachlor, and
oxychlordane), PCBs (Aroclors1  and dioxin-like PCBs), and chlorinated dioxin and furans.

Results

The fish tissue chemical concentrations were evaluated for each study site and for the whole
basin.  The results of the study showed that all species offish had some levels of toxic chemicals
in their tissues and in the eggs of chinook and coho salmon and steelhead. The fish tissue
chemical concentrations were variable within fish (duplicate fillets), across tissue type (whole
body and fillet), across species, and study sites. However, the chemical residues exhibited some
        1 Aroclors = commercial formulation of mixtures of PCB congeners; Aroclors 1242, 1254, and 1260 were
 the only aroclors detected in fish tissue in our study

                                              E-3

-------
trends in distribution across species and locations.  The concentration of organic chemicals in the
salmonids (chinook and coho salmon, rainbow and steelhead trout) and eulachon were lower than
any other species. The concentrations of organic chemicals in three species (white sturgeon,
mountain whitefish, largescale sucker) and Pacific lamprey  were higher than any other species.
The concentrations of metals were more variable, with maximum levels of occurring in different
species.

Of the  132 chemicals analyzed in this study, DDE,  Aroclors, zinc, and aluminum were detected
in the highest concentration in most of the fish tissues sampled throughout the basin.  The basin-
wide average concentrations for for the organic chemicals (DDE, Aroclors, chlorinated dioxins
and furans) ranged from non-detectable in the anadromous fish species to the highest levels in
resident species.  DDE, the most commonly found pesticide in fish tissue from our study, ranged
from a basin- wide average of 11 ppb2 in whole body eulachon to 620 ppb in whole body white
sturgeon.  The  sum of Aroclors ranged from non-detectable in eulachon to 190 ppb in mountain
whitefish fillets, sturgeon.  Chlorinated dioxins and furans were found at low concentrations in
fish species. The basin-wide average concentration of the sum of chlorinated dioxins and furans
ranged from 0.0001 ppb in the walleye, largescale sucker, coho, and steelhead fillets, fall
chinook salmon (whole body, fillet, egg) and steelhead eggs to 0.03 ppb in whole body white
sturgeon.

The concentration of metals did not show a distinct difference between anadromous and resident
fish species. The basin-wide average concentrations of arsenic ranged from non-detectable in
rainbow trout fillet to 890 ppb in whole body eulachon. Mercury ranged from non-detectable
levels in Pacific lamprey fillets and whole body eulachon to 240 ppb in largescale sucker.

The distribution across stations was variable although fish collected from the Hanford Reach of
the Columbia River and the Yakima River tended to have higher concentrations of organic
chemicals than other study sites.

The chemical concentrations in fish species measured in this study were generally lower than
levels reported in the literature from the early 1970's and similar to levels reported in the  late
1980's to the present.  The literature included studies from the Columbia River Basin as well as
other water bodies in the United States.
         ppb = parts per billion = ng/kg

                                             E-4

-------
EPA uses a risk model to characterize the possible
health effects associated with chemical exposure.
For this model, toxicity information is combined
with estimates of exposure to characterize cancer
risks and non-cancer health effects. Toxicity
information (reference doses and cancer slope
factors) used in this study was obtained from
USEPA databases.
               EPA's Risk Assessment Model
                                                                      RISK
                                                                - Increase in Cancer Risk
                                                               L - Non-Cancer Health Effects
 The EPA method to estimate exposure to chemicals in fish depends upon the chemical
 concentration in the fish tissue, the amount and types of fish eaten, how long and how often fish
 is eaten, and the body weight of the person eating the fish. For this assessment, exposures to
 chemicals were estimated for both adults and children of CRITFC's member tribes and the
 general population. In addition to estimating exposure for each site, exposures were also
 estimated for the basin wide average offish tissue.  In estimating these exposures, it was assumed
 that a person eats the same type offish for their lifetime.
                                                 Average and high (99th percentile) fish consumption rates
                                                 for CRITFC's member tribes and the general public.
Different fish ingestion rates were used
for the general public and for CRITFC's
member tribes. Fish consumption rates
for CRITFC's member tribes were based
upon data from the CRITFC fish
consumption survey (CRITFC, 1994)
while those for the general public were
based upon EPA analysis of national fish
consumption rates (USEPA, 2000b).
  400-

  350-

  300-

-g 250-|

^200-1
CO
I 150-|
D)
  100-

   50-
                 High
              Consumption
                 Rate
  Average
Consumption
   Rate
                                                    general   CRITFC's   general    CRITFC's
                                                     public    member    public    member
                                                              tribes               tribes
In conducting a risk assessment, EPA evaluates the potential for developing non-cancer health
effects such as immunological, reproductive, developmental, or nervous system disorders and for
increased cancer risk. Different methods are used to estimate non-cancer health effects and
cancer risks.
For non-cancer health effects, EPA assumes that a threshold of exposure exists below which

                                              E-5

-------
health effects are unlikely. To estimate non-cancer health effects, the estimated lifetime average
daily dose of a chemical is compared to its reference dose (RfD). The reference dose represents an
estimate of a daily exposure level that is likely to be without deleterious effects in a lifetime.  The
ratio of the exposure level in humans to the reference dose is called a  hazard quotient. To
account for the fact that fish contained multiple chemicals, the hazard quotients for the chemicals
which cause similar health effects were added to calculate a single hazard index for each type of
health effect. For exposures resulting in hazard indices equal to or less than one, health impacts
are unlikely. Generally, the higher hazard index is above one, the greater the level of concern for
health effects.

For cancer,  EPA assumes that any exposure to a carcinogen may increase the probability of
getting cancer.  Thus, the risk from exposure to a carcinogen is estimated as the increase in the
probability or chance  of developing  cancer over a lifetime as a result of exposure to that chemical
(e.g. an increased chance of 1 in 10,000). Cancer risks, which are calculated for adults only, are
estimated by multiplying the lifetime average daily intake of a chemical  by its cancer slope
factor. The estimated cancer risk from exposure to a mixture of carcinogens is estimated by
adding the cancer risks for each chemical in a mixture.  The cancer risk estimates which are based
on EPA's methodology are considered to be upper-bound estimates of risk or the most health-
protective estimate.  Due to our uncertainty in understanding the biological mechanisms which
cause cancer, the true risks may in fact be substantially lower than the number estimated with
EPA's risk assessment model.

In interpreting cancer risks, different federal and state agencies often have different levels of
concern for cancer risks based upon their laws and regulations.  EPA has not defined a level of
concern for cancer. However, regulatory actions are often taken when the probability of risk of
cancer is within the range of 1 in 1,000,000 to 1 in 10,000. Risk managers make their decisions
regarding which level within this range is a concern depending on the circumstances of the
particular exposure(s). A level of concern for cancer risk has not been defined for this risk
assessment.

Using EPA's risk assessment models, hazard indices and cancer risks  were estimated for people
who consume resident and anadromous fish from the whole Columbia River Basin and from each
study site in the basin. For adults, hazard indices and cancer risks were lowest for the general
public at the average ingestion rate and highest for CRITFC's member tribes at the high ingestion
rate.  For adults in the general public with an average fish ingestion rate of about a meal3 per
month (7.5 g/day), hazard indices were less than 1 and cancer risks were less than 1 in 10,000'
except for a few of the more  highly contaminated samples of mountain  whitefish and white
sturgeon.  For adults in CRITFC's member tribes,  at the highest fish ingestion rate at about 48
meals1 per month (389 g/day), hazard indices were greater than 1 for several species at some sites.
Hazard indices (less than or equal to 8 at most sites) and cancer risks  (7 in 10,000 to 2 in 1,000)
were lowest for salmon, steelhead, eulachon and rainbow trout and highest (hazard indices greater
than 100 and cancer risks up to 2 in 100 at some sites) for mountain whitefish and white sturgeon.
         3Meal = eight ounces offish
                                              E-f

-------
For the general public, the hazard indices for children at the average fish ingestion rate were less
for adults (0.9) at the average ingestion rate; the hazard indices for children at the high ingestion
rate were 1.3 times greater than those for adults at the high ingestion rate.  For CRITFC's member
tribes, the hazard indices for children at the average and high ingestion rates were 1.9 times
greater than those for adults in CRITFC's member tribes at the average and high ingestion rates,
respectively.

For both resident and anadromous species, the major contributors to the hazard indices were
PCBs (Aroclors) and mercury.  DDT and its structural analogs were also important contributors
for some resident species.  The chemicals and or chemical classes that contributed the most to
cancer risk for most of the resident fish were PCBs (Aroclors and dioxin-like PCBs), chlorinated
dioxins and furans, and a limited number of pesticides.  For most of the anadromous fish, the
chemicals that contributed the most to cancer risk were PCBs (Aroclors and dioxin-like PCBs),
chlorinated dioxins and furans, and arsenic.

In estimating hazard indices and cancer risks for people who eat a certain fish species, it is
assumed that they eat only that type offish for their lifetime. However, many people eat a variety
offish over a lifetime.  Hazard indices and cancer risks were also estimated using a hypothetical
multiple species diet. This hypothetical multiple species diet was based upon information from
the CRITFC fish consumption study (CRITFC, 1994).  The hazard indices and cancer risks for
the multiple species diet were lower than those for most contaminated  species offish and greater
than those for some of the least contaminated species. The risks for eating one type offish may
be an over or underestimate of the risks for consumers of a multiple-species diet depending upon
the types offish and concentration of chemicals in the fish which make up the diet.

The risk assessment model for assessing exposure to lead is different from  other chemicals.  Lead
risk is based on a bio-kinetic model which includes all routes of exposure (ingestion of food, soil,
water, and inhalation of dust).  Based on EPA's risk assessment model, the lead concentrations in
Columbia River Basin fish tissues were estimated to be unlikely to cause a human blood lead
level greater than 10 |ig/dl. The blood lead level of 10 jig/dl is the national  level of concern for
young children and fetuses (CDC, 1991).

In addition to the survey of the basin for the 131 chemicals, a special study of radionuclides was
completed for a limited number of samples. White sturgeon were collected from the Hanford
Reach of the Columbia River, artificial ponds on the Hanford Reservation,  and from the upper
Snake River and analyzed for radionuclides. The levels of radionculides in  fish tissue from
Hanford Reach of the Columbia River and the ponds on the Hanford Reservation were similar to
levels in fish from the Snake River.  Cancer risks were estimated for consumption offish which
were contaminated with radionuclides.  These risks estimates were not combined with the
potential risks from other chemicals at these study sites. The potential cancer risks from
consuming fish collected from Hanford Reach and the artificial ponds on the Hanford Reservation
were similar to cancer risks in fish collected from the upper Snake River.
                                             E-7

-------
Conclusions

The concentration of toxic chemicals found in fish from the Columbia River Basin may be a risk
to the health of people who eat them depending on:

       1)     the toxicity of the chemicals,

       2)     the concentration in the fish,

       3)     the species and tissue type of the fish, and

       3)      how much and how often fish is consumed
The chemicals which contribute the most to the hazard indices and cancer risks are the persistent
bioaccumulative chemicals (PCBs, DDE, chlorinated dioxins and furans) as well as some
naturally occurring chemicals (arsenic, mercury).  Some pollutants persist in the food chain
largely due to past practices in the United States and global dispersion from outside North
America. Although some of these chemicals
are no longer allowed to be used in the
United States, a survey of the literature
indicates that these chemical residues
continue to accumulate in a variety of foods
including fish.  Human activities can alter
the distribution of the naturally occurring
metals (e.g. mining, fuel combustion) and
Recommendations for eating fish
 EPA recommends that people follow the
general advice provided by the health
departments for preparing and cooking
fish;
thus increase the likelihood of exposure to
toxic levels of these chemicals through
inhalation or ingestion of food and water.

Many of the chemical residues in fish
identified in this study are not unlike levels
found in fish from other studies in
comparable aquatic environments in North
America. The concern raised in the
Columbia River Basin also gives rise to a
much broader issue for water bodies
throughout the United States. The results of
this study, therefore, have implications not
only for tribal members but also the general
public.

While contaminants remain in fish, it is
useful for people to consider ways to still
derive beneficial effects  of eating fish, while
*Remove fat and skin before cooking

*While cooking, allow fat and oil to
drain

These preparation and cooking methods
should help to reduce exposures to PCBs,
DDTs, dioxins, and furans, and other
organics which accumulate in the fatty
tissues offish.
       Note: It is also important to
consider the health benefits of eating fish.
While fish accumulate chemicals from the
environment they are also an excellent
source of protein that is low in saturated
fats, rich in vitamin D and omega-3 fatty
acids, as well as other nutrients.
                                             E-S

-------
at the same time reducing exposure to these chemicals. Fish are a good source of protein, low in
saturated fats, and contain oils which may prevent coronary heart disease.  Risks can be reduced
by decreasing the amount offish consumed, by preparing and cooking fish to reduce contaminant
levels, or by selecting fish species which tend to have lower concentrations  of contaminants.

The results of this study confirm the need for regulatory agencies to continue to pursue rigorous
controls on environmental pollutants and to continue to significantly reduce those pollutants
which have been dispersed into our ecosystems. Reducing dietary exposure through cooking or
by eating a variety of fish will not eliminate these chemicals from the environment. Elimination
of many of the man-made chemicals from the environment will take decades to centuries.
Regulatory limits for new waste streams and clean up of existing sources of chemical wastes can
help to reduce exposure. The exposure to naturally occurring chemicals can be reduced through
better management of our natural resources.

There are many uncertainties in this risk assessment which could result in alternate estimates of
risk.  These uncertainties include our limited knowledge of the mechanisms which cause disease,
the variability of contaminants in fish and fish ingestion rates, and the effects of food preparation.
The uncertainties in our estimates may increase or decrease the risk estimates reported in this
study.
                                             E-9

-------

-------
1.0    Introduction

1.1    Report Organization

This report presents the results of an assessment of chemicals in fish and the risk estimates from
consuming these fish based on data analysis and conclusions reached by EPA. It is organized into
five volumes.

The study results are presented in 10 sections in Volume 1.  Sections 1 and 2 describe the study
background, methods, and the chemical concentrations in fish tissues.  Sections 3,4, and 5
describe risk assessment methods. The risk characterization is presented in Section 6 for all
chemicals except lead and radionuclides. Lead and radionuclide risk characterizations are
presented in sections 7, and 8, respectively.  The fish tissue residues from this study are compared
to other fish contaminant studies as well as other food types in Section 9. Uncertainties in this
study are presented in Section 10.  The discussion of uncertainty includes all aspects of the risk
assessment as well as the sections on fish tissue concentrations (Section 2) and the comparisons
with other studies (Section 9). The uncertainty section contains additional calculations to show
how the characterization of cancer risk and non-cancer hazards would change if different values
had been used to estimate exposure or to characterize toxicity. Finally, conclusions for this study
are discussed in Section 11.

Volume 2 provides all the chemical data from the results of the study, as well as sex, length and
weight of the fish, and other descriptive data on fish collection.  Volume 3 is the Field Operations
Manager sampler's notebook(s) which provides a record for the collection of samples. Volume 4
is the Quality Assurance Report which includes a review of the field activities, sample
preparation, laboratory measurements, quality assurance procedures, system audits, corrective
actions, and the data quality assessment.  The appendices to this volume contain all the project
data including information about the field sampling locations. Volume 5 is the Quality Assurance
Project Plan which was prepared in  1996. The Quality Assurance Project Plan contains the
documentation for the study design, objectives, methods, and quality control procedures.

1.2    Study Background

After reviewing the results of the EPA 1989 national survey of pollutants in fish (USEPA,
1992a), EPA became concerned about the potential health threat to Native Americans who
consume large amounts offish from the Columbia River Basin.  The cause for concern for native
peoples in the Columbia River Basin was also raised by the Columbia River Intertribal Fish
Commission (CRITFC) and its member tribes4.

In order to evaluate the likelihood that tribal people  may be exposed to high levels of
        4A11 references to "tribes" in this report are only applicable to CRITFC's member tribes: Confederated
 Tribes of Warm Springs, Yakama Nation, Umatilla Confederated Tribes, Nez Perce Tribe. They are collectively
 referred to as CRITFC's member tribes.

                                              1-1

-------
contaminants in fish tissue EPA, CRITFC and its member tribes designed a study in two phases.
The first phase of this study was a fish consumption survey which was completed in 1994 by
CRITFC (CRITFC, 1994). The results of this survey documented the importance offish in the
diet and culture of CRITFC's member tribes. The types and amounts offish that were eaten by
the four CRITFC's member tribes were identified. The primary fish that were consumed by
CRITFC's member tribes were salmon and trout.  The survey also demonstrated that the average
daily fish consumption for adults (63.2 g/day) of CRITFC's member tribes was much higher than
the national average for adults (6.5 g/day)5. This survey accentuated the need to complete a
survey of contaminants in fish tissue to provide information on the quality of the fish being
consumed by CRITFC's member tribes.

The plans for the fish contaminant survey began with the formation of a multi-agency task force
with representatives from EPA, CRITFC, the Yakama Nation, the Umatilla Confederated Tribes,
the Nez Perce Tribe, the Warm Springs Tribe, the Washington Departments of Ecology and of
Health, the Oregon Departments of Environmental Quality and Health, the US Geological Survey
(USGS), and the US Fish and Wildlife Service.  A Memorandum of Agreement signed by EPA
and CRITFC in 1996 established the basis for the continued interaction of the EPA staff and tribal
members to complete the contaminant survey. With the help of members of CRITFC's member
tribes as well as state and federal fish hatchery personnel, sample collection took place between
1996 and 1998. Chemical analyses were completed in 1999. The analyses were done by EPA
and commercial laboratories.

This study was designed to estimate risks for a specific group of people (CRITFC's member
tribes). The CRITFC fish consumption survey combined information from all the member tribes
into a single distribution, therefore, the risk estimates in this study do not represent the risks of
any specific tribe.

The types offish, tissue types, and sampling locations were selected by the CRITFC's member
tribes.  Fish collection locations were selected because they were important to characterizing
risks to CRITFC's member tribes. Chemicals were chosen because they were identified in other
fish tissue surveys of the Columbia River Basin as well as being common contaminants found in
the environment.

This type of sampling is biased with unequal sample sizes and predetermined sample locations
rather random.  This bias is to be expected when attempting to provide information for
individuals or groups based on their  preferences.  The results of this survey should not be
extrapolated to any other fish or fish from other  locations.

The exposure assumptions used to estimate risk for CRITFC's member tribes were also
predetermined from CRITFC fish consumption  survey (CRITFC, 1994).  While the study was
designed to assess fish which were known to be important to CRITFC's member tribes, it was
         The average fish ingestion used by the EPA in risk assessments for the general public was changed from
 6.5 g/day to 7.5 g/day in 2000 (USEPA 2000a)

                                            1-2

-------
assumed that other people would be concerned about the contaminant levels in fish from the
Columbia River Basin. This decision to estimate risks for the general public was determined after
the chemical analyses were completed. Thus, the consumption patterns used this assessment for
the general public were not specific to people who eat fish from the Columbia River Basin.
However, the risk estimates provide a point of departure for discussions of levels of
contamination in the fish from this river basin.

The objectives of this study of chemical residues in the fish from the Columbia River Basin were
to determine:
              1)     if fish were contaminated with toxic chemicals,

              2)     the difference in chemical concentrations among fish species and study
                     sites, and

              3)     the potential human health risk due to consumption offish from the
                     Columbia River Basin.

This contaminant survey also provided information on those chemicals which were most likely to
be accumulated in fish tissue and therefore pose the greatest risks to people.

1.3    Study Area

The Columbia River Basin dominates more than a dozen ecological regions as it flows 1,950 km
from its source, Columbia Lake, located near the crest of the Rocky Mountains in British
Columbia, to the Pacific Ocean. The Columbia River drains an area of about 670,800 km2 of
which about fifteen percent is in Canada.  Eleven major tributaries enter the river: Cowlitz,
Lewis, Willamette, Deschutes, Snake, Yakima, Spokane, Pend Oreille, Wenatchee, Okanagan,
and Kootenay Rivers (Lang and Carriker, 1999). The study was confined to the Columbia Basin
below Grand Coulee to the north, the Clearwater River to the east, just below Bonneville Dam  to
the west and the Willamette River to the south(Figure 1-1).

1.4    Sampling Locations

One hundred and two fishing locations were identified by the Yakama, Nez Perce, Umatilla, and
Warm Springs tribal biologists.  Due to resource constraints, all of these sampling locations could
not be sampled.  The study design (Volume 5) presents in detail the process that was used to
reduce the number of sampling locations. Initially fishing locations that represented greater than
40% of each CRITFC' s member tribes' fishing use for resident and anadromous fish species were
identified. The number of fishing locations was further reduced by selecting sampling locations
at the base of a watershed to represent the entire watershed (98, 30,101, 96) and limiting the
number of sampling locations on the mainstream Columbia River to each of the dam reaches (6,
7,8,9,14). Additional sampling locations  (48,49) were added because they were near local
pollution sources. Sample location 49 on the Yakima River was also important for rainbow trout
spawning (personal communication CRITFC's member tribes). Other sampling locations (3,
21,21b, 62,63) were selected because of the concern for a particular fish species.

                                             1-3

-------
The final sampling locations were located on 16 rivers and creeks and the mainstream Columbia
(Figure 1-1, Table 1-1).  The actual sampling locations were variable within a study reach
because of the sampling techniques and/or mobility offish species.  To simplify the data analysis,
similar sampling locations within a study reach were combined to yield  one study site.  The river
miles for sampling locations are presented in Table 1-1.  The latitude and longitude for each
sampling location is presented in Volume n, Appendix A-2.
                                             1-4

-------
1-5

-------
Table 1-1. Description, study site, sampling location, and river mile for Columbia River Basin fish sampling 1996-1998. Some of the sampling
locations (S. Location) are combined into a single site for this study (SS = study site). Fish species are also listed. KM = river mile
Waterbody
Columbia River below Bonneville Dam
SS
3
Columbia River between Bonneville dam and Dalles dam6
Columbia River between Dalles dam and John Day dam
7
Columbia River between John Day dam and McNary dan8
Columbia River below confluence with Snake River
Columbia River (Hanford Reach)
Columbia River just below Priest Rapids Dam
Wind River
Little White Salmon River
Fifteen mile Creek
Hood River
Willamette Falls
MF Willamette River
Deschutes River
Umatilla River at the mouth
Umatilla River upper river
Thomas Creek
Meacham Creek
Yakima River below Roza Dam
Yakima River above Roza Dam
Klickitat River

Snake River below Hell's Canyon Dams
Snake River above Hell's Canyon Dams
Clearwater - Snake River
Looking Glass Creek - Grand Ronde
Icicle Creek - Wenatchee River
9L
9U
14
63
62
24
25
21
21B
98
30
101

48
49
56

13
93
96
94
51
S. Location
3B
6C
7B,D
7A
8B,D,E,F,G,H,I
9A,B,C,D
9 E,F A H, I,
9 N,0, P, Q
14 hatchery
63 hatchery
62 hatchery
24
25
21
2 IB-hatchery
98 A,B AD,E
30
30A , 30B
101,101A
101B
101C
48 F, G
48 H, I, J
49
56
56 A hatchery
56B,F
13C,D,E,F
93A hatchery
96 hatchery
94 hatchery
51 hatchery
KM
39-41
154-155
203-207
197.5
216-292
295-304
369-372
389-393
396
18
1
0.2-0.5
4
26.6
203.6
55-59
3
0-1
88.5-89.5
1.5-2.5
2-2.5
47.1
81-85
139-141
2.2
42.5
64-84
128-135
270
40.5
0.1
2.8
Fish Species
eulachon
white sturgeon
walleye
white sturgeon
largescale sucker, white sturgeon, fall chinook salmon, steelhead trout
white sturgeon
largescale sucker, white sturgeon
mountain whitefish
fall chinook salmon
spring chinook salmon
spring chinook salmon
Pacific lamprey
steelhead
Pacific lamprey
spring chinook salmon
mountain whitefish, rainbow trout, largescale sucker
spring chinook salmon, coho salmon, fall chinook salmon
largescale sucker, walleye,
mountain whitefish, rainbow trout
mountain whitefish, rainbow trout
rainbow trout
bridgelip sucker, largescale sucker, spring chinook salmon, fall chinook
salmon, steelhead, mountain whitefish, spring chinook salmon,
largescale sucker
largescale sucker, rainbow trout
fall chinook salmon, steelhead
spring chinook salmon
rainbow trout
largescale sucker, white sturgeon
steelhead
steelhead
spring chinook salmon
spring chinook salmon
                                                                      1-6

-------
1.5    Fish Species

A total of 281 fish samples were collected including 132 whole body, 129 fillet, 11 egg, and 9
field duplicates (Table l-2a,b). The fish species included anadromous fish species (Pacific
lamprey, eulachon, coho salmon, fall and spring chinook salmon, steelhead) and resident fish
species (largescale sucker, bridgelip sucker, mountain whitefish, rainbow trout, white sturgeon,
walleye). These species were selected because of their importance to CRITFC's member tribes.
Table l-2a. Resident fish species collected from the Columbia River Basin, 1996 -1998. The sample
location and identification number and number of replicates are given for each species.
Replicates
Fish species
White Sturgeon- Acipenser transmontanus
16 single fillets without skin, BW = 9,525g - 34,927 g
8 single whole body, BW = 8,108g - 22,380 g
4 duplicates of single fish each
White sturgeon samples were individual fish.
Rainbow Trout -Oncorhynchus mykiss
1 fillet composites with skin; BW = 318g - 551 g
Number in each composite = 7-11
12 whole body composites; BW = 47g - 475 g
Number in each composite = 7-30
Largescale Sucker - Catostomus macrocheilus
19 fillet composites with skin; BW = 809g- 1541 g
Number in each composite =4-12
23 whole body composites ; BW = 395g - 1,764 g
Number in each composite =5-12

Bridgelip sucker - Catostomus columbianus
3 whole body composites; BW = 588g - 637g;
Number in each composite = 7
Walleye -Stizostedion vitreum
3 fillet composites with skin; BW = 822g - 850g
Number in each composite = 8
3 whole body composites; BW = 749g - 1503g
Number in each composite = 4-8
Mountain Whitefish - Prosopium williamsoni
12 fillet composites with skin; BW = 247g - 517g
Number in each composite = 9-35
12 whole body composites; BW = 247g - 428 g
Number in each composite =9-35
1 duplicate composite
BW = Body weight; F= fillet WB = whole body ; Dup =
Studv Site
Columbia River - 6
Columbia River - 7
Columbia River - 8
Columbia River - 9L
Columbia River - 9U
Snake River - 1 3
Deschutes River - 98
Umatilla River -101
Yakima River - 49
Klickitat River -56
Columbia River - 8
Columbia River - 9 U
Umatilla River - 30
Deschutes River - 98
Yakima River - 48
Yakima -River - 49
Snake River - 1 3
Yakima River - 48
Columbia River - 7
Umatilla River - 30
Columbia River - 9U
Deschutes River - 98
Umatilla River - 101
Yakima River - 48
duplicate
F
3
3
3
3
1
3
4
3
3
4
3
3
3
3


3
3
3
3
3

W
3
3
2
3
4
3
2
2
3
3
3
6
3
3
3
2
1
3
3
3
3

Dup

1 fillet
1 fillet
1 fillet
1 fillet







1 fillet

                                             1-7

-------
  Table l-2b. Anadromous fish species collected from the Columbia River Basin, 1996 -1998. The sample
  location and identification number are given for each species. The number of replicates for each tissue type
  are listed after the location.
  Fish Species
Study Site
   Replicates    Dup

F    WB   Egg
  Coho salmon  - Oncorhynchus kisutch
  3 fillet with skin composites; BW = 3,647g -3,960g
      Number in each composite = 6
  3 whole body composite; BW = 2,855g - 3,455g
      Number in each composite = 4
  Fall chinook salmon - Oncorhynchus tshawytscha
  15 fillet composites with skin; BW = 3,790g - 10,970g
      Number in each composite = 4
  15 whole body composites; BW = 4,160g - 8,623g
      Number in each composite = 6
  1 egg composite ;
  2 duplicate fillet composites

  Spring chinook salmon - Oncorhynchus tshawytscha
  24 fillet composites with skin;  BW = 4536g - 9373g
      Number in each composite = 3-5
  24 whole body composites; BW = 4,292g - 7,058g
       Number in each composite = 5
  6 egg composites;
  1 duplicate composite
Umatilla River 30
Columbia River - 8               33
Columbia River-14*              3    3
Umatilla River - 30               33
Yakima River - 48                33
Klickitat River - 56               33
                 1 fillet
                 1 fillet
  Steelhead - Oncorhynchus mykiss
  21 fillet composite with skin; BW = l,784g - 5,537g
        Number in each composite = 3-4
  21 whole body composite; BW =  l,633g - 6,440g
        Number in each composite =3-8
   1 egg composite sample;
  1 duplicate composite
  Pacific Lamprey  - Lampetra tridentata
  3 fillet composites with skin; BW = 364g - 430g
        Number in each composite = 20
  9 whole body composites; BW =  334g - 463g
         Number in each composite = 10-20

  Eulachon - Thaleichthys pacificus
  3 whole body composites BW =  37g;
         Number in composite = 144
Little White Salmon River-62*     3    3
Wind River-63**               3    3
MF Willamette River - 21B * *      3    3
Umatilla River - 30               33
Yakima River - 48                33
Klickitat River-56*              3    3
Icicle Creek-51*                 3    3
Grand Ronde River - 94*          33

Columbia River- 8               66
Hood River - 25                  33
Yakima River - 48                33
Klickitat River - 56               33
Snake River-93*                3    3
Clearwater River - 96*             33

Fifteen mile Creek - 24                  3
Willamette Falls-21              3    6
Columbia River - 3
                                                                                                1 fillet
                 1 fillet
* Fish taken from hatchery Dup = duplicate; F= fillet; WB = whole body BW = average body weight of the fish in a composite

With the exception of walleye, all these fish are cold water native species which are stressed by
alteration of their natural habitat (Netboy, 1980; Dietrich, 1995; Close, et. al., 1995;  Musick, et.
al., 2000; DeVore, et. al., 1995; Beamesderfer, et. al.,1995; Coon ,1978; Lepla, 1994). Walleye
were introduced to the Columbia River Basin from the late 1800s to the early and mid 1900s and
are well established in some of the reservoirs (e.g., the John Day Reservoir).

In order to estimate risks for the general public, it was assumed that these species were also
consumed by other people in the basin. While there were no comprehensive surveys offish

-------
consumption by the general public in the Columbia River Basin at the time of this study, there
have been surveys in the Middle Fork Willamette River (EVS, 1998), lower Willamette River
(Adolfson Associates, Inc., 1996), and Lake Roosevelt (WDOH,1997).  The types offish
identified (Table 1-3) in these surveys include some of the same types listed in the CRITFC
consumption survey(CRITFC, 1994).
Table 1-3. Recent surveys of types offish consumed by the general public in the Columbia River Basin.

Location
Tissue Type

Fish Type











EVS 1998
Middle Willamette
primarily muscle some skin, eggs,
eyes
bullhead
carp
sucker
bass
northern pikeminnow
crappie
bluegill
trout
white sturgeon
lamprey
salmon
steelhead
Adolfson Associates
Lower Willamette
muscle

yellow perch
brown bullhead
northern pikeminnow
starry flounder
white sturgeon







WDOH 1997
Lake Roosevelt
fillets primarily some skin, eggs, fish
heads
rainbow trout
walleye
bass









1.6    Sampling Methods

Sampling methods (Volume 4, Appendix A) for fish included: electrofishing, hand collection,
hatchery collection, trapping at dams, dip netting, fish traps, and gill netting.  The preferred
method was dependent on the conditions at the sampling location, selected species, and legal
constraints. A global positioning system (GPS) was used to identify the latitude and longitude for
each sampling location (Volume 4, Appendix A).

After retrieval from sampling devices, each fish was identified to the species level by personnel
familiar with the taxonomy of the fish in the Columbia River Basin. The length and weight were
then measured for each fish to ensure that they met the size class as defined in the Quality
Assurance Project Plan (Volume 5).  The length and weight data are provided in Volume 2,
Appendix A.

Four types of samples were collected: whole-body with scales, fillet with skin and scales, fillet
without skin, and eggs. The white sturgeon is the only species where fillet without skin was
collected.  The armor-like skin of the white sturgeon was considered too tough for ingestion.
Whole-body samples were selected to maximize the chances of measuring detectable levels of
contaminants of concern and because data presented in the consumption study showed that
CRITFC's member tribes may consume several fish parts in addition to the fillet (CRITFC,
1994).  Eggs from spring chinook salmon, fall chinook salmon, and steelhead were measured
because consumption data show that their eggs were widely consumed by CRITFC's member
                                             1-9

-------
tribes.  The fish were not scaled as recommended in the EPA guidance (USEPA, 1998a).  Based
on conversations with CRITFC's member tribes, it was assumed that people consume the whole
body or fillet with scales intact.

The Columbia River Basin is very large and the number of samples which could be analyzed was
relatively small. Due to limited resources, composites were analyzed (with the exception of white
sturgeon) instead of individual fish as being a better estimate of the average concentrations of
chemicals from a study site. The number offish in each composite are listed in Volume n,
Appendix A-2.  It is assumed that by compositing, the error in representativeness would be
reduced. However, by using an average of individual fish the true variability in individual fish
tissue samples was lost.  Thus, the actual residues in  individual fish from the Columbia River
Basin may be higher or lower than the concentrations reported in this study. Due to the size and
difficulty of homogenization, composites were not taken for white sturgeon. Instead, individual
fish were sampled and analyzed from each sampling location. Since this study was designed for
fish consumption and people eat what they collect, random samples offish were selected for each
composite rather than predetermined age or gender.

An attempt was made to collect three replicate samples for each fish type from each study site to
estimate variability between study sites.  However, this was not always possible due to
availability offish and problems with sampling gear. The final number of replicates for each fish
species and tissue type are listed in Table 1-2 a,b. To reduce differences due to sampling error,
replicate samples were collected at the same time and study site.

1.7     Chemical Analysis

The homogenization of samples, the lipid analysis, and chemical analysis of chlorinated dioxins
and furans, and dioxin-like PCB congeners were conducted by AXYS Laboratory in Victoria,
Canada. The remaining analyses were performed by the EPA Region  10 laboratory at
Manchester, WA. Laboratory analytical protocols specified for this study are referenced in
Volumes 4 and 5.

Chemical analysis of the fish tissue was completed in 1999. The fish samples were analyzed for
132 different chemicals (Tables 1-4 a,b,c,d,e,f,g), including the following classes: semi-vocatives,
chlorinated dioxins and furans, dioxin-like PCB congeners, Aroclors, pesticides and selected trace
metals6.

Of the  132 compounds analyzed, 40 were not detected (Tables 1-4 a,b,c,d,e,f,g).  The individual
chemical analyses offish tissue samples are presented in Volume 2, and summarized in Volume
l,AppD.
        6 "Metals", as used in this report, also refers to metalloids or semi-metals. Antimony, selenium, boron, and
 arsenic are in the metalloid groups.

                                             1-10

-------
Table l-4a. 51 semi-volatile chemicals analyzed.
                                 Table l^lb.  26 pesticides analyzed.
22 detected
1,2-Diphenylhydrazine
2,6-Dinitro toluene
Acenaphthene
Ac enaphthy lene
Anthracene
Benz-a-anthracene
Benzo-a-pyrene
Benzo-b-fluoranthene
Benzo-k-fluoranthene
Chrysene
Dibenz [a,h] anthracene
Fluoranthene
Fluorene
Indeno( 1,2,3 -cd)py rene
Pyrene
Phenanthrene
Benzo(g,h,I)perylene
Naphthalene
1 -Methyl-naphthalene
2 -Me thy 1-naphthalene
Phenol
Retene
29 not detected
Nitrobenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
1,2,4-Trichlorobenzene
2,4-Dinitro toluene
2-Chloronaphthalene
4-Bromophenyl-phenylether
4-Chlorophenyl-phenylether
bis(2-Chloroisopropyl)ether
Hexachloro butadiene
Hexachloro ethane
Dibenzofuran
2-Chlorophenol
4-Chloro-3-methylphenol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4,5-Trichlorophenol
2,3,4,6-Tetrachlorophenol
2,4,6-Trichlorophenol
Pentachlorophenol
4-Chloroguaiacol
3,4-Dichloroguaiacol
4,5-Dichloroguaiacol
4,6-Dichloroguaiacol
3,4,5-Trichloroguaiacol
3,4,6-Trichloroguaiacol
4,5,6-Trichloroguaiacol
Tetrachloroguaiacol	
21 Detected
Aldrin
cis-Chlordane
gamma-Chlordane
oxy-Chlordane
cis-Nonachlor
trans-Nonachlor
alpha-Chlordene
o,p'DDT
p,p'DDT
o,p'DDE
p,p'DDE
o,p'DDE
p,p'DDE
DDMU
Endosulfan Sulfate
Hexachlorobenzene
Heptachlor Epoxide
Alpha BHC
Gamma-BHC (Lindane)
Mirex
Pentachloroanisole
5 Not Detected
gamma-Chlordene
Heptachlor
Delta-HCH
Beta-HCH
Toxaphene
Table We.
18 Metals analyzed.
16 detected
Aluminum
Arsenic
Barium
Beryllium
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Thallium
Vanadium
Zinc

2 not detected
Antimony
Silver






Table l-4d. 7
3 detected
Aroclor 1242
Aroclor 1254
Aroclor 1260





Aroclors analyzed
4 not detected
Aroclor 1016
Aroclor 1221
Aroclor 1232
Aroclor 1248





 Table l-4e. 13 Dioxin-like PCB
 congeners analyzed. All Detected
PCB 77
PCB 105
PCB 114
PCB 118
PCB 123
PCB 126
PCB 156
PCB 157
PCB 167
PCB 169
PCB 170*
PCB 180*
PCB 189

        Table l-4f. 7 chlorinated
        dioxins analyzed. All Detected
         Table l-4g. 10 chlorinated
         furans analyzed. All Detected
                                      2,3,7,8-TCDD
                                      1,2,3,7,8-PeCDD
                                      1,2,3,4,7,8-HxCDD
                                      1,2,3,6,7,8-HxCDD
                                      1,2,3,7,8,9-HxCDD
                                      1,2,3,4,6,7,8-HpCDD
                                      OCDD
                                          2,3,7,8-TCDF
                                          1,2,3,7,8-PeCDF
                                          2,3,4,7,8-PeCDF
                                          1,2,3,4,7,8-HxCDF
                                          1,2,3,6,7,8-HxCDF
                                          1,2,3,7,8,9-HxCDF
                                          2,3,4,6,7,8-HxCDF
                                          1,2,3,4,6,7,8-HpCDF
                                          1,2,3,4,7,8,9-HpCDF
                                          OCDF
                                                 1-11

-------
1.7.1  PCB analysis

Two methods were used for measuring PCB congeners: 1) congener analysis, and 2) Aroclor
analysis. PCB congeners are a group of synthetic organic chemicals that contain 209 individual
chlorinated biphenyl compounds. Each molecule of a PCB congener has 10 positions in its
ringed structure which can be occupied by a chlorine atom. The placement and number of
chlorine atoms into these positions determine the physical and chemical properties and the
lexicological significance of the specific PCB congener molecule in question. Each unique
arrangement is called a "PCB congener". The congeners which have chlorine atoms substituted
in the "para" and "meta" positions acquire a structure which is similar to chlorinated dioxins and
furans.

In the congener method only those congeners (Table l-4e) which are believed to have the same
lexicological mechanisms as 2,3,7,8 tetrachlordibenzodioxin (2,3,7,8-TCDD) were measured.
Of the 209 possible PCB congeners 13 were  analyzed.  Of these 13 congeners only 11 were
considered in the risk assessment. Two of the congeners  (PCB 180 and PCB 170) were included
because they were in the original EPA chemical method for measuring dioxin-like PCB
congeners. However, subsequent methods do not include these congeners because there was
"insufficient evidence on in vivo toxicity" to establish toxicity factors for these congeners (Van
den Berg, et al., 1998).  Although PCB 81 is considered to have the same toxicological
mechanism as 2,3,7,8-TCDD, EPA Method 1668 (USEPA, 1997a) did not list it as a target
compound. Therefore, it was not included in this study.

Commercially available PCB congener mixtures are known in the United States by their industrial
trade name, "Aroclor".  The last two digits indicate the percentage of chlorine in the compound
(i.e., 42% for Aroclor 1242 and 54% for Aroclor 1254). Each Aroclor mixture is further
identifiable by a specific number; i.e., "Aroclor 1242". The "12" portion of this designation
refers to the fact that the molecule contains 12 carbon atoms (bound together in two six-sided
phenyl rings; e.g., a "biphenyl"). The Aroclor analysis is the most common method for
measuring total PCBs.

1.7.2   Mercury and Arsenic analysis

Mercury and arsenic occur in organic and inorganic forms.  In this study, the chemical analyses
were as total mercury and total arsenic.  The fish tissue concentrations that are discussed in
Section 2 and Section 9 are based on the measured total mercury  and total arsenic.  For the
purposes of the risk assessment, the total mercury concentrations were assumed to be all
methymercury. Arsenic fish tissue concentrations was assumed to be 10% inorganic arsenic in
the anadromous fish tissue and 1% inorganic arsenic in the resident fish tissue.

1.7.3   Total Chlordane and Total DDT

The pesticides chlordane and DDT include a series of respective metabolites which are assumed
to act in the same manner with respect to human exposure  and toxicity. For this study, all forms
of chlordane (c/'s-chlordane, frvms-chlordane, c/'s-nonachlor, ^raws-nonachlor, and oxychlordane)

                                            1-12

-------
were summed as total chlordane to estimate tissue concentrations and risk estimates.

l,l,l-trichloro-2,2-te(p-chlorophenyl)ethane (DDT) and its structural analogs and breakdown
products: l,l-dichloro-2,2-&/XP'chl°r°phenyl)emylene (DDE), and l,l-dichloro-2,2-&/Xp-
chlorophenyl)ethane (DDD) are organo-chlorine pesticides.  DDT, DDE, and DDD also have two
isomers: the para (p,p) and ortho- para isomers (o,p). The p,p' and o,p' isomers of each DDT
structural analog (DDT, DDD, DDE) were combined into three concentration terms (DDT, DDD,
DDE) for fish tissue concentrations, and for the estimate of carcinogenic risks. All the DDT
structural analogs (p,p'-DDD, o,p'-DDD, o,p'-DDE, p,p'-DDE, o,p'-DDT, p,p'-DDT) were
summed into a single concentration (total DDT) term to estimate non-carcinogenic risks.

Although, l,l-te(p-chlorophenyl)2 chloro-ethylene (DDMU) is another structural analog or
breakdown of DDT it is not believed to exhibit the same toxicity as the other structural analogs.
Therefore it was not included in the sum of DDT for fish tissue concentrations and for the risk
assessment.

1.7.4. Lead Risk Characterization

Lead is not included in the risk characterization sections for other chemicals.  The methods for
assessing risks from exposure to lead are unique due to the ubiquitous nature of lead exposure and
the reliance upon blood lead concentrations to describe lead exposure, toxicity, and risks. Human
health risk assessment methods for lead also differ from other types of risk assessment because
they integrate all potential sources of exposure to predict a blood lead level.

1.7.5  Data Quality Validation of Chemical Analyses

A total of 93  data validation reports (Volume 4, Appendix B) were prepared detailing the quality
of project data.  Data quality assessment involved the following determinations:

        1)  whether the data met the assumptions under which the data quality objectives
       described in Volume 5 were developed, and

        2)  whether the total error in the data was small enough to allow the decision maker to
       use the  data.

No data were rejected in this study.

Nine field duplicate samples consisting of the opposite fillets of the same species and same type
of sample were collected to estimate the error in sample preparation and analysis (see Table l-2a-
b for list of field duplicates).  The range in duplicate concentrations is discussed in Section 10.

All the chemicals analyzed in fish tissue were within the requirements of the quality assurance
limits.  In the quality assurance review of the chemical data, certain chemical concentrations
were qualified with a "J". The "J" qualifier designates a concentration which is estimated.
Therefore, the analytical methodology suggests that the "J" qualified measurement may be

                                             1-13

-------
inaccurate. We chose to use these data in this study without conditions. No data were rejected.

1.7.6  Detection limits

The detection limits for chemicals were determined by performing a risk-based screening analysis
of tissue contaminant data collected within the Columbia River Basin during the last ten years
(1984-1994). The screening methods and quantitation limits are described in Volume 5.
The analytical methods were chosen to provide detection or quantitation limits which were as low
as possible within the constraints of available methods and resources.

The detection limits varied for each sample and each chemical.  The concentrations of chemicals
which are found at the detection limit could be treated as a zero; alternately they could also be
equal to the detection limit or somewhere in between.  For this study we assumed that the
concentration of a particular chemical was one half of the detection limit. For comparison, the
tissue chemical concentrations are presented in Appendix E assuming the concentration for a
particular chemical equals 1) zero, 2) the detection limit, or 3) 1A the detection limit

The following rules were used when calculating average chemical concentrations in fish tissue:

       1) If a chemical was not detected in any sample for a given fish species and sample type,
       it was assumed to not be present and was not evaluated.

       2) If a chemical was detected at least once in samples for a given fish species and sample
       type, a concentration equal to one-half the detection limit was assumed for values reported
       as not detected when calculating the average chemical concentration.

       3) The paired duplicate sample concentration for a fish at a site was averaged to obtain
       one concentration for that fish at that site.  In cases where one duplicate was reported as a
       measured concentration and the paired duplicate as a non-detected concentration, the
       measured concentration and one-half the detection limit for the non-detected value were
       averaged to obtain a single estimate of concentration.  In cases where both duplicate
       samples were not detected, one-half the detection limit for each sample was used as the
       mean chemical concentration.

1.7.7  Statistical Data Summaries

All fish residue data are presented on a wet weight basis.  All the data for each sample are
included in Volume n, Appendix C. The summary statistics (average, minimum, maximum, and
standard deviation) for each site and the basin are included in Volume 1, Appendix D.

The following statistical summaries include the non-detect rules described in  Section 1.7.6. The
data for each fish species were pooled and average chemical concentrations were calculated by
site and by basin:

       1) Site averages—All replicate samples for a given fish species and tissue type collected

                                              1-14

-------
       at a given site were pooled to obtain an estimate of the average chemical concentration at
       each site.

       2) Basin averages—All samples for a given fish species and tissue type collected during
       this study were pooled to obtain an estimate of the average chemical concentration within
       the basin.

1.8    Lipid Analysis

Most of the organic chemicals measured in this study were lipid soluble to a significant extent.
The lipid content of all samples was analyzed as a measure of the likelihood of bioaccumulation
of these types of organic chemicals. The percent lipid for each sample is given in Volume 4,
Appendix A.  The lipid normalized tissue concentrations are included in Volume 2, Appendix A.

Chemical residues were normalized to lipid using the following formula:

 (Equation 1-1)      ug chemical / kg lipid = (ug chemical/kg tissue x 100) ^ percent lipid

For example if wet weight concentration = 40 ug DDT/kg and the percent lipid = 5%
               (40 jug/kg x  100) - 5 = 800 ug DDT/kg lipid

The lipid normalized data were not used in the risk assessment.

1.9    Special Studies

Three additional studies were added after the original study was initiated:

       1) fish tissue chemical concentrations in channel catfish and smallmouth bass,

       2) exploratory study of acid-labile pesticide analysis using Gas Chromatograph/Atomic
       Emission Detector (GC/AED) methods for a limited number of samples, and

       3) radionuclide analysis for fish possibly exposed to potential releases from the Hartford
       Nuclear Facility.

1.9.1  Channel Catfish and  Smallmouth Bass

Due to interest in comparing the results  of this study with other Columbia River Basin surveys,
two additional species (channel catfish and smallmouth bass) were added to the initial study when
additional resources became available (Table 1-5).
                                             1-15

-------
      Table 1-5. Sampling study sites and numbers of replicates for survey of chemicals in tissues of
      smallmouth bass and channel catfish collected in the Columbia River Basin, 1996-1998.
                                                                      	Replicates
      Species	Study site	FS    WB	
      Channel Catfish - Ictaluruspunctatus                 Columbia River - 8    2     3
      5 fillet with skin composites; BW= l,236g-2,555g     Yakima River - 48    3     3
         Number in each composite = 2
      6 whole body composites; BW = 734g - l,135g
          Number in each composite = 5-6
      Smallmouth Bass -Micropterus dolomie               Yakima River -48     3     3
      3 fillet with skin composites; BW = l,413g - 1463g
          Number in ,each composite = 3
      3 whole body composites; BW = l,313g - l,487g
          Number in each composite = 3

       FS = fillet with skin; WB  = Whole body BW= average body weight offish in a composite

Since these were not species which were consumed in large amounts by CRITFC's member
tribes, the assessment of chemicals in these fish were not included in the discussion offish tissue
concentrations in Section 2 or  in the risk assessment (Sections 3-8). The results of chemical
analyses in these fish  are discussed in Section 9.

1.9.2  Acid-Labile Pesticides

In addition to the basic set of chemical analyses, EPA Region 10's laboratory measured 76 acid
labile pesticides using advanced EPA Gas Chromatography/Atomic Emission Detection
(GC/AED)  method 8085 (Volume 5, Table 12).  Of the 76 acid-labile pesticides measured only
17 were detected (Table 1-6).  Method 8085 is applicable to the screening of semi-volatile
organohalide, organophosphorus, organonitrogen, and organosulfur pesticides that are amenable
to gas chromatography.

The chemical analytical results are included in Appendix L.  Risk estimates were not completed
for the acid labile pesticides. These analyses were done to  ascertain only the presence or absence
of these chemicals.  A description of these chemicals is included in the toxicity profiles
(Appendix C).

    Table 1-6.  AED pesticides detected in fish tissue from the Columbia River Basin, 1996-1998.	
    Atrazine            DACTHAL-DCPA        Endosulfanll              Pentabromodiphenyl ether
    Bromacil            Dichlorobenzophenone    Endosulfan Sulfate          Propargite
    Chlorpyrifos         Dieldrin                Hexabromodiphenyl ether    Tetrabromodiphenyl ether
    Chlorpyrifos-methyl  Endosulfan I             Pendimethalin              Triallate
   	Trifluralin	

1.9.3  Radionuclide analyses

Due to the possibility of radionuclide contamination offish in the mainstream Columbia River a
subset offish samples was selected for radionuclide analysis.  These samples were collected in
the mainstream Columbia River (sites 7,  8, 9L, 9U) and cooling ponds (K ponds) on the Hanford
Reservation (Table 1-7).  Additional samples were collected from the Snake River (Study Site 13)
                                               1-16

-------
as a background or reference sample for the samples collected at or in the vicinity of the Hanford
Nuclear Facility.

 Table 1-7. Radionuclide fish tissue samples including study site, species, and number of replicates from the
 Columbia River Basin. 1996-1998.	
                                                            Replicates*
 Study Site	Fish species	F	WB     Duplicate
Columbia River 7
Columbia River 8


Columbia River 9 lower (L)
Columbia River 9 upper (U)


Hanford Reservation cooling ponds - 9K
Snake River 1 3
white sturgeon
white sturgeon
channel catfish
largescale sucker
white sturgeon
white sturgeon
mountain whitefish
largescale sucker
white sturgeon
white sturgeon
3
3
1

3
2
3
3

3

3
3
2
3
2
3
3
3





1 whole body
2 fillet
1 whole body


1 fillet
* each replicate was a composites of 4-35 fish except white sturgeon which were single fish; Fillets were with skin, except white
sturgeon which were fillets without skin; F - fillet; WB = whole body;

Radionuclides ( Table 1-8) were measured by EPA National Air and Radiation Environmental
Laboratory (NAERL) in Montgomery, Alabama, and a commercial laboratory (Barringer
Laboratory) in Golden, Colorado.

  Table 1-8. The radionuclides analyzed in fish tissue collected in the Columbia River Basin 1996-1998. _
  Uranium -234        Plutonium -239   Bismuth-214      Lead-212     Radon-224     Telllurium-208
  Uranium-235+D      Strontium- 90+D   Bismuth-212      Lead-214    Radon-226+D    Thorium-228+D
NAREL is a comprehensive environmental laboratory managed by the EPA Office of Radiation
and Indoor Air.  Among its responsibilities, NAREL conducts a national program for collecting
and analyzing environmental samples from a network of monitoring stations for the analysis of
radioactivity. This network has been used to track environmental releases of radioactivity from
nuclear weapons tests and nuclear accidents.

Quality assurance requirements for the 45 samples (see Volume 4, Appendix A, Table A-l)
selected for radionuclide measurements are described in the Quality Assurance Project Plan.. The
radionuclide data are reported in Volume 1, Appendix K.

The radionuclide fish tissue measurements and risk assessment are discussed in Section 8.
Radionuclides were not included with the other chemicals because radionuclides were not
analyzed in all fish tissues.  Although the method used to assess cancer risk from exposure to
radionuclides is similar to that for other chemicals in this risk assessment, there are some unique
aspects for radionuclides (e.g., analytical issues, estimation of risk  coefficients) that make a
separate discussion of them advantageous.
                                               1-17

-------
2.0    Fish Tissue Chemical Concentrations

In this section fish tissue chemical residues measured in this study are discussed.  The fish tissue
and egg samples were all composites with the exception of the white sturgeon which were
individual fish.  The concentrations discussed in this section include the rules for non-detected
chemicals described in Section 1.7.6.  In reviewing the results of this study the species were
evaluated in two groups: 1) resident fish species (white sturgeon, mountain whitefish, walleye,
bridgelip sucker, largescale sucker, rainbow trout) and the anadromous fish species (coho
salmon, spring and fall chinook salmon, steelhead,  pacific lamprey, eulachon). The resident fish
species spend their life cycle in the Columbia River and its tributaries. Their exposure and uptake
of chemicals will occur in fresh water in the vicinity of the locations where they were collected.
The anadromous species spend most of their life cycle in open ocean. They reproduce in fresh
water, but feed at sea. Therefore, their uptake of chemicals is likely to occur  at sea rather than at
the site where they were collected.

There were not equal numbers of samples offish species or tissue types (Table l-2a,b).  In
particular, the bridgelip sucker, coho salmon and eulachon were each collected at only one
location; Pacific lamprey and walleye at only two locations. Thus the data reported for these
species were not indicative of concentrations throughout the basin.  Bridgelip sucker and
eulachon were only collected as whole body fish tissue. Bridgelip  sucker were collected
opportunistically at this particular site.  However, they were not part of the original study design.
The eulachon were small fish. Therefore, it was necessary to collect 144 individual fish for each
composite to obtain enough tissue  for analysis.  It was also impractical to attempt to fillet these
fish.  Therefore only whole body samples were collected.  Despite these many variables,  general
trends in the monitoring of pollutants in these various species and tissues were evident.

he method for combining duplicate samples in this study was to average the duplicates. Thus, the
two measurements would be treated as one number for the purposes of this assessment.   The non-
detects were included in the data summaries at /^ their detection limits.  The actual  detection limit
is noted on the tables and in the text with a symbol for less than (<).  See Sections 1.7.6  and 1.7.7
for a detailed description of these methods.

The basin-wide and study site specific average chemical concentrations reported in this section
were used as the exposure concentrations in the estimation of risks discussed in Section 6.

2.1     Percent Lipid

The egg samples from the chinook salmon, and steelhead, had the highest percent lipid of all the
fish tissue samples (Figure 2-1).  The whole body and fillet tissues of Pacific lamprey and spring
chinook salmon, and the whole body eulachon had higher percent lipid than the whole body or
fillet tissues of any other species. Coho salmon, rainbow trout, walleye fillets,  and  largescale
sucker had the lowest percent lipid.

With the exception of the walleye samples there was not a large difference in lipid  content of
whole body and fillet samples.  The average whole body walleye samples contained 8% lipid as

                                             2-18

-------
compared to the 1.5% from the walleye fillets.  The technique used to fillet the samples was to
keep as much of the skin and associated fatty tissue (lipid) intact.  Thus, the chance of finding a
clear differentiation between fillet and whole body was not preserved.
                         Pacific lamprey
                             eulachon
                         spring Chinook
                            fall Chinook
                             steelhead
                                coho
                       mountain whitefish
                         bridgelip sucker
                              walleye
                         white sturgeon
                           rainbow trout
                       largescale sucker
          ANADROMOUS
RESIDENT
  All fish samples were composites except
  white sturgeon which were individual fish.
  Fillets were with skin except white sturgeon
  which were without skin.
                                                    10       15
                                                    Percent Lipid
                20
25
                  Figure 2-1.  Basin-wide average percent lipid in fish collected from the
                  Columbia River Basin. Study sites are described in Table 1-1. Sample numbers
                  for each species are listed in Table l-2.a,b

2.2     Semi-Volatile Chemicals
The semi-volatile chemicals include the guaicols, ethers, phenols, and polynuclear aromatic
hydrocarbons (PAH). The number of samples with detectable levels of the semi-volatile
chemicals was quite low (Table 2-la,b). The guiacols and ethers were not detected in any
sample. There were no semi-volatile chemicals detected in the fall chinook salmon or coho
salmon tissue samples.  The phenols were detected in only one white sturgeon sample from the
main-stem Columbia River (study site 8).  Many of these semi-volatile chemicals were not
detected because they were not in the fish tissue, the detection limits were too high, or the
chemicals may have been metabolized or otherwise degraded to chemicals which were not
included in this survey.
The average concentrations for the PAHs were quite similar across species and chemicals. Of the
PAHs, 2-methyl naphthalene (Table 2-la,b) had the highest detection frequency.  Pyrene was
found at the highest concentrations of all the PAHs (450 ppb) in a rainbow trout collected from
the upper Yakima River (study site 49). The largescale sucker was the fish species with the most
frequent detection of PAHs. This may be due to the large number of largescale sucker samples
rather than some unique exposure.
                                               2-19

-------
Table 2-la. Basin-wide composite concentrations* of semi-volatile chemicals detected in resident fish species
lig/kg US/kg
Species/Chemical T
bridgelip sucker
1 ,2-Diphenylhydrazine WB
Naphthalene, 1 -methyl- WB
Naphthalene, 2-methyl- WB
largescale sucker
1 ,2-Diphenylhydrazine WB
9H-Fluorene WB
Acenaphthene WB
Acenaphthylene WB
Benzo(a)anthracene FS
Benzo(a)pyrene FS
Benzo(g,h,i)perylene FS
Benzo[b]Fluoranthene FS
Benzo[k]fluoranthene FS
Chrysene FS
Dibenz[a,h]anthracene FS
Indeno(l,2,3-cd)pyrene FS
Naphthalene WB
Naphthalene, 1 -methyl- WB
Naphthalene, 2-methyl- FS
Naphthalene, 2-methyl- WB
Phenanthrene WB
Pyrene WB
Retene WB
N

3
3
3

23
23
23
23
19
19
19
19
19
19
19
19
23
23
19
23
23
23
23
F Max Ave

;
i
3

1
1
1
2
1
1
1
1
1
1
1
1
1
2
2
J
1
2
2

14
10
20

120
26
53
26
24
24
47
24
24
24
47
47
67
26
24
26
95
53
200

7
5
16

12
5
11
5
5
J
10
5
5
5
10
10
12
5
5
8
7
10
16
Species/Chemical
rainbow trout
Anthracene
Fluoranthene
Naphthalene, 2-methyl-
Naphthalene, 2-methyl-
phenanthrene
Pyrene
Retene
walleye
Naphthalene, 1 -methyl-
Naphthalene, 2-methyl-
Naphthalene, 2-methyl-
white sturgeon
Naphthalene, 1 -methyl-
Naphthalene, 2-methyl-
Phenol
mountain whitefish
2,6-Dinitrotoluene
Acenaphthene
Naphthalene, 2-methyl-




T N F

WB 12 1
WB 12 1
FS 7 3
WB 12 1
WB 12 1
WB 12 1
WB 12 1

WB 3 1
FS 32
WB 3 1

FW 16 1
FW 16 1
WB 8 1

WB 12 1
WB 12 1
WB 12 3




Max

























27
53
11
27
50
450
53

10
10
16

15
25
530

40
31
10




Ave

5
12
5
6
9
46
12

6
6
9

4
5
230

16
9
5




Table 2-lb. Basin-wide composite concentrations* of semi-volatile chemicals detected in anadromous
fish species from the Columbia River Basin,

Fish Species
eulachon
9H-Fluorene
Naphthalene, 2- methyl
Phenanthrene
Pacific lamprey
Fluoranthene
Naphthalene, 1- methyl
Naphthalene, 2- methyl
Naphthalene, 2- methyl
Phenanthrene
spring chinook salmon
Acenaphthene
Naphthalene, 2-methyl
Naphthalene, 2-methyl
Pyrene
steelhead
1 ,2-Diphenylhydrazine
1 ,2-Diphenylhydrazine
2,4-Dinitrotoluene
2,4-Dinitrotoluene
Benzo(a)pyrene














































1996-1998.
























T

WB
WB
WB

WB
WB
FS
WB
WB

WB
FS
WB
WB

FS
WB
FS
WB
FS

N F

3 1
3 1
3 1

9 1
9 4
3 1
9 4
9 3

24 1
24 4
24 5
24 2

21 1
21 1
21 2
21 1
21 1
Mg/kg



Max Ave

170
11
170

50
25
77
44
25

81
29
40
120

100
26
48
52
24

56
6
60

14
12
42
22
10

13
6
8
18

7
6
9
12
5










































. * All samples were composites except white sturgeon which were individual fish;
T= tissue type; N= number of samples; F = detection frequency; FS = fillet with skin; FW= fillet without skin; WB = whole body;
Ave= average; Max = Maximum
                                                                  2-20

-------
2.3    Pesticides

Of the 26 pesticides that were analyzed the most frequently observed pesticides were
hexachlorobenzene, mirex, pentachloronanisole, chlordane and related compounds, and the DDT
series of structural analogs (DDT,DDE,DDD).
   white sturgeon
  bridgelip sucker
       walleye
 largescale sucker
mountain whitefish
    rainbowtrout
      steelhead
   spring chinook
      eulachon
    coho salmon
     fall chinook
   pacific lamprey
                                                                                  RESIDENT
                                                                ANADROMOUS
                                                                 Composfte samples except white
                                                               sturgeon which were indivudals; Whfte
                                                                 sturgeon fillets were without skin
•eggs

Dwhole body

•fillet with
  skin
                                                                    MO   400   500   600   700  MO
                                                                   Total pestcides ug/kg
                                         Figure 2-2.  Basin-wide average concentrations of total pesticides in
                                         composite fish tissue collected from Columbia River Basin.  Study sites
                                         are described in Table 1-1. Sample numbers are given in Table l-2a,b.
The basin-wide average concentrations
of all pesticide residues were compared
across fish species. With the exception
of rainbow trout and walleye fillets, the
average pesticide residue levels in the
resident fish species were higher than in
the anadromous fish species (Figure 2-
2).  The average concentrations of total
pesticide residues were highest in white
sturgeon (Figure 2-2).

Of the anadromous fish species, Pacific
lamprey had the highest basin-wide
average concentrations of total
pesticides.  Pacific lamprey also had the
highest lipid content of any anadromous
fish species (Figure 2-1). The
concentrations of pesticides in the
Pacific lamprey may have been due to this high lipid content. However, egg samples which had
high lipid concentrations (Figure 2-1) did not have high pesticide concentrations as one would
expect for lipophilic compounds.

2.3.1  DDMU, Hexachlorobenzene, Aldrin, Pentachloroanisole, and Mirex

DDMU, Aldrin, pentachloroanisole,  and mirex were detected infrequently.  The highest
concentration (40 |ig/kg) of DDMU was in fish tissue from largescale sucker and mountain
whitefish. Aldrin was detected in only 2 species: mountain whitefish and white sturgeon (Table
2-2a). The maximum concentration (6 |ig/kg) of aldrin occurred in mountain whitefish from the
Hanford Reach of the Columbia River (study site 9U).  The maximum concentration of
pentachloroanisole occurred in largescale sucker (5 |ig/kg).  Mirex was only detected 9 times in
all the fish tissue from this study. The maximum concentration of mirex (13 |ig/kg) was detected
in mountain whitefish. Hexachlorobenzene was detected over 100 times; most frequently in
white sturgeon, spring and fall chinook salmon, and steelhead (Table 2-2a,b). The maximum
concentration of hexachlorobenzene (19 |ig/kg) occurred in white sturgeon (Table 2-2a).
                                                                                               900
                                               2-21

-------
 Table 2.2a.  Basin-wide concentrations of pesticides in resident fish tissue from the Columbia River Basin,
 1996-1998.
Species/Chemicals
                                N  F
Max     Aw     Species/Chemicals
                                                                                       T   N   F
Max
Aw
bridgelip sucker
Endosulfan Sulfate
largescale sucker
Pentachloroanisole
Pentachloroanisole
Mirex
Mirex
Hexachlorobenzene
Endosulfan Sulfate
Endosulfan Sulfate
DDMU
DDMU
mountain whitefish
Pentachloroanisole
Pentachloroanisole
Mirex
Mirex
Hexachlorobenzene
Hexachlorobenzene
DDMU
DDMU
Alpha-BHC
Aldrin
AlHrin

WB
WB
FS
WB
FS
WB
WB
FS
WB
FS

WB
FS
FS
WB
WB
FS
FS
WB
WB
FS
WR

3
23
19
23
19
23
23
19
23
19

12
12
12
12
12
12
12
12
12
12
17

3
4
2
3
4
2
3
13
8

3
2
3
3
6
3
6
6
3
1
3

5.4
5.0
2.6
5.0
2.6
5.0
6.5
2.6
40.0
19.0

3.0
2.4
13.0
6.0
3.0
2.4
40.0
31.0
3.0
6.0
•>, n

4.6
1.1
1.0
1.2
1.1
1.3
1.5
1.3
8.8
4.5

1.3
1.1
2.9
2.1
1.4
1.0
14.0
13.9
1.2
1.4
white sturgeon
Hexachlorobenzene
Hexachlorobenzene
Heptachlor Epoxide
DDMU
Alpha-Chlordene
Aldrin
Aldrin
walleye
Mirex
Hexachlorobenzene
DDMU
rainbow trout
Pentachloroanisole









WB
FW
FW
WB
FW
WB
FW

WB
WB
WB

WB









8
16
16
8
16
8
16

3
3
2

12









J
16
1
6
1
4
4

2
2
2

2









19.0
13.0
2.0
16.0
2.4
2.0
2.0

4.1
3.8
8.3

5.4









9.3
5.5
1.0
7.8
1.0
1.1
1.0

2.8
2.3
8.1

1.1








 * All fish samples were composites except white sturgeon which were individual fish.  T= tissue type; N = number of samples; F= detection
frequency; Max = maximum; Ave = average; FS= fillet with skin; FW = fillet without skin; WB = whole body
                                                         2-22

-------
       Table 2.2b.  Basin-wide concentrations of pesticides in anadromous fish tissue from the
       Columbia River Basin, 1996-1998. All anadromous fish samples were composites.
Mg/kg
Species/Chemicals
coho salmon
Hexachlorobenzene
fall chinook salmon
Hexachlorobenzene
Hexachlorobenzene
DDMU
DDMU
spring chinook salmon
Pentachloroanisole
Pentachloroanisole
Hexachlorobenzene
Hexachlorobenzene
DDMU
DDMU
steelhead
Hexachlorobenzene
Hexachlorobenzene
DDMU
Endosulfan Sulfate
Heptachlor Epoxide
Pentachloroanisole
Endosulfan Sulfate
DDMU
pacific lamprey
Hexachlorobenzene
Hexachlorobenzene
DDMU
DDMU
Pentachloroanisole
Pentachloroanisole
Tissue Type

WB

WB
FS
WB
FS
WB
FS
WB
FS
WB
FS

WB
FS
WB
WB
WB
WB
FS
FS
WB
FS
WB
FS
WB
FS
N

3

15
15
15
15
24
24
24
24
24
24

21
21
21
21
21
21
21
21
9
3
9
3
9
3
F

3

1
1
2
2
6
1
1
1
2
2

2
1
9
3
3
2
3
5
6
3
6
3
6
3
Max

1.2

4.5
3.4
2.4
2.0
4.2
3.8
3.8
3.5
4.2
3.8

3.2
2.8
2.4
2.1
2.1
2.1
2.1
2.0
11.0
8.0
6.9
5.6
3.6
1.7
Ave

1.2

3.0
2.1
1.1
1.0
1.1
1.1
2.3
2.1
1.2
1.1

2.2
1.6
1.3
1.0
1.0
1.0
1.0
1.1
6.3
7.6
3.9
4.5
1.4
1.6
         T= tissue type; N = number of samples; F= detection frequency; Max = maximum; Ave = average; FS= fillet with skin; FW = fillet
       without skin; WB = whole body

2.3.2  Total Chlordane

Total chlordane is a mixture of several chemically related compounds (oxy-chlordane, gamma,
beta and alpha chlordane, cis and trans nonachlor).

The fillet or whole body samples of bridgelip sucker, rainbow trout, eulachon, and coho salmon
had no detectable concentrations of any of the chlordane compounds.  The highest concentrations
of total chlordane were in egg samples from the spring chinook salmon and the fillet and whole
body Pacific lamprey.

The total chlordane concentrations in the whole body fish tissue samples were generally equal to
or greater than the fillet samples with the exception of the Pacific lamprey where the fillet
samples were slightly higher than the whole body samples (Table 2-3).  The walleye samples had
the most variation between whole body and fillet.
                                               2-23

-------
Table 2-3 . Basin-wide average concentrations of total chlordane (oxy-chlordane, gamma, beta and
alpha chlordane, cis and tram nonachlor) in fish from the Columbia River Basin, 1996-1998.
Fillet with skin
Resident species
white sturgeon*
walleye
mountain whitefish
largescale sucker
rainbow trout
bridgelip sucker
Anadromous species
Pacific lamprey
eulachon
spring chinook salmon
fall chinook salmon
steelhead
coho salmon
N
16
3
12
19
7
NS
3
NS
24
15
21
3
ug/kg
23
6
11
6
<5
43
NS
7
7
6
<5
Whole body
N
8
3
12
23
12
3
9
3
24
15
21
3
ug/kg
29
20
12
8
<7
<8
33
8
8
7
<5
Eggs
N ug/kg





6 66
1 15
1 15
3 33
       * white sturgeon were single fish and fillets without skin
       N = number of samples; NS= not sampled; Ave = average; < = chemicals not detected

2.3.3  Total DDT

Total DDT is the sum of the DDT structural analogs and breakdown products: p,p' and o,p' DDT,
p,p' and o,p' ODD, and p,p'and o,p' DDE.  DDMU is also a breakdown product of DDT which is
not believed to exhibit the same toxicity as the other breakdown products. Therefore it was not
included in the total DDT concentrations  for fish tissue concentrations.

The concentrations of total DDT (Table 2-4) in the salmonids (chinook, coho, rainbow, and
steelhead) and eulachon were much lower than in white sturgeon, largescale sucker, whole body
walleye, and mountain whitefish.  The Pacific lamprey DDT concentrations were higher than the
salmonids but 3 to 8 times lower than the resident species.  White sturgeon had the highest
concentrations followed by bridgelip sucker.  This is the same pattern observed with the total
pesticides (Figure 2-2).  The concentration of total DDT in walleye fillet was much less than in
the whole body, similar to the distribution seen with total chlordane.

The concentrations in egg samples were much lower than the fish tissue of the white sturgeon,
bridgelip and largescale suckers, whole body walleye, and mountain whitefish.  The
concentrations in egg samples from steelhead were higher than the other egg samples and fish
tissues of the anadromous species and rainbow trout.
                                             2-24

-------
Table 2-4. Basin-wide average concentrations of total DDT (DDT, DDE, ODD) hi composite fish
tissue samples from the Columbia River Basin, 1996-1998.
Fillet with skin
Resident Species
white sturgeon*
bridgelip sucker
walleye
largescale sucker
mountain whitefish
rainbow trout**
Anadromous Species
pacific lamprey
coho salmon***
steelhead***
spring chinook salmon
fall chinook salmon****
eulachon****
N
16
NS
3
19
12
7
3
3
21
24
15
NS
US/kg
578
NS
59
241
424
29
95
41
21
22
21
NS
Whole body
N
8
3
3
23
12
12
9
3
21
24
15
3
US/kg
787
529
489
450
405
38
90
42
27
27
25
21
Eggs
N ug/kg





3 39
1 14
6 24
1 14

       N= number of samples; NS = not sampled  * white sturgeon were individual fish and fillets without skin;
       ** p,p'-DDE and p,p'-DDT were the only isomers detected; *** p,p'-DDD and p,p'-DDE were the only isomers
       detected; ****p,p'-DDE was the only isomer detected

DDT found in the environment gradually degrades to DDE. Because of it is ubiquitous,
lipophilic, and persistent, DDE can be a useful surrogate in comparing fish species and study sites
in terms of estimating general trends of "relative loading" from persistent and agriculturally
derived organochlorines. p,p'DDE was the pesticide measured at the highest concentrations of all
the DDT structural analogs in fish tissues from this study (Figure 2-3).
                                             DDE-o.p'
                        Figure 2-3. Percent contribution of DDT structural analogs to
                        total DDT concentration in whole body largescale sucker. Basin-
                        wide average of 23 fish tissue samples.
With the exception of walleye and rainbow trout fillet samples, the maximum concentrations of
p,p'-DDE were higher in the resident fish species than the anadromous fish species (Table 2-5).
The maximum concentrations were measured in the white sturgeon fillet (1400 |ig/kg) and whole
body largescale sucker (1300 |ig/kg). The maximum concentration in the anadromous fish
species was in the whole body Pacific lamprey (77 ng/kg).
                                              2-25

-------
Table 2-5. Basin-wide average and maximum concentrations of p,p'DDE hi composite samples offish from
the Columbia River Basin. 1996-1998.
Fillet With Skin
MS/kg

Resident Species
white sturgeon*
largescale sucker
mountain whitefish
walleye
rainbow trout
bridgelip
Anadromous Species
Pacific lamprey
fall chinook salmon
coho salmon
steelhead
spring chinook salmon
eulachon
N

16
19
12
3
7
NS

3
15
3
21
24
NS
F

16
19
12
3
7


3
15
3
21
24

ranee

100-1400
14-740
8-910
44-52
4-54
NS

46-55
4-26
29-35
5-28
6-18
NS
Ave

470
200
360
47
22
NS

50
12
33
11
12
NS
N

8
23
12
3
12
3

9
15
3
21
24
3
F

8
23
12
3
12
3

9
15
3
21
24
3
Whole Body
MS/kg
ranee

400-1100
28-1300
13-770
350-440
3-84
310-560

35-77
5-53
31-37
5-33
11-22
10-11


Ave

620
370
340
410
29
400

53
15
35
15
15
11
Egg
Mg/kg
N F ranee Ave









1 1 6.6
3 3 31-33 32
1 1 6.5
6 6 10-16 12

NS = not sampled: N = number of samples; F = detection frequency; Ave= average * White sturgeon samples were single fish and fillets without
skin

The chemical concentrations in replicate fish tissue samples were compared across study sites for
white sturgeon, largescale sucker, and mountain whitefish (Figure 2-4).

The concentrations across study sites were extremely variable for the three fish species.  The
highest concentrations of p,p'DDE  observed in white sturgeon were from the Hartford Reach of
the Columbia River (study site 9U; Figure 2-4a). These samples were duplicate fillets from
opposite sides of the same fish. The duplicate sample concentrations were similar (1300 |ig/kg
and 1400 |ig/kg). The concentrations of p,p'DDE in the two whole body samples from this site
were much lower: 540 |ig/kg and 640 |ig/kg. The size of the fish from which the fillets (34,927g)
were collected was greater than the two whole body fish samples (~ 10,000 and 20,000g). This
may account for the difference in p,p'DDE concentrations between the whole body and fillets at
study site 9U. The fillet samples from study site 9U were quite different than the other sites on
the main-stem Columbia and Snake Rivers where white sturgeon were sampled.  The duplicate
samples from the lower Columbia River (study site 9L; 590 |ig/kg,  630 |ig/kg), main-stem
Columbia River (study site 6; 410 |ig/kg, 590 |ig/kg) and the Snake River (380 |ig/kg, 420 |ig/kg)
were similar to each other.

The maximum concentration (1300 |ig/kg) for the whole body largescale sucker  was from the
Yakima River below Roza Dam (study site 48; Figure 2-4b). The concentrations of p,p'DDE in
whole body largescale sucker from this site ranged from 390 to 1300 |ig/kg while the fillets
ranged from 430- 680  |ig/kg. The largescale sucker composite samples from this study site (48)
included 6 replicates. The number of replicates of the largescale suckers may have accounted for
the range in concentrations.

Mountain whitefish p,p'DDE concentrations were lower than the white sturgeon and largescale
sucker (Figure 2-4c).  The highest concentrations occurred in the Hanford Reach of the  Columbia
River (study site 9U) and Yakima River (study site 48) similar to the largescale sucker and white
sturgeon. The p,p'DDE fish tissue concentrations in the Deschutes and Umatilla River sites were
                                             2-26

-------
much lower than those in the Columbia or Yakima Rivers. The concentrations of p,p' DDE in
duplicate fillet samples from the Deschutes River were similar (6.6 |ig/kg and 9.4 |ig/kg) to each
other.
            LEGEND
          FW = fillet without
          skin
          FS = fillet with skin
          WB = whole body

          Study sites are listed
          by number and name
          and described in
          Table 1-1.
          Concentration points
          on graphs include
          each duplicate and
          chemicals at their
9U, Columbia River, WB'

9U, Columbia River, FW

9L, Columbia River, WB

9L, Columbia River, FW

 8, Columbia River, WB"

 8, Columbia River, FW

 7, Columbia River, FW

 6, Columbia River, FW

  13, Snake River, FW
                                                                       400        800        1200
                                                                     White Sturgeon, p,p'-DDE, ug/kg
                                      Figure 2-4a.  Study site specific concentrations of p,p' DDE in white sturgeon
                                      individual fish tissue samples in the Columbia River Basin.  Duplicate fillets
                                      were collected from study sites 9U, 9L, 6, and 13.

-------
    9U, Columbia River, WB
     9U, Columbia River, FS
    98, Deschutes River, WB
     98, Deschutes River, FS
      8, Columbia River, WB
       49, Yakima River, WB
       49, Yakima River, FS
       48, Yakima River, WB
       48, Yakima River, FS
      30, Umatilla River, WB
      30, Umatilla River, FS
        13, Snake River, WB
        13, Snake River, FS
                              0            400           800           1200
                                        Largescale Sucker, p,p' DDE, ug/kg
  Figure 2-4b.  Study site specific concentrations of p,p DDE in largescale sucker
  composite fish tissue samples from the Columbia River Basin.
   9U, Columbia River, WB
   9U, Columbia River, FS

   98, Deschutes River, WB

   98, Deschutes River, FS

     48, Yakima River, WB

     48, Yakima River, FS

   101, Umatilla River, WB
   101, Umatilla River, FS
                                   200       400       600       800
                                   Mountain Whitefish, p,p' DDE, ug/kg
Figure 2-4c.   Study site specific concentrations of p,p DDE in mountain whitefish
composite fish tissue samples from the Columbia River Basin.  Study site 98
includes duplicate fillet samples.
                                           2-28

-------
 2.4   Aroclors

Of the seven Aroclors analyzed in this study (Aroclors: 1016,1221,1232,1248,1242,1254,1260)
Aroclor 1016, Aroclor 1221, Aroclor 1232, and Aroclor 1248 never detected (Table l-4d).  The
most frequently observed Aroclors were 1254 and 1260. Aroclor 1242 was only detected in the
mountain whitefish samples.

The white sturgeon, mountain whitefish, whole body walleye, and Pacific lamprey had the
highest concentrations of Aroclors (Table 2-6).  The whole body concentrations of Aroclors in the
walleye were higher than the concentrations in fillets. There were no Aroclors detected in the
eulachon.  The concentrations in the egg samples were similar to the anadromous fish fillet and
whole body samples and less than the levels all the resident fish species except rainbow trout.
Table 2-6. Basin-wide average concentrations of total Aroclors (1242, 1254,1260) detected* in
composite fish tissue samples from the Columbia River Basin.
Fillet with skin
Resident Species
white sturgeon**
walleye
mountain whitefish
largescale sucker
bridgelip sucker
rainbow trout
Anadromous Species
pacific lamprey
eulachon
spring chinook salmon
fall chinook salmon
coho salmon
steelhead
N
16
3
12
19
NS
7

3
NS
24
15
3
21
Jig/kg
120
30
190
52
NS
33

106
NS
38
37
35
34
Whole body
N
8
3
12
23
3
12

9
3
24
15
3
21
Mg/kg
173
135
123
78
70
32

114
<57
40
40
38
37
Eggs
N ug/kg









6 43
1 31
3 34
1 35
     < = detection limitN= number of samples: NS= not sampled.\
     *Aroclor 1242 was only detected in mountain whitefish; aroclors 1016, 1221, 1232, and 1248 were not detected in any
   fish or egg samples
     "White sturgeon samples are individual fish and fillets without skin


Aroclors 1254 and 1260 were compared across study sites for white sturgeon (Figure 2-5a,b),
largescale sucker (Figure 2-6 a,b), and mountain whitefish (Figure 2-7 a,b).

 The maximum concentration for Aroclor 1254 was in the mountain whitefish (930 |ig/kg) fillet
sample from the Hanford Reach of the Columbia River (study site 9U; Figure 2-7a). The white
sturgeon fillet samples from the Hanford Reach of the Columbia River (study site 9U) had the
highest concentration (200 |ig/kg) of Aroclor 1260 for all species and all sites (Figure 2-5b).

Aroclor 1254 and 1260 were quite similar in white sturgeon samples (Figure 2-5a,b).  The highest
concentrations for both Aroclors occurred in the fillet samples from the Hanford Reach of the
Columbia River (study site 9U). Aroclor 1254 concentrations in the duplicate fillet samples from
study site 9U were 170 |ig/kg and 210 |ig/kg.  The whole body concentrations from this study site


                                             2-29

-------
were much lower (65 |ig/kg in both samples).  Aroclor 1260 concentrations were 190 |ig/kg and
210 |ig/kg in the duplicate fillets from study site 9U and 65 |ig/kg in the whole body samples.
The differences in sizes of the fillet and whole body fish (discussed in Section 2.3.3) from study
site 9U, may account for the difference in PCB concentrations in the fillet and whole body
samples.
The next highest Aroclor 1254 concentrations were from the main-stem Columbia River (study
site 6 ) where the duplicate concentrations were quite different (47|ig/kg and 160 |ig/kg;
 Figure 2-5a).  The percent lipid
(4.8%) of the duplicate with the
higher Aroclor 1254
concentration was higher than         au, Columbia River, we
percent lipid (3.1%) in the
opposite fillet.  Thus, the lipid
may account for the difference in
tissue  levels.  However, the
concentration of Aroclor 1260 in
the duplicate fillets from this site
were similar (43 |ig/kg and 40
jig/kg) to each other (Figure 2-
5b).
    9U, Columbia River, FW

    9L, Columbia River, WB

    9L, Columbia River, FW

     8, Columbia River, WB

     8, Columbia River, FW

     7, Columbia River, FW

     6, Columbia River, FW

       13, Snake River, FW
The Aroclor concentrations in the
duplicate fillets for Snake River
(study site 13) and for the lower
Columbia River (study site 9L)
were similar to each other
(Figure 2-5a,b).
                                                                  White Sturgeon, Aroclor 1254, ug/kg
  Figure 2-5a.  Study site concentrations of Aroclor 1254 in white sturgeon
  individual fish tissue samples from the Columbia River Basin.
             LEGEND

         FW = fillet without
         skin
         WB = whole body
         Study sites are listed
         by number and name
         and described in
         Table 1-1.
         Study sites 9u, 9L 6,
         and 13 include
         duplicate fillet
         samples.
         Concentration points
         on graphs include
         duplicate fillets and
         chemicals at their
         detection limits.
  9U, Columbia River, WB

  9U, Columbia River, FW

  9L, Columbia River, WB

  9L, Columbia River, FW

   8, Columbia River, WB

   8, Columbia River, FW

   7, Columbia River, FW

   6, Columbia River, FW

     13, Snake River, FW
                              50      100       150
                             White Sturgeon, Aroclor 1260, ug/kg
Figure 2-5b. Study site specific concentrations of Aroclor 1260 in white sturgeon
individual fish tissue samples from the Columbia River Basin.
                                                  2-30

-------
The concentrations of Aroclor 1254 and 1260 were variable in  largescale sucker. Aroclor 1254
ranged from <18 |ig/kg in the fillet composite from the Umatilla River to 65 |ig/kg in the whole
body sample from the Hanford Reach of the Columbia River (study site 9U; Figure 2-6a).
 Aroclor 1260 concentrations ranged from <19 |ig/kg in the  Snake River (study site 13)  and
Deschutes River (study site
98) to 100 |ig/kg in several
whole body samples from the
Hanford Reach of the
Columbia River 9study  site
9U) and the Yakima River
(study  site 48) (Figure 2-6b).
              LEGEND
       FS = fillet with skin
       WB = whole body
       Study sites are listed by
       number and name and
       described in Table 1-1.
       Concentration points on
       graphs include chemicals
       at their detection limits.
9U, Columbia River, WB
 9U, Columbia River, FS
98, Deschutes River, WB
 98, Deschutes River, FS
  8, Columbia River, WB
   49, Yakima River, WB
   49, Yakima River, FS
   48, Yakima River, WB
   48, Yakima River, FS
  30, Umatilla River, WB
  30,  Umatilla River, FS
   13, Snake River, WB
   13, Snake River, FS
                                                                   20           40
                                                                 Largescale Sucker, Aroclor 1254, ug/kg
                                   Figure 2-6a. Concentration of Aroclor 1254 in largescale sucker composite fish tissue
                                   samples from the Columbia River Basin.
9U, Columbia River, WB
9U, Columbia River, FS
98, Deschutes River, WB
98, Deschutes River, FS
 8, Columbia River, WB
  49, Yakima River, WB
  49, Yakima River, FS
  48, Yakima River, WB
  48, Yakima River, FS
 30, Umatilla River, WB
  30, Umatilla River, FS
   13, Snake RN/er, WB
   13, Snake River, FS
                                                              20      40       60       80       100
                                                                 Largescale Sucker, Aroclor 1260, ug/kg
                                   Figure 2-6b.  Concentration of Aroclor 1260 in largescale sucker composite fish
                                   tissue samples from the Columbia River Basin.
                                                    2-31

-------
In the mountain whitefish samples Aroclor concentrations from the Deschutes and the Umatilla
River sites were low with <17 |ig/kg for Aroclor 1254 in the Umatilla River and <16 |ig/kg for
Aroclor 1260 in the Deschutes River (Figure 2-7a,b).  The duplicate fillet samples from the
Deschutes River were equal or similar to each other. The maximum Aroclor 1254 concentration
of 930 |ig/kg in the fillet fish tissue from the Hartford Reach of the Columbia River was much
higher than the other fillet and whole body samples from this study site(Figure 2-7a).  The three
fillet samples from this study site had the same number offish per composite (35), approximately
the same weight (448-515g), length (352-369 mm) and percent lipid (7.9-7.7%). Thus, there was
nothing in the fish size or lipid
content which could account for
the differences in concentrations.
The maximum Aroclor 1260 in
the mountain whitefish fillet
(190 |ig/kg) was from the
Yakima River (study site 48;
Figure 2-7b).
         LEGEND
      FS = fillet with skin
      WB = whole body
      Study sites are listed
      by number and name
      and described in
      Table 1-1
      Study site 98 includes
      duplicate fillet
      samples.
      Concentration points
      on graphs include
      duplicate fillets and
      chemicals on their
      detection limits.  .
 9U, Columbia River, WB

 9U, Columbia River, FS

 98, Deschutes River, WB

198, Deschutes River, FS

-  48, Yakima River, WB

   48, Yakima River, FS

 101, Umatilla River, WB

 101,  Umatilla River, FS
                                                                 200     400      600     800
                                                                Mountain Whitefish, Aroclor 1254, ug/kg
                                     Figure 2-7a.  Concentration of Aroclor 1254 in mountain whitefish composite
                                     fish tissue samples from the Columbia River Basin.
9U, Columbia River, WB

9U, Columbia River, FS

98, Deschutes River, WB

98, Deschutes River, FS

  48, Yakima River, WB

  48, Yakima River, FS

101, Umatilla River, WB

101, Umatilla River, FS
                                                                  50       100       150       200
                                                                Mountain Whitefish, Aroclor 1260, ug/kg
                                    Figure 2-7b.  Concentration of Aroclor 1260 in mountain whitefish composite fish
                                    tissue samples from the Columbia River Basin.
                                                  2-32

-------
2.5    Dioxin-Like PCB congeners

When compared across all fish species, mountain whitefish fillet had the highest average
concentration (25  |ig/kg) of dioxin-like PCB congeners followed by the whole body walleye (11.7
|ig/kg, Table 2-7).

There was considerable difference between the whole body walleye samples and the fillets. This
was similar to the pattern observed in the walleye for DDT, chlordane, and Aroclors.  This may
be related to the amount of lipid in the whole body sample since dioxin-like PCB congeners are
also lipid soluble similar to the pesticides.

The concentrations of dioxin-like PCB congeners (Table 2-7) in the egg samples from the
anadromous fish were similar to the fillet and whole body samples of the coho salmon, eulachon,
spring and fall chinook salmon, and steelhead.

        Table 2-7. Basin-wide average concentrations of the sum of dioxin-like PCB congeners in
        composite fish samples from the Columbia River Basin. 1996-1998.
                                         Fillet With         Whole Body           Eggs
Resident Species

mountain whitefish
walleye
white sturgeon*
largescale sucker
bridgelip sucker
rainbow trout
Anadromous species
Pacific Lamprey
coho salmon
steelhead
fall chinook salmon
spring chinook salmon
eulachon
N

12
3
16
19
NS
7

3
3
21
15
24
NS
US/kg
ave
25.0
1.2
6.5
3.1

2.0

5.5
1.3
1.0
0.9
0.8

N

12
3
8
23
3
12

9
3
21
15
24
3
US/kg
ave
10.2
11.7
10.0
5.1
2.3
1.6

5.5
1.3
1.1
1.0
1.0
0.5
N ug/kg
ave








3 1.2
1 0.6
1 0.4
6 0.8

      N= number of samples; NS = not sampled. * white sturgeon were individual fish; fillets without skin

The concentrations of dioxin-like PCB congeners 118 and 105 were the major contributors to the
total dioxin-like PCB congeners (Figure 2-8a,b) for resident and anadromous fish species.   PCB
congeners 126,169, and 189 each contributed less than 1% to the total dioxin-like PCB congeners
in mountain whitefish (Figure 2-8a) and spring chinook (Figure 2-8b). PCB 126, the most toxic
dioxin-like PCB congener, was at quite low concentrations with a range of
0.0006-0.096 |ig/kg in mountain whitefish fillets and 0.00081- 0.028 |ig/kg in whole body.
PCB 126 was not detected in 5 of the 12 samples in mountain whitefish.  The range of PCB 126
concentrations in spring chinook was 0.00081-0.0046 |ig/kg in fillets and 0.00052-0.0047  |ig/kg
in whole body.  Of the 24 samples of spring chinook, 7 fillet and 8 whole body samples were not
detectable.
                                             2-33

-------
                      PCB167
                PCB157 2%    PCB77

            PCB156.2%
             9%
                                                                       PCB167
         PCS 123
           1%
                                                                                         PCB114
  Figure 2-8a. Percent contribution of dioxin-like PCB   Figure 2-8b. Percent contribution of dioxin-like PCB congeners
  congeners in mountain whitefish composite fillet samplein spring chinook salmon composite fillet samples from the
  from the Columbia River Basin.                     Columbia River Basin.
The concentrations of dioxin-like PCB congeners (Figure 2-9) were compared across study sites
for white sturgeon and mountain whitefish.  The average concentrations in mountain whitefish
and white sturgeon fillets from the Hanford Reach of the Columbia River (study site 9U) were the
highest of all the stations sampled.  The levels in the lower Columbia River (study site 9L),
Deschutes River, and Umatilla River were lower.  The concentrations of dioxin-like PCB
congeners in the white sturgeon and mountain whitefish (Figure 2-9) were consistent with the
Aroclor tissue residues (Figure 2-5, 2-6, and 2-7). The white sturgeon fillet from the Hanford
Reach of the Columbia River was an average of two fillets from the same fish.

The mountain whitefish were an average of three replicate composite samples with 35 fish per
composite. The variability of dioxin-like PCB congener concentrations in the mountain whitefish
fillets was similar to the distribution of Aroclors (Table 2-6).  The mountain whitefish fillet from
the Hanford Reach of the Columbia River (study site 9U) had a higher concentration (186 |ig/kg)
of dioxin-like PCB  congeners than other replicates from that site  (29|ig/kg,
36 jig/kg).
                                              2-34

-------
               Columbia R (9U)

               Columbia R (9L)

                 Snake R (13)

                Columbia R (6)

                Columbia R (8)

                Columbia R (7)

                          !


               Columbia R (9U)

                  Yakima (48)

               Deschuts R (98)

               Umatilla R (101)
      WHITE STURGEON
               Single fish
             Fillet without skin
             and whole body
MOUNTAIN  WHITEFISH
      Composite fish
 fillet with skin and whole body
                               • fillet
                               Dwhole body
                                                      40      50
                                                  dioxin-like PCBs ug/kg
              Figure 2-9.  Study site average dioxin-like PCB congeners in white sturgeon and mountain
              whitefish samples from the Columbia River Basin. Study sites are described in Table 1-1.
              Sample numbers are listed in Table l-2a,b.
The dioxin-like PCB congeners were highly
correlated with Aroclors in whole body
samples offish tissue (Figure 2-10). The
coefficient of determination (R2) for these
two variables was 0.94. The coefficient of
determination is a measure of the degree of
association of two variables.  It can range
from zero to 1, with  1 being a perfect
association (Sokal and Rohlf 1981). The
two variables are not dependent upon each
other, it is simply that they are both effects
of a common cause (Sokal and Rohlf,
1981).  It is also evident from this graph
that the white sturgeon, walleye, and
mountain whitefish had the highest average
concentrations of dioxin-like PCB
congeners and Aroclors.

2.6     Chlorinated Dioxins and Furans
     25
« 201
b
c
HI
D) 1c
C 13
O
O
m
o 10
          pacific lamprey
                                   white sturgeon


                      mountain whitefisfi   /'walleye




                        Hargescale sucker
                "bridgelip sucker

               .spring and fall Chinook
                                all samples are whole body composites
                                except white sturgeon which were single
                                        fish.

                                    R2 = 0.9421
                   50
                              100
                          Aroclors ug/kg
                                          150
                                                      200
 Figure 2-10. Correlation of basin-wide average concentrations of
 Aroclors 1242,1254,1260 (x axis) with dioxins like PCB congeners
 (y axis).
The average concentrations of chlorinated dioxins and furans in white sturgeon were higher than
the all other fish by an order-of-magnitude (Table 2-8).  The next highest average concentration
was in the mountain whitefish.  Coho salmon had the highest average concentrations of
chlorinated  dioxins and furans for the anadromous fish species although the levels were an order
                                                 2-35

-------
of magnitude lower than the highest white sturgeon concentrations measured in this study. The
egg samples from the steelhead and fall chinook were lower than the fillet or whole body fish
tissues of all species.  The egg samples from the coho salmon were higher than the other egg
samples, as well as the fish tissue of spring and fall chinook salmon, steelhead, largescale sucker,
and rainbow trout.
Table 2-8. Basin-wide average concentrations of the sum of chlorinated dioxins and furans in composite
fish samples from the Columbia River Basin. 1996-1998.
Fillet with skin
Resident Species
white sturgeon*
walleye
mountain whitefish
bridgelip sucker
largescale sucker
rainbow trout
Anadromous Species
eulachon
pacific lamprey
spring chinook salmon
steelhead
fall chinook salmon
coho salmon
N
16
3
12
NS
19
7

NS
3
24
21
15
3
Mg/kg
0.020
0.001
0.006
NS
0.001
0.002

NS
0.003
0.002
0.001
0.001
0.001
Whole body
N
8
3
12
3
23
12

3
9
24
21
15
3
Mg/kg
0.030
0.007
0.006
0.003
0.002
0.002

0.004
0.004
0.002
0.002
0.001
0.008
Eggs
N jig/kg









6 0.002
1 0.0008
1 0.0009
3 0.003
N = number of samples; NS = not sampled . *white sturgeon were individual fish; fillets without skin

Chlorinated dioxins and furans concentrations were compared across study sites for mountain
whitefish, white sturgeon, and largescale sucker (Figure 2-11). The  largescale sucker samples
were quite low compared to the mountain whitefish and the white sturgeon.  The largescale
sucker concentrations of chlorinated dioxins and furans (Figure 2-11), similar to the Aroclors
(Figure 2-6a.b), were much lower than the levels observed in mountain whitefish or white
sturgeon. However, the largescale sucker p,p'DDE concentrations  (Figure 2-4b) were equal to
the levels found in white sturgeon and mountain whitefish.

The total chlorinated dioxins  and furans were highest in the white sturgeon fillet from the lower
Columbia River (study site 9L, Figure 2-11).  The distribution of dioxins and furans in white
sturgeon across sites was different than the p,p' DDE (Figure 2-4a) and Aroclor (Figure 2-5a,b)
fish tissue residue distribution.  The p,p' DDE and Aroclor levels were higher in the Hartford
Reach (study site 9U) and study sites 6 and 8 in the Columbia River.

The mountain whitefish chlorinated dioxins and furans concentrations were highest in the
Hanford Reach of the Columbia River followed by the concentrations in the Yakima River
(Figure 2- 11).  This distribution was similar to the p,p' DDE (Figure 2-4c) and Aroclor 1260
levels (Figure 2-7b).
                                              2-36

-------
      Columbia (9U)
       Yakima (48)
        Deschutes
      Umatilla (101)

      Columbia (9L)
      Columbia (9U)
       Columbia(S)
       Columbia (6)
       Columbia(7)
        Snake(13)

       Columbia(S)
        Snake(13)
      Columbia (9U)
      Umatilla (30)
       Yakima (49)
        Deschutes
       Yakima (48)
              mountain whitefish
                   Composite samples; fillet with skin
                        white  sturgeon
                 Each sample is a single fish; fillet without skin
largescale sucker
 Composite samples; fillet with skin
                       0.005
                                 0.01
                                         0.015      0.02     0.025      0.03
                                       chlorinated dioxins & furans ug/kg
                                                                             0.035
                                                                                       0.04
                                                                                                0.045
     Figure 2-11.  Study site average concentrations of chlorinated dioxins and furans in mountain whitefish, white sturgeon,
     and largescale sucker from study sites in the Columbia River Basin. Study sites are described in Table 1-1). The
     number of samples are listed in Table 1-2.
2,3,7,8-TCDD, the most commonly studied chlorinated dioxin was generally found at the lowest
concentrations in all the samples.  The most frequently detected and the highest concentrations of
chlorinated dioxins and furans in fish tissue from this study were 2,3,7,8-TCDF  and OCDD
(Figure 2-12).
                                        OCDF  2,3,7,8-TCDD
                                                       1,2,3,7,8-PeCDD
                                                           4%
                    Figure 2-12. Percent contribution of each chlorinated dioxin and furan in
                    largescale sucker. Basin-wide average of 23 composite whole body fish
                    tissue samples. Only those congeners which exceed 1% of total
                    chlorinated dioxin and furan concentrations are shown on the figure.
                                                 2-37

-------
The maximum concentration of 2,3,7,8-TCDF was in the white sturgeon (Table 2-9). The fish
species tended to cluster into three groups:
       1) < 0.001 ng/kg = all the egg samples; walleye fillets, rainbow trout, spring chinook
       salmon fillets, steelhead, coho salmon, eulachon,
       2) > 0.001 to < 0.010 |ig/kg = largescale sucker, whole body walleye, bridgelip sucker,
       Pacific lamprey, fall chinook salmon, and whole body spring chinook salmon, and
       3) > 0.010 |ig/kg = white sturgeon and mountain whitefish.
Table 2-9a. Basin-wide concentrations of 2,3,7 ,8-TCDF in composite samples offish tissue from the
Columbia River Basin. 1996-1998.

Fillet
Whole Body
ug/kg
NF
Resident species
white sturgeon* 16 16
mountain whitefish 12 12
largescale sucker 19 18
walleye 3 3
rainbow trout 7 7
bridgelip sucker NS
Anadromous species
Pacific lamprey 3 3
fall chinook salmon 15 14
spring chinook salmon24 24
eulachon NS
steelhead 21 21
coho salmon 3 3
N = number of samples; F = detection frequency;
range

0.0025 -
0.00014-
O.0001 - 0
0.0006 - 0
0.0001 - 0


0.0012-0


0.054
0.014
.0015
.0008
.0003


.0017
O.0003 - 0.0014
0.0004 - 0

0.0002 - 0
0.0004 - 0
NS=not sampled;
.0007

.0007
.0005
Ave

0.017
0.0045
0.0004
0.0007
0.0002


0.0014
0.0007
0.0006

0.0004
0.0005
N

8
12
23
3
12
3

9
15
24
3
21
3
F

8
12
23
3
11
3

9
15
24
3
21
3
ug/kg
range

0.008
0.0002
0.0008 -
0.0038 -
0.0004 -
0.0008

0.0011-
0.0004 -
0.0006 -
0.0006 -
0.0003 -
0.0004 -
Ave

- 0.047
-0.012
0.0036
0.0055
0.0005
-0.001

0.0032
0.0014
0.0011
0.0008
0.0006
0.0005


0
0
0
0


0
0
0
0
0
0

0.021
.0044
.0009
.0046
.0002
0.001

.0020
.0008
.0007
.0007
.0004
.0004
< = detection limit
*white sturgeon were individual fish and fillets without skin
Table 2-9b. Basin-wide concentrations of 2,3,7,8-TCDF in composite samples of eggs

1996-1998.
Egg
Ug/kg

fall chinook salmon
N F
1 1


spring chinook salmon 6 6
steelhead
coho salmon
1 1
3 3


range
0.00043
0.0004 - 0.0007
0.0002
0.0003 - 0.0007










Ave

0.0005

0.0005















        N = number of samples; F = detection frequency
                                             2-38

-------
2.7     Toxicity Equivalence Concentrations of Chlorinated Dioxins and Furans, and
        Dioxin-Like PCB congeners

Chlorinated dioxins and furans are found in the environment together with other structurally-
related chlorinated chemicals, such as some of the various dioxin-like PCB congeners. Therefore,
people and other organisms are generally exposed to mixtures of these structurally similar
compounds, rather than to a single chlorinated dioxin or furan, or dioxin-like PCB congener.

In order to estimate risks for exposure to dioxin-like chemicals (Table l-4e,f,g) a method was
developed to estimate a toxicity equivalence concentration (Van den Berg et al, 1998).  In this
methodology the toxicity equivalence factor for 2,3,7,8-TCDD is equal to 1; all other dioxin,
furan, and dioxin-like PCB congeners  are calculated as some relative percent of 1.  The toxicity
equivalence factors (Table 2-10) were  derived by a panel of experts using careful scientific
judgment after considering all available relative potency data (Van den Berg et al.,  1998).
Dioxin-like congener-specific toxicity  equivalence factors (Table 2-10) are used to convert
individual dioxin-like congener concentrations to 2,3,7,8-TCDD equivalents.

    Table 2-10. Toxicity Equivalence Factors (TEF) for dioxin-like PCB congeners, dioxins, and furans
    (from Van den Berg et al.. 1998).
PCBs
PCB 126
PCB 169
PCB 157
PCB 156
PCB 114
PCB 77
PCB 189
PCB 123
PCB 118
PCB 105
PPR 1 6,1
TEF
0.1
0.01
0.0005
0.0005
0.0005
0.0001
0.0001
0.0001
0.0001
0.0001
n nnnm
Dioxins
2,3,7,8-TCDD
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1, 2,3,4 ,6,7,8-HpCDD
OCDD


TEF
1
1
0.1
0.1
0.1
0.01
0.0001


Furans
2,3,4,7,8-PeCDF
2,3,7,8-TCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,7,8-PeCDF
1, 2,3,4 ,6,7,8-HpCDF
1,2,3,4,6,7,8,9-HpCDD
OCDF
TEF
0.5
0.1
0.1
0.1
0.1
0.1
0.05
0.01
0.01
0.0001
The toxicity equivalence concentration is the product of the toxicity equivalence factor multiplied
by the concentration for an individual dioxin-like congener as shown in
Equation 2-1:

Equation 2-1)                 TEC=(TEFix [congenerfish tissue concentration]'J
               TEF = Toxicity equivalence factor
               TEC = toxicity equivalence concentration
The toxicity equivalence concentrations for each dioxin, furan, and dioxin-like PCB congener are
then summed to determine the total toxicity equivalence concentration.

The mountain whitefish fillet sample had the highest toxicity equivalence concentration
(0.0063 |ig/kg) followed by the white sturgeon (Table 2-11).  The primary contributors to the
mountain whitefish toxicity equivalence concentration were 2,3,7,8-TCDF and dioxin-like PCB
congeners (118,126,156).  The primary contributor to the high white sturgeon toxicity
equivalence concentration was 2,3,7,8-TCDF and dioxin-like PCB congeners (105,118,156). The

                                              2-39

-------
Pacific lamprey had the highest concentration of toxicity equivalence concentrations of all the
anadromous species. The concentrations 2,3,7,8 TCDF (Table 2-9), dioxinlike PCBs (Table 2-7)
Aroclors (Table 2-6, and total pesticides (Figure 2-2) were also higher in Pacific lamprey than in
any of the anadromous species.

 Table 2-11. Basin-wide average concentrations of the toxicity equivalence concentrations for composite fish
 samples from the Columbia River Basin. 1996-1998.
Fillet

Resident Species
white sturgeon*
walleye
mountain whitefish
largescale sucker
bridgelip sucker
rainbow trout
N

16
3
12
19
_ug/kg

0
0.
0
0

.0043
00049
.0063
.0009
NS
7
0
.0008
Whole
N

8
3
12
23
3
12
body

	 ug/kg

0,
0,
0,
0,
0,
0,

.0051
.0036
.0033
.0016
.0013
.0009
Anadromous Species
Pacific lamprey
spring chinook salmon
steelhead
eulachon
coho salmon
fall chinook salmon

N

3
24
21

3
15
Fillet
Mg/kg

0.0027
0.0006
0.0.0009
NS
0.0.0004
0.0.0004
Whole body
N

9
24
21
3
3
15
US/kg

0.0035
0.0009
0.0009
0.0007
0.0006
0.0005
      N = number of samples: NS = not sampled.; *white sturgeon were individual fish and fillets without skin

2.8    Metals

Of the sixteen metals analyzed, antimony and silver were not detected. Thallium was only
detected once in a mountain whitefish. Unlike the organic chemicals the high metal
concentrations did not appear to be associated with certain species or locations.

The percent contribution of each of the metals to the sum of metals was compared in fillet
samples of largescale sucker (Figure 2-13 a) and spring chinook salmon (Figure 2-13b).  While
there was considerable variability in the percent contribution in fish tissue, zinc and aluminum
were found at the highest concentrations in all species (Figures 2-13a,b).  Arsenic was generally
higher in the anadromous fish species than  in the resident fish species.
                      Figure 2-13a. Basin-wide average percent of individual metals in
                      largescale sucker fillets. N=23.
                                               2-40

-------
                                          Aluminum
                                                         Barium
                                                          1 %   Chromium
                                                                 2%
                                                                    Manganese
                                                                      1%
                                                                   Mercury
                                                                    1%
                     Figure 2-13b. Basin-wide percent of individual metals in spring
                     chinook salmon fillets. N=24.

Basin-wide concentrations of metals were compared across species (Table 2-12, 2-13, 2-14).  The
maximum concentrations of individual metals (Table 2-12) were generally higher in the whole
body fish samples with the exception of arsenic, copper, mercury, selenium, and zinc. Arsenic
and mercury were higher in fillet samples while copper, selenium, and zinc were higher in the egg
samples from the anadromous fish.  The maximum concentrations of barium, cadmium, and
manganese were in whole body largescale sucker samples from the Hartford Reach of the
Columbia River (study site 9U). The maximum concentrations of chromium and cobalt were
measured in the whole body white sturgeon from the main-stem Columbia River (study site  8).

  Table 2-12. Basin-wide maximum concentrations * of metals in composite fish tissues measured in the
  Columbian River Basin. 1996 -1998.
Chemical
Aluminum
Arsenic
Barium
Cadmium
Chromium
Copper
Copper
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Selenium
Vanadium
Zinc
Zinc
Species
Largescale sucker
Steelhead
Largescale sucker
Largescale sucker
White sturgeon
Steelhead
Fall chinook
White sturgeon
Fall chinook
Largescale sucker
Spring chinooksalmon
Steelhead
Spring chinooksalmon
White sturgeon
Rainbow trout
Steelhead
Mountain whitefish
N
2
3
3
3
3
1
3
3
3
3
3
3
3
1
4
1
3
Tissue tvne
WB
FS
WB
WB
WB
Egg
WB
WB
WB
WB
FS
WB
egg
FW
WB
egg
WB
ua/k2
190000
1500
4700
250
1000
18000
14000
420
1200
21000
510
17000
5500
2700
770
76000
40000
Study Site**
Columbia River (8)
Hood River (25)
Columbia River (9U)
Columbia River (9U)
Columbia River (8)
Snake River (96)
Columbia River (14)
Columbia River (8)
Columbia River (14)
Columbia River (9U)
Klickitat River (56)
Klickitat River (56)
Umatilla River (30)
Columbia River (9U)
Umatilla River (101)
Snake River (96)
Deschutes (98)
 * All samples were composites except white sturgeon which were individual fish.; * *study site name with study site number in parentheses
 N = number of samples; FS = fillet with skin; FW = fillet without skin; WB = whole body.
                                               2-41

-------
Mercury was not detected in any anadromous egg sample (Table 2-13).  The concentrations of
copper, manganese, selenium and zinc were higher in the egg samples than any of the
anadromous fish tissue samples (Table 2-12;Table 2-14).

     Table 2-13. Basin-wide average concentrations of metals in samples of eggs from anadromous fish
     collected in the Columbia River Basin, 1996-1998. Barium and beryllium were not detected in any
     egg samples.
Chemical
Number of samples
fall chinook salmon
1
spring chinook salmon
6
coho salmon
3
steelhead
1
Concentration (jig/kg)
Aluminum
Arsenic
Cadmium
Chromium
Cobalt
Copper
Lead
Manganese
Mercury
Nickel
Selenium
Vanadium
Zinc
500
240
<4
<100
35
5800
<10
960
<50
54
2400
19
36000
950
460
35
100
43
6200
14
1500
<79
78
4200
13
43000
850
330
<4
<100
12
4500
<10
700
<100
84
1200
28
31000
4500
25
34
220
170
18000
41
2200
<43
520
4500
110
76000
       < = detection limit

Largescale sucker had the highest basin-wide average concentrations (Table 2-14) of aluminum
(69,000 ng/kg), barium (2,300 |ig/kg), manganese (14,000 |ig/kg), mercury (240 |ig/kg), and
vanadium (310 |ig/kg). White sturgeon had the highest basin-wide average concentrations of
beryllium (8 |ig/kg), chromium (360 |ig/kg), cobalt (260 |ig/kg), and selenium (1,100 |ig/kg).

The basin-wide average whole body concentrations of cadmium, chromium, cobalt, copper, lead,
manganese, nickel, vanadium, and zinc were higher than the fillet concentrations (Table 2-14).
This may be due to the concentrations of these chemicals in the internal organs, bones, and skin
of the fish.  Selenium was generally higher in the whole body fish tissue with the exception of the
white sturgeon. The concentrations of barium and aluminum were higher in the whole body
tissue of resident fish species. In the anadromous fish species the whole body aluminum and
barium concentrations were equal to or less than the fillet.
                                             2-42

-------
Table 2-14. Basin-wide average concentrations of metals


Chemical
N-FS
N-WB

Aluminum
Aluminum
Arsenic
Arsenic
Barium
Barium
Beryllium
Beryllium
Cadmium
Cadmium
Chromium
Chromium
Cobalt
Cobalt
Copper
Copper
Lead
Lead
Manganese
Manganese
Mercury
Mercury
Nickel
Nickel
Selenium
Selenium
Vanadium
Vanadium
Zinc
Zinc

Tissue
Tvne



FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
fall
chinook
salmon
15
15
Mg/kg
630
510
810
860
130
110
2
2
<4
6
71
100
47
140
640
3400
7
220
87
320
84
77
75
130
330
470
6
24
6700
27000
spring
chinook
salmon
24
24
Mg/kg
790
610
850
830
100
110
2
2
10
120
180
210
21
110
790
1400
14
21
90
370
100
64
63
270
350
530
5
17
6300
25000

coho
salmon
3
3
Mg/kg
<1000
<1000
540
500
160
140
2
2
<4
22
140
130
120
120
1700
1300
81
15
190
500
120
100
54
1200
290
360
7
38
7100
30000
in composite samples offish from the Columbia River Basin, 1996-1998.


steelhead
21
21
Mg/kg
1200
550
560
580
220
220
2
3
6
57
81
140
57
150
720
3200
8
45
150
460
120
100
44
900
330
650
14
66
7900
22000

Pacific
lamnrev
3
9
Mg/kg
500
1200
310
260
100
100
2
2
24
110
80
100
33
96
1200
4500
<10
16
380
390
<110
120
15
110
430
580
10
40
20000
22000


eulachon
NS
3
Mg/kg

8800

890

180

2

9

<100

7

940

500

500

<35

50

290

17

14000

largescale
sucker
19
23
Mg/kg
2400
69000
70
160
800
2300
3
5
5
55
120
310
65
170
550
1400
29
170
2700
14000
240
130
110
1100
260
310
11
310
20000
23000

*white
stureeon
16
8
Mg/kg
3800
48100
300
370
250
1900
2
8
2
42
65
360
27
260
250
990
8
120
260
2700
150
140
56
410
1100
650
9
220
3800
8200

mountain
whitefish
12
12
Mg/kg
2600
11100
100
140
280
700
2
2
7
28
130
120
51
110
620
1200
15
35
840
3400
80
67
76
280
510
960
29
160
15000
27500


walleve
3
3
Mg/kg
2500
2400
360
490
240
670
2
2
<4
7
90
110
8
56
570
2500
<10
190
370
950
180
180
260
260
390
470
5
14
8700
14000

rainbow
trout
7
12
Mg/kg
1100
27000
<50
120
390
1200
5
3
2
12
70
93
28
88
500
1800
<10
26
450
3200
77
73
59
330
220
360
17
190
12000
29000

bridgelip
sucker
NS
3
Mg/kg

37000

280

2000

5

29

180

96

1200

54

18000

32

400

280
29
190

20000
* white sturgeon were single fish; fillets were without skin N= Number of samples; FS = fillet with skin; WB = whole body; < = detection limit
                                                                                                  2-43

-------
2.8.1  Arsenic

Arsenic and mercury are discussed in detail in this report because of their contribution to risk.
They are often primary components of risk because of their toxicity as well as their ubiquitous
distribution in the environment as natural minerals in soil and from mining activities, smelting
(arsenic) and fossil fuel burning (mercury).

With the exception of Pacific lamprey, anadromous fish had higher arsenic concentrations than
resident fish (Table 2-14).  The whole body concentrations of arsenic were uniformly higher than
the fillet concentrations in the resident fish species (Table 2-14). However, there was no
consistent pattern in the whole body versus fillet arsenic concentrations in the anadromous fish
species (Table 2-14). Pacific lamprey had the lowest arsenic concentrations of all the
anadromous species, which was the inverse of the relationship for organic chemicals, where
Pacific lamprey had the highest concentrations. The average concentrations ( 240 - 460 |ig/kg) of
arsenic in the egg samples (Table 2-14) was similar to the whole body and fillet fish tissue
concentrations (70-860 |ig/kg) except for the steelhead  eggs  (25 |ig/kg) and rainbow trout fillets
(<50) which had the lowest concentrations of all the samples.

Arsenic concentrations were compared across sites for white sturgeon (2-14a) largescale sucker
(Figures 2-14b), mountain whitefish (2-14c), spring chinook (2-15 a) and steelhead (2-15b)

White sturgeon arsenic concentrations were generally consistent within sites but with
considerable variability across sites (Figure 2-14a).  For instance, the concentration in whole body
samples ranged from 240 |ig/kg in the white sturgeon from the Hanford Reach of the Columbia
River (study site 9U) to 660 |ig/kg in the white sturgeon from the main-stem Columbia River
(study site 8).  The fillet samples ranged from 150 |ig/kg in the Snake River (study site 13) to 640
jig/kg in the fillet sample from main-stem Columbia River (study site 7). The maximum
concentration occurred in the whole body  sample from the main-stem Columbia River (660
jig/kg; study site 8).   The arsenic concentrations in the duplicate fillets were equal or similar to
each other.

The highest arsenic concentrations of largescale sucker were measured in whole body and fillet
samples from the main-stem Columbia River (200-320  |ig/kg; study sites 9U, 8) and the whole
body samples from the Snake River (study site 13; 200-270 |ig/kg; Figure 2-14b).  The lower
concentrations ranged from 50-150 |ig/kg in whole body and fillet fish tissues from the
Deschutes, Yakima, Umatilla Rivers and the fillet fish tissues from Snake River (Figure 2-14b).

Mountain whitefish arsenic concentrations ranged from  100 to 140 |ig/kg with the maximum at
180 |ig/kg in the whole body sample from the Umatilla River (Figure 2-14c).  The lowest
concentrations were measured in the Deschutes River fillet samples. There was some variability
between fillet and whole body with the whole body samples being higher than the fillet samples
from Umatilla River and Deschutes River. The arsenic concentrations in the duplicate fillets
from the Deschutes River were similar to each other.

The concentrations of arsenic in spring chinook salmon showed no consistent trend within

                                             2-44

-------
stations or across stations (Figure 2-15a).  The highest concentrations were in the whole body
(1200 ng/kg) and fillet (1100 |ig/kg)from the Little White Salmon River and the whole body
(1100 |ig/kg)and fillet (1200 |ig/kg )from the Middle Fork of the Willamette River.  The arsenic
concentrations in the duplicate fillet samples from Looking Glass Creek (study site 94) were
similar (777 |ig/kg, 783 |ig/kg) to each other.

The maximum concentration (1500 |ig/kg)  of arsenic in all the fish samples was in the fillet
sample from the Hood River (Table 1-12 and Figure 2-15b). The maximum whole body
concentration from the Hood River was 1200 |ig/kg. However there was considerable variability
in the replicates  for this site with most whole body and fillet samples at about 430 |ig/kg.  The
samples from the other sites were between 290 and 800 |ig/kg (Figure 2-15b). The duplicate fillet
samples from the Clearwater River were not the same (480 |ig/kg, 582 |ig/kg) with the higher
concentration  (582  |ig/kg) falling outside the range of the other samples from this site but lower
than the maximum  observed in the Hood River.
           LEGEND
     FW = fillet without skin
     FS = fillet with skin
     WB = whole body
     Study sites are listed by
     number and name and
     described in Table 1-1
     Concentration pints on
     the graphs include
     duplicate fillets and
     chemicals at their
     detection limits.
9U, Columbia River, WB

9U, Columbia RN/er, FW

9L, Columbia RN/er, WB'

9L, Columbia River, FW-

 8, Columbia River, WB"

 8, Columbia River, FW

 7, Columbia River, FW-

 6, Columbia River, FW

  13, Snake RN/er, FW-
                                                      100    200    300    400    500    600
                                                               White Sturgeon, Arsenic, ug/kg
                                  Figure 2-14a. Site specific concentrations of arsenic in white sturgeon individual
                                  fish tissue samples from the Columbia River Basin.  Study sites 9U, 9L, 6, and 13
                                  include duplicate fillet samples.
                                                2-45

-------
     9U, Columbia River, W&
      9U, Columbia River, FS"
    98, Deschutes River, WB-
     98, Deschutes River, FS'
      8, Columbia River, WB-
      49, Yakima River, WB'
       49, Yakima River, FS
      48, Yakima River, WB'
       48, Yakima River, F&
      30, Umatilla River, WB-
       30, Umatilla River, FS
        13, Snake River, WB1
        13, Snake River, FS-
                        10
                                       110             210
                                      Largescale Sucker, Arsenic, ug/kg
                                                                       310
Figure 2-14b. Site specific concentration of arsenic in largescale sucker composite fish
tissue samples from the Columbia River Basin.
       9U, Columbia River, WB"

        9U, Columbia River, FS'

      98, Deschutes River, WB-

       98, Deschutes River,  FS •

         48, Yakima River,  WB"

          48, Yakima River, FS"

        101 , Umatilla River, WB-

         101, Umatilla River, FS-
                                           70               120
                                          Mountain Whitefish, Arsenic, ug/kg
  Figure 2-14c. Site specific concentration of arsenic in mountain whitefish composite fish
  tissue samples from the Columbia River Basin.  Study site 98 includes duplicate fillet
  samples.
                                           2-46

-------
LEGEND

FS = fillet with skin
WB = whole body
Study sites are listed
by number and name
and described in Tabl
1-1.
Concentration points
on graphs include
duplicate fillets and
chemicals at their
detection limits.
      94, Looking Gtes Creek, FS


            63,WindRwer,FS


    62, Lite White Salmon Rwer, FS


           56, KfckitatRrver, FS


           51,lcdeCieeKFS


           43, Yakima F»er, FS


           30, UmatibRrver, FS


  21, MWte Fork Willamette Rwer, FS
                          500
                                     700         900
                                   Spring Chinook; Arsenic, i
                                                           1100
Figure 2-15a. Study site concentrations of arsenic in spring chinook
composite samples from the Columbia River Basin. Study site 94 includes
duplicate fillet samples.
                                  96, Clearwater River, WB"

                                   96, Clearwater River, FS"

                                      93, Snake River, WB"

                                      93, Snake River, FS-

                                    8, Columbia River, WB"

                                     8, Columbia River, FS-

                                    56, Klickitat River, WB-

                                     56, Klickitat River, FS-

                                     48, Yakima River, WB-

                                      48, Yakima River, FS-

                                      25, Hood River, WB-

                                      25, Hood River, FS-
                                                     0
                                                                400         800
                                                                    Steelhead, Arsenic, uq
                           Figure 2-15b. Site specific concentrations of arsenic in steelhead composite fish
                           tissue samples from the Columbia River Basin. Study site 96 includes duplicate
                           fillet samples.
                                                    2-47

-------
2.8.2  Mercury

The mercury levels in fish samples were extremely variable. The maximum concentration of
mercury (510 |ig/kg ) was in the fillet sample of spring chinook salmon from the Klickitat River
(Table 2-12).

There was no consistent pattern in mercury concentrations between whole body and fillet samples
in the basin-wide average concentrations (Table 2-14). The average concentrations in fillet
samples ranged from <91  |ig/kg in the Pacific lamprey to 240 |ig/kg in the largescale sucker.  The
whole body average concentrations ranged from <35 |ig/kg in the eulachon to 180 |ig/kg in the
walleye.

Mercury concentrations were compared across study sites for white sturgeon, largescale sucker,
mountain whitefish, spring chinook salmon, and steelhead  (Figures 2-16a,b,c and 2-17a,b).

The maximum concentration (617 |ig/kg) for white sturgeon was measured in the duplicate fillet
from the Snake River (Figure 2-16a). The mercury concentrations in duplicate fillets from the
Snake River were quite different from each other (617 |ig/kg, 353 |ig/kg) and the whole body
samples (100  |ig/kg) from this site. Since, the duplicate fillets from the same fish were averaged
(430 |ig/kg) in the data-set for this report, the maximum level of mercury for this study was
reported as 510 |ig/kg for spring chinook  (Table 2-12). The concentrations in the duplicate fillets
from study sites 9L, 6, and 13 were similar to each other.

The largescale sucker mercury concentrations were extremely variable across and within study
sites.  There was no distinct maximum although the fillet samples for the Umatilla and Snake
Rivers were higher than the whole body samples from these study sites.

The mountain whitefish mercury concentrations were also variable.  The maximum
concentrations occurred in the Yakima, and Deschutes Rivers, although there was no difference in
average concentrations. The duplicate fillets from the Deschutes River were equal to each other
(71 Jig/kg).

The concentrations of mercury in spring chinook salmon samples were at or near non-detectable
levels, with the exception of the fillet samples from the Klickitat River, where the maximum
concentration (510 |ig/kg) was measured.  This fillet sample also appeared to be an outlier for
spring chinook salmon within this site and across all sites.  The duplicate fillets from Looking
Glass Creek were equal to each other (100 |ig/kg).

The maximum concentration (420 |ig/kg)  was a single whole body sample from the Clearwater
River. Except for the whole body sample from the Clearwater River,  Steelhead mercury
concentrations were all  less than 180 |ig/kg, with most samples in the 50-110 |ig/kg range.  The
duplicate fillets from the Clearwater River were equal to each other.
                                             2-48

-------
       9U, Columbia River, WB"

       9U, Columbia River, FW

       9L, Columbia River, WB"

       9L, Columbia River, FW "

        8, Columbia River, WB'

        8, Columbia River, FW

        7, Columbia River, FW

        6, Columbia River, FW-

          1 3, Snake River, FW
                                      White Sturgeon, Mercury, ug/kg
   Figure 2-16a.  Site specific concentrations of mercury in white sturgeon fish tissue
   samples from the Columbia River Basin.  Study sites 9U, 9L, 13, and 6 include
   duplicate fillet samples.
    9U, Columbia River, WB'

     9U, Columbia River, FS'

    98, Deschutes River, WB'

    98, Deschutes River, FS'
      48, Yakima River, WB-
       48, Yakima River, FS'

     101,  Umatilla River, WB'

      101, Umatilla River, FS-
                                 60       80       100      120
                                   Mountain Whitefish, Mercury, ug/kg
Figure 2-16c. Site specific concentrations of mercury in mountain whitefish
composite fish tissue samples from the Columbia River Basin. Study site 98 includes
duplicate fillet samples.








48, Yakima River, WB"



















                                                                                                                     100        200        300
                                                                                                                       Largescale Sucker, Mercury, ug/kg
                                                                                                                                                   400
Figure 2-16b.  Site specific concentrations of mercury in largescale sucker
composite fish tissue samples from the Columbia River Basin.
                         LEGEND
      FW = fillet without skin
      FS =  fillet with skin
      WB = whole body
      Data points represent composite samples offish tissu :
      except white sturgeon which are individual fish
      Study sites are listed by name and number and descr bed
      in Table 1-1.

      Concentration points on graphs include duplicate filk ts
      and chemicals at their detection limits.
                                                                            2-49

-------
                                        94, Looking Glass Creek, WB •
                                        94, Looking Glass Creek, FS •
                                              63, Wind River, WB'
                                               63, Wind River, FS •
                                     62, Little White Salmon River, WB •
                                     62, Little White Salmon River,  FS •
                                            56, Klickitat River, WB •
                                             56, Klickitat River, FS •
                                              51,  Icicle Creek, WB •
                                              51, Icicle Creek, FS •
                                             48, Yakima River,  WB'
                                             48, Yakima River, FS •
                                            30, Umatilla River,  WB-
                                             30, Umatilla River, FS •
                                  21, Middle Fork Willamette River, WB •
                                   21, Middle Fork Willamette River, FS •
                                                                     100       200       300       400
                                                                          Spring Chinook, Mercury, ug/kg
                                 Figure 2-17a.  Site specific concentrations of mercury in spring chinook salmon
                                 composite fish tissue samples from the Columbia River Basin. Study site 94 includes
                                 duplicate fillet samples.
               LEGEND
FS = fillet with skin
WB = whole body
Study sites are listed by name and numl
and described in Table
.Concentration points on graphs includt
duplicate fillets and chemicals at their
detection limits.
                                                  96, Clearwater Rwer, WB"
                                                   96, Clearwater River, FS-
                                                     93, Snake River, WB-
                                                      93, Snake Rwer, FS-
                                                    8, Columbia River, WB1
                                                     8, Columbia Rwer, FS-
                                                    56, Klickitat Rwer, WB-
                                                     56, KlicWtat River, FS"
                                                     48, YaWma Rwer, WB"
                                                     48, Yakima Rwer, FS-
                                                      25, Hood River, WB-
                                                      25, Hood Rwer, FS-
                                                                            100       200       MO
                                                                                 Steelhead, Mercury, ug/kg
                                                                                                         400
                                            Figure 2-17b.  Site specific concentrations of mercury in steelhead
                                            composite fish tissue samples from the Columbia River Basin.  Study site
                                            96 includes duplicate fillet samples.
                                                        2-50

-------
3.0    Human Health Risk Assessment

EPA uses risk assessment to characterize the potential cancer risks and non-cancer hazards for
individuals exposed to contaminants in environmental media. A systematic framework for risk
assessment was first outlined by the National Academy of Sciences (NAS, 1983). Building upon
this foundation, EPA has developed risk assessment guidance (e.g., USEPA, 1984, USEPA, 1989;
USEPA,  1995) that consists of the following components:

•      Data Collection and Analysis - involves gathering data to define the nature and extent of
       contamination in the environmental media of concern.
•      Exposure Assessment - characterizes how people may be exposed to environmental
       contaminants and estimates the magnitude of these exposures.
•      Toxicity Assessment - examines the types of adverse health effects associated with
       chemical exposure, and the relationship of the magnitude of exposure and the health
       response.
•      Risk Characterization - estimates the potential for adverse health effects (both cancer risk
       and non-cancer hazards) by integrating the information on toxicity and exposure.

The data  collection and analysis step for this study have been previously discussed in Section  1.
Section 2 provides information on contaminant levels in fish tissues.  Section 4 (Exposure
Assessment) describes how these contaminant levels are used with other exposure information
(e.g. how much fish people eat) to  estimate the magnitude of exposure for people consuming fish
from the Columbia River Basin. Section 5 (Toxicity Assessment) provides the toxicity
information that is used with the exposure estimates to characterize cancer risks and non-cancer
hazards in Section 6 (Risk Characterization).
                                            3-51

-------
4.0    Exposure Assessment

The objective of this exposure assessment is to estimate the amount of contamination that a
person may be exposed to from eating fish caught as a part of this study.

4.1    Identification of Exposed Populations

The potentially exposed populations for this risk assessment include (1) individuals within the
general public, and (2) CRITFC's member tribes.

As previously discussed in Section 1 of this report, the basis for the design of this fish study was
the fish consumption survey conducted by CRITFC (CRITFC, 1994), which targeted members of
the Nez Perce, Umatilla, Yakama, and Warm Springs Tribes (Appendix A). The CRITFC study
is the only comprehensive survey offish consumption that has been conducted for the Columbia
Basin and was used to develop tribal fish ingestion rates for this risk assessment.

Three other recent fish consumption surveys have been conducted in the Columbia River Basin:
in the middle Willamette River (EVS, 1998), lower Willamette River (Adolfson Associates, Inc.,
1996), and in Lake Roosevelt (WDOH, 1997).  These three studies are limited in scope and
focused on specific regions or populations within the Columbia River Basin. Therefore, the data
from them was not used to develop fish ingestion rates for this risk assessment. However, these
three surveys as well as the CRITFC survey are discussed in Section 4.5 (Fish Ingestion Rates)
because all the surveys illustrate the point that fish consumption practices can vary greatly
depending upon the age, gender, cultural practices, and/or  socioeconomic status of the anglers
surveyed.  These variations can include the types and amounts offish eaten, the frequencies of
meals, the portions of the fish that are eaten, and the preparation methods (USEPA, 1998a).

4.2    Exposure Pathway

An exposure pathway describes the course a chemical or physical agent takes from the source to
the exposed individual. A complete description of an exposure pathway involves four elements:
1)  a source and mechanism of chemical release, 2) movement of the chemical through the
environment resulting in contamination of environmental media, 3) a point of potential human
contact with these contaminated media (referred to as the exposure point), and 4) an exposure
route, such as ingestion, at the point of contact with these media (USEPA, 1989). While several
different exposure pathways could conceivably result in human exposure to chemical
contaminants within the Columbia River Basin, this risk assessment evaluates only part of one
pathway - exposure from  consumption offish. Data on contaminant levels in fish were gathered
and potential exposures through fish consumption estimated, but the source of these contaminants
and their subsequent movement through the environment into fish were not evaluated.
                                            4-52

-------
4.3    Quantification Of Exposure

To characterize the risk from consuming fish, an estimate of the amount of contaminant ingested
from eating fish must be estimated. This exposure is estimated using Equation 4-1:

                                   C x  CF x IRx EF x ED
     (Equation 4-1)      ADD  =	BW * AT	

where:
       ADD          =      Average daily dose of a specific chemical (mg/kg-day)
       C             =      Chemical concentrations in fish tissue (mg/kg)
       CF            =      Conversion factor (kg/g)
       IR             =      Ingestion (consumption) rate (g/day)
       EF             =      Exposure frequency (days/year)
       ED            =      Exposure duration (years)
       BW           =      Body weight (kg)
       AT            =      Averaging time for exposure duration (days)

As can be seen from this equation, an individual's exposure (average daily dose) depends upon
several factors including: the concentrations of contaminants in fish; the amount offish eaten;
how often and how long fish are eaten; and body weight.  Because this exposure occurs over time,
the total exposure is divided by a time period of interest (the averaging time) to obtain an average
exposure rate per unit time. When this average rate is expressed as a function of body weight, the
resulting exposure rate is referred to as the average daily dose (ADD) expressed in milligrams of
a chemical taken into the body per kilogram body weight per day  (mg/kg/day).

As can be seen from Equation 4-1, one individual's exposure may differ from another's because
of differences in these exposure factors.  Thus, in a population  offish consumers, a wide range of
individual exposures would be expected, from those individuals who have little exposure (e.g.,
because they don't eat much fish and/or eat fish that have low contaminant concentrations) to
those who have high exposure (e.g., because they eat highly contaminated fish and/or eat large
amounts offish). For this risk assessment, several of the exposure factors (fish ingestion rate,
exposure duration, and body weight) were varied to estimate a possible range in exposures among
individual fish consumers  (adults and children). For example, the  use of average exposure factors
in Equation 4-1  is expected to result in a daily dose that is more representative of the average
exposure in a population while the use of a mixture of average  and high-end exposure factors is
more representative of those members of the population who have higher exposures.  The
selection of these exposure parameters was made to ensure that, at a minimum, cancer risks and
non-cancer health impacts for those individuals with more average exposures as well as those
with much higher exposures are calculated.

For this risk assessment, exposures were estimated for adults and  children for both the general
public and CRITFC's member tribes. The exposure values selected for estimating exposure with
Equation 4-1 are shown in Table 4-1 (non-cancer) and Table 4-2 (cancer) and are discussed in
more detail in Sections 4.4 through 4.9.  The same tissue chemical concentrations are used to

                                            4-53

-------
estimate exposure for all of the populations, for cancer and non-cancer endpoints. However,
other exposure parameters differ. For example, cancer risks are estimated for lifetime exposures
only. Therefore, only exposure parameters for adults are included in Table 4-2. Four different
fish ingestion rates were used for adults (for estimating both cancer risks and non-cancer hazards)
and four for children (for estimating non-cancer hazards). These rates were based on two surveys
discussed in Section 4.5. The body weights used for each population correspond to the age of the
person for which consumption data was obtained in the two fish consumption surveys. For adults
for both cancer and non-cancer endpoints, a 70 kilogram body weight is used. However, data
were collected on children of different ages in the two surveys (children less than 15 years of age
for the survey used for the general public and children less than 6 years of age for the survey used
for CRITFC's member tribes), so the body weights also differ.
                                             4-54

-------
  Table 4-1. Exposure parameters used to calculate average daily dose for assessing noncarcinogenic health
  effects for potentially exposed populations
Potentially Exposed Population
Exnosure Parameter
Tissue chemical concentration
Ingestion rate of fish tissue (g/day)
Adults
Children <15
Children <6
Exposure frequency (days/yr)
Exposure duration (yrs)
Adults
Children <15
Children <6
Body weight (kg)
Adults
Children <15
Children <6
Averaging time (days)
Adults

Children <15
Children <6
General
Abbreviation AFC
C Average
IR
7.5"
2.83"
-
EF 365
ED
30V70f
15
-
BW
70s
30h
-
AT
10,9507
25,550
5,475
-
Public
HFC
Average

142.4"
77.95"
-
365

30V70f
15
-

70s
30h
-

10,9507
25,550
5,475
-
CRITFC's
AFC
Average

63.2C
-
24.8C
365

30V70f
-
6

70s
-
15s

10,9507
25,550
-
2,190
member tribes
HFC
Average

389"
-
162d
365

30e/70f
-
6

70s
-
15s

10,950/25,550

-
2,190
AFC - average fish consumption ; HFC - high fish consumption
"Mean U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA, 2000b).
" 99th percentile U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA ,2000b).
c Mean consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the Columbia
River Basin (CRITFC, 1994)
d 99th percentile consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the
Columbia River Basin (CRITFC, 1994).
e 90th percentile length of time  an individual stays at one residence (USEPA, 1997b)
f Average life expectancy of the general public (USEPA, 1989).
g Average body weight for adults (male and female) in the general public (USEPA, 1989).
11 Average body weight for children of both sexes of age 6 months to 15 years in the general public (USEPA, 1997c).  Corresponds
to ingestion rate data for children taken from USEPA 2000b.
'Average body weight for children of both sexes frm the age of 6 months through 5 years in the general public (USEPA, 1997c).
Corresponds to ingestion rate data for children in CRITFC, 1994.
                                                  4-55

-------
    Table 4-2. Exposure parameters used to calculate average daily dose for assessing carcinogenic risks
    for potentially exposed populations.

                                                              Potentially Exposed Population
                                                           General Public
                                   CRITFC's member
                                         tribes
    Exposure Parameter
Abbreviation
  AFC
 HFC
 AFC
HFC
    Tissue chemical concentration


    Ingestion rate offish tissue (g/day)

                             Adults


    Exposure frequency (days/yr)


    Exposure duration (yrs)
    Body weight (kg)
    Averaging time (days)
                             Adults
                             Adults
     C


    IR



    EF


    ED



    BW



    AT
Average   Average    Average   Average
  7.5*
142.4"
 63.2C
389d
  365       365
            365       365
 30V70f     30V70f      30V70f     30V70f
  70s
 25,550
25,550
            70s
25,550
           70s
   AFC - average fish consumption ; HFC - high fish consumption
   "Mean U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA, 2000b).
   b 99th percentile U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA ,2000b).
   c Mean consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the Columbia
  River Basin (CRITFC, 1994)
   d 99th percentile consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the
   Columbia River Basin (CRITFC, 1994).
   e 90th percentile length of time an individual stays at one residence (USEPA, 1997b)
   f Average life expectancy of the general public (USEPA, 1989).
  s Average body weight for adults (male and female) in the general public (USEPA, 1989).


4.4     Exposure Point Concentrations (Chemical Concentrations in Fish)


The exposure point concentrations for this risk assessment are the average chemical
concentrations in uncooked fish tissue. Exposure point concentrations for fish tissue or shellfish
are commonly based on average concentrations (USEPA, 1989).  The average concentrations are
assumed to be representative of the chemical concentrations to which fish consumers would most
likely be exposed over the long exposure durations being used in this risk assessment.


Ideally, the concentrations used as the exposure point concentrations for an individual should
represent the average chemical concentrations in fish found at study sites  where fish are collected
for consumption during  the exposure duration.  Fishing study  site preferences within the
Columbia River Basin are available for members of the Nez Perce, Umatilla, Yakama, and Warm
Springs Tribes (CRITFC, 1994); these preferences were used in designing the sampling plan for
this study. However, similar information is not available for the general public. To try and
maximize the information conveyed in this risk assessment and allow individuals to assess their
own risks based on their fishing practices, the data for each fish species were pooled by (1) study
                                               4-56

-------
site - all replicate samples for a given fish species and tissue type collected at a study site were
averaged to produce a "study site" average and (2) basin-wide all samples for a given fish species
and tissue type collected in the Columbia River Basin during this study were averaged to
calculate the "basin-wide" averages.  The calculation of these study site and basin-wide averages
were previously discussed in Section 1.

4.5    Fish Ingestion Rates

4.5.1   Fish Ingestion Rates for the General Population

Three fish consumption surveys were completed in the Columbia River Basin: two for the
Willamette River, Oregon and one for Lake Roosevelt, Washington (EVS, 1998; Adolfson
Associates, Inc., 1996; WDOH, 1997).  A brief description of these surveys is presented in this
section. Although these three surveys do not provide fish ingestion rates that can be used for this
risk assessment, they do provide useful information on the species offish consumed in different
parts of the basin and on the parts of the fish that are eaten.

In 1998, EVS Environment Consultants (EVS, 1998) conducted a qualitative fish consumption
survey for a 45-mile stretch of the Willamette River extending downstream from  Wheatiand
Ferry to the Willamette Falls near Oregon City, Oregon. Information on fish consumption was
obtained by conducting phone interviews with individuals representing various community
centers, fishing guide services, ethnic associations, fishing-related government agencies and
businesses.  The survey indicated that anglers are consuming bullhead, carp, sucker, bass,
northern pikeminnow, crappie, bluegill, trout, white sturgeon, lamprey, salmon, and  steelhead
from this section of the Willamette River.  All respondents indicated that muscle tissue was the
most commonly consumed portion of the fish, although some respondents indicated that the skin,
eggs, eyes, and the entire fish were being consumed (EVS, 1998).

In 1995, Adolfson Associates (Adolfson Associates, Inc., 1996) conducted a fish consumption
survey by interviewing anglers along the Columbia Slough and Sauvie Island at the mouth of the
Willamette River, Oregon This survey found that Caucasians made up the majority of individuals
consuming fish from these locations.  The ethnic descent of Columbia Slough anglers was 47%
Caucasians of eastern European descent, 22% Hispanic, 19% African American, 8% Caucasian
(excluding eastern Europeans), and 3% Asian.  The most commonly caught fish  was carp,
followed by yellow perch and banded sculpin.  The ethnic descent of Sauvie Island anglers was
67% Caucasian (excluding eastern Europeans), 16% Asian, 8% African American,  and 2%
Hispanic. The most commonly caught fish was yellow perch, followed by brown bullhead,
northern pikeminnow, starry flounder, and white sturgeon. Anglers from both locations indicated
the most commonly consumed portion offish was muscle tissue.

In 1994, the Washington State Department of Health (WDOH, 1997), in cooperation with the
Spokane Tribe of Indians, conducted a fish consumption survey of anglers fishing within Lake
Roosevelt, Washington,  a 151-mile stretch of water extending upstream from the Grand Coulee
Dam on the Columbia River to the United States-Canada border.  Fish consumption data were
collected using a survey form and from creel surveys. The majority of anglers surveyed consisted

                                            4-57

-------
of individuals who repeatedly fish from Lake Roosevelt.  Surveyed anglers were mainly male
(90%), Caucasian (97%), and over fifty years of age (60%).  The most frequently consumed
species were rainbow trout, followed by walleye, kokanee, and bass. The average annual number
offish meals consumed by respondents was 42 meals per year.  Assuming a typical meal size of 8
ounces, this average consumption rate corresponds to a daily fish consumption rate of 26 g/day.
Fillets were the primary portion of the fish consumed; few anglers consumed fish skin, eggs, or
fish head.

Because these three studies provide only a limited amount of information on fish consumption
rates for the general public within the Columbia River Basin, a recent EPA fish consumption
report (USEPA, 2000b) was used to select the fish consumption rates for this risk assessment that
may be representative of adults and children within the general public that consume average and
high amounts offish. The fish consumption rates reported by EPA are based on data collected
from the combined 1994, 1995, and 1996 Continuing Survey of Food Intakes by Individuals
(CSFn), conducted annually in all 50 states by the United States Department of Agriculture. The
CSFII was conducted by interviewing over 15,000 respondents according to a stratified design
that accounted for geographic location, degree of urbanization, and socioeconomics. Eligibility
for the survey was limited to households with gross incomes at or less than 130% of the federal
poverty guidelines. The mean daily average per capita (fish consumers and non-consumers) fish
consumption rates of freshwater and estuarine fish (uncooked) reported by EPA (USEPA, 2000b)
for adults (7.5 g/day) and children (14 years of age and younger, 2.83 g/day) were selected to be
representative of average fish consumption by the general public within the Columbia River
Basin.  The 99th percentile per capita fish consumption rates of freshwater and estuarine fish
(uncooked) reported by EPA (USEPA, 2000b) for adults  (142.4 g/day) and children (14 years of
age and younger, 77.95 g/day) were selected to be representative of high fish consumption by the
general public within the Columbia River Basin.

4.5.2  Fish Ingestion Rates for CRITFC's Member Tribes

During 1991-1992, CRITFC conducted a comprehensive survey offish  consumption by members
of the Nez Perce, Umatilla, Yakama, and Warm  Springs Tribes that possess fishing rights to
harvest anadromous fish and resident fish species originating in streams and lakes flowing
throughout the Columbia River Basin (CRITFC,  1994). The survey data were collected by
interviewing a total of 513 adult tribal members.  Information obtained in this survey included
age-specific fish consumption rates, the fish species and parts of the fish consumed, and the
methods used to prepare the fish for consumption. Salmon and steelhead were consumed by the
largest number of adult respondents followed by trout, lamprey and smelt.  The survey
determined that the average consumption rate offish by adults and children (5 years of age and
younger) who consume fish was 63.2 g/day and 24.8 g/day, respectively. The 99th percentile fish
consumption rates of adults and children (5 years of age and younger) who consume fish was 389
g/day and 162 g/day, respectively. The average and 99th percentile fish  consumption rates were
selected as representative values for average and high fish consumption by CRITFC's member
tribes.

The fish consumption survey conducted by CRITFC (1994) showed that fish consumption by

                                            4-58

-------
CRITFC's member tribes is considerably higher than that of the general public.  The average and
99th percentile fish consumption rates for adults in CRITFC's member tribes are higher by factors
of 8.4 and 2.7, respectively, than the corresponding per capita fish consumption rates reported for
the general public by EPA (USEPA, 2000b).  It should be noted that Harris and Harper (1997)
have suggested that a fish consumption rate of 540 g/day represents a reasonable subsistence fish
consumption rate for CRITFC's member tribes who pursue a traditional lifestyle.  The value of
540 g/day was based  on the authors' review of several non-subsistence Native American studies,
two subsistence studies, and personal interviews (by the authors or others) of members of the
Umatilla and Yakama Tribes.  This value of 540 g/day is  1.4 times the 99th percentile fish
consumption rate reported by CRITFC (1994) which is used as the high-end consumption rate for
CRITFC's member tribes in this risk assessment.

Some individuals may find it difficult to assess their fish consumption in terms of grams per day.
Two other common ways to present this information is in terms of 8-ounce fish meals over some
period of time or in terms of pounds per year.  An 8-ounce meal size is the value recommended
by EPA (USEPA, 2000a) for fish meals. This meal size was also the most commonly selected
(48.5%) serving size for adult fish meals based on the CRITFC (1994) survey of its member
tribes.

Table 4-3 shows the fish consumption rates used in this risk assessment expressed in different
units.

     Table 4-3. Fish consumption rates expressed in alternative units.
                                                             Consumption Rate Units
Target Population
General public - average fish consumption
Adults
Children <15
General public - high fish consumption
Adults
Children <15
CRITFC's member tribes - average fish consumption
Adults
Children <6
CRITFC's member tribes - high fish consumption
Adults
Children <6
g/day

7.5s
2.83"

142.4"
77.95"

63 .2C
24. 8C

389d
162d
8-oz Meals

12 meals/year
5 meals/year

1 9 meals/month
1 1 meals/month

2 meals/week
40 meals/year

12 meals/week
5 meals/week
Lbs/yr

6.0
2.3

114.6
62.7

50.8
20.0

313
131
  "Mean U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA, 2000b).
  " 99th percentile U.S. per capita consumption rate of uncooked freshwater and estuarine fish (USEPA , 2000b).
  c Mean consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the Columbia
  River Basin (CRITFC, 1994)
  d 99th percentile consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the
  Columbia River Basin (CRITFC, 1994).

As  discussed in Section 1 of this report, a small number of egg samples were collected for some


                                            4-59

-------
of the anadromous fish species. There are no studies for the Columbia River Basin with
quantitative ingestion rates for eggs. Therefore, a risk characterization for eggs was not included
in the Risk Characterization Section (Section 6) of this report. However, an example risk
characterization for eggs is presented in the Uncertainty Section (Section 10). This example for
eggs is very uncertain but serves as a useful comparison to the results for fish tissue.

4.6    Exposure Frequency

An exposure frequency of 365 days per year was assumed for calculation of the average daily
dose. While not all fish species analyzed for this risk assessment can be collected by anglers
throughout the year, an exposure frequency of 365 days per year was assumed for all fish species
since anglers might catch and freeze fish for later consumption or receive fish for consumption
from other anglers.

4.7    Exposure Duration

The  exposure duration is the length of time over which exposure occurs at the concentrations and
ingestion rates specified by the other parameters in Equation 4-1.  Specific information on the
length of time over which the general public or CRITFC's member tribes may be consuming fish
from the Columbia River Basin is not available. Therefore estimates of exposure duration were
made for this risk assessment.

4.7.1  Adults

Two exposure durations, 30 years and 70 years, were assumed for calculations of the adult
average daily intake in this risk assessment. Thirty years is the national 90th percentile length of
time that an individual stays at one residence (USEPA,  1997b).  This value is  recommended by
EPA (USEPA, 1989) as a reasonable maximum exposure duration when assessing the potential
health risks for a residential exposure scenario.

A 70-year exposure duration was selected to assess the potential health risk of a lifetime exposure
to chemicals detected in fish tissue. The average life expectancy  of the general population in the
United States is 72 years for males and 79 years for females (USEPA, 1997c). EPA (USEPA,
1997c) suggests that 75 years is an appropriate value to reflect the average life expectancy of the
general population. A value of 70 years was selected as a lifetime exposure duration in this risk
assessment because this value has been commonly used in other regional human health risk
assessments offish consumption (e.g., Tetra Tech, 1996; EVS, 2000) to represent the exposure
duration for those individuals (e.g., tribal members) who fish from one area their entire life.  In
addition,  since a 70-year lifetime is used to derive cancer slope factors (USEPA, 2000c), the use
of 70 years avoids the necessity of having to adjust the cancer slope factors used in this risk
assessment.
                                             4-60

-------
4.7.2  Children

An exposure duration of 15 years was used to estimate the average daily dose for children in the
general public. This exposure duration was selected for children because it corresponds to the
age range for which the fish consumption rate data were developed for children in the CSFn
Survey (USEPA, 2000b).

An exposure duration of 6 years was used to estimate the average daily dose of children for
CRITFC's member tribes.  This exposure duration was selected because it corresponds to the age
range for which fish consumption data were reported by CRITFC (1994) for children up to 6
years of age.

4.8    Body Weight

The value for body weight in Equation 4-1 is the average body weight over the exposure period.
Information on the body weights of the individuals reported in the  CRITFC consumption survey
(CRITFC, 1994) and the CSFH consumption survey (USEPA, 2000b) were not available,
therefore data from the studies, discussed in the following sections, were used.

4.8.1  Adults

Existing EPA guidance (USEPA, 1989, USEPA, 2000a) recommends the use of a body weight of
70 kg (kilograms) to calculate adult exposures.  A 70 kg adult body weight is assumed for the
derivation of cancer slope factors in IRIS.  However, a more recent survey data of the population
in the United States suggests that a body weight of 71.8 kg may be more appropriate for adults
(USEPA, 1997c).

For this risk assessment, a 70 kg body weight was assumed for adults because its use is consistent
with EPA risk assessment guidance (USEPA, 2000f), it avoids the necessity of having to adjust
cancer slope factors to accommodate the 71.8 kg average body weight, and allows for
comparisons with other regional human health risk assessments offish consumption that also
used 70 kg as the adult body weight.

4.8.2  Children

A body weight of 30 kg was used to calculate the average daily dose of children in the general
public.  This body weight corresponds to the average weight of female and male children ages 6
months through age 14 (USEPA, 1997c). Six months through the age of age 14 is the age group
for which fish consumption data were collected in the CSFn Survey.

A body weight of 15 kg was used to calculate the average daily dose of children for the Columbia
River Basin tribes. This body weight corresponds to the average weight of female and male
children ages 6 months through age 5 (USEPA, 1997c). Six months through age 5 years is the
age group for which fish consumption data were collected in the CRITFC fish consumption
survey.

                                           4-61

-------
4.9    Averaging Time

As discussed earlier, exposure to contaminants in fish occurs over time.  Therefore the total
exposure is divided by the time period of interest (the averaging time) to obtain an average
exposure rate per unit time. When this average rate is expressed as a function of body weight, the
resulting exposure rate is referred to as the average daily dose (ADD) expressed in milligrams of
a chemical taken into the body per kilogram body weight per day (mg/kg/day).

The averaging time selected depends upon the type of toxic effect being assessed. When
evaluating exposures to non-cancer effects, exposures (dose) are calculated by averaging dose
over the period of exposure (for this risk assessment - 30 or 70 years for adults; 6 or 15 years for
children).  Since the averaging time (AT) is always the same as the time period over which
exposure occurs for non-cancer effects, exposure duration (ED), the exposure (dose) in
mg/kg/day is the same for both exposure durations within a target populations (e.g. the same for
both 30 and 70 years exposure duration for general public adults).

For evaluating cancer risks for adults, exposures are calculated by prorating the total dose over a
lifetime (70 years). The exposures calculated for cancer risk assuming 30 or 70 years exposure
duration are different from each other because the averaging time is always a lifetime or 25,550
days, but the exposure durations assumed for this report for adults are either 30 (10,950 days) or
70 years (25,550 days). Thus, in this report, cancer risks for both exposure durations (30 and 70
years) are presented.

4.10   Multiple-Species Diet Exposures

The cancer risk and non-cancer hazards that are discussed in most of Section 6 assume that
people eat only one species offish. For example, for estimating the cancer risk from consuming
white sturgeon, it is assumed that the adults in the general public, with high fish consumption
(142.4 g/day), consume 142.4 grams a day of white sturgeon for either 30 years or 70 years.

However, it is likely that many individuals consume more than one species offish from the
Columbia River Basin. When an individual consumes multiple fish species, additional exposure
information is needed on the relative amounts of different species in that individual's diet to
obtain an estimate of the individual's potential overall health risk. Because fish consumption
practices, including the types and amounts offish eaten, can vary greatly among individuals,
within populations because of differences in age, gender, cultural practices, and/or socioeconomic
status, it is difficult to generalize about the potential risk of an individual diet that includes the
consumption of multiple species. This  section includes the methods and the assumptions used in
the example of a multiple-species diet. This  example is intended to assist individuals to use the
data for individual fish species presented in this report to estimate their own risks when
consuming multiple species.

The example selected to illustrate the risk associated with consuming multiple species is based on
information obtained during the 1991-1992 survey offish consumption by members of the Nez
Perce, Umatilla, Yakama, and Warm Springs Tribes (CRITFC,  1994). The survey included 513

                                             4-62

-------
adult participants.  The percentage of these adults that consumed 10 fish species were also
presented in this survey (CRITFC, 1994; Table 17).  These percentages are included in this
section in Table 4-4, column A. To simplify the calculations, the responses from the CRITFC
survey for fall chinook salmon, spring chinook salmon, coho salmon, and steelhead were
combined into one category, salmon. To estimate the hypothetical diet, it was assumed that the
data in the CRITFC survey on percentages of adults consuming different fish species could be
used to estimate the percent that each fish species contributes to the hypothetical diet.  Table 4-4,
Column B, shows the percentage of the diet assumed for each fish species. Each species value in
Column B was calculated by dividing the percentage of each fish species consumed (based on the
CRITFC study and shown in Column A) by the sum of the percentages for all species in Column
A.  For example, the value of 27.7% shown for salmon in Table 4-4 (Column B) was obtained by
dividing the percentage of adults that consume salmon (92.4 in Column A) by the sum of the
percentages of consumption for all species (333.5 in Column A) and multiplying the result by 100
to express the fraction as a percentage:

(Equation  4-2)

Percent of diet composed   =   percentage of adults that consume salmon   x 100
 of salmon                     sum of the percentages for all species

                 27.7%  =      92.4   x 100
                                333.5

In Table 4-4, a consumption rate of 63.2 g/day (the average ingestion rate reported for adults in
CRITFC's member tribes (CRITFC, 1994), is used along with the percentages offish in the
hypothetical diet to calculate the consumption rates for each species in the hypothetical multiple
diet of an adult in CRITFC's member tribes with average fish consumption. Consumption rates
for  each species were calculated by multiplying 63.2 g/day by the percentage assumed in the
hypothetical diet for that species. For example, the consumption rate  of 17.5 g/day shown for
salmon in Table 4-4 (Column C) was obtained by multiplying the total average consumption rate
(63.2 g/day) for adults in CRITFC's member tribes by the percent that salmon was calculated to
represent (27.7%) in this multiple-species diet.

 (Equation 4-3)

Consumption rate for   =  Percent of hypothetical diet  X  Average adult ingestion
     salmon                  composed of salmon              rate for all species
 (g/day)

          17.5 g/day    =   27.7%  X 63.2 g/day

This multiple-species diet methodology was used to estimate exposure and to calculate cancer
risks and non-cancer hazards for adults in the general public and CRITFC member tribes in
Section 6.2.5 for both the average and high fish ingestion rates.  The hypothetical diet of multiple-
species based on the CRITFC fish consumption study was used for all of the adult populations.

                                            4-63

-------
The exposure due to ingestion of each species in the hypothetical diet was calculated by using the
same exposure parameters described for adults in Tables 4-1 and 4-2 except that the fish
consumption rates for the multiple-species diet scenario replaced those in the tables. For the
adults in CRITFC's member tribes with an average fish consumption rate, those ingestion rates in
Table 4-4 (Column C) were used. For the other 3 adult populations assessed (high fish
consumption rates for adults in CRITFC's member tribes; average and high fish consumption
rates for general public adults), species specific consumption rates were calculated using the
multiple diet method just described but using total fish consumption rates for that population and
the hypothetical multiple-species diet shown in Table 4-4.  Exposure for the hypothetical mixed
diet is the sum of all of the exposures calculated for each of the eight species that had ingestion
rates calculated in Table 4-4.

 Table 4^1. Description of the methodology used to calculate exposure for a multiple-species diet.
                                     A
                           Percentage of Adults that               B                         C
                                  Consume          Percentage of Hypothetical      Consumption Rate0
Species
Salmon"
Rainbow trout
Mountain whitefish
Smelt
Pacific lamprey
Walleye
White sturgeon
Sucker
Totals
Species
92.4
70.2
22.8
52.1
54.2
9.3
24.8
7.7
333.5"
Diet
27.7
21.0
6.8
15.6
16.3
2.8
7.4
2.3
100.0
(drams/davl
17.5
13.3
4.3
9.9
10.3
1.8
4.7
1.5
63.2
'This category includes spring chinook salmon, fall chinook salmon, steelhead and coho  salmon.
b Although shad and pikeminnow were included in the CRITFC fish consumption survey (CRITFC ,1994), this total does not
include values for these species because these two  species were not sampled in this study.
0 a consumption rate of 63.2 g/day (the average ingestion rate reported for adults in CRITFC's member tribes (CRITFC, 1994), is used along with
the percentages offish in the hypothetical diet to calculate the consumption rates for each species
                                                4-64

-------
5.0    Toxicity Assessment

The toxicity assessment for a chemical is done in two steps. The first step, hazard identification,
summarizes and weighs the available evidence regarding a chemical's potential to cause adverse
health effects, such as cancer, birth defects, or organ damage. The second step, dose-response
evaluation, provides an estimate of the relationship between the extent of exposure to the
contaminant and the likelihood of these adverse effects occurring. As part of the dose-response
assessment, toxicity values - reference doses (RfD) and cancer slope factors (CSFs) - are derived.
These toxicity factors are used with the exposures calculated using methods described in Section
4 to estimate cancer risks and non-cancer hazards.

For most environmental contaminants of concern, EPA has already performed the toxicity
evaluation and has made the results available in databases. For the risk characterization in this
section, all of the toxicity information, including the reference doses and cancer slope factors,
was obtained from three EPA toxicity databases. Information was preferentially obtained from
IRIS (USEPA, 2000c). If data were not available in IRIS, they were obtained from the fiscal year
1997 Health Effects Assessment Summary Tables (HEAST) (USEPA, 1997d), and finally, from
the EPA National Center for Environmental Assessment (NCEA).

A toxicity value  has not been developed for all chemicals analyzed in this study. Chemicals
currently without toxicity values are listed in Table 5-1. The potential health risks associated
with exposure to these chemicals were not evaluated.

            Table 5-1. Chemicals without oral reference doses and cancer slope factors. (Source:
            IRK, NCEA, USEPA, 2000c; USEPA, 1997d)
            Acenaphthylene                       1 -methy 1-Naphthalene
            alpha-Chlordene                        2-methy 1-Naphthalene
            Benzo(ghi)perylene                     4-Bromophenyl-Phenylether
            DDMU                               4-Chloroguaiacol
            delta-HCC                            4-Chlorophenyl-Phenylether
            Dibenzofuran                          3,4-Dichloroguaiacol
            gamma-Chlordene                      4-Chloro-3-methylphenol
            Pentachloroanisole                      4,5-Dichloroguaiacol
            Phenanthrene                          4,6-Dichloroguaiacol
            Retene                               3,4,5-Trichloroguaiacol
            Tetrachloroguaiacol                    3,4,6-Trichloroguaiacol
                                                 4,5,6-Trichloroguaiacol

Of the 23 chemicals listed in Table 5-1, only two, 2-methyl naphthalene and pentachloroanisole,
were detected in fish at greater than a 10% frequency.  Table 1-4 in Section 1 shows both the
detected and non-detected chemicals in this study. It should also be noted that although lead does
not have toxicity values (RfD,  CSF), lead toxicity is well characterized and is discussed in detail
in Section 7.

The remainder of this section is divided into three parts. First, the methods used to assess toxicity
data and develop reference doses for non-cancer effects are summarized in Section 5.1.  Next, the
methods used to assess carcinogenicity data and develop cancer slope factors are summarized in

                                               5-65

-------
Section 5.2. Finally, those chemicals for which unique assumptions and/or methods were used to
estimate the study site and basin-wide averages due to toxicological considerations are discussed
in Section 5.3.

5.1     Summary of Toxicity Assessment for Non-Cancer Health Effects

Summaries of the available toxicity information (e.g., results of animal tests and/or human
occupational studies) for each chemical are provided in IRIS, HEAST or by NCEA.  For those
chemicals that were analyzed for in fish in this study and that have toxicity values, a summary of
the types of non-cancer effects caused by that chemical is provided in Table 5-2.

In Table 5-2, the effects that can potentially result from exposure to each of these chemicals are
designated with a check or a closed circle.  For most chemicals, there is more than one type of
non-cancer health effect (e.g., effects on metabolism, effects on the immune system) that can
result from exposure to that chemical. The number of effects seen and the severity of a given
effect depend upon the level of exposure to that chemical, with both the number and severity of
effects usually increasing as exposure increases.

The RED is an estimate (with uncertainty spanning perhaps an order of magnitude or greater) of
the daily exposure to the human population, including sensitive sub-populations, that is likely to
be without an appreciable risk of deleterious effects during a lifetime (USEPA, 2000c).  To derive
the RfD, all available studies are first reviewed. If adequate human data are available, this
information is used as the basis of the RfD. Otherwise, animal studies are the basis of the RfD. If
several animal studies are available, the study on the most sensitive species (the species showing
the toxic effect at the lowest dose) is selected as the critical study for the basis of the RfD.  The
effect associated with the lowest dose which resulted in an observed adverse effect is referred to
as the "critical toxic effect". After the critical study and critical toxic effect have been selected,
the experimental exposure level at which no adverse effect is demonstrated  (the no-observable-
adverse-effect-level) for that effect is then defined.  The  no-observable-adverse-effect-level is
used as the basis for deriving the RfD and is in part based upon the assumption that if the critical
toxic effect is prevented then all toxic effects will be prevented. For example, for total Aroclors,
the RfD was based upon a rhesus monkey  study. This  study was designated as the critical study
and the RfD is based on the critical toxic  effects on the immune system that were found in the
study.  For some chemicals (e.g., methyl mercury), the  RfD may be based on more than one
critical toxic effect (central nervous system and developmental/reproductive effects).  Table 5-2
also contains information on critical health endpoints used to derive the RfD as well as other
adverse health effects.

To develop the RfD, the no-observable-adverse-effect-level, or the lowest-observed-adverse-
effect-level if no-observable-adverse-effect-level can be determined from the studies, is divided
by uncertainty factors and a modifying factor. These factors, which usually consist of multiples
of 10 or lower, are applied to account for the different areas of uncertainty and variability that are
inherent in the toxicological data.  They include:
                                              5-t

-------
       An uncertainty factor to account for variations in the sensitivity of the general population.
       This factor is intended to protect sensitive subpopulations (e.g., the elderly and children).

•      An uncertainty factor to extrapolate from animals to humans when animal data is used.

•      An uncertainty factor to account for the uncertainty if only a lowest-observed-adverse-
       effect-level instead of a no-observable-adverse-effect-level is available.

       An uncertainty factor if data from only short term rather than lifetime studies are
       available.

•      A modifying factor to account for additional uncertainties not already addressed (e.g., if
       there is a lack of data on reproductive or developmental effects in the experimental data).

For each chemical with non-cancer effects, Table 5-3 presents the oral reference dose for that
chemical, the confidence in the reference dose, the uncertainty factors and the modifying factor
associated with the reference dose, and the toxic effect from the critical study that the reference
dose was based upon.  For many chemicals, both oral and inhalation reference doses have been
developed and are included in EPA toxicity databases.  However, because the exposures assessed
in this study result from ingestion offish, only oral reference doses were used.
                                               5-67

-------
TABLE 5-2. CHEMICALS CONTRIBUTING TO NON-CANCER HAZARD INDICES (WITH TOXIC EFFECTS OF EACH CHEMICAL DENOTED BY • AND
GROUP
Metals
















Semivolatiles







ANALYTE
Aluminum
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium (VI)
Cobalt
Copper
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
2-Chloronaphthalene
2,4-Dinitro toluene
2, 6-Dinitro toluene
1 ,2,4-Trichlorobenzene
Acenaphthene
Anthracene
Benzene, 1,2-dichloro-
Benzene, 1,3-dichloro-
Metabolism

•









•


•

•








Blood and blood
formation







(• )




•





•
•





Immune system
Cardiovascular

; •
: •
1 (' )










: •










>,
0
C
~O
^


•
(' )

•







•
•




•





0
>
_l












•
•
•


•
•
•

•


(' )
Central Nervous System
(' )

•


•



•
•

•

•



•
•





Reproduction/developme
nt
Gastrointestinal or
intestinal lesions




: •

1 (' )



• :

• :












ro
>,
O)
<













•











-D
'o
>,
.C
1-
























(' )
0
_C
O











•


•


•







Adrenal gland




















•




Clinical signs

























Selenosis












•












Hyperpigmentation/kerat
osis


•






















No clear critical toxicity
endpoint








(' )






(' )






(' )
(' )

5-68

-------
TABLE 5-2. CHEMICALS CONTRIBUTING TO NON-CANCER HAZARD INDICES (WITH TOXIC EFFECTS OF EACH CHEMICAL DENOTED BY • AND
GROUP









Guaiacols/
Phenols




Pesticides








ANALYTE
Benzene, 1,4-dichloro-
bis(2-Chloroisopropyl)ethei
Fluoranthene
9H-Fluorene
Hexachloroethane
Hexachlorobutadiene
Naphthalene
Nitrobenzene
Pyrene
2-Chlorophenol
2,3,4,6-Tetrachlorophenol
2 ,4-Dichlorophenol
2,4-Dimethylphenol
2,4, 5 -Trichlorophenol
Pentachlorophenol
Phenol
Aldrin
Chlordane (total)
DDT'
Endosulfan sulfate
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
gamma-HCH
Mirex
Metabolism






•







•
•

•

•





1 Blood and blood
formation

•
•
•



•




•












Immune system
Cardiovascular











• ;







j •




: •
>,
0
C
~O
^


•

•
•

•
•




•
•
•



•



•
•
0
>
_l
C )

•

•


•

•
•
•

•
•

•
•
•

•
•
•
•
•
Central Nervous System











•
•

•

•
•
•
•
•
•
•
•
•
Reproduction/developme
nt
Gastrointestinal or
intestinal lesions
(• );



• :




• j





• :


• :
• j

• :



ro
>,
D)
<

























-D
'o
>,
.C
1-
























•
0
_c
O




















•

•


Adrenal gland







•

















Clinical signs












•












Selenosis

























IHyperpigmentation/kerat
osis

























No clear critical toxicity
endpoint

























5-69

-------
TABLE 5-2. CHEMICALS CONTRIBUTING TO NON-CANCER HAZARD INDICES (WITH TOXIC EFFECTS OF EACH CHEMICAL DENOTED BY • AND













GROUP
PCBs













ANALYTE
Total Aroclors b








E
(ft

o

s
0
5







0
o
^
ra °

~o ro
0 £
5 £







E ! -
0 : ra
l\ 8
0 i >
C : O
3 : 	
I i "H
C : TO
i ; o
• :











0
c
5













0
5
•
E
0

(f)

CO
3
0
0
^
ro

"c
0
O
•
0 ;
Q. :
O |

0 ! ±_
> | O yj
5 - "™ o
C : C 	
° r-K s
0 1 » =
^ E c ro


S. i to »
.S^iS^
• :










ro

D)













0

H













0
O








"O
c
"5)


c
0
<








(f)
C
D)
	

O
'c
O









w
'w
0

_0
0
CO

ro
0


c
0
"ro
•^
0
O)
Q_

Q. (/)
X 0

^

'o

X
o
"ro
0
o
j— c


° -D
Z S

• - Chronic oral reference dose for this chemical is based on this health endpoint (critical effect). All chemicals with a • for a given health endpoint were summed to obtain an
estimate of the hazard index.
(• ) - Chronic oral reference dose has been developed for this chemical but the critical effect used is not clear. Although hazard quotients were calculated for these chemicals and
summed into the total hazard index, these chemicals were not summed into endpoint-specific hazard indices.
• -  Other observed health endpoints
'Comprised of DDE, ODD, and DDT.
b For each species, total Aroclors is the sum of detected Aroclors, which includes at least one of the following: Aroclor  1242, Aroclor 1254, and Aroclor 1260.
                                                                        5-70

-------
Table 5-3. Oral reference doses (RfDs) used in this assessment, including the level of confidence in the RfD, uncertainl
factors (UF) and modifying factor (MF) used to develop the RfD, and the toxic effect(s) from the critical study that the
RfD was based upon.
Chemical
1 ,2,4-Trichlorobenzene
2,3,4,6- Tetrachlorophenol

2,4,5-Trichlorophenol
2-Chloronaphthalene

2-Chlorophenol
2,4-Dichlorophenol

2,4-Dimethylphenol

2,4-Dinitrotoluene

2 , 6 -D initro to luene


Acenaphthene
Aldrin
Aluminum
Anthracene



Antimony
Total Aroclor a




Arsenic, inorganic b

Barium
Benzene, 1 ,2-dichloro-
Benzene, 1,3-dichloro-

Benzene, 1 ,4-dichloro-
Beryllium
bis(2-
Chloroisopropyl)ether
Cadmium
Chlordane (total) c
Chromium (VI)
Cobalt
Copper
DDTd
Oral RfD
(mg/kg-day)
l.OxlO'2
3.0 xlO-2

l.OxlO4
8.0 x 10-2

5.0 xlO-3
3.0 xlO'3

2.0 x 10-2

2.0 x 10-3

l.OxlO-3


6.0 x 10-2
3.0 xlO'5
1.0
3.0xl04



4.0 xlO4
2.0 x 10-5




3.0 xlO4

7.0 x ID'2
9.0 x 10-2
9.0 xlO4

3.0 xlO-2
2.0xlO-3
4.0 x 10-2

l.OxlO-3
5.0 xlO4
3.0 xlO-3
6.0 x ID'2
3.7 xlO-2
5.0 xlO4
Confidence
Medium
Medium

Low
Low

Low
Low

Low

High

_


Low
Medium
-
Low



Low
Medium




Medium

Medium
Low
_

-
Lowto Medium
Low

High
Medium
Low
-
-
Medium
UF/MF
1000/1
1000/1

1000/1
3000/1

1000/1
100/1

3000/1

100/1

3000


3000/1
1000/1
-
3000/1



1000/1
300/1




3/1

3/1
1000/1
_

-
300/1
1000/1

10/1
300/1
300/3
-
-
100/1
Critical Effect
Increased adrenal weight
Increased liver weights and centrilobular
hypertrophy
Liver and kidney pathology
Dyspnea, abnormal appearance, liver
enlargement
Reproductive effects
Decreased delayed hypersensitivity
response
Clinical signs (lethargy, prostration, and
ataxia) and hematological changes
Neurotoxicity, Heinz bodies and biliary
tract hyp erplasia
Mortality, neurotoxicity, Heinz bodies
effects, methemoglobinemia, bile duct
hyperplasia, and kidney histopathology
Hepatotoxicity
Liver toxicity
Minimal neurotoxicity
No treatment-related specific
toxicological endpoints observed in mice
at the doses administered in laboratory
studies
Longevity, blood glucose, cholesterol
Ocular exudate, inflamed and prominent
Meibomian glands, distorted growth of
finger- and toenails; decreased antibody
(IgG and IgM) response to sheep
erythrocytes
Hyperpigmentation/keratosis and possible
vascular complications
Hypertension and kidney effects
None identified
No identified critical toxicological
endpoint
Liver and reproductive effects
Small intestinal lesions
Decrease in hemoglobin and possible
erythrocyte destruction
Significant proteinuria
Hepatic necrosis
Gastrointestinal effects
Polycytemia - too many red blood cells
Unspecified
Liver lesions
Source
USEPA, 2000c
USEPA, 2000c

USEPA, 2000c
USEPA, 2000c

USEPA, 2000c
USEPA, 2000c

USEPA, 2000c

USEPA, 2000c

USEPA 1997e


USEPA, 2000c
USEPA, 2000c
NCEA
USEPA, 2000c



USEPA, 2000c
USEPA, 2000c




USEPA, 2000c

USEPA, 2000c
USEPA, 2000c
NCEA

NCEA
USEPA, 2000c
USEPA, 2000c

USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
NCEA
USEPA 1997e
USEPA, 2000c
5-71

-------
  Table 5-3.  Oral reference doses (RfDs) used in this assessment, including the level of confidence in the RfD, uncertainty
  factors (UF) and modifying factor (MF) used to develop the RfD, and the toxic effect(s) from the critical study that the
  RfD was based upon.
  Chemical
 Oral RfD
(mg/kg-day)
Confidence     UF/MF  Critical Effect
Source
  Endosulfan sulfate
  Fluoranthene
  Fluorene
  6.0 xlO'3        Medium       100/1   Reduced body wt. gain, increased            USEPA, 2000c
                                           incidence of marked progressive
                                           glomerulonephrosis in males

  4.0 xlO'2          LOW        3000/1  Nephropathy, increased liver weights,        USEPA, 2000c
                                           hematological alterations, and clinical
                                           effects

  4.0 Xl0'2          LOW        3000/1  Decreased red blood cell, packed cell          USEPA, 2000c
                                           volume and hemoglobin
gamma-HCH (Lindane)
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
Manganese
Methylmercury e
Mirex
Naphthalene
Nickel, soluble salts
Nitrobenzene
Pentachlorophenol
Phenol
Pyrene
Selenium
Silver
Thallium f
Vanadium
Zinc
3.0 xlO4
5.0 xlO4
1.3xlO-5
8.0 xlO4
2.0 xlO4
l.OxlO-3
1.4xl04
l.OxlO4
2.0 xlO4
2.0 x 10-2
2.0 x 10-2
5.0 xlO4
3.0 xlO-2
6.0 x 10-1
3.0 xlO-2
S.OxlO-3
5.0 xlO-3
9.0 x 10-5
7.0 x 10-3
3.0xl04
Medium
Low
Low
Medium
-
Medium
-
Medium
High
Low
Medium
Low
Medium
Low
Low
High
Low
Low
-
Medium
1000/1
300/1
1000/1
100/1
1000
1000/1
1/1
10/1
300/1
3000/1
300/1
10,000/1
100/1
100/1
3000/1
3/1
3/1
3000/1
100
3/1
Liver and kidney toxicity
Liver weight increases in males
Increased liver-to-body weight ratio in
both males and females
Liver effects
Renal tube regeneration
Atrophy and degeneration of the renal
tubules
CNS effects
Developmental neurological abnormalities
in human infants
Liver cytomegaly, fatty metamorphosis,
angiectasis; thyroid cystic follicles
Decreased average terminal body weight
in males
Decreased body and organ weights
Hematologic, adrenal, renal and hepatic
lesions
Liver and kidney pathology
Reduced fetal body weight
Kidney effects (renal tubular pathology,
decreased kidney weights)
Clinical selenosis, liver dysfunction
Argyria
Increased levels of SGOT8 and LDHh
Unspecified
47% decrease in erythrocyte superoxide
dismutase (ESOD) concentration in adult
females after 10 weeks of zinc exposure
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA 1997e
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
" For each fish species, total Aroclors is the sum of detected Aroclors, which includes at least one of the following: Aroclor 1242, Aroclor 1254,
andAroclor 1260. The toxicity value for Aroclor 1254 was used.
b Total arsenic was measured. Inorganic arsenic was assumed to represent 10% of the total arsenic concentration (see Section 5.3.3).
cChlordane (total) is the sum of cis-chlordane, cis-nonachlor, oxychlordane, trans-chlordane, and trans-nonachlor.
d Toxicity value for p,p'-DDT used.
'Reported as mercury in data set.
'Toxicity value based on thallium nitrate.
8Seram glutamic oxaloacetic transaminase.
h LDH-lactate dehydrogenase.

-------
5.2    Summary of Toxicity Assessment for Cancer

In the hazard identification step for cancer, summaries of the available toxicity information (e.g.,
results of animal tests and/or human occupational studies) on a chemical are reviewed. For
cancer, this review is done to determine if that chemical is likely to cause cancer in humans.
Based upon this evaluation, a chemical is classified into one of five weight-of-evidence classes
that have been developed by EPA.  These classes, shown in Table 5-4, define the potential for a
chemical to cause cancer in humans.

           Table 5^4. EPA weight-of-evidence classifications for carcinogens. (USEPA, 2000c).
              Weight-of-Evidence
                 Classification                            Category
                     A            Human carcinogen
                     B            Probable human carcinogen
                     C            Possible human carcinogen
                     D            Not classifiable as a human carcinogen
                     E            Evidence of noncarcinogenicity in humans

In the second part of the toxicity assessment, the dose-response assessment, the toxicity values
(CSFs) used to estimate cancer risk are developed. Based upon the manner in which some
chemicals are thought to cause cancer, no exposure is thought to be without risk. Therefore, in
evaluating cancer risks, a "safe" level of exposure cannot be estimated. To develop toxicity
values for carcinogens, mathematical models are used to extrapolate from high levels of exposure
where effects have been seen in animal studies or human studies to the lower exposures expected
for human contact in the environment.  The result of this extrapolation is a dose-response line
whose slope is known as the cancer slope factor.

Table 5-5 shows the cancer slope factors for the 23 chemicals evaluated for cancer in this risk
assessment. Because of the method used to develop these cancer slope factors, they are
considered to be a plausible upper-bound estimate of the cancer potency of a  chemical. By using
these upper-bound estimates for the cancer slope factors, there is reasonable confidence that the
actual cancer risks will not exceed the estimated risks calculated with these slope factors and may
actually be lower. Table 5-5 also includes the weight-of-evidence classification for each
carcinogen, the type of tumor that the cancer slope factor was based upon, and the source of this
information. As previously discussed with reference doses, for many chemicals, both oral and
inhalation cancer slope factors have been developed and are included in EPA toxicity databases.
However, because the exposures assessed in this study result from ingestion offish, only oral
cancer slope factors were used.
                                             5-73

-------
Table 5-5. Oral cancer slope factors with their weight of evidence classification with the type(s) of tumor the
cancer slope factor is based upon.
Cancer Slope
Factor
Chemical (kg-d/mg)
2,3,7,8-TCDD
1 ,2-Dipheny Ihydrazine
2,4,6-Trichlorophenol
Aldrin
alpha-HCH (alpha-BHC)
Adjusted Aroclors a
Arsenic, inorganic
1 ,4-dichlorobenzene
Benzo(a)pyrene
beta-HCH (beta-BHC)
bis(2-Chloroisopropyl)ether
Chlordane (total)b
DDD (total)0
DDE (total)0
DDT (total)c
gamma-HCH (Lindane)
Heptachlor

Heptachlor epoxide
Hexachlorobenzene
Hexachlorobutadiene

Hexachloroethane
Pentachlorophenol
l.SxlO5
8.0
l.lxlO'2
1.7x10'
6.3
2.0
1.5
2.40 x lO'2
7.3
1.8
7.0 x 10-2
3.5 x 104
2.4 x 104
3.4 xlO4
3.4 xlO4
1.3
4.5

9.1
1.6
7.8 x 10-2

1.4xlO-2
1.2xlOJ
Weight of
Evidence Tumor type
B2
B2
B2
B2
B2
B2
A
C
B2
C
C
B2
B2
B2
B2
B2-C
B2

B2
B2
C

C
B2
Respiratory system and liver tumors
Hepatocellular carcinomas and
neoplastic liver nodules
Leukemia
Liver carcinoma
Liver tumors
Hepatocellular carcinomas
Skin cancer, internal organs (liver,
kidney, lung, bladder)
Liver tumors
Forestomach, squamous cell
papillomas and carcinomas
Benign liver tumors
Liver and lung tumors
Non-Hodgkin"s lymphoma and
liver tumors
Lung, liver, and thyroid tumors
Liver and thyroid tumors
Liver
Liver tumors
Hepatic nodules and hepatocellular
carcinomas
Liver carcinoma
Liver, thyroid, kidney tumors
Renal tubular adenomas and
adenocarcinomas
Hepatocellular carcinomas
Hepatocellular adenoma/carcinoma,
Source
USEPA, 1997d
USEPA, 2000c
USEPA, 2000c
USEPA , 2000c
USEPA, 2000c
USEPA, 1996
USEPA, 2000c
USEPA, 1997d
USEPA, 2000c
USEPA, 2000c
USEPA, 1997d
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 2000c
USEPA, 1997d
USEPA, 2000c

USEPA, 2000c
USEPA, 2000c
USEPA, 2000c

USEPA, 2000c
USEPA, 2000c
  Toxaphene
1.1
B2
pheochromocytoma/malignant
pheochromocytoma,
hemang io sarcoma/hemang ioma
Hepatocellular carcinoma and
neoplastic nodules
USEPA, 2000c
aFor each fish snecies. adjusted Aroclors is the sum of detected Aroclors less the sum of detected PCB congeners. Detected Aroclors included at
least one of the following: Aroclor 1242, Aroclor 1254, and Aroclor 1260.
b Chlordane (total) is the sum of alpha-chlordane, cis-nonachlor, gamma-chlordane, oxychlordane, and trans-nonachlor.
c Slope factor for DDD (total), DDE (total), and DDT (total) based on the p,p' isomers.
                                                           5-74

-------
5.3    Special Assumptions and Methods Used For Selected Chemicals

The average study site and basin fish contaminant levels for some of the chemicals in this risk
characterization were calculated using unique assumptions.  The need for these assumptions
results from the lack of non-cancer toxicity values (reference doses) for each of the isomers of
chlordane; for DDE and DDD; and for Aroclors 1242 and 1260 (Section 5.3.1); special methods
for calculating cancer risks for chlorinated dioxins/furans, Aroclors and dioxin-like PCB
congeners, and PAHs (Section 5.3.2); and the differential toxicity among arsenic species (Section
5.3.3).

5.3.1   Non-Cancer Toxicity Values for Chlordanes, DDT/DDE/DDD, and Aroclors

For non-cancer effects for chlordanes, DDT/DDE/DDD, and Aroclors, the average fish
contaminant levels were calculated as summed quantities of individual chemicals in the class of
chemicals.  This summation methodology was applied to these three classes of chemicals because
toxicity values were not available for all individual chemicals in these three classes and these
chemicals were commonly detected in fish tissue. Use of this methodology assumes that the
mechanisms of action for all of the chemicals in a class of chemicals are the same.

•      Total  chlordane was calculated as the sum of c/s-chlordane, trans-chlordane, cis-
       nonachlor, frvms-nonachlor, and oxychlordane.  Non-cancer health effects for total
       chlordane were based on the reference dose for technical chlordane (USEPA, 2000c).
       Technical chlordane is not a single chemical, but is a mixture of several closely related
       chemicals, which consist of some of the various chlordane isomers and metabolites,
       including: cis-chlordane, trans-chlordane, cis-nonachlor, trans-nonachlor, and chlordenes,
       and other compounds.

       Total  DDT was calculated by summing the ortho-para and para-para isomers of DDT,
       DDD, and DDE. IRIS contains a reference dose for DDT, but there are no specific
       reference doses for DDE or DDD.  However, because the structures and toxicities of DDD
       and DDE closely resemble that of DDT (see Toxicity Profiles in Appendix B), for
       purposes of this risk characterization, it was assumed that they (and their various ortho-
       and para-isomers) have the same reference dose as DDT.

•      Although PCB congeners were analyzed using two different methods: 1) Aroclors and 2)
       individual PCB congeners, non-cancer health effects were estimated only for Aroclors as
       EPA has not established an oral reference dose for individual PCBs congeners (USEPA,
       2000c). Three Aroclors were detected in fish tissues, depending on the particular fish
       species, study site, and tissue type: Aroclor 1242, Aroclor 1254, and Aroclor 1260. The
       types  and amounts of specific PCB congeners (each of which have their individual
       associated toxicity) differ in these three Aroclor mixtures.  Only one of the Aroclors
       detected in this study has an oral reference dose, Aroclor 1254. Therefore, to provide a
       health protective estimate of non-cancer health impacts, the oral reference dose for
       Aroclor 1254 was also used for Aroclor 1242 and Aroclor 1260.
                                             5-75

-------
5.3.2  Cancer Toxicity for Chlorinated Dioxins/Furans, Dioxin-Like PCB congeners, and
       PAHs

The toxicity of the chlorinated dioxins/furans and dioxin-like PCB congeners were evaluated
using toxicity equivalence factors recommended by WHO (Van den Berg et al,  1998). Table 2-
10 (Section 2.7) listed the seventeen 2,3,7,8-substituted dioxin and furan congeners and 11
dioxin-like PCB congeners with 2,3,7,8-TCDD toxicity equivalence factor values. The toxicity
equivalence factors were developed using careful scientific judgement after considering all
available scientific data and are an order-of-magnitude estimate of the toxicity of these
compounds relative to 2,3,7,8-TCDD.

Cancer risks from exposure to polycyclic aromatic hydrocarbons (PAHs) found in fish tissue in
this study that are thought to be carcinogens were estimated from methods described in EPA
guidance (USEPA, 1993). A cancer slope factor is available for one PAH only, benzo(a)pyrene.
Relative potency factors have been developed for six PAHs (benz(a)anthracene,
benzo(b)fluoranthene, benzo(k)fluoranthene, chrysene, dibenz(ah)anthracene, indeno(l,2,3-cd)
pyrene) relative to benzo(a)pyrene (see Table 5-6) (USEPA,  1993). These relative potency
factors are used to convert the concentrations of the six PAHs into benzo(a)pyrene equivalent
concentrations. As with the toxicity equivalence factors for chlorinated dioxins and furans and
dioxin-like PCB congeners, these relative potency factors are order-of-magnitude estimates and,
therefore, have inherent uncertainties.  However, unlike the toxicity equivalence factors, these
relative potency factors for the PAHs are to be considered as  an "estimated order of potential
potency" because they do not meet all of the guiding criteria for the toxicity equivalence method
described by EPA for PCB mixtures (USEPA, 1991).

            Table 5-6. Relative potency factors for PAHs (USEPA,1993).
            Chemical	Relative Potency Factors	
            Benz(a)anthracene                                           0.1
            Benzo(a)pyrene                                              1
            Benzo(b)fluoranthene                                         0.1
            Benzo(k)fluoranthene                                         0.01
            Chrysene                                                   0.001
            Dibenz(ah)anthracene                                         1
            Indeno(l,2,3-cd)pyrene                                        0.1

A methodology recommended by  EPA for Aroclors was used to calculate cancer risk estimates
for study site and basin-wide average fish concentrations (USEPA, 1996a).  Because Aroclors
consist of a mixture of both dioxin-like and non-dioxin-like congeners, calculating a cancer risk
estimate for PCB congeners by summing the risk of both Aroclors and individual dioxin-like PCB
congeners would overestimate cancer risk. To reduce this bias, the total Aroclor concentrations
were "adjusted" by subtracting the total concentrations of dioxin-like congeners for each sample
as shown in Equation 5-1.

(Equation  5-1)  adjusted Aroclors = £Mass of Aroclors - £Mass of PCB congeners
                                             5-76

-------
The resulting adjusted Aroclor concentrations were used in association with a cancer slope factor
for Aroclor mixtures to estimate the cancer risk associated with Aroclors detected in the fish
samples (USEPA, 1996a). The cancer risk of dioxin-like PCB congeners was determined using
the cancer slope factor for 2,3,7,8-TCDD and toxicity equivalence factors for PCB congeners.
The cancer risks attributable to total PCBs were estimated by summing the risk estimates based
on adjusted Aroclor concentrations and PCB congeners. While this method still likely
overestimates the cancer risk of PCB congeners because the cancer slope  factors developed for
Aroclors include an unknown  contribution from dioxin-like PCB congeners, the approach
attempts to reduce the bias of double-counting the PCB risk (USEPA, 1996a).

5.3.3  Arsenic Toxicity

Arsenic exists in many chemical forms (chemical species), both organic and inorganic. These
chemical species have varying toxicities ranging from practically non-toxic to very toxic.
Organic arsenic species (those with carbon molecules bonded to the arsenic) are less toxic and the
inorganic arsenic species (those in which the arsenic atom has a 3+ or 5+ charge and no carbon
molecules;  denoted  as As3+ or As5+, respectively) are more toxic.  EPA considers inorganic
arsenic to be a human carcinogen (see Table 5-5 for the oral CSF for inorganic arsenic). An oral
RfD for the non-cancer health endpoints of inorganic arsenic has also been developed (see Table
5-3). EPA  consensus toxicity  values for organic arsenic species are not available at this time.

Fish contain both organic and  inorganic arsenic species, with the organic arsenic species
predominating.  The organic arsenic species identified in fish include arsenobetaine,
arsenocholine, arsenosugars, dimethyarsenic (DMA) and monomethylarsenic (MMA) For this
risk assessment, fish tissue were analyzed for total (inorganic and organic) arsenic.  Since toxicity
values are only available for inorganic arsenic, to estimate the cancer risk and potential non-
cancer health impacts from exposure to arsenic in this report, an estimate of the percentage of
inorganic arsenic in  fish had to be made. Of the many studies that have been done worldwide to
measure the levels of arsenic in fish, several have included analyses of the various organic and
inorganic species (ICF Kaiser, 1996). Most of these studies have been done with saltwater
species and report inorganic arsenic levels in fish from zero to a few percent; however, some
higher percentages of inorganic arsenic have also been found (e.g., 3.6% for herring, hairtail and
saury, and 9.5% for shark). There are very few studies in which inorganic  arsenic species have
been determined in freshwater fish tissues (ICF Kaiser, 1996).

Inorganic arsenic results are available from two studies in fish from the Columbia River Basin -
one in the Lower Columbia River Bi-State Water Quality Program (Tetra Tech, 1996) and a
more recent one done on the Willamette River.

In the Lower Columbia River study (Tetra Tech, 1996), composites offish were collected in 1995
from the mouth of the Columbia River to below the Bonneville Dam on the Columbia River (at
River Mle  146) and analyzed for a large suite of chemicals, including inorganic arsenic.
Sturgeon samples were skinned and analyzed as individual fish; all other fish were composites of
fillets with skin. Table 5-7a shows a summary of the arsenic data from the  six fish species
collected as a part of this study (coho salmon, chinook salmon, sturgeon, sucker, carp and

                                             5-77

-------
steelhead).  Analyses were done for total arsenic, inorganic arsenic, and the methylated species
(MMA, DMA). The percent of inorganic arsenic and the percent of the sum of DMA and MMA
were calculated and are also shown in the table.

The percent inorganic arsenic ranged from a low of 0.1% in two of the steelhead composites and
one chinook composite (2 of the 3 values of 0.1% are based on non-detect values) to a high of
26.6% in a sucker composite (Table 5-7a). Within the same species the variation between
different composite samples was large.  For example, percent inorganic arsenic in the sucker
composites ranged from 0.6% (based upon a nondetected value) to 26.6%. Individual sturgeon
ranged from 1.9% to 18.2% .  The average percent inorganic arsenic by species ranged from 0.5%
in carp to 9.2% in sturgeon (Table 5-7c) with an overall arithmetic average for all composites of
6.5% (see Table 5-7b).

Average percent inorganic arsenic was also estimated for anadromous fish versus resident fish
species (Table 5-7d).  As can be seen from this table, the average percent inorganic arsenic in
anadromous fish species is about 1% while that from resident fish species is about 9%.
                                             5-78

-------
 Table 5-7a. Results of arsenic (As) analyses from Lower Columbia River Bi-State Water Quality Program
 (Source: Tetra Tech, 1996).

Species/Sample
Coho/HCMPl
Coho/HCMP2
Coho/HCMP3
Chinook/KCMPl
Chinook/KCMP2
Chinook/KCMP3
Sturgeon/SINDl
Sturgeon/SIND2
Sturgeon/SIND3
Sturgeon/SIND4
Sturgeon/SIND5
Sturgeon/SIND6
Sturgeon/SINDV
Sturgeon/SINDS
Sturgeon/SIND9
Sturgeon/SINDIO
Sturgeon/SINDll
Sturgeon/SINDl 2
Sucker/LSCMPl-1
Sucker/LSCMPl-2
Sucker/LSCMPl-3
Sucker/LSCMP2-l
Sucker/LSCMP2-2
Sucker/LSCMP2-3
Sucker/LSCMP3-l
Sucker/LSCMP3-2
Sucker/LSCMP3-3
Carp/CCMPl
Steelhead/DCMPl
Steelhead/DCMP2
Steelhead/DCMP3
Total As
(ug/gWW)
0.415
0.344
0.361
1.235
0.884
0.760
1.793
0.563
0.558
0.533
0.275
0.485
0.395
0.357
0.669
0.748
0.24
0.311
0.151
0.133
0.143
0.113
0.181
0.17
0.098
0.178
0.168
0.221
0.677
0.753
0.703
Inorganic As
(ug/gWW)
0.001
0.007
0.001
0.023
0.001
0.015
0.034
0.011
0.047
0.045
0.05
0.047
0.039
0.04
0.043
0.033
0.039
0.041
0.017
0.024
0.038
0.012
0.008
0.004
0.006
0.001
0.003
0.001
0.018
0.001
0.001
Q* Percent DMA&MMA Q* Percent
Inorganic As
UJ 0.2%
J 2.0%
UJ 0.3%
J 1.9%
UJ 0.1%
J 2.0%
1.9%
2.0%
8.4%
8.4%
18.2%
9.7%
9.9%
11.2%
6.4%
4.4%
16.3%
13.2%
11.3%
18.0%
26.6%
10.6%
4.4%
2.4%
6.1%
U 0.6%
1.8%
0.5%
2.7%
0.1%
U 0.1%
(ug/gWW)
0.056
0.029
0.039
0.038
0.078
0.034
0.038
0.023
0.019
0.013
0.007
0.009
0.01
0.003
0.01
0.13
0.009
0.01
0.007
0.004
0.007
0.004
0.007
0.011
0.001
0.011
0.007
0.02
0.021
0.033
0.031
DMA&MMA
13.5%
8.4%
10.8%
3.1%
8.8%
4.5%
2.1%
4.1%
3.4%
2.4%
2.5%
1.9%
2.5%
0.8%
1.5%
17.4%
3.8%
3.2%
4.6%
3.0%
4.9%
3.5%
3.9%
6.5%
U 1.0%
6.2%
4.2%
9.0%
3.1%
4.4%
4.4%

Table 5-7b. Mean concentrations** of arsenicfAs) in
all fish species combined
Total As Inorganic As

Arithmetic mean
Geometric mean
(ug/g WW
0.47
0.36
(ug/gWW)
0.02
0.01
Percent Inorganic As
6.5%
2.9%

DMA&MMA
(ug/gWW)
0.02
0.01

Percent
DMA& MMA
5.0%
3.9%

Table 5-7c. Arithmetic means** of percent
inorganic arsenic by species.
Species
coho
chinook
sturgeon
sucker
carp
steelhead







Mean
0.9%
1.3%
9.2%
9.1%
0.5%
1.0%
Table 5-7 d. Arithmetic means ** of percent inorganic
arsenic - resident fish versus anadromous fish species.
Species
Anadromous only
Resident only




% Inorganic As






1.0%
9.1%




WW = wet weight; As = arsenic; MMA = momomethylarsenic; DMA = dimethylarsenic
*Q = data qualifiers; Blanks indicate data was not qualified; U = not detected; J= estimated;
**calculations based on Tetra Tech, 1996.
coho/HCMP=coho/coho composite; chinook/KCMP = chinook/chinook composite;
sturgeon/SIND = sturgeon/sturgeon individual; sucker/LSCMP = sucker/largescale sucker composite;
carp/CCMP= carp/carp composite; steelhead/DCMP = steelhead/steelhead composite
                                                               5-79

-------
For the middle Willamette River study (EVS, 2000), composites offish (largescale sucker, carp,
smallmouth bass, and northern pikeminnow) were collected from a 45-mile section of the
Willamette River extending from the Willamette Falls near Oregon City (River Mile 26.5) to
Wheatland Ferry (River Mile 72). Total arsenic and inorganic arsenic concentrations were
determined in each of the composite fish samples. These samples included composites of whole
body, composites of fillet with skin, and composites of that portion of the fish remaining after
removing fillets from both sides of the fish. A summary of the arsenic data for whole  body and
fillet with skin samples is shown in Table 5-8. Percent inorganic arsenic in the individual
composites ranged from 2% (carp) to 13.3% (sucker). Only two species had multiple composite
samples analyzed for the same body type, whole body for carp and fillet for smallmouth bass.
The average percent of inorganic arsenic was 4.2% for the carp (range of 2 to 6.9% in the four
whole body composites) and 3.8% for the smallmouth bass (2.7% (not detected) and 6.3% in two
fillet composites).
Table 5-8. Summary of Willamette River, speciated arsenic data ( EVS, 2000).



Sucker/ Comp 1
Sucker/ Comp 12
Carp/ Comp 3
Carp/ Comp 4
Carp/ Comp 5
Carp/ Comp 14
Carp/ Comp 9
Bass/ Comp 6
Bass/ Comp 7
Pikeminnow/ Comp 1 3
Pikeminnow/ Comp 10



Total As
Tissue Tyne fue/ke WW) O
F
WB
WB
WB
WB
WB
F
F
F
WB
F
Comp = composite; F= fillet; WW= wet weight; WB =
Q = data qualifier; U = not detected;
afor whole body carp; bfor bass fillet
0.08
0.12
0.16
0.13
0.15
0.15
0.12
0.11
0.08
0.05 U
0.05 U
whole body

Inorganic As
fue/keWW)
0.004
0.016
0.007
0.009
0.005
0.003
0.003
0.003
0.005
0.003
0.003


Percent
O Inorganic As
5.0%
13.3%
4.4%
6.9%
3.3%
2.0%
U 2.5%
U 2.7%
6.3%
U 6.0%
U 6.0%

Average
Percent
O Inorganic As





4.2%"
U
U
3.8%"
U
U

blanks indicate that data was not qualified





Only two species, carp and sucker, were analyzed for inorganic arsenic and total arsenic in both
the Lower Columbia River and Willamette River studies. For carp, one composite sample of
fillet with skin was analyzed in each of the studies giving inorganic arsenic percentages of 2.5%
(Willamette, based on a non-detected value) and 0.5% (Lower Columbia River). For sucker
composites, the average for percent inorganic arsenic in the Lower Columbia River study (fillet
with skin, 9 composites) is 9.1% compared to that for the one fillet sample from the Willamette of
5.0%. The range of values for the 9 sucker composites from the Lower Columbia River study is
large (0.6% to 26.6%).

In deciding what value to assume for inorganic arsenic in fish in this assessment, consideration
was given to the Lower Columbia River and Willamette River inorganic arsenic data cited in this
study as well as to uncertainties related to 1) arsenic toxicity (i.e., from DMA) and 2) arsenic
analyses in fish tissue:
                                            5-80

-------
(1) Arsenic toxicity - Because arsenobetaine and arsenocholine are readily absorbed from the
human digestive tract and excreted in urine rapidly and unchanged, these arsenic species are
considered virtually  non-toxic.  In contrast, arsenosugars are apparently metabolized in the human
body to DMA which is then excreted in urine (Ma and Le, 1998).  EPA has classified DMA as a
category B2 carcinogen (probable human carcinogen based on sufficient animal but insufficient
human evidence) based on tumors in rodents (USEPA,  2001). However, no EPA consensus
toxicity values are available for DMA.

Although DMA may be toxic, no DMA data is available on the fish samples collected as a part of
this Columbia River Basin study. In addition information on the concentrations of DMA in
freshwater fish from other studies are limited. Concentrations of DMA and MMA, combined, are
available from the Lower Columbia River Bi-State Water Quality Program (Tetra Tech, 1996)
and are shown in Tables 5-7a and 5-7b. The percent of DMA and MMA combined ranged from
0.8% to 17.4% among the composites.  The arithmetic mean for the combined levels of MMA
and DMA among all six of the fish species analyzed was about 5% (Table 5-7b). However, the
values for DMA alone are not available.

Thus, although DMA may be an arsenic species of concern in fish or of concern as a result of
metabolism of arsenosugars, it is not possible to evaluate the potential impact on the risk
characterization that this compound would have in this study.

(2) Analysis for arsenic in fish - the identity of the chemical species of arsenic in aquatic species
is currently an area of active research and rapidly advancing knowledge. Existing analytical
methods for the chemical speciation of arsenic have several limitations  including, but not limited
to, a lack of data on  the efficiencies of recovery of arsenic  species during analysis, the possible
inter-conversion of arsenical species during extraction and  analyses and the lack of native
standard reference materials for use in determining accuracy, precision and reproducibility.

In the estimating non-cancer hazards and cancer risks from exposure to arsenic in fish tissue
(Sections 6.2.1 and 6.2.2) it was assumed that 10% of total arsenic is inorganic arsenic.  The
value of 10% was chosen after considering:

       1) the wide range found in percent inorganic arsenic among the freshwater samples of a
       given species in the Lower Columbia River and Willamette River studies,
       2) the limited data base  on concentrations of inorganic arsenic in freshwater fish,
       3) the uncertainties in the toxicity and concentrations of DMA in fish, and
       4) the uncertainties in the analytical techniques used for the chemical  speciation of
       arsenic.

This value of 10% is expected to result in a health protective estimate of the potential health
effects from arsenic in fish.
However, the inorganic arsenic data for anadromous fish species in the Lower Columbia River

                                             5-81

-------
study suggest that the assumption of a lower percentage (i.e., about 1%, see Table 5-8d) of
inorganic arsenic in these anadromous fish species may also be appropriate. This is also
consistent with the literature on saltwater species which show inorganic arsenic levels in the low
percentages for most saltwater fish.  Therefore, in Section 6.2.6 the analyses of cancer risk and
non-cancer hazards were presented assuming that inorganic arsenic is only 1% of the total arsenic
in anadromous fish species.

Using a range of assumptions for percent inorganic arsenic in anadromous fish species provides
information on the potential uncertainties in the risk characterization.
                                              5-82

-------
6.0    Risk Characterization

Risk characterization is the final step in the risk assessment process.  It combines the information
from the Exposure Assessment (Section 4) and Toxicity Assessment (Section 5) to estimate non-
cancer hazards and cancer risks. In addition, risk characterization addresses the uncertainties
underlying the risk assessment process (Section 10, Uncertainty Evaluation). This risk
characterization was prepared in accordance with the EPA guidance on risk characterization
(USEPA, 1992b;USEPA, 1995).

The methodology used to quantify potential non-cancer health effects and cancer risks is
described in Section 6.1.  The estimated non-cancer health hazards are discussed in detail in
Section 6.2.1. and the estimated cancer risks in Section 6.2.2.  Cancer and non-cancer results are
summarized in Section 6.2.3. In Section 6.2.4 the differences in cancer risks and non-cancer
hazards are compared between whole body and fillet fish samples collected from each site in the
Columbia River Basin. Section 6.2.5 discusses the results of the multiple-species diet calculation,
and; Section 6.2.6 shows how assumptions of percent inorganic arsenic impact the risk
characterization.

Non-cancer health hazards and cancer risk estimates are calculated  separately and reported
separately. Because EPA uses  different methods to evaluate these endpoints, non-cancer and
cancer estimates cannot be combined.

6.1    Risk Characterization Methodology

6.1.1  Non-Cancer Health Effects

For non-cancer health effects, it is assumed that there is an exposure threshold below which
adverse effects are unlikely to occur. In this assessment, the evaluation of non-cancer health
effects involved a comparison of average daily exposure to chemicals in fish tissue with the EPA
reference doses discussed in Section 5.  The reference dose is an estimate of the daily exposure to
a chemical that is unlikely to cause toxic effects. Potential health hazards from non-cancer effects
for a specific chemical are expressed as a hazard quotient (HQ), which is the ratio of the
calculated exposure (Section 4) to the reference dose for that chemical.

Both the estimated average daily doses from consuming fish and the reference doses are
expressed in units of amount (in milligrams) of a chemical ingested per kilogram of body weight
per day (mg/kg-day) (USEPA, 1989):


(Equation 6-1)                     HQ =	—-—
                                            RJD
Where:
       HQ  = Chemical-specific hazard quotient (unitiess)
       ADD = Average daily dose (mg/kg-day)
       RfD  = Chemical-specific oral  reference dose (mg/kg-day)
                                             6-83

-------
In this risk assessment, hazard quotients were first calculated for individual chemicals in each
species at each study site and for the basin. These results are found in Appendices Gl and G2.
However, because the fish collected for this study contain more than one contaminant, estimating
non-cancer hazard by considering only one chemical at a time might significantly underestimate
the non-cancer effects associated with simultaneous exposures to several chemicals.  Therefore,
to assess the overall potential for non-cancer hazards posed by multiple chemicals, the procedures
recommended by EPA for dealing with mixtures were applied (USEPA, 1986a; USEPA, 1989).

EPA recommends that a total hazard index value  first be calculated by summing all hazard
quotients for individual chemicals regardless of the type of health effect that each chemical
causes.  This approach to assessing mixtures - adding the hazard quotients - is known as dose
addition. Dose addition assumes that all compounds in a mixture have similar uptake,
pharmacokinetics (absorption, distribution, and elimination in the body), and lexicological
processes; and that dose-response curves of the components have similar shapes.  Thus,
calculating a total hazard index (adding all of the hazard quotients for all of the chemicals in a
fish sample regardless of their health endpoint) has several uncertainties since it results in
combining chemicals with reference doses that are based upon very different critical effects,
levels of confidence, and uncertainty/modifying factors.  Because the assumption of dose
additivity is most properly applied to compounds that induce the same effect by the same
mechanism of action, summing the hazard quotients for all chemicals to calculate a total hazard
index could overestimate the potential for effects, and is therefore, only the first step in assessing
non-cancer effects from a mixture.

If the total hazard index calculated is greater than one, EPA recommends that the hazard quotient
values for chemicals with similar target organs or mechanisms of action (health endpoints) be
summed to calculate a hazard index specific for each health endpoint (USEPA, 1986a).  If an
endpoint specific hazard index is greater than 1, unacceptable exposures  may be occurring, and
there may be concern for potential  non-cancer effects. Generally, the greater the magnitude of the
hazard index greater than 1, the greater the level of concern  for non-cancer health effects.

For this risk assessment, both the total hazard index and endpoint specific hazard indices were
calculated for each study site and for the basin. As previously discussed in Section 5, a total of
seventeen non-cancer health endpoints were considered in developing endpoint specific hazard
indices.  Hazard indices are presented by species in Appendices O (resident fish species) and P
(anadromous fish species).  The  non-cancer hazard discussion in this section (Section 6) further
summarizes the information in these appendices, focusing on the range in total and endpoint
specific  hazard indices among the species and on  the chemicals which contribute the most to non-
cancer hazards.

6.1.2  Cancer Risk Assessment

The potential cancer risk from exposure to a carcinogen is estimated as the incremental increase
in the probability of an individual developing cancer over a lifetime as a result of exposure to that
carcinogen (USEPA, 1989). The term "incremental" means the risk due to environmental
chemical exposure above the background cancer  risk experienced by all individuals in a course of

                                             6-84

-------
a lifetime. Approximately one out of every two American men and one out of every three
American women will have some type of cancer during their lifetime (American Cancer Society,
2002). The risk characterization in this report estimates the cancer risk that may result from only
one source - exposure to contaminants as a result of eating fish from the Columbia River Basin.
Other cancer risks (i.e., "background" cancer risks) are not evaluated.

Under current risk assessment guidelines, EPA assumes that a threshold dose does not exist for
carcinogens and that any dose can contribute to cancer risks (USEPA, 1986b). In other words,
the risk of cancer is proportional to exposure and there is never a zero probability of cancer risk
when exposure to a carcinogenic chemical occurs. Cancer risk probabilities were estimated by
multiplying the estimated exposure level  (average daily dose in mg/kg-day, discussed in Section
4) by the cancer slope factor (SF) for each chemical. The cancer slope factors used in this risk
characterization were developed by EPA and are discussed in Section 5 and shown in Table 5-5.
Cancer slope factors are expressed in units that are the reciprocal of those for exposure (i.e.,
(mg/kg-day)"1). The cancer risk calculated for a chemical using this method represents the upper-
bound incremental cancer risk that an individual has of developing cancer in their lifetime due to
exposure to that chemical.

(Equation 6-2)          Risk = ADD x SF

Where:
       Risk   = Estimated chemical-specific individual excess lifetime cancer risk
                  (probability; unit-less)
       ADD  = Chemical-specific average daily dose (mg/kg-day)
       SF      = Chemical-specific oral cancer slope factor  (kg-day/mg)"1

The excess cancer risk estimates in this report are shown in scientific notation format.  These
values should be interpreted as the upper-bound estimates of the increased risk of developing
cancer over a lifetime. For example, 1 X 10~6 or 1E-06 (E=exponent of base 10) is the estimated
upper-bound lifetime cancer risk of 1 in  1 million.  Because these are upper-bound estimates, the
true risks could be lower.

Because the fish collected for this study contain more than one carcinogen, estimating cancer
risks by considering only one carcinogen  at a time might significantly under-estimate the cancer
risk associated with simultaneous exposures to several chemicals.  Therefore, to assess the overall
potential for cancer risks from exposure to multiple chemicals, the procedure recommended by
EPA for dealing with mixtures were applied (USEPA, 1986a; USEPA, 1989).

EPA recommends that to assess the risk posed by simultaneous exposure to multiple carcinogenic
chemicals, the  excess cancer risk for all carcinogenic chemicals be summed to calculate a total
cancer risk.  This summing approach for carcinogens, also called response addition, assumes
independence of action by the carcinogens in a mixture.  The assumption in applying this method
is that there are no synergistic or antagonistic interactions among the carcinogens in fish and that
all chemicals produce the same effect, which in this case is cancer.
                                             6-85

-------
In interpreting cancer risks, different federal and state agencies often have different levels of
concern for cancer risks based upon their laws and regulations.  EPA has not defined a level of
concern for cancer.  However, regulatory actions are often taken when the risk of cancer exceeds
a probability of 1 in 1,000,000 to 10,000 (i.e., 1 x 10'6 to 1 x lQ-4). A level of concern for cancer
risk has not been defined for this risk assessment.

For this risk assessment, the cancer risks for each chemical for a given species and study site were
calculated (Appendix I).  The cancer risks for each chemical were then summed to calculate the
total cancer risks for each study site and for the basin.  Appendices O (resident fish species) and P
(anadromous fish species) show these total cancer risks by species as well as the contaminants
with risks equal to or greater than 1 X 10"5 for CRITFC's member tribal adults (average fish
consumption, 70 years exposure duration).  The cancer risk discussion in this section (Section 6)
further summarizes the information in the Appendices focusing on the range in total cancer risk
among the species and on the chemicals which contribute the most to cancer risks.

6.1.3  Chemicals Not Evaluated

As previously discussed in Section  1 of this report, a total of 132 chemicals were selected for
analyses in all fish in this study. Forty (30%) of these chemicals, including 29 semivolatiles, 5
pesticides, 4 Aroclors, and 2 metals, were never detected in the tissue of any fish samples at the
detection limits achieved for this study (Tablel-4a-g).  Twenty-three chemicals that were
analyzed for did not have reference doses or cancer slope factors (see Section 5.0) so that cancer
risks and non-cancer hazards using the methods described in Section 6.1.2 and 6.1.3 could not be
estimated. A risk characterization was done for only the detected chemicals with toxicity values;
a total of 82 chemicals.

6.1.4  Arsenic

As was previously discussed in Section 5.3.3, the non-cancer hazards and cancer risks discussed
in Section 6.2.1 and 6.2.2, respectively, and the results presented in the appendices assume that
for all fish species (resident fish and anadromous fish) caught in this study, 10% of the total
arsenic is inorganic arsenic.  Section 6.2.6 includes risk characterization results (using basin-wide
data) assuming the alternative assumption that inorganic arsenic is only 1% of total arsenic for
anadromous fish species.

6.1.5  Sample Type

In the CRITFC fish  consumption study (CRITFC, 1994), respondents were asked to identify the
fish parts they consume for each species.  For most of the fish species sampled as a part of this
study, the majority of the respondents said that they consume fish fillet with skin. However, a
smaller proportion consumed other fish parts as well (head, eggs, bones and organs).

Information on the portions offish that are consumed by the general public is not available.
However, as previously discussed in the Exposure Section, respondents to the qualitative fish
consumption survey conducted by EVS (EVS, 1998) for the Wheatland Ferry-Willamette Falls

-------
Reach of the Willamette River, which is a part of the Columbia River Basin, indicated that all
ethnic groups consume fillet tissue; however, other parts of the fish (eyes, eggs and skin) are also
consumed as are whole body fish.

For this study, whole body samples as well as fillets were collected when possible,  since the fish
consumption surveys show that fillets as well as other body parts may be eaten. Both whole fish
and fillet with  skin samples were analyzed for all species except white sturgeon, bridgelip sucker,
and eulachon.  Sturgeon were analyzed as whole fish and fillet without skin (since it is unlikely
that sturgeon skin is eaten).  For bridgelip sucker and eulachon only whole body samples were
collected.

Some of the risk characterization results summarized in Sections 6.2.1 and 6.2.2 are presented for
fillet and whole body samples, and others only for fillet with skin samples (except for those
species for which fillet with skin data were not available).  However, non-cancer hazards and
cancer risks were calculated for all samples collected and  are included in the Appendices of this
report. In addition, the impacts of sample type on the risk characterization results are discussed in
more detail in  Section 6.2.4, where the risk characterization results for whole body and fillet fish
samples are compared using site specific data.

6.2    Risk Characterization Results

A summary and discussion of the non-cancer hazards (for adults and children for both the general
public and CRITFC's member tribes) and excess cancer risks (for adults for the general public
and CRITFC's member tribes) are presented in this section.  More detailed information on the
risk characterization results are presented in Appendices G through J and Appendices M through
P for each fish species and tissue type analyzed in this study, for both individual study sites and
for the Columbia River Basin:

       Appendix Gl: Hazard quotients for individual chemicals for adults
       Appendix G2: Hazard quotients for individual chemicals for children
•      Appendix HI: Percent contribution from individual chemicals to the total hazard index
•      Appendix H2: Percent contribution from individual chemicals to endpoint-specific hazard
       indices
       Appendix II:  Estimated cancer  risks for individual chemicals for adults, assuming 30
       years exposure
       Appendix 12:  Estimated cancer  risks for individual chemicals for adults, assuming 70
       years exposure
•      Appendix J: Percent contribution of individual chemicals to total estimated cancer risk
•      Appendix M:  Comparison of the total and endpoint  specific hazard indices across sites
       for a CRTTFC tribal child (high fish consumption rate).
       Appendix N:  Cancer risks across a range of consumption rates, by site and species
       Appendix O:  Summary of risk characterization results (hazard indices and estimated
       cancer risks) for resident species
•      Appendix P:  Summary  of risk characterization results (hazard indices and estimated
       cancer risks) for anadromous species

                                             6-87

-------
6.2.1  Non-Cancer Hazard Evaluation

6.2.1.1  Non-Cancer Hazard Evaluation for Resident Fish

Six species of resident fish were sampled in the Columbia River Basin: bridgelip sucker,
largescale sucker, mountain whitefish, white sturgeon, walleye, and rainbow trout. Because of
the large amounts of data that are presented in the appendices on the risk characterization for
these species, one species (white sturgeon) was chosen as an example species to be discussed in
detail. Data for the other resident fish species will be summarized. Tables 6-1 and 6-2 are
identical to Tables 4.1 and 4.2,  respectively, in Appendix O for sturgeon.

As previously discussed in Section 1, white sturgeon were collected from six study sites in the
Columbia River Basin: 5 study  sites in the main-stem Columbia River (study sites 6, 7, 8, 9L, and
9U) and in the Snake River (study site 13). Chemical analyses were performed on two tissue
types, fillet without skin and whole body.

Table 6-1 summarizes both  the total and end-point specific hazard indices calculated for white
sturgeon.  Results are presented for each of the six study sites that white sturgeon were caught as
well as for the basin.

-------
Table 6-1. Total hazard indices (HI) and end point specific hazard indices (at or greater than 1.0) for white
sturgeon.
Hazard Index
Consumption Rate/
Tissue Type Health Endpoint
Study site6
CR-6
CR-7
CR-8
CR-9L
CR-9U
SR-13
Basin
Average
General Public - Adult1 b
AFC

AFC

HFC




HFC




FW

WB

FW




WB




Immune system
Total ffl
Immune system
Total HI
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
-
0.8
na
na
2.3
2.4
9.9
2.4
15
na
na
na
na
na
-
0.6
na
na
2.1
2.2
5.9
2.2
11
na
na
na
na
na
-
0.6
1.1
1.5
2.2
1.0
7.1
1.0
11
4.0
3.5
20
3.5
29
-
1.2
-
1.0
4.0
2.2
16
2.2
23
3.2
2.7
13
2.6
20
2.1
2.9
-
1.2
7.7
7.3
40
7.3
55
3.8
1.9
16
1.9
23
-
0.9
na
na
2.5
6.2
7.9
6.2
17
na
na
na
na
na
0.6
0.9
0.9
1.3
3.1
3.1
11
3.1
18
3.8
2.8
17
2.7
24
General Public - Child" b
AFC

AFC
HFC




HFC




CRITFC's
AFC




AFC




HFC





HFC
FW

WB
FW




WB




Immune system
Total ffl
Total ffl
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
-
0.7
na
2.9
3.1
13
3.1
19
na
na
na
na
na
-
0.5
na
2.6
2.9
7.6
2.9
14
na
na
na
na
na
-
0.5
1.3
2.8
1.3
9.1
1.3
14
5.1
4.5
26
4.4
37
-
1.1
0.9
5.1
2.8
21
2.8
29
4.1
3.4
16
3.3
25
1.8
2.6
1.1
9.8
9.4
51
9.4
70
4.9
2.4
21
2.4
29
-
0.8
na
3.2
7.9
10
7.9
22
na
na
na
na
na
0.5
0.8
1.1
4.0
4.0
14
4.0
23
4.9
3.9
22
3.8
31
Member Tribes - Adulf d
FW




WB




FW





WB
Liver
Central nervous system
Immune system
Reproduction/development
Total HI
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
Liver
Central nervous system
Immune system
Reproduction/development
Selenosis
Total ffl
Liver
1.0
1.1
4.4
1.1
6.6
na
na
na
na
na
6.2
6.6
27
6.6
-
40
na
-
-
2.6
-
4.7
na
na
na
na
na
5.6
6.1
16
6.1
1.3
29
na
-
-
3.1
-
4.7
1.8
1.6
9.0
1.5
13
6.1
2.8
19
2.8
1.5
29
11
1.8
-
7.2
-
10
1.4
1.2
5.7
1.2
8.8
11
6.0
44
6.0
2.0
62
8.8
3.4
3.3
18
3.3
24
1.7
-
7.3
-
10
21
20
108
20
-
150
10
1.1
2.8
3.5
2.8
7.5
na
na
na
na
na
6.8
17
22
17
-
46
na
1.4
1.4
5.0
1.4
7.9
1.7
1.2
7.4
1.2
11
8.5
8.5
31
8.5
1.2
49
10
6-89

-------
Table 6-1. Total hazard indices (HI) and end point specific hazard indices (at or greater than 1.0) for white
sturgeon.
Hazard Index
Consumption Rate/
Tissue Type




CRITFC's Member
AFC FW




AFC WB




HFC FW







HFC WB








Study site6
Health Endpoint
Central nervous system
Immune system
Reproduction/development
Total ffl
Tribes - Child0 d
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
Liver
Central nervous system
Immune system
Reproduction/development
Total ffl
Liver
Cardiovascular
Central nervous system
Immune system
Reproduction/development
Hyperpigmentation/keratosis
Selenosis
Total ffl
Liver
Cardiovascular
Central nervous system
Immune system
Reproduction/development
Hyperpigmentation/keratosis
Selenosis
Gastrointestinal
Total ffl
CR-6
na
na
na
na

1.8
2.0
8.0
2.0
12
na
na
na
na
na
12
1.1
13
52
13
1.1
-
79
na
na
na
na
na
na
na
na
na
CR-7
na
na
na
na

1.7
1.8
4.8
1.8
8.6
na
na
na
na
na
11
1.2
12
32
12
1.2
2.6
56
na
na
na
na
na
na
na
na
na
CR-8
9.6
56
9.5
79

1.8
-
5.8
-
8.6
3.2
2.9
17
2.8
24
12
1.2
5.5
38
5.5
1.2
2.9
56
21
1.8
19
110
18
1.8
1.1
1.1
150
CR-9L
7.2
35
7.1
54

3.2
1.8
13
1.8
18
2.6
2.2
10
2.1
16
21
1.2
12
86
12
1.2
3.8
120
17
1.1
14
69
14
1.1
1.7
1.8
110
CR-9U
5.1
45
5.1
62

6.2
6.0
32
6.0
45
3.1
1.5
13
1.5
18
41

39
210
39
-
1.4
290
20
1.0
10
87
9.9
1.0
1.4
-
120
SR-13
na
na
na
na

2.0
5.1
6.4
5.1
14
na
na
na
na
na
13

33
42
33
-
1.5
89
na
na
na
na
na
na
na
na
na
Basin
Average
7.6
45
7.5
66

2.5
2.5
9.2
2.5
14
3.1
2.5
14
2.4
20
16
1.1
16
60
16
1.1
2.3
94
20
1.4
16
91
16
1.4
1.3
1.1
130
AFC = average fish consumption                na =not applicable; sample type not analyzed at this study site
HFC = high fish consumption       - = health endpoint <1.0 at that study site
Total Ffl = the sum of hazard quotients regardless of health endpoint                  FW - fillet without skin; WB - whole body
" AFC risk based on average U.S. per capita consumption rate of uncooked freshwater and estuarine fish for general public (adult) of 7.5 g/day, or 1
8-oz meal per month, and for general public (child) of 2.83 g/day, or 0.4 8-oz meal per month (USEPA, 2000b).
b FIFC risk based on 99th percentile U.S. per capita consumption rate of uncooked freshwater and estuarine fish for general public of 142.4 g/day,
or 19 8-oz meals per month, and for general public (child) of 77.95 g/day, or 11 8-oz meals per month (USEPA, 2000b).
c AFC risk based on average consumption rate for adult fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the
Columbia River Basin of 63.2 g/day, or 9 8-oz meals per month, and for child fish consumers of 24.8 g/day, or 3 8-oz meals per month (CRITFC
1994).
d FfFC risk based on 99th percentile consumption rate for adult fish consumers  in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of
the Columbia River Basin of 389 g/day, or 53 8-oz meals per month, and for child fish consumers of 162 g/day, or 22 8-oz meals per month
(CRITFC 1994).
" Study sites are described in Table 1-1. CR = Columbia River; SR = Snake River
                                                                     6-90

-------
For white sturgeon, the endpoints which had hazard indices greater than 1 for most of the
populations were the immune system, liver, central nervous system, and
reproduction/developmental, with the immune system endpoint having a higher hazard index than
the other endpoints (Table 6-1).  At the lowest (average) fish ingestion rates for the general public
(average fish consumption, adults and children), only the immune endpoint exceeds a hazard
index of 1 (high of 2.1). At the higher fish ingestion rates (e.g., the high ingestion rates for
CRITFC's member tribal child), other endpoints with hazard indices greater than 1 begin to
appear: liver, central nervous system, reproductive/developmental,  cardiovascular,
hyperpigmentation/keratosis, selenosis, and gastrointestinal.

Table 6-1 also shows that, as expected, the magnitude of both the end-point specific and total
hazard indices increases proportionally to the estimated exposure for that population. For adults,
the only differences in exposure for the four adult populations (general public, average and high
fish consumption; CRITFC's member tribes, average and high fish  consumption) are due to the
different fish ingestion rates used. Thus, the hazard index increases proportionally to the fish
ingestion rate. All other exposure parameters either remain constant for all four adult populations
(fish contaminant levels, exposure frequency, body weight) or do not impact the exposure
(exposure duration and averaging time) for the reasons discussed in Section 4.9 (Averaging
Time).  This direct relationship between the hazard index and the fish ingestion rates for adults is
shown in Figure 6-1 and Table 6-2.
             3
             0 20 J
                                                           389 grams/day
                                      142.4arams/dav
                            63.2 grams/day

                     "7.5 grams/day
                                    Fish Ingestion Rate (grams/day)

            Figure 6-1. Total hazard index versus fish consumption rate for adults. White
            sturgeon, Columbia River Basin-wide average concentrations (fillet without skin).
                                               6-91

-------
    Table 6-2. Comparison of Estimated Total Hazard Indices Among Adult Populations.
             White sturgeon (whole body) from Columbia River, study site 8
Ingestion rate
Ponulation (e/dav)
General public
average fish consumption 7.5
high fish consumption 142.4
CRITFC's member tribal
average fish consumption 63.2
high fish consumption 389
Approximate ratio of hazard
index to that of general public
Total hazard adult with average fish
index consumntion

1.5
29

13
79

1
19

9
50
Table 6-2 shows the total hazard indices estimated for adults consuming sturgeon at Columbia
River study site 8 (whole body samples) at each ingestion rate. Also shown is the ratio of the
total hazard indices for CRITFC's member tribes (average and high fish consumption) and the
general public (high fish consumption) to that for the general public, average fish consumption.
The ingestion rate and exposure for adults is lowest at the average fish consumption rate for the
general public and increases proportionally for the other populations as their ingestion rates
increase. For example, the ingestion rate for the high fish consumers, general public,  is about 19
times higher than that for the average fish consumer. Thus, the exposure estimated and the total
hazard indices calculated for the general public, high fish consumer would be expected to be 19
times higher that those calculated for the general public, average fish consumer. This relationship
also holds true for the endpoint specific hazard indices calculated for each study site and the
basin. The hazard index for the immune system (Table 6-1) was about 1 at Columbia River study
site 8  for the general public, average fish consumption (whole body fish) and 20 for the high fish
consumption, general public - approximately a 20 fold difference (not exactly 19  fold as shown in
the Table 6-2 due to rounding of hazard indices).

A similar comparison can be made for the populations of children assessed  in this risk
assessment.  However, as discussed in Section 4.3, for children, exposures vary by ingestion rate
as well as by body weight and exposure duration.  This is because of the difference in the ages of
the children in the two different fish consumption studies used to estimate fish ingestion rates for
children (general public children versus CRITFC's member tribal children). Table 6-3 shows the
ratio of hazard indices for three of the child populations (general public, high fish consumption;
CRITFC's member tribes, average and high fish consumption) compared to that of the general
public child with average fish consumption using data for the Columbia River (study site 8),
whole body  sturgeon. As can be seen from this table, the hazard indices estimated for CRITFC's
member tribal children at the high ingestion rate were  over 100 times those estimated for general
public children at the average ingestion rate.
                                             6-92

-------
  Table 6-3. Comparison of Estimated Total Hazard Indices Among Child Populations
             White sturgeon (whole body) from Columbia River, study site 8
Ratio of hazard index to that of
Ingestion rate general public with average fish
Ponulation (e/dav) Total hazard index consumntion
General public
average fish consumption
high fish consumption
CRITFC's member tribal
average fish consumption
high fish consumption

2.83
77.95
24.8
162

1.3
37
24
150

1
28
18
115
A review of Table 6-1 also shows that for the general public at the average ingestion rate, the
hazard indices for children were about 0.9 of those for adults; the hazard indices for general
public children at the high ingestion rate were about 1.3 times those for general public adults,
high ingestion rate.  For example, the basin-wide total hazard index was 23 at the high fish
consumption rate (77.95 grams/day) assumed for the general public child compared to 18 for the
high fish consumption rate (142.2 grams/day) assumed for the general public adult. For
CRITFC's member tribes, the hazard indices for children at the average and high fish ingestion
rates were both about 2 times those for CRITFC's member tribal adults at the average and high
ingestion rates, respectively.

The differences in hazard indices between adults and children as well as the differences among
sites and at different fish ingestion rates is shown in Figures 6-2a-d. These figures show a
comparison of the total hazard indices for sturgeon (fillet without skin) across sites for both adults
and children at different fish ingestion rates (note that the scale of the Y axis increases from
Figure 6-2a through Figure 6-2d).  Figure 6-2a compares the total hazard indices for general
public adults and children at the average fish ingestion rate. The hazard index varies by site with
the Hanford Reach of the Columbia River (study site 9U) having the highest values (hazard
indices of 2.9 for adults and 2.6 for children). At a given site, the total hazard index for a child is
about 0.9 that of that for an adult at the average fish ingestion rate for the general public.  Figure
6-2d compares the results for CRITFC tribal adults and children at the high ingestion rate. Again,
the total hazard index varies across sites with the Hanford Reach of the Columbia River (study
site 9U) having the highest values (hazard indices of 150 for adults and 290 for children).  At a
given site, the total hazard index for a child is about 2 times that for those of adults at the high
fish ingestion rate for CRITFC tribal adults and children.

The chemicals which had hazard quotients at or greater than 1.0 (i.e., exposures for that chemical
were greater than the reference  dose) for sturgeon for most populations were total  Aroclors, total
DDT, and mercury (Table 6-4,  same as Table O-4.2 in Appendix O).  Selenium, arsenic, and
chromium were generally greater than 1.0 only at the highest exposures (high fish consumption
rates for CRITFC's member tribal adults and children).  It is useful to compare the  chemicals
contributing the most to non-cancer hazard for sturgeon (Table 6-4) with the hazard indices for
each endpoint (Table 6-1). Aroclors, which had the highest hazard quotients (Table 6-4) were
also the only chemicals contributing to the endpoint of immunotoxicity.  Thus the endpoint
specific hazard indices for immunotoxicity were also the highest of all hazard indices (Table 6-1).

                                              6-93

-------
Mercury was the major contributor to the endpoints of central nervous system and
reproduction/developmental, and DDT to the liver endpoint. Thus the hazard quotients calculated
for Aroclors, mercury, and DDT (Table 6-4) were the major contributors to (and often equal or
close to) the hazard indices for the endpoints of immunotoxicity, central nervous system and
reproduction/development, and liver, respectively (Table 6-1). The hazard indices greater than
1.0 for the cardiovascular and hyperpigmentation endpoints (Table  6-1) were primarily a result of
exposures greater than the reference dose for arsenic. Selenosis was a result of exposures greater
than the reference dose for selenium, and gastrointestinal effects were a result of exposures
greater than the reference dose for chromium.
                                              6-94

-------
                          3K
                                                X
                                                X
                                             CHILDREN
Figure 6-2a. Hazard indices for general public adults and children,
average fish consumption rate of white sturgeon fillets. Note that hazard
indices are the same at study site 7 and 13.




1 30<
fl 20.










X
X
X *
9
^^
ADULTS CHILDREN







• S 7
AS 8
XS 9L
JKS 9U
• S 13












Figure 6-2b. Hazard indices for CRITFC's member tribal adults and
children, average fish consumption rate for white sturgeon fillets. Note that
hazard indices are the same at study site s 7 and 13.



•8
S 40'
s







X


^
X • • Si 6
A V • Si 7
Y A ASi 8
!Ksi pu
• Si 13
ADULTS CHILDREN










  Figure 6-2c. Hazard indices for general public adults and children, high fish
  consumption rate of white sturgeon fillets.  Note that hazard indices are the
  same for study sites 7 and 13.




1




X


X
X A
£

• Sit 6
•Sit 7
A Sit X
XSit 9L
XSit 9U
• Sit 13
ADULTS CHILDREN

 Figure 6-2d. Hazard indices for CRITFC's member tribal adults and
 children, high fish consumption rate of white sturgeon fillets. Note that
 hazard indices are the same at study sites 7 ad 13.
                                                                     6-95

-------
It is important to point out that there are no reference doses available for dioxins, furans and
dioxin-like PCB congeners.  Therefore, hazard quotients could not be calculated for these classes
of chemicals and their potential impact on the magnitude of non-cancer hazards (i.e., endpoint
specific hazard indices and total hazard indices) could not be evaluated.
  Table 6-4.  Chemicals having hazard quotients at or greater than 1.0 in white sturgeon.
                                Adults
                                                  Children
  Tissue Type       Hazard Quotient    y j   >.       Chemical        Hazard Quotient
                                                                                   Study Sites" with
                    AFC   HFC
                                                AFC
                                                 HFC
  General Public
  Fillet without skin
     Total Aroclors
2.1
5.9-40  6b,7b,8b,9Lb,9U,13b
Total Aroclors
1.8
7.6-51  6b,7b,8b,9Lb,9U513b
Total DDT
Mercury
Whole body
Total Aroclors 1.1
Total DDT
Mercury -
CRITFC's Tribal Members
Fillet without skin
Total Aroclors 2.6-18
Total DDT 1.3-3.2
Mercury 1.0-3.3
Selenium
Whole body
Total Aroclors 5.7-9.0
Total DDT 1.2-1.6
Mercury 1.2-1.5


1.5-7.1
1.0-7.3
13-20
2.6-3.7
1.9-3.5

16-110
4.1-20
2.8-20
1.3-2.0
35-56
7.8-10
5.1-9.5


6,7,8,9L,9U,13
6,7,8,9L,9U,13
8,9L»,9Ub
8,9L,9U
8,9L,9U

6b,7b,8b,9L,9U,13b
6,7,8,9L59U
6,7,8b,9Lb,9U,13
7,8,9L
8,9L,9U
8,9L,9U
8,9L,9Ub


Total DDT
Mercury
Total Aroclors
Total DDT
Mercury

Total Aroclors
Total DDT
Arsenic
Mercury
Selenium
Total Aroclors
Total DDT
Arsenic
Chromium
Mercury
Selenium
—

4.8-32
1.2-5.8
1.8-6.0
11-17
2.1-3.0
-
1.5-2.8
-
1.9-9.1
1.3-9.4
17-26
3.4-4.7
2.4-4.4

32-210
8.0-38
1.1-1.2
5.5-39
1.4-3.8
69-110
14-20
1.0-1.8
1.1-1.8
9.9-19
1.1-1.7
6,7,8,9L,9U,13
6,7,859L,9U,13
8,9L,9U
8,9L,9U
8,9L,9U

6,7,8,9L,9U,13
6,7,8,9L,9U,13
6,7,8,9L
6,7,8b,9L,9U,13
7,8,9L,9U,13
8,9L,9U
8,9L,9U
8,9L,9U
8,9L
8,9L,9U
8,9L,9U
AFC = average fish consumption; HFC = high fish consumption;
- = <1; Astudy sites are described in Table 1-1. BHFC only

The summary of the results of the non-cancer hazard evaluation for the other resident fish species
are shown in Appendix O by species.  Summaries of the endpoint specific and total hazard indices
and of the chemicals having hazard quotients at or greater than 1 are shown in Tables 1.1 and 1.2
(bridgelip sucker), 2.1 and 2.2 (largescale sucker), 3.1 and 3.2 (mountain whitefish), 4.1 and 4.2
(white sturgeon), 5.1 and 5.2 (walleye), and 6.1 and 6.2 (rainbow trout). A review of these tables
shows that:

        The total hazard indices and endpoint specific hazard indices increase among the general
        public and CRITFC's member tribal populations as the exposures for that population
        increase:
                                                6-96

-------
       The endpoints which are more frequently greater than a hazard index of 1 are immune
       system (due to Aroclors), liver (due primarily to DDE for most species), and central
       nervous system and reproduction/developmental (due primarily to methyl mercury), with
       the immune system endpoint usually having a higher hazard index than the other
       endpoints.  These hazard indices vary among sites for a given species and among species;

•      At the lowest (average) fish ingestion rates for the general public (adults and children), the
       endpoint-specific hazard indices were at or less than 1 for all of the resident fish with the
       exception of sturgeon and whitefish at the Hanford Reach of the Columbia River
       (9U) where hazard indices for immunotoxicity were greater than 1 (high of 3 for
       whitefish).

•      For the more highly exposed populations (e.g., at the high fish ingestion rates for
       CRITFC's member tribes), endpoint specific hazard indices for reproduction/development
       and central nervous system, immunotoxicity, and liver are greater than 1 at most sites for
       most species.  For mountain whitefish and white sturgeon, hazard indices for the most
       contaminated study site (Columbia River, study site 9U) were greater than 100 for the
       immunotoxicity endpoint.

•      At these highest ingestion rates for CRITFC's member tribal adults and children, other
       endpoints with hazard indices greater than 1 begin to appear for some species.  These
       endpoints include cardiovascular and hyperpigmentation/keratosis, selenosis,
       gastrointestinal, kidney, and metabolism.  These effects were primarily the result of
       exposures greater than the reference dose for arsenic; selenium; chromium; cadmium; and
       nickel  and zinc, respectively. For walleye, thallium also contributes to the overall hazard
       index calculated for liver. The highest endpoint-specific hazard index for these endpoints
       was approximately 4.0.

Table 6-5 is a summary of the ranges in endpoint  specific hazard indices across study sites for
each resident fish species. Results are shown for both average and high fish consumption rates for
the general public and CRITFC tribal member adults. Hazard indices are shown only for those
endpoints that most frequently exceed a hazard index of 1 (reproduction/development and the
central nervous system, immunotoxicity, and liver). It should be kept in mind that not all fish
species were caught at the same sites and that sample numbers varied by species.
                                              6-97

-------
Table 6-5  Summary of ranges in endpoint specific hazard indices across study sites for adults who
consume resident fish from the Columbia River Basin.
                                      Non-cancer endpoints which most frequently exceed a hazard index of 1
                                                              for all species
Species
General Public - Adult
Average Fish Consumption
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
High Fish Consumption
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
CRITFC's Member Tribal Adult
Average Fish Consumption
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
High Fish Consumption
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
Reproductive/ Developmental And
N Central Nervous System Immunotoxicty
3
19
12
16
3
7

3
19
12
16
3
7


3
19
12
16
3
7

3
19
12
16
3
7
<1
<1
<1
<1
<1
<1

<1
2 to 7
<1 to 3
1 to 7
4
1 to 2


<1
<1 to 3
<1 to 1
<1 to 3
2
<1

2
5 to 20
<1 to 7
3 to 20
10
4 to 5
<1
<1
<1 to 3
<1 to 2
<1
<1

6
1 to 8
1 to 50
6 to 40
1
1 to 2


3
<1 to 3
<1 to 22
3 to 18
<1
<1

17
<1 to 21
4 to 140
16 to 108
4
3 to 4
Liver
<1
<1
<1
<1
<1
<1

2

-------
average ingestion rate (7.5 g/day).
                   70
                   50
                   40-
                   30-
        ^—. Hazard In dec =65 far consumption
             tt mountain whitetish assumluj huh
             rejection rate by adults r\ CPU PC's
             member trfees
                       White
                       Sturgeon
                       (n=16)
Mountain
Whitefish
 (n=12)
Bridgelip
 Sucker
 (n=3)
Largescale
 Sucker
 (n=19)
                                              Species
           * Fillet with skin samples except for sturgeon (fillet without skin)
                   and bridgelip sucker (whole body)

            n = number of samples
                                                                  P
Rainbow
 Trout
                                          1
Walleye
 (n=3)
General Public
• Average Fish
Consumption
1 	 1 HighFish



CRITFC
Member Tribes
1 1 HighFish
      Figure 6-3. Adult total non-cancer hazard indices for resident fish species* using basin-wide average data.

For a more detailed comparison of the total and endpoint specific hazard indices, see Appendix
M, where hazard indices are compared for all resident species across study sites for CRITFC's
member tribal children with a high fish consumption rate (162 g/day or 5 meals per week).

The contribution from specific chemicals and classes of chemicals to the overall non-cancer
hazard for resident fish species is shown in Table 6-6. These results were calculated using
Columbia River Basin average concentrations for fillet without skin samples, except for those
species where such sample types were not available (bridgelip sucker, whole body; white
sturgeon, fillet without skin). The number of samples used to compute the basin-wide averages
vary among species, and for some species represent only a few samples (e.g.,  3  samples for
walleye and bridgelip sucker). The results in Table 6-6, which are also depicted in the charts in
Figures 6-4 through 6-9, show that the percent contribution of specific chemicals to the total
hazard index differs among the resident fish species. For example, Aroclors contribute 83% to
the total non-cancer hazard for mountain whitefish, but only 20% for walleye.  Total DDT
contribution to the total hazard index ranges from 3-21% among the species and methyl mercury
from about 6-54%.  Except for thallium for walleye (percent contribution of 14%), the only
chemicals contributing greater than 5% to the non-cancer hazards for resident fish species are
Aroclors, total DDT, and mercury.
                                               6-99

-------
Table 6-6. Percent contribution of contaminant groups to total non-cancer hazards for resident fish
species. Based on Columbia River Basin-wide averages.

Tissue Type
Number of samples
Total metals
Mercury
Arsenic
Chromium
Manganese
Selenium
Thallium
Zinc
Other Metals
Total Aroclors
Total Pesticides
Total DDT
Other Pesticides
white sturgeon
FW
16
22
17
1
<1
<1
2
ND
<1
<1
63
15
13
2
bridgelip
sucker
WE
3
18
6
2
1
3
1
ND
1
4
60
21
21
<1
largescale
sucker
FS
19
50
45
<1
1
<1
1
ND
1
1
40
10
9
<1
mountain
whitefish
FS
12
9
7
<1
<1
<1
1
ND
<1
<1
83
8
7
1
walleye
FS
3
77
54
4
1
<1
2
14
1
1
20
3
3
ND
rainbow
trout
FS
7
55
46
ND
1
<1
3
ND
2
2
42
3
3
ND
FW = fillet without skin; FS = fillet with skin; WB = whole body; ND = Not Detected
                                                             6-100

-------
                                           Mercury
                                             17%
                         Total Aroclors
                            63%
Figure 6-4. Percent contribution of basin-wide average chemical
concentrations to non-cancer hazards from consumption of white sturgeon
fillet without skin.  Number of samples =16.
                                Other Pesticides
                                    o.:
                 Total DDT
                    9%
        Total Aroclors,
            40%
                                                Mercury
                                                 46%
                                   Other Metals
                                      5%
 Figure 6-5.  Percent contribution of basin-wide average chemical concentrations of
 non-cancer hazards from consumption of largescale sucker fillets with skin. Number
 of samples =19.
                                    6-101

-------
                       Endosulfan Sulfate
                           0.02%
Mercury
  6%
                                                 Other Metals
                                                     12%
                             Total Aroclors
                                60.4%


Figure 6-6. Percent contribution of basin-wide average chemical concentrations to
non-cancer hazards from consumption of whole body bridgelip sucker. Number of
samples = 3.
                               Total DDT
                                   3%
         Total Aroclors
              42%
              Mercury
                46%
                                          Other Metals
                                               9%
 Figure 6-7. Percent contribution of basin-wide average chemical concentrations to non-
 cancer hazards from consumption of rainbow trout fillet with skin. Number of samples = 7.
                                        6-102

-------
                                   Total DDT
                                      3%
                 Total Aroclors
                     20%
                  Other
                Inorganics
                   5%
                      Thallium
                        14%
      Mercury
       54%
                            Arsenic
                              4%
     Figure 6-8.  Percent contribution of basin-wide average chemical concentrations to non-
     cancer hazards from consumption of walleye fillet with skin. Number of samples = 3.
                    Total DDT
                       7%
                                   Other
                                 Pesticides
                                    1%
Mercury
  7%

  Other Metals
      2%
                                 Total Aroclors
                                      83%
Figure 6-9.  Percent contribution of basin-wide chemical concentrations to non-cancer hazards
from consumption of mountain whitefish fillet with skin.  Number of samples =12.
                                      6-103

-------
6.2.1.2  Non-cancer Hazard Evaluation for Anadromous Fish

The anadromous fish sampled in the Columbia River Basin were coho salmon, fall chinook
salmon, spring chinook salmon, steelhead, eulachon, and Pacific lamprey. The summary of the
results of the non-cancer hazard evaluation for these anadromous fish species are shown in
Appendix P by species.  Summaries of the endpoint-specific and total hazard indices and of the
chemicals having hazard quotients greater than 1 are shown in Tables 1.1 and 1.2 (coho salmon),
2.1 and 2.2 (fall chinook salmon), 3.1 and 3.2 (spring chinook salmon), 4.1 and 4.2 (steelhead), -
5.1 and 5.2 (eulachon), and 6.1 and 6.2 (Pacific lamprey).  As with the resident fish species, the
values of the total hazard indices and endpoint-specific hazard indices increase among all of the
populations as the exposure to that population increases.

Because the results for coho salmon, fall chinook, spring chinook, and steelhead were similar,
they are summarized as a group.  The results for eulachon and lamprey are discussed separately.

Tables 1.1 and 1.2 (coho salmon), 2.1  and 2.2 (fall chinook salmon ), 3.1 and 3.2 (spring chinook
salmon), and 4.1 and 4.2 (steelhead) show that:

•       At the average fish ingestion rates for the general public, adults and children, the endpoint
        specific hazard indices were less than 1.0.

        The endpoints which had hazard indices greater than 1 most frequently for salmon and
        steelhead were immunotoxicity (due to Aroclors) and reproductive/developmental and
        central nervous system (due primarily to mercury).   In general, the hazard indices for the
        immunotoxicity endpoint for salmon and steelhead were much lower and did not vary as
        much across study sites as those for the resident fish species with the highest contaminant
        levels (largescale sucker, mountain whitefish, and white sturgeon).

        As exposures increase, other endpoints with hazard indices greater than 1 begin to appear.
        These include: cardiovascular and hyperpigmentation/keratosis; metabolism; selenosis;
        gastrointestinal; and kidney, resulting primarily from exposures greater than the reference
        dose to arsenic; nickel and zinc; selenium; chromium; and cadmium, respectively. The
        highest hazard indices for these endpoints at the highest ingestion rates were at or less
        than 4. At these exposures, hazard indices for immunotoxicity,
        reproduction/development, and central nervous system are greater than 1 for most sites.

Pacific lamprey were collected at 2 study sites, Willamette Falls (study site 21) and Fifteen Mile
Creek (study site 24). Pacific lamprey results were similar to those for salmon and steelhead in
that, at the average fish ingestion rates for the general public, adults and children, the endpoint
specific hazard indices never exceed 1.0.  In examining endpoint specific hazard indices with
increasing exposure, the immune  system hazard index is exceeded first. The estimated endpoint
specific hazard index for immunotoxicity, which is the largest contributor to the total hazard
index for Pacific lamprey is due to exposures greater than the reference dose for Aroclors.  At the
same  ingestion rates, the endpoint specific  hazard indices for immunotoxicity were higher for
lamprey than for salmon and steelhead.

                                             6-104

-------
Eulachon (smelt) were caught at only one study site, Columbia River study site 3, and analyzed as
whole body samples. Two endpoint specific hazard indices were exceeded (cardiovascular and
hyperpigmentation/keratosis) at the high fish consumption rates for CRITFC's member tribal
adults (hazard index of 1.7) and children (hazard index of 3.2) (see Table 5.1). These
exceedances were a result of arsenic exposures greater than the reference dose (Table 5.2).

Table 6-7 is a summary of the ranges in endpoint specific hazard indices across study sites for
anadromous fish. Results are shown for both average and high fish consumption rates for the
general public and CRITFC tribal member adults. Hazard indices are shown only for the three
endpoints which frequently exceeded a hazard index of 1: reproduction/development and the
central nervous system, immunotoxicity, and liver. It should be kept in mind that not all  species
were caught at the same study sites and that sample numbers varied by species.

Figure 6-10 shows the relative differences in total hazard indices in the Columbia River Basin for
anadromous fish species using average and high fish consumption rates for general public adults
and for CRITFC's member tribal adults. The total hazard index is highest for lamprey,  followed
by salmon and steelhead, which are in the same range, and then eulachon.

For a more detailed comparison of the total and endpoint specific hazard indices  across study
sites for anadromous fish species, see Appendix M. In this appendix, hazard indices are
compared for the population with the highest exposure and non-cancer hazards - CRITFC's
member tribal children with a high fish consumption rate (162 grams/day or about 5 meals per
week).
                                            6-105

-------
  Table 6-7  Summary of ranges in endpoint specific hazard indices across study sites for adults who
  consume anadromous fish species from the Columbia River Basin.

                                             Non-cancerendpoints which most frequently exceed a hazard index of 1
                	for all species	
                                              Reproductive/ Developmental And
Species
General Public-
Average Fish Consumption
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey
High Fish Consumption
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey
CRITFC's Member Tribal
Average Fish Consumption
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey
High Fish Consumption
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey
N


3
15
24
21
3
3

3
15
24
21
3
3


3
15
24
21
3
3

3
15
24
21
3
3
Central Nervous System


<1
<1
<1
<1
<1
<1

2
1 to 2
<1 to 6
1 to 3
<1
<1


1
<1 tol
<1 to 3
<1 to 1
<1
<1

7
3 to 6
<1 to 17
4 to 8
<1
<1
Immunotoxicty Liver


<1 <1
<1 <1
<1 <1
<1 <1
<1 <1
<1 <1

3 <1
<1 to 3 <1
1 to 2 <1
1 to 2 <1
<1 <1
9 <1


1 <1
1 <1
<1 <1
<1 to 1 <1
<1 <1
4 <1

7 <1
<1 to 8 <1
3 to 6 <1
3 to 6 <1
<1 <1
24 2
N= number of samples; All samples are fillet with skin except white sturgeon which is fillet without skin. Bridgelip sucker and eulachon are whole
body fish samples.
                                                     6-106

-------
                 30
                 25'
                 20
              M  15
Hasa-d I nctx = 28 for consumption of Pacifc
Ismprc/b/aLlts in CRITPC's mem her tribes
3t the hi^ rigsetion rde
                      Spring
                     Chinook
                      (n=24)
   Fall
  Chinook
  (n=15)
 Pacific
Lamprey
 (n=3)
                                               Species
              * Fillet with skin samples except for eulachon (whole body)
              n = number of samples
General Public

• Average Fish
Consumption
I I High Fish
I 	 1 Consumption




CRITFC
Member Tribes
D cZS™
I I High Fish
      Figure 6.10 Adult total non-cancer indices for anadromous fish species*. Average concentrations for the
      Columbia River Basin.

Table 6-8 and Figures 6-11 through 6-16 show the major chemicals contributing to the total
hazard index for each anadromous fish species (shown for basin-wide data, fillet with skin for all
species except eulachon which was whole body). Aroclors and mercury were the primary
chemicals of concern for non-cancer hazards for anadromous fish species, followed by arsenic.
For eulachon, arsenic was the major contributor to non-cancer hazard.  For Pacific lamprey,
Aroclors contributed almost 87% to the non-cancer health effects.
                                               6-107

-------
Table 6-8. Percent contribution of contaminant groups to total non-cancer hazards for
anadromous fish species. Based on Columbia River Basin-wide averages.

Number of samples
Tissue type
Total Metals
Mercury
Aluminum
Arsenic
Cadmium
Chromium
Copper
Selenium
Zinc
Other Metals
Total Aroclors
Total Pesticides
Chlordane (total)
Total DDT
Hexachlorobenzene
spring
chinook
24
FS
65
43
<1
12
<1
3
1
3
1
2
34
2
<1
2
<1
coho
salmon
3
FS
54
41
ND
6
ND
2
2
2
1
<1
45
1
<1
1
ND
eulachon
3
WB
95
ND
2
62
2
ND
5
12
9
2
ND
4
ND
4
ND
fall chinook
15
FS
58
39
<1
12
ND
1
1
3
1
<1
40
2
<1
2
<1
Pacific
lamnrev
3
FS
7
ND
ND
2
1
1
1
2
1
<1
87
6
2
4
<1
steelhead
21
FS
55
43
<1
7
<1
1
1
2
1
<1
43
2
<1
1
<1
FS = fillet with skin; FW = fillet without skin; WB = whole body; ND= not detected
                                                           6-108

-------
                      Total DDT
                         2%
pther Pesticides
    0.4%
       Total Aroclors
           34%
           Other Metals.
               5%
              Mercury
               44%
                  Selenium
                           Chromium
                             2.5%
     Arsenic
      12%
Figure 6-11. Percent contribution of basin-wide average chemical
concentrations to non-cancer hazards from consumption of spring chinook fillet
with skin.  Number of samples = 24.
                       Total Chlordane
                          0.1 %
          Total Aroclors
              45%
                                                        Mercury
                                                         41%
                              Other Metals
                                  7%
            Arsenic
             6%
  Figure 6-12. Percent contribution of basin-wide chemical concentrations to non-cancer
  hazards from consumption of coho salmon. Number of samples = 3.
                                      6-109

-------
                       Total DDT
                          2%
            Total Aroclors
                 40%
                    Other Metals
                         3%
          Mercury
           39%
                                 Selenium
                                    3%
Arsenic
  12%
Figure 6-13. Percent contribution of basin-wide average chemical concentrations to non-cancer
hazards from consumption of fall chinook fillet with skin. Number of samples =15.
                            Total DDT

                              1.3%
                                              Other Pesticides
                                                 0.3%
                Total Aroclors
                    43%
                          Other Metals
                              3%
         Figure 6-14. Percent contribution of basin-wide average chemical concentrations
         to non-cancer hazards from consumption of steelhead fillet with skin. Number of
         samples = 21.
                                      6-110

-------
                           Other
                         Pesticides
                            2%
Arsenic
  2%
             Total DDT
                4%
        Other Metals
           5%
                              Total Aroclors
                                  87%
Figure 6-15. Percent contribution of basin-wide average chemical concentrations to
non-cancer hazards from consumption of Pacific lamprey fillet with skin.  Number of
samples = 3.
                     9H-Fluorene Total DDT
                        0.3%   ~1  4%
             Other Metals
                 2%
             Zinc
             9%
Aluminum
   2%
              Copper
               5%
               Cadmium
                 2%
Figure 6-16.  Percent contribution of basin-wide average chemical concentrations
to non-cancer hazards from consumption of whole body eulachon. Number of
samples =3.
                             6-111

-------
6.2.1.3 Comparisons Between Anadromous Fish and Resident Fish Species

A comparison of the total hazard indices, endpoint specific hazard indices, and chemicals with
hazard quotients greater than 1.0 among all of the fish species (resident fish and anadromous fish)
can be made using the summary tables in Appendices O and P. The conclusions from these
comparisons, are limited by the fact that different species were caught at different study sites and
that sample numbers and sample types for each species varied.

       The endpoint specific hazard indices that were greater than 1 the most often and that had
       the highest values for all of the resident fish species were immunotoxicity, central nervous
       system, reproduction/developmental, and liver, with immunotoxicity usually having the
       highest endpoint specific hazard index. For resident fish species, endpoint specific hazard
       indices were rarely greater than 1 for children and adults in the general population with an
       average fish ingestion rate. The exceptions to this were white sturgeon and mountain
       whitefish caught in the Hanford Reach of the Columbia River (study site 9U), where
       endpoint specific hazard indices were greater than 1 (high of 2.7) for the endpoint of
       immunotoxicity.  This was due to exposures to Aroclor greater than its reference dose.

•      For salmon and steelhead, three of these endpoints were also the ones that also had the
       highest hazard indices: immunotoxicity, central nervous system, and
       reproduction/developmental, with most endpoints  specific hazard indices being within a
       small range among the three salmon and  steelhead (the exception is for the Klickitat due
       to mercury levels in spring chinook). No endpoint specific hazard indices were greater
       than 1 for children or adults in the general population with an average fish ingestion rate.

•      For Pacific lamprey fillet with skin, the major contributor to non-cancer hazards was due
       to immunotoxicity; for whole body lamprey, it was immunotoxicity as well as central
       nervous system and reproduction/development endpoints  (due to higher levels of mercury
       in whole body samples of lamprey).  There were no endpoint specific hazard indices
       greater than  1 for the general population (adults or children) with an average fish
       consumption rate.

•      For eulachon, only the endpoints of cardiovascular and hyperpigmentation/keratosis had
       hazard indices greater than 1 and only at the highest exposures (CRITFC's member tribal
       adults and children, high fish consumption).

Hazard indices greater than 1 for specific endpoints were primarily a result of elevated hazard
quotients for a few chemicals: total Aroclors (immunotoxicity), mercury (central nervous system,
and reproduction/developmental), total DDTs (liver), and arsenic  (cardiovascular and
hyperpigmentation/keratosis). This can be seen in the figures previously discussed for resident
fish species (Figures 6-4 to 6-9) and anadromous fish species (Figures 6-11 to 6-16).
Although similar endpoint specific hazard indices were exceeded for many of the fish species
tested, the magnitude of both the endpoint specific and total hazard indices vary substantially

                                             6-112

-------
among the species.  Table 6-9 shows a summary of the non-cancer results across all species at the
high fish consumption rate for CRITIC'S member tribal adults.  All of the non-cancer endpoints
that exceed 1.0 are shown for each species as are the range in total hazard indices across study
sites and the total hazard index for the basin. For this table, fillet with skin data were used except
for the species that had no fillet with skin samples (fillet without skin data for sturgeon and whole
body for bridgelip sucker and eulachon).

 Table 6-9.  Summary of end point specific hazard indices and total hazard indices (by study site and basin-
 wide) for CRTTFC's tribal member adult, high fish consumption.
Non-cancer endpoints


Sample
Species N
Resident Species
Bridgelip sucker 3
Largescale 19
Mt. whitefish 12
White sturgeon 16
Walleye 3
Rainbow trout 7
Anadromous species
Coho salmon 3
Fall chinook 15
Spring chinook 24
Steelhead 21
Eulachon 3
Pacific lamprey 3
type

WB
FS
FS
FW
FS
FS

FS
FS
FS
FS
WB
FS
Central
nervous
system

2
5-20
<1 - 7
3-20
10
4, 5

7
3-6
<1 - 17
4-8
<1
<1

Reproduction/
developmental

2
5-20
<1 - 7
3-20
10
4,5

7
3-6
<1 - 17
4-8
<1
<1

Immuno- Cardio- Hyperpig-
toxicity Liver vascular mentation

17 6 <1 <1
<1 -21 1-7 <1 <1
4- 140 <1 - <1 <1
16-108 6-21 <1 <1
4 4 <1 <1
3,4 <1 <1 <1

7 <1 <1 <1
<1 - 8 <1 1-2 1-2
3-6 <1 2 2
3-6 <1 1-2 1-2
<1 <1 2 2
24 2 <1 <1
Range in
study site
total
hazard
indices

27
10-45
9-150
29-150
18
8,10

16
6-16
6-24
9-15
3
28
Total
basin
hazard
index

27*
29
65
49
18*
9

16*
12
13
16
3*
28*
N= Number of samples; FW = fillet without skin; FS = fillet with skin, WB = whole body
"Columbia River Basin index based on study site.

A review of Table 6-9 (reference to study site specific information can be found in the tables in
Appendices O and P) suggests that:

•      For eulachon, all  of the endpoint specific hazard indices were equal to or less than 2. The
       endpoint specific hazard indices were at or less than 2 for Pacific lamprey with the
       exception of a value of 24 for immunotoxicity.  This was due to exposures greater than the
       reference dose for Aroclors.  Total basin-wide hazard indices were 3 and 28, respectively,
       for eulachon and lamprey.

•      For the salmon and steelhead, all of the study site endpoint specific hazard indices were 8
       or less, except for one study site/species (hazard index of 17 for spring chinook for
       reproduction/development and central nervous system due to mercury in the sample from
       the Klickitat River). The total basin-wide hazard indices range from  12 to 16 for salmon
       and steelhead.
       For two of the resident fish species, walleye and rainbow trout., the endpoint specific
                                              6-11:

-------
       hazard indices were at or less than 10. The endpoint specific hazard index for bridgelip
       sucker were less than 6, with the exception of immunotoxicity which had a value of 17.
       The total basin-wide hazard indices were 9, 18 and 27 for rainbow trout, walleye and
       bridgelip sucker, respectively.

•      For large scale sucker the endpoint specific hazard indices for the central nervous system
       and reproductive/development range from 5 to 20 and for immunotoxicity from <1 to 21.
       The study site total hazard indices were from 10 to 45 with five of the six study site total
       hazard indices being greater than 20.

•      The resident fish species, mountain whitefish and sturgeon, had the highest total study site
       hazard indices which ranged from 9 to 150 and 29 to 150, respectively.  For the whitefish.,
       total hazard indices were 9  (Umatilla), 13 (Deschutes), 72 (Yakima), and 150 (Hanford
       Reach of the Columbia, study site 9U)(see Table 3.1).  The two highest values (72 for the
       Yakima and 150 for the Columbia at 9U) were due primarily to the high endpoint specific
       hazard indices for immunotoxicity (due to Aroclors) at these study sites.  For sturgeon, all
       of the study site total hazard indices were greater than 20: hazard indices of 29 (Columbia
       at study sites 7 and 8);  40 (Columbia, study site 6); 46 (Snake, study site!3); 62
       (Columbia,  study site 9L); and 150 (Columbia, study site 9U)(see Table 4.1).  The high
       values for sturgeon were also in large part also due to exposures greater than the reference
       dose for Aroclors resulting in high endpoint specific hazard indices for immunotoxicity.
       It is obvious from Table 6-9 that for these 2 species (whitefish and sturgeon), their high
       endpoint specific hazard indices for immunotoxicity (due to total Aroclors) at some study
       sites tend to distinguish them from the other species.

Figure 6-17 is a summary of the total hazard indices for each species for all four ingestion rates
for adults (general public adult, average and high fish consumption; CRITFC's member tribal
adult, average and high fish consumption). Basin-wide fillet with skin data were used for this
figure, except for those species that had only whole body samples (bridgelip sucker and eulachon)
or fillet without skin (sturgeon) data. As can be seen from this table, the total hazard indices vary
by species with white sturgeon and mountain whitefish having the highest total hazard indices
among the  12 fish sampled. Largescale  sucker, lamprey, and bridgelip sucker had similar but
lower total  hazard indices followed by the salmon, steelhead, and walleye, then rainbow trout and
eulachon.
                                             6-114

-------
70-
60-
50-
40-
30-
20-
10-
0-
Hazsrdlnitex = 65 1br consumption of \k
mountain whtdfch b/adutte in CRITFC •*
member trfcee ^ the hi^i hgretion rate
Anadromous



JJrDroLnJ
	

-











Resident



I


-


D
PI

-

_
U
       * Fillet of skin samples except for \/\/hite sturgeon
       (fillet without skin) and bridgelip sucker and
       eulachon (whole body)
Species
   Figure 6-17. Adult total non-cancer hazard indices across all species*. Columbia River Basin data.

As was previously discussed for white sturgeon (Figures 6-2a-d), the estimated hazard indices for
children were different than those for adults. For the general public, the hazard indices for
children at the average fish ingestion were about 0.9 of those for adults at the average ingestion
rate; the hazard indices for children at the high ingestion rate were about 1.3 times those for
adults at the high ingestion rate.  For CRITFC's member tribes, the hazard indices for children at
the average and high ingestion rates were both about 1.9 times those for CRITFC's member tribal
adults at the average and high ingestion rates, respectively.

Appendix M contains a comparison of the total and endpoint specific hazard indices across sites
(anadromous and resident fish species) for CRITFC's member tribal children with a high
ingestion rate. This was the population with the highest exposures and hazard indices.

6.2.2   Cancer Risk Evaluation

Because the incremental increase in cancer risks resulting from ingestion offish was calculated
for adults only, only four populations had cancer risk estimates: average and high fish
consumption for both the general public adult and CRITFC's member tribal  adult.  However, for
                                              6-115

-------
cancer risk, exposure duration does have an impact on the calculations.  Therefore, risks were
estimated for both 30 and 70 year exposure durations. This results in eight separate cancer risk
calculations per study site and in the basin:

Average Fish Consumption
General public adult, 30 years         CRITFC's member tribal adult, 30 years
General public adult, 70 years         CRITFC's member tribal adult, 70 years
High Fish Consumption
General public adult, 30 years         CRITFC's member tribal adult, 30 years
General public adult, 70 years         CRITFC's member tribal adult, 70 years

The cancer risks calculated for each chemical for each study site are shown in Appendices II
(general public and CRITFC's member tribal adults, 30 year exposure) and 12 (general public and
CRITFC's member tribal adults, 70 year exposure).  Appendix N shows the species specific
cancer risks by study site over a range offish ingestion rates. Appendices O and P, which were
previously used for discussion of the non-cancer results, include summary results for the total
cancer risk estimates by fish species and tissue type.  Included in Appendices O and P are:  (1)
tables showing the total cancer risks by study site and basin for all 8 separate cancer risk
calculations, and (2) tables showing the cancer risks by study site for those chemicals that were at
or greater than a cancer risk of 1 X 10 ~5 for one population, CRITFC's member tribal adults,
average fish consumption, 70 years exposure.

As with the non-cancer summary, a more detailed discussion of cancer risk will be done with one
species, white sturgeon. This will be followed by a summary of the cancer risks for the rest of the
resident fish species, the anadromous fish species, and finally, a summary across all species.

As previously discussed in Section 6.1.2, all of the cancer risks discussed in this risk
characterization should be considered to be upper bound estimates of the increased risk of
developing cancer as a result offish consumption.

6.2.2.1 Cancer Risk Evaluation for Resident Fish

The potential cancer risks associated with consumption of fillet without skin and whole body
white sturgeon were assessed by first calculating the risk for all detected chemicals with cancer
slope factors (see Appendix I).  These chemical specific risks in each sample were then summed
to estimate the total cancer risk for a study site and for the basin.  For sturgeon, these results are
shown in Table 6-10.
                                             6-116

-------
Table 6-10. Summary of total estimated cancer risks for white sturgeon.
Total Excess Cancer Risk
Consumption Rate/
Exposure Duration
General Public" b
AFC/30-yr

HFC/30-yr

AFC/70-yr

HFC/70-yr

Tissue
Type

FW
WB
FW
WB
FW
WB
FW
WB
Study Site6
CR-6

4X10'5
na
8X104
na
9X10-5
na
2X10'3
na
CR-7

3X1 0'5
na
6X104
na
7X10-5
na
ixio-3
na
CR-8

4X1 0'5
7X10-5
7X104
ixio-3
8X10-5
2X104
2X1 0'3
3X1 0-3
CR-9L

8X1 0'5
6X1 0-5
ixio-3
ixio-3
2X104
IXIO4
3X1 0-3
3X1 0-3
CR-9U

IXIO4
7X10-5
2X1 0-3
ixio-3
3X104
2X104
5X1 0-3
3X1 0-3
SR-13

3X10'5
na
6X104
na
7X10-5
na
ixio-3
na
Basin
Average

5X1 0-5
7X1 0-5
9X104
ixio-3
IXIO4
2X104
2X1 0-3
3X1 0-3
CRITFC's Tribal Member1"
AFC/30-yr

HFC/30-yr

AFC/70-yr

HFC/70-yr

FW
WB
FW
WB
FW
WB
FW
WB
3X104
na
2X10-3
na
8X104
na
5X10-3
na
3X104
na
2X10-3
na
6X104
na
4X10-3
na
3X104
6X104
2X10-3
4X10-3
7X104
ixio-3
4X10-3
9X1 0'3
6X104
5X104
4X1 0-3
3X1 0-3
ixio-3
ixio-3
9X1 0-3
7X1 0-3
ixio-3
6X104
6X10-3
4X10-3
2X1 0-3
ixio-3
ixio-2
8X1 0-3
3X104
na
2X10-3
na
6X104
na
4X10-3
na
4X104
6X104
3X1 0-3
3X1 0-3
ixio-3
ixio-3
6X1 0-3
8X1 0-3
AFC - average fish consumption HFC - high fish consumption F W - fillet without skin WB - whole body
na - not applicable; sample type not analyzed at this study site
"AFC risk based on average U.S. per capita consumption rate of uncooked freshwater and estuarine fish for general public of 7.5 g/day, or 1 8-oz
meal per month (USEPA, 2000a).
bF£FC risk based on 99th percentile U.S. per capita consumption rate of uncooked freshwater and estuarine fish for general public of 142.4 g/day, or
19 8-oz meals per month (USEPA, 2000a).
c AFC risk based on average consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the Columbia
River Basin of 63.2 g/day, or 9 8-oz meals per month (CRITFC 1994).
dHFC risk based on 99th percentile consumption rate for fish consumers in the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the
Columbia River Basin of 389 g/day, or 53 8-oz meals per month (CRITFC 1994).
e Study site descriptions are in Table 1.1. CR = Columbia River; SR = Snake River


As can be seen from Table 6-10, for white sturgeon the total excess cancer risks range from a low

of 3 X 10~5 in fillet without skin samples from the Columbia River (study site 7) and the Snake

River (study site 13) assuming an average fish consumption rate and a 30 year exposure for the

general population adult to a high of 1 X  10~2 in fillet without skin  samples from the Columbia
(study site 9U) assuming a high fish consumption rate and a 70 year exposure  duration for

CRITFC's member tribal adults.
The estimated upper bound cancer risks differ by study site for sturgeon since contaminant levels
vary by study site (Table 6-10).  For example, for one exposure - CRITFC's member tribal adult,
average fish consumption, 30 year exposure - the ingestion of sturgeon (fillet without skin) from

                                                    6-117

-------
the Columbia River (study sites 6, 7 and 8) and the Snake River (study site 13) results in the same
estimated cancer risk, 3 X 10~4, while the risks estimated from consuming fish from the Columbia
River, study site 9L (6 X 10~4) and study site 9U (1 X 10~3) were higher. This same difference was
seen across all study sites (within a given sample type) for each of the exposure groups evaluated
for cancer risk.

As previously discussed for non-cancer effects, the cancer risk at a given study site increases
proportionally with increasing exposure.  For cancer risks, exposures were lowest for the general
public adult, average fish consumption, 30 years exposure and highest for CRITFC's member
tribal adult, high fish consumption, 70 years exposure and depend both upon the exposure
duration (30 or 70 year) and fish consumption rate. Table 6-11 shows the total cancer risks for all
adult populations for white sturgeon (whole body) caught in the Columbia River at study site 8.
Also shown are the ratios of the total cancer risks for the general public,  average fish
consumption at 30 years exposure to that of the other groups assessed in this risk assessment:
CRITFC's member tribal adults with average and high fish consumption at both 30 and 70 years
exposure; the general public adults with high fish consumption at 30 years exposure, and; the
general public adults with average and high fish ingestion at 70 years exposure. As can be seen
from this table, for whole body samples of sturgeon at Columbia River study site 8, the estimated
upper bound cancer risk from eating fish was 7 X 10"5 for the general public, average fish
consumption and 30 years exposure and 1 X 10~3 for the general public,  high fish consumption
and 30 years exposure. This was a difference of about 19 fold (when the rounding of the values
in this table are accounted for).  Likewise, the risks from eating sturgeon for the general public,
average fish consumption and 70 years exposure  was about 2 times higher than that for general
public, average fish consumption and 30 years exposure.

Figure 6-18  shows the differences in cancer risks across sites for sturgeon (fillet without skin) for
CRITFC member tribal adults and general public adults at the high fish consumption for both 30
and 70 year exposures. As can be seen, the cancer risks vary by site with the Hanford Reach  of
the Columbia River (site 9U) having the highest estimated risks.

  Table 6-11. Comparison of estimated total cancer risks among adult populations





General public
General public
CRITFC's member tribe
CRITFC's member tribe
General public
General public
CRITFC's member tribe
CRITFC's member tribe



Fish ingestion rate
(grams/day)
average (7.5)
high (142. 4)
average (63.2)
high (389)
average (7.5)
high (142. 4)
average (63.2)
high (389)


Exposure
duration
(years)
30
30
30
30
70
70
70
70
Total cancer risk for
adults for white
sturgeon at Columbia
River, study site 8
(whole body samples)
7X10-5
1 X 10-3
6X104
4 X 10-3
2X104
3 X 10-3
1 X 10-3
9 X 10-3
Approximate ratio of
estimated cancer risks to
that of general public
with average fish
consumption, 30 years
exposure
1
19
8
52
2
44
20
121
                                             6-118

-------
1H-U1 -
.S3 tE-Oz
K
o
§
o
"3
H tE-03

/
X
XX
!

30 years
exposure
/\
t

General
Public



N. /
x x
X
ft &

CRITFC General
Tribes Public

70 years
exposure
z\




J



1 Site 6
• Site 7
A Site 8
X Site 9L
X Site 9U
• Site 13

CRITFC
Tribes

      Figure 6-18. Comparison of estimated total cancer risks for consumption of white sturgeon across study
      sites for adults in the general public and CRITFC's member tribes at high consumption rates. Note that
      cancer risks for consumption of white sturgeon are the same for study sites 7 and 13.
Figure 6-19 shows the linear relationship between fish ingestion rate and estimated upper bound
basin-wide cancer risk for adults for basin-wide average concentration of chemicals in white
sturgeon fillet samples from the Columbia River Basin assuming both 30 and 70 years exposure
duration.  It also shows that cancer risks for a 70 year exposure were about 2 fold (i.e., 70
years/30 years = 2.3) higher than those for a 30 year exposure (see Appendix N for similar figures
by study site and species).
                                              6-119

-------
              1.0E-02
                                       Fish hgesdon Rate (gramsWay)

           Figure 6-19.  Total cancer risks versus fish consumption rate for adults. White sturgeon,
           basin-wide data (fillet with skin).
In the previous discussion on non-cancer results, it was shown that a small number of chemicals
were responsible for most of the non-cancer health hazards from consuming fish. Tables 6-12
(fillet without skin) and Table 6-13 (whole body)  show the chemicals with cancer risks at or
greater than 1  X 10~5 for sturgeon for CRITFC's member tribal adults, average fish consumption
and 70 years exposure duration. For cancer risks, a limited (but larger) number of chemicals were
responsible for the majority of the cancer risk. These chemicals are:

       PCBs, including both Aroclors and dioxin-like PCB congeners,

•      chlorinated dioxins and furans, with 2,3,7,8,-TCDF having the highest risk among the
       congeners,

       the pesticides aldrin, chlordane (total), DDD, DDE, and hexachlorobenzene, with DDE
       having the highest risk, and

       one metal, arsenic.
Not all chemicals were detected at every study site. For example, in the table with fillet without
skin results (Table 6-12), Aroclors and PCB congeners 105, 118 and 156 were detected in all of
the study site samples while other PCB congeners were detected at only one or two study sites.
                                             6-120

-------
Table 6-12. Chemicals with estimated cancer risks at or greater than 1 X IQr5 for white sturgeon, fillet
without skin. CRTTFC's member tribal adult, average fish consumption, 70 years exposure.
Study Site*

PCBs
Total Aroclors**
PCB 105
PCB 114
PCB 118
PCB 126
PCB 156
PCB 157
Dioxin/furans
1,2,3,7,8-PeCDD
2,3,4,7,8-PeCDF
2,3,7,8-TCDD
2,3,7,8-TCDF
Pesticides
Aldrin
Chlordane (total)
DDD
DDE
Hexachlorobenzene
Metals
Arsenic
Total Cancer Risk for All Chemicals
"<" means that estimated cancer risk was less than 1 X 10"5
CR-6 CR-7

2X104 1X104
3 XIO-5 2 XIO-5
i xio-5 <
3 XIO-5 2 XIO-5
< 2 XIO-5
4 XIO-5 3 XIO-5
< <

1X10-5 2 XIO-5
< 1 X ID'5
4 XIO-5 5 XIO-5
2 XIO4 2 XIO4

< <
< <
1 X 10-5 1 X 10-5
1 X 104 IX 104
< <

4 X 10-5 5 X 10-5
8 XIO4 6 XIO4
*Study site descriptions are in Table 1.
CR-8 SR-13

1 X 104 IX 104
2 XIO-5 3 XIO-5
< 1 X ID'5
2 XIO-5 4 XIO-5
< <
3 XIO-5 5 XIO-5
< <

2 XIO-5 1X10-5
2 XIO-5 <
6 XIO-5 5 XIO-5
2 XIO4 6 XIO-5

< <
< <
1 X 10-5 1 X 10-5
1X104 1X104
< <

5 X 10-5 3 X 10-5
7 XIO4 6 XIO4
CR-9L

3 XIO4
4 XIO-5
2 XIO-5
5 X ID'5
<
9 XIO-5
2 XIO-5

<
2 XIO-5
1X104
5 XIO4

2 XIO-5
1 X 10-5
4 XIO-5
2 XIO4
2 XIO5

5 X 10-5
1 X 10-3
CR-9TJ

7 XIO4
1X104
5 XIO-5
2 XIO4
<
2 XIO4
5 X ID'5

<
2 XIO-5
3 X ID'5
3 XIO4

1 X ID'5
2 XIO-5
8 XIO-5
4 XIO4
<

4 XIO-5
2 XIO-3
1. CR = Columbia River; SR = Snake River
* * Based on "adjusted" Aroclor concentration (see Section 5.3.2)
Table 6-13. Chemicals with estimated cancer risks at or greater than 1 X 10 5 for white sturgeon,
whole body. CRTTFC's member tribal adult, average fish consumption. 70 years exposure.


PCBs
Total Aroclors**
PCB 105
PCB 114
PCB 118
PCB 156
PCB 157
Dioxin/furans
2,3,4,7,8-PeCDF
2,3,7,8-TCDD
2,3,7,8-TCDF
Pesticides
Aldrin
Chlordane (total)
DDD
DDE
Hexachlorobenzene
Metals
Arsenic
Total Cancer Risk for All Chemicals

CR-8

3 XIO4
6 XIO-5
2 XIO-5
7 XIO-5
1X104
2 XIO-5

2 XIO-5
9 XIO-5
3 XIO4

<
<
2 XIO-5
2 XIO4
<

7 XIO-5
1 X 10-3
Study Site*
CR-9L

2 XIO4
4 X 10-5
2 X 10-5
5 X 10-5
9 X 10-5
2 X 10-5

3 X 10-5
1X104
3 XIO4

2 X 10-5
1 X 10-5
3 X 10-5
2 XIO4
2 X 10-5

4 X 10-5
1 X 10-3

CR-9U

3 XIO4
5 X 10-5
2 X 10-5
5 X 10-5
9 X 10-5
2 X 10-5

2 X 10-5
9 X 10-5
4 XIO4

2 X 10-5
<
5 X 10-5
2 XIO4
1 X 10-5

4 X 10-5
1 X 10-3






















"<" means that estimated cancer risk was less than 1 X 10"5.  CR = Columbia River
 *Study site descriptions are in Table 1-1.  "Based on "adjusted Aroclor concentration (see Section 5.3.2)
                                              6-121

-------
The total cancer risk estimates and the summary of chemicals with risks at or greater than
1 X 10~5 for other resident fish species are provided in Appendix O by species: Tables 1.3 and 1.4
(bridgelip sucker), 2.3 and 2.4 (largescale sucker), 3.3 and 3.4 (mountain whitefish), 4.3 and 4.4
(white sturgeon), 5.3 and 5.4 (walleye), and 6.3 and 6.4 (rainbow trout). Table 6-14 shows a
summary of the total cancer risk estimates for the resident fish species for one adult population -
CRITFC's member tribal adults with an average fish consumption and 70 years exposure.
Results of the fillet with skin samples are shown, except for sturgeon (only fillet without skin
sampled) and bridgelip sucker (only whole body sampled).
Table 6-14. Summary of estimated total cancer risks by study site and basin-wide, resident fish species.
CRITFC's tribal member adult, average fish consumption, 70 years exposure
Snecies
Bridgelip sucker
Largescale sucker





Mountain whitefish



White sturgeon





Walleye
Rainbow trout

Sample Study site
N tvne name
3 WB Yakima
19 FS Columbia
Deschutes
Umatilla
Snake
Yakima
Yakima
12 FS Columbia
Deschutes
Umatilla
Yakima
16 FW Columbia
Columbia
Columbia
Columbia
Columbia
Snake
3 FS Umatilla
7 FS Deschutes
Yakima
Study
Site
48
9U
98
30
13
48
49
9U
98
101
48
6
7
8
9L
9U
13
30
98
49
Study site
cancer risk
5X104
6X104
1X104
2X104
2X104
4X104
3X104
4 X 10-3
3X104
1X104
1 X ID'3
8X104
6X104
7X104
1 X 10-3
2 X 10-3
6X104
2X104
2X104
2X104
Range in study site Basin
cancer risks cancer risk
5 X 104 5 X 10-4*
lto6X104 4X104





ixio-4to4xio-3 ixio-3



6X1°-4to2X10-3 1X1(>3





2X104 2X10'4*
2X104 2X104

       N= number of samples; WB = whole body; FS = fillet with skin; FW= fillet without skin
       * Basin-wide cancer risk based on one study site

White sturgeon and mountain whitefish had the highest estimated basin-wide cancer risks at 1 X
10~3 (Table 6-14). All of the white sturgeon study site cancer risks were at or greater than 6 X 10~4
with a high of 2 X 10~3.  The highest cancer risks for sturgeon were from consuming fish from the
Columbia River at study sites 9L (1 X 10'3) and 9U (2 X 10'3). The four mountain whitefish
study sites span more than an order of magnitude in cancer risk -1 X 10~4 for the Umatilla (study
site 101),  3 X 10-4 for the Deschutes (study site 98), 1 X 10'3 for the Yakima (study site 48), and 4
X 10~3 for the Columbia River (study site 9U).  Cancer risks were highest for the Yakima (study
site 48) and Columbia River (study site 9U) for whitefish and for the Columbia River at study
sites 9U and 9L for sturgeon.

Bridgelip sucker (one study site at 5 X 10~4) and largescale sucker (six study sites ranging from 1
to 6 X 10~4) had the next highest basin-wide cancer risks, 5 X 10~4 and 4 X  10~4, respectively.
Walleye (one study site at 2 X 10~4) and rainbow trout (two study sites at 2 X 10~4) had the lowest
basin-wide cancer risks.

                                             6-122

-------
Figure 6-20 summarizes the total basin-wide cancer risks for resident fish species for adults using
high and average fish consumption rates for the general public and for CRITFC's member tribal
populations assuming 70 years exposure duration.  Note that the Y axis is on a logarithmic scale
and that each bar begins at 0 on the Y axis. For example, the cancer risk for mountain whitefish
for the general public adult, high fish consumption for 70 years, is 3 X 10"3; for CRITFC member
tribal adults, high fish consumption for 70 years, the cancer risk estimates is 8 X 10"3.  As with
Table 6-14, this figure shows that consumption of mountain whitefish and white sturgeon result in
the highest cancer risks, followed by sucker, rainbow trout, and walleye.  It also shows that for all
species, the total cancer risks were highest for CRITFC's member tribal adults at the high fish
ingestion rates (389 g/day) followed by the general public adult, high ingestion rate (142.4 g/day);
CRITFC's member tribal adult, average ingestion  rate (63.2 g/day); and general public adult,
average ingestion rate (7.5 g/day).

For a more detailed comparison of cancer risks across resident fish species for each study site, see
Appendix N. In this appendix, cancer risks are shown over a range of ingestion rates for all
species caught at a study site.
          HI
          u
          c
          re
          O
                1.E-04
                       White
                      Sturgeon
                       (n=16)
Largescale
 Sucker
 (n=19)
Rainbow
 Trout
 (n=7)
Walleye
 (n=3)
                                               Species
            n =number of samples
            * Fillet of skin samples except
            for sturgeon (whole body)
  Figure 6-20. Adult cancer risks for resident fish species*. Columbia River Basin data (70 years exposure).
                                              6-123

-------
The chemicals with cancer risks equal to or greater than 1 X 10~5 for resident fish species are
shown in Appendix O for CRITFC's member tribal adults for the average fish consumption rate
and 70 years exposure (Tables 1.4 (bridgelip sucker), 2.4.1  and 2.4.2 (largescale sucker), 3.4.1
and 3.4.2 (mountain whitefish), 4.4.1 and 4.4.2 (white sturgeon), 5.4.1 and 5.4.2 (walleye), and
6.4.1 and 6.4.2 (rainbow trout).

In general, four chemical classes (PCBs, chlorinated dioxins and furans, pesticides and metals)
were responsible for the cancer risks at or greater than 1 X 10~5 for all of the resident fish species.
The exception to this was two study site samples for largescale sucker: the Snake River (study
site 13, fillet with skin) had 2 semivolatiles at or greater than a 1 X 10~5 cancer risk,
dibenz(a,h)anthracene and benzo(a)pyrene, and the Yakima River (study site 49, whole body) had
one, 1,2-diphenylhydrazine.

For the metals, only one of the contaminants detected, inorganic arsenic, had an oral cancer slope
factor.  Thus, inorganic arsenic was the only detected metal for which cancer risks were
estimated.

For the three other classes of chemicals contributing the most to the cancer risk (PCBs,
dioxins/furans, and pesticides), the chemicals within each class that were at or greater than 1 X
10~5 vary among species and sometimes among different sample types of the same species.  For
example, the pesticide, hexachlorobenzene, was found at a level greater than 1 X 10~5 risk in only
three white sturgeon samples: at Columbia River study site 9L for fillet without skin and at
Columbia River study sites 9L and 9U for whole body samples. Aldrin was found at a cancer risk
greater than 1 X 10~5 in only 2 species: at the Columbia River, study sites 9L and 9U, for both
types of sturgeon samples (fillet without skin and whole body); and at Columbia River study site
9U for whitefish samples (whole body and fillet with skin).

All study sites and species had total Aroclors at or greater than a risk of 1 X 10~5 except for the
Snake River (study site 13) for largescale sucker (fillet with skin). Up to seven different PCB
congeners (105, 114,  118, 126, 156,  157 and 169) were found at or greater than a risk of 1 X lO'5
with the number per study site varying from zero to seven at different study sites. Up to four
dioxins/furans (2,3,7,8-TCDF, 2,3,4,7,8-PCDF, 2,3,7,8-TCDD and 1,2,3,7,8-PCDD) were at or
greater than a cancer risk of 1 X 10~5 with the number varying from two to four per study site.

Table 6-15 and Figures 6-21 through 6-26 show the percent contribution to total cancer risk from
each chemical and class of chemical using the basin-wide cancer risk data for resident fish (fillet
with skin for all species except sturgeon (fillet without skin) and bridgelip sucker (whole body).
                                             6-124

-------
Table 6-15. Percent contribution of contaminant groups to estimated cancer risks for resident fish species.
Based on Columbia River Basin-wide averages.
White
Sturgeon
Tissue Type
Number of Samples
Total Metals
Arsenic
Total PCBs/Aroclors
PCB 105
PCB 114
PCB 118
PCB 126
PCB 156
PCB 157
PCB 169
Other PCBs
Total Aroclors*
Total Semi- Vocatives
1 ,2-Dipheny Ihydrazine
Benzo(a)pyrene
Dibenz [a,h] anthracene
Indeno(l,2,3-cd)pyrene
Other Semi- Vocatives
Total Pesticides
Aldrin
DDD
DDE
DDT
Heptachlor Epoxide
Hexachlorobenzene
Other Pesticides
Total Dioxins/Furans
2,3,4,6,7,8-HxCDF
2,3,4,7,8-PeCDF
2,3,7,8-TCDD
2,3,7,8-TCDF
OCDD
OCDF
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
other dioxins
FW
16
4
4
39
3
1
4
2
6
1
ND
<1
21
ND
ND
ND
ND
ND
ND
23
2
2
15
<1
1
1
2
36
<1
1
7
26
<1
<1
1
<1
1
Largescale
Sucker
FS
19
2
2
46
2
1
6
9
6
1
2
<1
19
28
ND
8
17
2
2
21
ND
1
16
2
ND
ND
2
5
<1
<1
1
1
<1
<1
2
<1
1
Mountain
Whiteflsh
FS
12
1
1
83
6
2
15
18
12
2
<1
1
26
ND
ND
ND
ND
ND
ND
10
2
1
8
<1
ND
<1
<1
8
<1
1
1
5
<1
<1
2
<1
<1
Walleve
FS
3
33
33
31
3
1
6
ND
6
ND
ND
<1
15
ND
ND
ND
ND
ND
ND
11
ND
1
10
<1
ND
ND
ND
26
1
1
7
6
<1
ND
7
1
2
Rainbow
Trout
FS
7
ND
ND
68
4
2
9
29
8
2
ND
<1
15
ND
ND
ND
ND
ND
ND
5
ND
<1
4
1
ND
ND
<1
29
2
2
6
2
<1
<1
13
1
4
Bridgelip
Sucker
WE
3
8
8
46
2
1
3
14
4
ND
1
<1
22
1
1
ND
ND
ND
ND
32
ND
3
25
3
ND
ND
<1
13
<1
2
2
3
<1
<1
5
<1
1
ND=Not detected; *Based on adjusted Aroclor concentration (See Section 5.3.2)
                                                 6-125

-------
                  Other Dioxins
                   and Furans
                      4%
                              Arsenic
                                4%
        PCB105
          3%
          2,3,7,8-TCDF
             26%
         2,3,7,8-TCDD
              7%
                    Other
                   Pesticides
                     4%
Total DDT
  18%
                 PCB 118
                   4%

                  PCB 156
                     6%
                     Other PCB
                     Congeners
                        4%
                   Total Aroclors
                       20%
Figure 6-21. Percent contribution of basin-wide average chemical concentrations to
cancer risk from consumption of white sturgeon fillet without skin. Number of samples
= 16.
                      Other
                    Pesticides
                      2%
                          Dioxins and
                            Furans
                              6%
     Arsenic
       2%   PCB 118
           "   5%
                   PCB 126
                     9%
             Total DDT
               18%
          Other Semi-
           Volatiles
             3%
         Dibenz[a,h]anthracene
                       PCB 156
                          6%

                      Other PCB
                      Congeners
                         7%
                            Benzo(a)pyrene
                                 8%
             Total Aroclors
                 19%
Figure 6-22. Percent contribution of basin-wide average chemical concentrations to
cancer risk from consumption of largescale sucker fillet with skin. Number of samples :
19.
                                   6-126

-------
                         Other
                     Dioxins/Furans
                           8%
Arsenic
  8%
PCB118
   3%
                1,2,3,7,8-
                 PeCDD
                  5%
             Total DDT
                32%
                 PCB126
                   14%
                                                            PCB156
                                                              4%
               Other PCBs
                   3%
                               1,2-
                          Diphenylhydrazi
                                ne
                                1%
                                                  Total Aroclors
                                                      22%
Figure 6-23.  Percent contribution of basin-wide average chemical concentrations to cancer risk from
consumption of whole body bridgelip sucker. Number of samples = 3.
                             Other
                          Dioxins/Furans
                              10%
 PCB 105
   4%   PCB 118
          9%
                       1,2,3,7,8-
                        PeCDD
                         13%


                   2,3,7,8-TCDD
                       6%
                        Total Aroclors
                            15%
      PCB 156
        8%
           Figure 6-24. Percent contribution of basin-wide average chemical
           concentrations to cancer risk from consumption of rainbow trout fillet with skin.
           Number of samples = 7.
                                       6-127

-------
                      1,2,3,7,8
                       PeCDD
                         7%
                                    Other
                                Dioxins/Furans
                                     5%
          2,3,7,8-TCDF
               6%
        2,3,7,8-TCDD
             7%
               Total DDT
                  11%
                                               Arsenic
                                                 33%
                                              PCB118
                                                 6%
                   Total Aroclors
                       15%
                                           PCB156
                                             6%
                                       Other PCBs
                                           5%
Figure 6-25.  Percent contribution of basin-wide average chemical concentrations to cancer
risk from consumption of walleye fillet with skin. Number of samples =3.
                     Other
                   Pesticides
                      2%
                          Dioxins and
                            Furans
                                    PCB105
                                      6%
               Total DDT
                 9%
            Total Aroclors
                26%
                                                 PCB118
                                                   15%
                                                    PCB126
                                                      19%
                                            PCB156
                                              12%
                     Other PCBs
                     Congeners-
                        5%
Figure 6-26.  Percent contribution of basin-wide average chemical concentrations to
cancer risk from consumption of mountain whitefish fillet with skin. Number of
samples =12.
                                  6-128

-------
For all of the resident fish species except walleye, the majority of the cancer risk was from
dioxins and furans, a small number of pesticides and PCBs. (Table 6-15 and Figures 6-21 through
6-26). Inorganic arsenic contributes to about 33% of the cancer risk for walleye.

•      Chlorinated dioxins and furans contribute from 5% of the total cancer risk for largescale
       sucker to 36% for sturgeon. For sturgeon, 2,3,7,8-TCDF  was by far the largest contributor
       of the dioxins/furans. For some of the other species, other congeners (e.g., 2,3,7,8-TCDD
       and  1,2,3,7,8-PeCDD) were contributors to the dioxin/furan cancer risk.

       Pesticides contribute from about 5% to 32% of the total cancer risk, with DDE
       contributing more than any other pesticide.

•      PCBs (both total Aroclors and dioxin-like congeners) contribute from 31% to 83% of the
       total cancer risk.  The contribution from  Aroclors (primarily 1254 and  1260) to the cancer
       risk for this class of chemicals was approximately 15% for rainbow trout, 26% for
       mountain whitefish, 19% for largescale sucker, 22% for bridgelip sucker, 15% for
       walleye, and 21% for sturgeon. The contribution to PCB cancer risk from the dioxin-like
       PCB congeners ranges from a low of 17% for walleye to a high of 56% for mountain
       whitefish.

       The  contribution from inorganic arsenic to total cancer risk was from 0% (not detected in
       rainbow trout fillets) to 33% for the resident fish species.  For most species, the value was
       less than 8%. The exception was walleye at 33%.

6.2.2.2 Cancer Risk Evaluation for Anadromous Fish

The total cancer risk estimates for the anadromous fish species are provided in Appendix P by
species: Tables 1.3 (coho salmon), 2.3 (fall chinook salmon), 3.3 (spring chinook salmon), 4.3
(steelhead), 5.3 (eulachon), and 6.3 (Pacific lamprey).

Table 6-16 summarizes the estimates of the total cancer risks for anadromous fish species by
study site and by basin for CRITFC's member tribal adults, average consumption rate (63.2
g/day), and 70 years exposure. Fillet with skin data are shown except for eulachon, which had
only whole body samples collected. Figure 6-27 shows the relative differences in cancer risks for
anadromous  fish species using average and high fish consumption rates for the general public and
CRITFC's member tribal adult assuming 70 years exposure. Note that the Y axis is on a
logarithmic scale and that all of the bars begin at 0 on the Y axis.  For example, the cancer risk for
Pacific lamprey for the general public adult, high fish consumption for 70 years,  is slightly
greater than  1 X 10~3;  for CRITFC member tribal adults, high fish consumption for 70 years, the
cancer risk estimates is 4 X 10~3. Columbia River Basin data are shown for all species (for coho
salmon, eulachon and Pacific lamprey, only one study site was sampled).
                                            6-129

-------
Table 6-16. Summary of estimated total cancer risks by study site and basin-wide, anadromous fish species
CRTTFC's tribal member adult, average fish consumption, 70 years exposure
Snecies
Coho salmon
Fall chinook salmon




Spring chinook salmon







Steelhead





Eulachon
Pacific lamprey
Sample
N tvne Study site name
3 FS Umatilla
15 FS Columbia
Columbia
Klickitat
Umatilla
Yakima
24 FS Willamette
Wind River
Little White Salmon
Klickitat
Looking Glass Creek
Umatilla
Yakima
Icicle Creek
21 FS Columbia
Hood River
Klickitat
Snake River
Clearwater
Yakima
3 WE Columbia
3 FS Willamette
Study
site#
30
8
14
56
30
48
21
63
62
56
94
30
48
51
8
25
56
93
96
48
3
21
Study site
cancer risk
2X104
2X104
2X104
2X104
1X104
2X104
2X104
2X104
2X104
2X104
2X104
3X104
2X104
2X104
1X104
3X104
2X104
2X104
3X104
2X104
2X104
6X104
Range in
study site
cancer risks
2X104
1 to2X104




2to3X104







ItoSXIO4





2X104
6X104
Basin
cancer
risk
2X104*
2X104




2X104







2X104





2X104*
6X104*
N= Number of Samples WB = whole body; FS = fillet with skin
* Basin-wide cancer risks based on one study site
                                                    6-130

-------
                                      Basin Data (70 years exposure)
            E
                 1.E-02
                 1.E-03
                 1.E-04--
                 1.E-05
                     Spring Chinook Fall Chinook  Coho Salmon  Steelhead
                        (n=24)      (n=15)       (n=3)       (n=21)
Eulachon
 (n=3)
Pacific Lamprey
   (n=3)
         n = number of samples
         * Fillet of skin samples except
         for eulachon (whole body)
                                               Species
                                                                    General Public
     Consumption
     High Fish
     Consumption
      CRTTFC
      Member Tribes
        I Average Fish
      I	| Consumption

      I  I  HighFish
      !	| Consumption
     Figure 6-27. Adult cancer risks for anadromous fish species*. Columbia River Basin-wide average data (70 years
     exposure).
For coho salmon, fall chinook salmon, spring chinook salmon, steelhead and eulachon, the study
site cancer risks were all within a range of 1 X 10~4 to 3 X 10~4 and the basin-wide risks were at
approximately 2 X 10~4. The estimated cancer risk from consumption of Pacific lamprey was 6 X
10-4 (Table 6-16).

For all species, the total cancer risks were highest for CRITFC's member tribal adults at the high
fish ingestion rates (389 g/day) followed by the general public, high ingestion rate (142.4 g/day);
CRITFC's member tribal adult, average ingestion rate (63.2 g/day); and general public, average
ingestion rate (7.5 g/day) (Figure 6-27).

For a more detailed comparison of cancer risks across anadromous fish species for each study
site, see Appendix N.  In this appendix, estimated cancer risks are shown for all species caught at
a study site for a range of ingestion rates.
                                                6-131

-------
The chemicals with risks at or greater than 1 X 10~5 for each species for CRITFC's member tribal
adults with average fish consumption and 70 years exposure are summarized in Appendix P by
species. A review of this appendix shows that:

•      For steelhead, spring chinook salmon, and fall chinook salmon, the same three chemical
       classes (PCBs, dioxins/furans, and one inorganic, arsenic) were responsible for the
       majority of the risks at or greater than 1  X 10~5.  Fillet with skin and whole body samples
       of coho had no risks greater than 10~5 for dioxins and furans while whole body samples
       had a 1 X 10~5 risk for DDE.  For spring and fall chinook salmon and steelhead, which had
       dioxins and furans risks at or greater thanl X 10~5, three congeners were greater than this
       risk level - 1,2,3,7,8-PCDD; 2,3,4,7,8-PCDF; and 2,3,7,8-TCDF. For steelhead and all
       three salmon, Aroclors and PCB congeners 126 and 118 were found at all study sites at or
       greater than 1 X 10~5, as was inorganic arsenic.

       Eulachon was sampled at only one site (Columbia River, study site 3). Risks from
       consumption of the whole body composite sample were at or greater than 1 X 10~5 for two
       chemicals, arsenic and 1,2,3,7,8-PCDD.

•      Pacific lamprey collected at two sites -Willamette Falls (21) and Fifteen Mile Creek (24)
       - had risks at or greater than 1 X 10~5 for four classes of chemicals: PCBs (Aroclors as
       well as PCBs 105,114,118,126, and 156); chlorinated dioxins/furans (1,2,3,7,8-PCDD and
       2,3,7,8-TCDF); metals (inorganic arsenic); and pesticides (total chlordane, DDT, DDE
       and hexachlorobenzene).

Tables 6-17 and Figures 6-28 through 6-33 show the percent contribution to total cancer risk for
each chemical and/or chemical class using basin-wide cancer risk data (based on fillet of skin
data for all species except eulachon which was whole body).

A review of Table 6-17 and Figures 6-28 through 6-33 shows that:

•      Arsenic contributes from 33 to 54% of the total cancer risk for salmon and steelhead; 58%
       for eulachon; and only about 7% for lamprey.

•      PCBs (Aroclors and dioxin-like congeners) contribute from 32 to 50% of the total cancer
       risk for the salmon and steelhead, 77% for lamprey, and only  4% for eulachon. For the
       salmon, steelhead, and lamprey, Aroclors contribute from 12  to 28% of the total  cancer
       risk. Aroclors were not detected in eulachon. Nine different  PCB congeners were
       detected with PCB  126 contributing the most to total cancer risk (from 6 to 35%) for all
       species except eulachon. PCB 126 was not detected in eulachon.

       The percent contribution from all pesticides was from about 1 to 9% of the risk.

•      The contribution to total cancer risk for chlorinated dioxins and furans was from
       9 to 14% for all species except eulachon. For eulachon, the percent contribution to total
       cancer risk is about 36%.

                                           6-132

-------
        Salmon and steelhead look very similar in that arsenic and PCBs were the major
        contributors to cancer risk followed by dioxin/furans and then pesticides. For Pacific
        lamprey, PCBs were the major risk contributor at 77% with the rest of the risk split
        between arsenic, dioxin/furans and pesticides. Most of the risk for eulachon is from
        arsenic, then dioxins/furans with less than 4% from PCBs and pesticides combined.
Table 6-17. Percent contribution of contaminant groups to cancer risk for anadromous fish species.
Based on Columbia River Basin-wide averages.

Tissue Type
Number of samples
Total Metals
Arsenic
Total PCB/Aroclors
PCB 105
PCB 114
PCB 118
PCB 123
PCB 126
PCB 156
PCB 157
PCB 169
Other PCBs
Total Aroclors**
Total Pesticides
Aldrin
Chlordane total
DDD
DDE
DDT
Heptachlor Epoxide
Hexachlorobenzene
Total Dioxins/Furans
2,3,4,6,7,8-HxCDF
2,3,4,7,8-PeCDF
2,3,7,8-TCDD
2,3,7,8-TCDF
OCDD
OCDF
1,2,3,7,8-PeCDD
1,2,3,4,7,8-HxCDD
Other dioxins
Spring
Chinook
Salmon
FS
24
50
50
32
1
1
3
<1
14
1
<1
ND
<1
12
4
ND
1
<1
2
1
ND
1
14
<1
4
1
4
<1
<1
4
<1
1
Coho Salmon
FS
15
45
45
43
3
1
ND
<1
6
5
ND
ND
<1
28
1
ND
<1
<1
<1
<1
ND
ND
11
ND
2
1
4
<1
<1
3
ND
1
Fall Chinook
Salmon
FS
3
54
54
32
2
1
4
<1
10
1
<1
ND
<1
15
4
ND
1
<1
2
<1
ND
1
9
ND
1
1
5
<1
<1
2
ND
<1
Steelhead
FS
21
33
33
50
1
1
3
<1
24
2
<1
<1
<1
19
4
ND
1
<1
2
<1
ND
1
14
<1
6
1
2
<1
<1
4
<1
1
Pacific
Lamp rev
FS
3
1
1
77
3
2
8
<1
35
3
1
ND
<1
25
9
ND
2
<1
3
2
ND
2
9
<1
1
1
3
<1
ND
2
<1
1
Eulachon
WB
3
58
58
4
1
<1
2
<1
ND
1
<1
ND
<1
ND
2
ND
ND
ND
2
ND
ND
ND
36
1
4
5
5
<1
<1
16
1
5
* Number in parenthesis is number of samples in basin data ** Based on adjusted Aroclor concentration (see Section 5.3.2)
ND = not detected
                                               6-133

-------
             2,3,4,7,8-
             PeCDF
               4%

         Total DDT
            2%
                Other
              Pesticides
                 2%
         Total Aroclors
             12%

            Other PCB
            Congeners
               4%
                         2,3,7,8-TCDF
   1,2,3,7,8-
    PeCDD
     4%
            Other Dioxins
             and Furans
                 3%
                Arsenic
                  50%
                       PCB 126
                         14%
  PCB 118
    3%
Figure 6-28.  Percent contribution of basin-wide average chemical concentrations to cancer
risk from consumption of spring chinook fillet with skin.  Number of samples = 8.
                      2,3,7,8-TCDF
                          4%
                  Total DDT
                     1%

                 Total Chlordane
                    0.3%
                Total Aroclors
                    28%
1,2,3,7,8-
 PeCDD
  30/0     Other Dioxins
          and Furans
              4%
                 Arsenic
                  45%
                        Other PCB
                        Congeners-1 RGB 156
                           4%       5%
        PCB 126
          6%
      Figure 6-29.  Percent contribution of basin-wide average chemical concentrations
      to cancer risk from consumption of coho salmon fillet with skin.  Number of
      samples =3.
                                          6-134

-------
                      2,3,7,8-TCDF
                           5%
                 Total DDT
                    2%
                 Other
               Pesticides'
                  2%
            Other Dioxins
             and Furans
                 4%
          Total Aroclors
              15%
            Other PCB
            Congeners
               4%
                          Arsenic
                           54%
                    PCB 126
                      10%
     PCB 118
        4%
Figure 6-30.  Percent contribution of basin-wide average chemical concentrations to
cancer risk from consumption of fall chinook salmon fillet with skin.  Number of samples
= 15.
               2,3,4,7,8-
                PeCDF
                  5%
         Total DDT
            2%
             Other
           Pesticides
              1%

         Total Aroclors
             19%
1,2,3,7,8-
 PeCDD
   4%
Other Dioxins
 and Furans
     5%
                        Arsenic
                         33%
                           PCB 118
                             3%
                  Other PCB
                  Congeners
                     5%
                PCB 126
                 23%
 Figure 6-31.  Percent contribution of basin-wide average chemical concentrations to
 cancer risk from consumption of steelhead fillet with skin.  Number of samples = 21.
                                   6-135

-------
                                        Other Dioxins
                                         and Furans
                                             5%
                                                       Arsenic
                                                         33%
          Total Aroclors
              1 9%
                   PCB 118

                      3%
                    Other PCB
                    Congeners
                        5%
Figure 6-32.  Percent contribution of basin-wide average chemical concentrations to cancer risk from
consumption of Pacific lamprey fillet with skin. Number of samples =3.
                      1,2,3,7,8-
                       PeCDD
                        16%
            2,3,7,8-TCDF
                5%
               2,3,7,8-TCDD
                   5%
           2,3,4,7,8-
            PeCDF 	
              4%     2,3,4,6,7,8-
                      HxCDF
                        1%
Other Dioxins
 and Furans
     5%
                      Arsenic
                       57%
Total
DDTV
2%
/ ' >
Other PCB
Congeners
3%
LPCB118
2%
      Figure 6-33. Percent contribution of basin-wide average chemical concentrations to cancer
      risk from consumption of whole body eulachon.  Number of samples = 3.
                                       6-136

-------
6.2.2.3  Comparisons of Cancer Risks Between Anadromous Fish and Resident Fish Species

Table 6-18 shows a summary of the estimated total upper bound cancer risks for the basin and
across study sites for all species at the high fish consumption rate for CRITFC's member tribal
adults, 70 years exposure. It should be noted that the cancer risk estimates in Table 6-18 were
calculated using high fish ingestion rates for CRITFC's member tribal adults, 70 years of
exposure, while the results previously discussed for resident fish species in Table 6-14 and for
anadromous fish species in Table 6-16 were calculated using average fish ingestion rates for
CRITFC's member tribal adults, 70 years exposure.  Conclusions from the comparisons in Table
6-18 are limited by the fact that different species were caught at different study sites and that
sample numbers and types for each species varied.

Table 6-18 and the study site specific data in the tables in Appendices O and P show that for
CRITFC's member tribal adults consuming fish at the high ingestion rate for 70 years:

•      The basin-wide risks for rainbow trout and five of the anadromous fish (coho, spring, and
      fall chinook salmon, steelhead, and eulachori) were all estimated to be 1 X 10"3. The
       range in the study site risks for the four species that had multiple study sites sampled was
       generally small: less  than 2 fold for rainbow trout, fall chinook, and spring chinook.
       Steelhead had a slightly larger range (7 X 10~4 to 2 X 10~3) due primarily to an estimated
       cancer risk of 7 X 10"4 at the Columbia River (study site 8); the estimated cancer risks for
       the other 5 study sites were at 1 or 2 X 10~3.

•      The basin-wide risk  for walleye was 9  X 10"4. The cancer risk for this one sample was
       within the range of study site  risks for the species discussed in the previous bullet
       (rainbow trout, eulachon, the three salmon, and steelhead).

•      The estimated basin-wide risks for high ingestion by  adults in CRITFC's member tribes
       were greater than 1  X 10~3 among the remaining five species, with mountain whitefish and
       white sturgeon having the highest estimated basin-wide risks: largescale sucker (2X10"
       3); bridgelip sucker (3 X  10'3); lamprey (4 X 10'3); sturgeon (6 X 10'3), and; whitefish (8
       X 10"3). Three of these species had more than one study site used in the calculation of the
       basin-wide cancer risks, largescale sucker, sturgeon and whitefish. The range in cancer
       risks among the study sites sampled for sturgeon was about three-fold; for largescale
       sucker, about five-fold, and; for whitefish, about twenty-eight fold. The large difference
       in risk among study  sites for whitefish was due to the low estimate of cancer risk of 7 X
       10"4 for samples from the Umatilla (study site 101) and the high estimate of cancer risk of
       2 X  10"2 at the Hanford Reach of the Columbia River (study site 9U). For sturgeon, no
       study site  risk was less than 4 X 10"3; the study site with the highest estimated cancer risk
       was the Columbia River at study site 9U.
                                            6-137

-------
   Table 6-18.  Summary of estimated total cancer risks by study site and basin-wide, all species.  CRTTFC's
   tribal member adult, high fish consumption, 70 years exposure
Species
Resident species
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
Anadromous species
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey
N

3
19
12
16
3
7

3
15
24
21
3
3
Sample
tvne

WB
FS
FS
FW
FS
FS

FS
FS
FS
FS
WB
FS
Range in study site cancer risks

3 X 1C'3
8 X 104 to 4 X lO-3
7X104 to 2X 10-2
4X10'3 to IXIO'2
9X104
1 X 10-3, 1 X 1O3

1 X lO'3
9 X 104 to IX 10-3
1 to 2 X 10-3
7X104 to 2X10-3
1 X 10-3
4 X lO'3
Basin cancer risk

3 X 1C'3*
2X10-3
8X10-3
6X10'3
9 X 104*
1 X 1O3

1 X lO'3*
1 X 10-3
1 X 10-3
1 X 10-3
1 X 10-3*
4 X lO'3*
 WB = whole body; FS = fillet with skin; FW = fillet without skin; N = number of samples
 * Basin-wide cancer risks based on one study site

Figure 6-34 is a summary of the cancer risks estimated to result from consumption of the resident
fish and anadromous fish at all four ingestion rates for adults: general public adult, average and
high fish consumption; CRTTFC's member tribal adult, average and high fish consumption,
assuming 70 years exposure.  (Note that the Y axis is on a logarithmic scale). Basin-wide fillet
with skin data were used for this figure, except for those species that had only whole body
samples (bridgelip sucker and eulachon) or fillet without skin samples (sturgeon).  The basin-
wide cancer risks vary by species, with mountain whitefish having the highest estimated cancer
risks and white sturgeon having the second highest among the species sampled. Lamprey,
bridgelip sucker and largescale sucker were the next highest followed by the remaining seven
species - the three salmon, steelhead, eulachon, rainbow trout, and walleye.
                                              6-138

-------
        1.E-02
        1.E-03
    E

     I
     ra
    O
        1.E-04 -•
        1.E-05
                     Anadromous
Resident
                                                                            I
        "Fillet of skin data except for white sturgeon
        (fillet without skin) and bridgelip sucker and
        eulachon (whole body)
                                           Species
  Figure 6-34. Adult estimated total cancer risks across all fish species sampled. Columbia River Basin-wide
  average data (70 years exposure).
For a more detailed comparison of cancer risks for anadromous fish and resident fish species for
each study site, see Appendix N. In this appendix, estimated cancer risks are shown for all
species caught at a sampling site using a range offish ingestion rates.

The percent contribution of the chemicals and chemical classes to total cancer risk were shown in
Tables 6-15 (resident fish species) and 6-17 (anadromous fish species) and in Figures 6-21 to 6-
26 (resident fish species) and Figures 6-28 thru 6-33 (anadromous fish species). Fillet with skin
data were used for these tables and figures except for sturgeon, for which fillet without skin data
were used, and eulachon and bridgelip sucker, for which whole body data were used. A
comparison of these tables and figures show that:

•      Arsenic - For anadromous fish species, arsenic was a major contributor to cancer risk for
       all of the salmon and steelhead (33 to 54% for steelhead, fall and spring chinook, and
                                               6-139

-------
       coho salmon), and eulachon (58%), but contributes only 7% to the total cancer risk for
       lamprey.  For resident fish, such a large contribution from arsenic was seen only for
       walleye (33%) and less so for bridgelip sucker (8%).  As discussed in Section 4, it was
       assumed that 10% of the total arsenic measured in fish was inorganic.  The impact of this
       assumption on the characterization of risk is discussed more in Section 6.2.6.

•      PCBs - dioxin-like PCB congeners and Aroclors contribute from 32% to 82% of the total
       cancer risk for the resident fish; and from 32% to 77% for five of the anadromous fish, the
       exception being eulachon. For eulachon, dioxin-like PCB congeners/Aroclors contribute
       only 4% to the total cancer risk. For those  11 fish where dioxin-like PCB
       congeners/Aroclors were major contributors to risk, Aroclors 1254/1260 and, in general,
       dioxin-like PCBs 118, 126, and 156, contribute the most to the total dioxin-like PCB
       congener/Aroclor risk.

       Semi-volatiles - Semi-volatiles, including, PAHs, contribute little to the total risk. The
       exception was largescale sucker, where the contribution to the basin-wide average was
       17% for dibenz(a,h,)anthracene and 8% for benzo(a)pyrene. This was misleading,
       however, because these two contaminants were found only at one of the six study sites
       where largescale sucker fillet were sampled, the Snake River at study site 13.

       Pesticides - For resident fish species, pesticides contribute from about 5% (for rainbow
       trout) to 32% (for bridgelip sucker) of the total cancer risk. For anadromous fish species,
       the percent contribution from pesticides was lower, from 1% (for coho salmon) to 9% (for
       lamprey). DDE was by far the major component of the pesticide cancer risk for resident
       fish species.

•      Chlorinated Dioxins/Furans - Chlorinated dioxins/furans contribute from 5% (for
       largescale sucker) to 36% (for sturgeon) of the total cancer risk for resident fish species.
       Dioxins/furans contribute 36% to the eulachon cancer risk, but only 9% for lamprey and
       chinook salmon, 11% for coho, and 14% for steelhead and spring chinook. For resident
       fish species, 2,3,7,8-TCDF, 1,2,3,7,8-PCDD, and 2,3,7,8-TCDD were the major
       contributors to the dioxin/furan cancer risk. For the anadromous fish species, 2,3,7,8-
       TCDF, 1,2,3,7,8-PCDD, and 2,3,4,7,8-PCDF were the major contributors.

6.2.3  Summary of Non-Cancer Hazards and Cancer Risks for All Species

Tables 6-19 through 6-22 are a summary of the range in endpoint specific hazard indices and
cancer risks across study sites for each species at the four fish ingestion rates used for adults.
Hazard indices are shown only for those endpoints that most frequently exceeded a hazard index
of 1. These endpoints are for reproduction/development and the central nervous system,
immunotoxicity, and liver resulting primarily from exposures greater than the reference dose for
methyl mercury, Aroclors, and DDT, DDE and DDD. Cancer risks are those estimated assuming
a 70 year exposure duration.
                                           6-140

-------
       Hazard indices and cancer risks were lowest for the general public adult at the average
       ingestion rate and highest for CRITFC's member tribal adults at the high ingestion rate.
       For the general public with an average fish ingestion (7.5 g/day or about a meal per
       month), hazard indices were less than 1 and cancer risks are less than 1 X 10~4 except for a
       few of the more highly contaminated samples of mountain whitefish and white sturgeon
       (Table 6-19).

       For CRITFC's member tribal adults at the highest fish ingestion rates (389 g/day or about
       48 meals per month), hazard indices were greater than 1 for several species at some study
       sites.  Hazard indices (less than or equal to 8 at most study sites) and cancer risks (ranging
       from 7 X 10~4  to 2 X 10~3) were lowest for salmon, steelhead, eulachon and rainbow trout
       and highest (hazard indices greater than 100 and cancer risks up to 2 X 10"2 at some study
       sites) for mountain whitefish and white sturgeon (Table 6-22).

       As discussed previously in Section 6.2.1, for the general public, the hazard indices  for
       children at the average fish ingestion rate were about 0.9 those for adults at the average
       ingestion rate;  the hazard indices for children at the high ingestion rate were about 1.3
       times those for adults at the high ingestion rate. For CRITFC's member tribes, the hazard
       indices for children at the average and high ingestion rates were both about 1.9 times
       those for CRITFC's member tribal adults at the average and high ingestion rates,
       respectively.

Table 6-19. Summary of Hazard Indices and Cancer Risks Across Study sites. General Public Adult,
average fish consumption (7.5 grams/day or 1 meal per month).
                         Non-cancer endpoints which most frequently exceed a hazard  Cancer Risks (70 years
Species*	N*	index of one for all species	exposure)

Resident species
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
Anadromous species
coho salmon
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey
Reproductive/ Developmental Immunotoxicty Liver
And Central Nervous Svstem

3 <1 <1 <1
19 <1 <1 <1
12 <1 <1 to 3 <1
16 <1 
-------
Table 6-20.  Summary of Hazard Indices and Cancer Risks Across Study sites. General Public Adult, high
fish consumption (142.4 g/day or 19 meals per month).	
Species*
N*
 Non-cancer endpoints which most frequently exceed a hazard
	index of one for all species	
    Cancer Risks (70 years
   	exposure)	
                             Reproductive/ Developmental
                             and Central Nervous system
                                  Immunotoxicty
                                                  Liver
Resident species
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
Anadromous species
coho salmon
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey

3
19
12
16
3
7

3
15
24
21
3
3


-------
  Table 6-22. Summary of Hazard Indices and Cancer Risks Across Study sites.  CRTTFC's Member Adult,
  high fish consumption (389 grams/day or 48 meal per month)	
  Species*
N*
 Non-cancer endpoints which most frequently exceed a
	hazard index of one for all species	
  Cancer Risks
(70 years exposure)
                                  Reproductive/
                            Developmental and Central
                                Nervous System
                              Immunotoxicty    Liver
Resident species
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
Anadromous species
coho salmon
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey

3
19
12
16
3
7
3
15
24
21
3
3

2
5 to 20

-------
species. For reproductive effects, the ratios of the hazard indices for reproductive effects in
whole body to fillet samples appear to be less than 1 more frequently than those for the other
hazard indices or cancer risks. This may be because the hazard index for reproductive effects is
based largely upon the contaminant mercury which is not lipophilic and binds strongly to protein
(e.g., muscle tissue). However, any conclusions on the results of whole body to fillet samples are
limited by the small sample sizes (usually 3) at each site and by the fact that whole body samples
were always from a composite offish different than those used for the whole body analysis (i.e.,
fillet and whole body samples are not from the same fish).

Table 6-23. Comparison of site specific non-cancer hazard indices (for CRTTFC's member tribal children)
and cancer risks (for CRTTFC's member tribal adults) from consuming whole body versus fillet for different
fish species.	
                                         Hazard Indices (1)
                      Immunotoxicitv
                     Reproductive
                        Effects
                    Total Hazard Index
                          Cancer Risk (2)
Species
Range in ratios of
hazard indices for
whole body/fillet
   across sites
Range in ratios of
hazard indices for
whole body/fillet
   across sites
 Range of ratios of
total hazard indices
for whole body/fillet
    across sites
  Range of ratios of
cancer risks for whole
     bodv/fillet

coho
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout

1.1
0.9-6.6
0.9-1.6
1.1-1.4
na
1
na
0.6-3.3
0.4-2.1
0.4-2.9
1.8
1.2-1.2
F
(1/1)
(3/5)
(4/8)
(6/6)
na
(0/1)
na
(3/5)
(2/4)
(1/3)
(1/1)
(2/2)

0.8
0.7-1.1
0.3-1.1
0.6-1.6
na
na
na
0.2-1.3
0.7-0.9
0.3-3.3
1
0.7-1.7
F
(0/1)
(1/5)
(1/3)
(1/6)
na
na
na
(1/6)
(0/3)
(2/3)
(0/1)
e/2)

1.1
1.0-1.6
0.6-1.6
0.9-1.5
na
1.2
na
0.5-2.2
0.8-1.6
0.4-2.7
1
1.1-1.5
F
(1/1)
(3/5)
(4/8)
(4/6)
na
(1/1)
na
(3/6)
(2/4)
(1/3)
(0/1)
(2/2)

1
1-2
1-2
0.5-2.0
na
1
na
0.7-2.5
0.5-1.4
0.8-2.3
1
1.0-1.0
F
(0/1)
(2/5)
(3/8)
(2/6)
na
(0/1)
na
(3/6)
(1/4)
(1/3)
(1/1)
(0/2)
F=Frequency of number of whole body to fillet ratios greater than 1 divided by the total number of whole body to fillet ratios for that species.
na = Not applicable; ratios could not be calculated because chemicals (Aroclors, mercury) were less than detection limits or because fillet data were
not available (I.e., for bridgelip sucker and eulachon)
(1) Hazard indices used are those calculated for CRITFC's tribal member children, high fish consumption rate
(2) Cancer risk are those calculated for CRITFC's tribal member adults, 70 years exposure, high fish consumption

6.2.5   Risk Characterization Using a Multiple-species Diet

As discussed in Section 4.10, a hypothetical diet consisting of multiple fish species was
developed based on information  obtained during the 1991-1992 survey offish consumption by
members of the Nez Perce, Umatilla, Yakama, and Warm Springs Tribes (CRITFC, 1994). The
percentage of the hypothetical diet assumed for each fish species and the resulting  species
specific ingestion rates (assuming a total fish ingestion rate of 63.2 g/day, the average for
CRITFC's tribal members adults) were shown previously in Table 4-4.
                                                 6-144

-------
Table 6-24 shows the resulting cancer risks and total non-cancer hazard indices calculated using
this hypothetical diet and the average fish consumption rate (63.2 grams/day) for CRITFC's
member tribal adult fish consumers.  Cancer risk estimates for individual species were highest for
lamprey fillets (1.0 X 10~4) and lowest for walleye fillets (4.2 X 10~6). The total excess cancer risk
for consuming the fish used in this example was 4.0 X 10"4. Total hazard indices for individual
species were highest for lamprey and mountain whitefish fillets (0.7) and lowest for eulachon and
largescale sucker fillets (0.1). The total hazard index for consuming the fish used in this example
was 3.2.
   Table 6-24. Estimate cancer risks and non-cancer health effects for a hypothetical multiple-species diet
   based upon CRITFC's member average adult fish consumption (CRTTFC, 1994)
Species
Salmonb-c-d
Rainbow Troutd
Mountain Whitefishd
Eulachon'
Pacific lampreyd
Walleye"
White Sturgeon5
Largescale Sucker11
Percentage of
Hypothetical
27.7
21.0
6.8
15.6
16.3
2.8
7.4
2.3
Consumption Rate
fe/dav)
17.5
13.3
4.3
9.9
10.3
1.8
4.7
1.5
Cancer
Risk"
5.8X10-5
3.5X10'5
9.3 X 10-5
3.3X10-5
1.0 X104
4.2 X10-6
7.1X10'5
9.3 X 10-6
Non-cancer
Effects"
0.6
0.3
0.7
0.1
0.7
0.1
0.6
0.1
   Totals                        100.0                   63.2              4.0 X 104            3.2
aRisk estimates assume fish consumption by a 70 kg CRITFC's tribal member adult at the specified rate 365 days per year for 70 years
bCancer risk estimates for salmon are the average of estimates for spring chinook (6.4 X 10"5), fall chinook (5.7 X 10"5), coho (4.5 X 10"5), and
steelhead (6.4 X1Q-5).
cNoncancer hazard indices for salmon are the average of estimates for spring chinook (0.6), fall chinook (0.5), coho (0.7), and steelhead (0.7).
dRisk estimates are based on analysis of uncooked composite samples of fillets with skin.
eRisk estimates are based on analysis of uncooked composite samples of whole body fish.
fRisk estimates are based on analysis of uncooked composite samples of fillets without skin.

Figure 6-35 shows the total non-cancer hazard indices and Figure 6-36 shows the total cancer
risks (70 years exposure) across all species with the results for the multiple-species diet shown for
comparison. The results  for both general public adult (average and high fish consumption) and
CRITFC's member tribal adults (average and high fish consumption) using basin-wide data are
included. For all four populations, the hypothetical diet of multiple species based on CRITFC's
fish  consumption survey  was used.  The non-cancer hazards and cancer risks for the multiple-
species diet were lower than those for the most contaminated species (e.g., sturgeon and
whitefish) and higher than those estimated for some of the least contaminated species (e.g.,
salmon, steelhead, rainbow trout, and eulachon).

These results demonstrate that the non-cancer hazards and cancer risks previously discussed in
Sections 6.2.1 and 6.2.2 for individual species may not adequately reflect the cancer risks and
non-cancer hazards  for CRITFC's member tribes or other individuals from the general public
whose diets are composed of a mixture offish types from the Columbia River Basin.
                                                6-145

-------

A"J"mn-



_
ij-j

i

r
/y>V
~


r
-




r

1
i
*****





"

-


r
[
////////

_J






Multiple
Species
Diet

General Public
• Average Fish
Consumption
1 	 1 High Fish

CRITFC Member
Tribes
I I Average Fish
1 1 High Fish

      *Fillet with skin samples except for white
      sturgeon (fillet without skin) and bridgelip
      sucker and eulachon (whole body)
                                              Species
Figure 6-35. Adult total hazard indices for all fish species, with multiple-species diet results. Basin-wide average
data.


1.E-03-
1.EOt-

Anadromous

r-.








-.





-


i-i








-








1 | | Resident
r















-
-








i_

-






















-

-


GereralPuHic
• Average Ksh
Ccnsumplicn
I 	 IH^iFHi
| 	 | Ccnsumplicn

CRirECMaita
Tnte






Consumpticn
CaBumplicn
                                                                                   Multiple
                                                                                   Species
                                                                                     Det
  "Rltet with skin samples except for while sturgeon (fillet without skin) and
  bridgelip sicker and eutehon fyvhote bod/)

 Figure 6-36.  Adult cancer risks for all species, with multiple-species diet results.  Columbia River Basin-
 wide average chemical concentration data.  70 years exposure.
                                                    6-146

-------
6.2.6  Risk Characterization Using Different Assumptions for Percent of Inorganic Arsenic

As discussed in Section 5.3.3, total arsenic was measured in fish tissue samples in this study.
Because a reference dose and cancer slope factor are available for only inorganic arsenic, an
assumption about the percent of inorganic arsenic in fish had to be made to estimate the non-
cancer hazards and cancer risks from consuming fish. The non-cancer hazards and cancer risks
discussed in Section 6.2.1 and 6.2.2, respectively, assumed that for all fish species (resident fish
and anadromous fish) caught in  this study, 10% of the total arsenic was inorganic arsenic. The
studies used to derive this value of 10% and the rationale for its selection were discussed in
Section 5.3.3.  The data in Section 5.3.3 also suggests that an alternative assumption for
anadromous fish species could be considered - the assumption that 1% of the total arsenic was
inorganic.  Therefore, the non-cancer hazards and cancer risk were recalculated for anadromous
fish species using basin-wide data assuming that 1% of the total arsenic was inorganic. The
assumption of 1% inorganic arsenic for anadromous  fish species in effect results in a contaminant
level for arsenic that one tenth of that assuming that  10% was inorganic arsenic.

Table 6-25 shows the impact of the two different assumption (10% and 1% inorganic) on the
estimated total  hazard indices for anadromous fish species using basin-wide data. These results
are shown for general public and CRITFC's member tribal adults at both the average and high
fish consumption rates.  As can  be seen from this table and from Figure 6-37, assuming that 1%
of total arsenic was inorganic rather than 10%, the total hazard indices were reduced by 2% for
lamprey, 6% for coho and steelhead, and 11% for spring and fall chinook. However, for
eulachon, the assumption of 1% inorganic arsenic reduces the total basin-wide hazard index for
this fish species by 56%. The effect of this assumption on risks due to ingestion of eulachon was
consistent with the data in Table 6-7 which showed the percent contribution of different
contaminants on the basin-wide total hazard indices  for  anadromous fish species. Arsenic
contributed from about 2% to 13% to the total hazard index for salmon, steelhead, and lamprey
but about 60% to that for eulachon. Thus, assuming that inorganic arsenic represents 1% rather
than 10% of total arsenic had the largest impact on the total non-cancer hazards for eulachon (a
56% reduction in the total hazard index) and less of an impact on the other anadromous fish
species.
                                            6-147

-------
Table 6-25. Total hazard indices (His) for adults assuming that total arsenic is 1% versus 10% inorganic
arsenic. Exposure concentrations used to estimate risks are Columbia River Basin-wide averages of fish
tissue samples
Average Fish Consumer




Snecies
coho salmon

spring chinook

fall chinook

steelhead

eulachon

Pacific lamprey





N
3

24

15

21

3

3




Tissue
Tvne
FS

FS

FS

FS

WB

FS

Percent
Inorganic
Arsenic as
Total
Arsenic
10
1
10
1
10
1
10
1
10
1
10
1
Percent
Decrease In
Total ffl
Assuming '
1%
Inorganic
Arsenic

6

11

11

6

56

2
Hieh Fish Consumer
Total ffl


general
nublic
0.3
0.3
0.3
0.2
0.2
0.2
0.3
0.3
0.1
0.0
0.5
0.5


CRTTFC
member tribe
2.5
2.4
2.1
1.9
2.0
1.7
2.6
2.4
0.4
0.2
4.5
4.4


general
nublic
5.7
5.4
4.8
4.2
4.4
3.9
5.7
5.4
1.0
0.4
10.1
9.9


CRITFC
member tribe
15.7
14.8
13.0
11.6
12.0
10.7
15.7
14.8
2.7
1.2
27.7
27.1
N= Number of samples; FS = fillet with skin; WB = whole body
Total HI is determined by summing all hazard quotients regardless of health endpoint.

X
0)
•o
c _„
Hazard
Ji C
1








10%

IV



—



Coho
10°
—



—



Spring
Chinook
10°/



	
Fall
Chinook



10%

Stee

-

h
—
5ad


10%
10V
i
-


n






^
Eulachon Lamprey



General Public
• Average Fish
Consumption
1 	 1 High Fish

CRITFC
Member Tribes
1 	 1 Average Fish
I 1 High Fish

                                                      Species
        1% -  One percent of total arsenic is inorganic arsenic
        10% - Ten percent of total arsenic is inorganic arsenic
        *Fillet with skin samples except for eulachon (whole body)
    Figure 6-37.  Impact of percent inorganic arsenic on total hazard index. Basin-wide data for
    anadromous fish species*.
                                                           6-148

-------
Tables 6-26 and Figure 6-38 show the impact of the two different assumptions (10% and 1%
inorganic arsenic as total arsenic) on the estimated total cancer risks for anadromous fish species
using basin-wide data. These results are shown for general public and CRITFC's member tribal
adults at both the average and high fish consumption rates and 70 years of exposure. Assuming
that 1% of total arsenic was inorganic versus 10%, the cancer risks were reduced about 6% for
lamprey, 29% for steelhead, and between 40% to 52% for coho, spring chinook, fall chinook and
eulachon.  These results are consistent with those previously discussed for Table 6-17 (percent
contribution of different contaminants on the basin-wide total cancer risk for anadromous fish
species) which showed that arsenic was a major contributor to the total cancer risks for all
anadromous fish species except Pacific lamprey.

 Table 6-26. Estimated total cancer risks for adults assuming that total arsenic was 1% versus 10%
 inorganic arsenic 70 years exposure.  Exposure concentrations used to estimate risks are Columbia River
 Basin-wide averages of fish tissue samples.	
                                                     Average Fish Consumer    High Fish Consumer
Species
coho salmon

spring chinook
fall chinook

steelhead

eulachon

Pacific lamprey

N
3

24
15

21

3

3

Tissue
Tvne
FS

FS
FS

FS

WB

FS

Percent
Inorganic
Arsenic as
Total
Arsenic
10
1
10
1
10
1
10
1
10
1
10
1
Percent Decrease In
Total Cancer Risk
Assuming 1%
Inorganic Arsenic

40.4
44.6

48.4

29.3

52.0

6.1
Total Cancer Risk
general
public
1.9E-05
1.1E-05
2.8E-05
1.5E-05
2.4E-05
1.2E-05
2.8E-05
2.0E-05
2.5E-05
1.2E-05
7.4E-05
6.9E-05
CRITFC
member
tribe
1.6E-04
9.7E-05
2.3E-04
1.3E-04
2.0E-04
1.1E-04
2.3E-04
1.7E-04
2.1E-04
l.OE-04
6.2E-04
5.8E-04
general
public
3.7E-04
2.2E-04
5.2E-04
2.9E-04
4.6E-04
2.4E-04
5.3E-04
3.7E-04
4.7E-04
2.3E-04
1.4E-03
1.3E-03
CRITFC
member
tribe
l.OE-03
6.0E-04
1.4E-03
7.9E-04
1.3E-03
6.5E-04
1.4E-03
l.OE-03
1.3E-03
6.2E-04
3.8E-03
3.6E-03
N = Number of samples; FS = fillet with skin; WB = whole body

This comparison of the results from using the two different assumptions (1% versus 10%) for
inorganic arsenic in fish shows that the reduction on the total non-cancer hazards was less than
12% for all anadromous fish species, except eulachon which had about a 50% reduction.
However, the impact was greater on the estimates of cancer risk. With the exception of lamprey
for which cancer risks were reduced by only 6%, the reductions in cancer risks for steelhead was
about 29% and for the other anadromous fish species ranged from about 40 to 50%.
                                             6-149

-------
                Coho
 F;
Chinook
                                                          Steelhead
                                                                          Eulachon
                                                                                        Lamprey
                                                    Species


   1% - One percent of total arsenic is inorganic arsenic
   10% - Ten percent of total arsenic is inorganic arsenic
   •Fillet with skin samples except for eulachon (whole body)


Figure 6-38. Impact of percent inorganic arsenic on cancer risks.  Basin-wide data for anadromous fish

species.
                                                      6-150

-------
7.0    Lead Risk Assessment

Lead health risks are presented separately because lead health risk methods are unique owing to
the ubiquitous nature of lead exposures and the reliance on blood lead concentrations to describe
lead exposure and toxicity. Lead risks are characterized by predicting blood lead levels with
models and guidance developed by EPA available from the following web site:
http://www.epa.gov/superfund/programs/lead/prods.htm - software.  In this assessment, lead
exposure from fish consumption is added to all other likely sources of lead exposure to predict a
blood lead level. Both the Integrated Exposure Uptake Biokinetic Model (IEUBK) for children
and the EPA Adult Lead Model for the fetus predict blood lead levels from a given set of input
parameters. There is no other model for lead exposures except the Adult Lead Model, so it is
used for children and fetuses.

In contrast to risk assessments for cancer or non-cancer risks, lead risk assessments typically use
central tendency exposure values to predict a central tendency (geometric mean) blood lead level.
The predicted geometric mean blood lead level is then used in conjunction with a modeled log-
normal distribution to estimate the probability of exceeding a target blood lead level of 10 |ig/dl.
Blood lead levels are a measure of internal dose that has been related to many adverse health
effects (NRC, 1993). The emphasis on blood lead integrates exposure, toxicity and risk, which
are more distinct in other types of risk assessment. For other chemicals, risk is described in terms
of an external dose (e.g. mg/kg-day).

The IEUBK Model was used to predict blood lead levels in children up to 72 months of age
(USEPA, 1994a,b). The EPA Adult Lead Model was used to predict blood lead levels in fetuses
(USEPA, 1996b).  This section on lead risk assessment is organized into separate discussions of
the two lead models. Each of the two lead models was run using both central tendency and high
end rates offish ingestion.  Central tendency rates offish ingestion were used to predict both
geometric mean blood lead levels and the probability of exceeding a blood lead  level of 10 |ig/dl
in both children and fetuses. For the high end fish ingestion rates, only the most likely blood
level could be predicted; it is not appropriate to predict the probability of exceeding 10 jig/dl
associated with high end fish consumption.

7.1    Lead Concentrations in Fish

Study sites, collection methods, analytical methods, and quality assurance plans are discussed in
Section 1; concentrations of lead in fish are discussed in Section 2. Whole fish had substantially
higher lead levels because lead tends to concentrate in the bones and gills (Ay et al, 1999). Note
that the maximum in the concentration scale for whole fish is 500 |ig/kg and 100 |ig/kg for fillets
(Table 2-14).  The highest individual sample was 1200 |ig/kg in a fall  chinook salmon taken from
Station 14 on the Columbia River.  For fish tissue samples with undetected lead concentrations, a
value of half the detection limit was used (5 |ig/kg) in all risk estimates.
                                            7-151

-------
7.2    Overview of Lead Risk Assessment Approach

Risk assessment methods for lead differ from other types of risk assessment because they
integrate all potential sources of exposure to predict a blood lead level. Lead in the blood reflects
all sources of lead exposure, regardless of its origin. Lead risk assessments reflect the widespread
distribution of lead in the environment.  Common sources of lead in the environment include
residual contamination from past uses of lead in gasoline, paint, agricultural chemicals, and
industrial sources including lead mining and smelting (NRC, 1993). People are exposed to lead
through ingestion of soil and dust, inhalation of lead from the air, and consuming food with
background concentrations of lead.  Lead can enter drinking water through contamination of
surface and groundwater as well as leaching from lead pipes and solder in plumbing systems. All
of these sources and exposure pathways are included in the models used to assess lead risks.  The
IEUBK model is used to simulate lead exposures from air, water, diet, soil, and house dust. The
Adult Lead Model accounts for the same sources of lead exposure by using a baseline blood lead
level derived from the National Health and Nutrition Examination Survey (USEPA, 1996b).

Risk assessment methodologies for substances other than lead utilize a combination of central
tendency and high end exposure values to estimate an aggregate reasonable maximum exposure
scenario. A point value for risk derived using a reasonable maximum exposure scenario is
accepted as being protective of public health.  Public health protection using lead risk assessment
methodology derives from a limit on the acceptable predicted blood lead values. An acceptable
risk for lead exposure typically equates to a predicted probability of no more than 5% greater than
the 10 |ig/dl level (USEPA, 1998b)

Risk, expressed as predicted blood lead levels, was calculated in two ways for  children and
fetuses. The first, and more typical, method used median fish ingestion rates to predict: 1) a
geometric mean blood lead level and 2) the corresponding risk of exceeding a blood lead level of
10 |ig/dl.  The probability of exceeding 10 |ig/dl was calculated with a log-normal risk model
based on the model's output (the geometric  mean blood lead level) and an assumed geometric
standard deviation. In the second method, high-end fish ingestion rates were used to predict
blood lead levels for children or mothers who consume large amounts  offish. Because the
resultant high-end fish ingestion prediction does not represent a geometric mean blood lead level,
the geometric standard deviation could not be applied to predict the probability of exceeding 10
Hg/dl. Predicted blood lead levels resulting  from high-end fish consumption scenarios represent
the most likely blood lead levels associated  with high-end consumption rates.

The adverse health effects of lead have been related to blood lead concentrations in units of
micrograms of lead per deciliter of whole blood (|ig/dl). As a result, blood lead levels have
evolved as measures of exposure, risk, and toxicity.  Since 1991, the national level of concern for
young children and fetuses has been 10 |ig/dl (CDC, 1991). An analogous level has not been
defined for other groups, but children and the developing fetus are accepted as being especially
vulnerable to lead because lead interferes with the development of the central nervous system
(NRC, 1993).  Lead risks were evaluated by comparing predicted blood lead levels to the 10 |ig/dl
standard and by determining the expected percentage to exceed the 10 |ig/dl criterion.
                                             7-152

-------
Adverse health effects observed at a blood lead level of 10 jig/dl are sub-clinical, meaning that,
these effects cannot be diagnosed in an individual. The adverse health effects include cognitive
deficits in IQ and learning, based on numerous scientific studies involving comparisons of large
groups of children to control for confounding factors and account for the natural variability in
cognitive function (NRC, 1993; USDHHS, 1999; CDC, 1991). The studies have incorporated
both cross-sectional and longitudinal designs. The importance of primary prevention of lead
exposure has been highlighted by recent studies suggesting adverse health effects at blood lead
levels less than 10 |ig/dl and the failure of chelation treatment to prevent cognitive impairments in
treated children (Lanphear  et al., 2000; Rogan et al., 2001; Rosen and Mushak, 2001).

Children are the population of greatest concern for lead exposure.  Blood lead levels tend to peak
in children as they become more mobile and begin to explore their surroundings.  Blood lead
levels normally peak at approximately 30 months of age when children are especially vulnerable
to neuro-behavioral deficits (Rodier, 1995;Goldstein, 1990). The adverse effects of low-level
lead poisoning can result from relatively  short-term  exposures on the order of months, as opposed
to periods of years or longer for other chemicals. The fetus is vulnerable to the same
developmental and neuro-behavioral effects as children. Although lead is harmful to fetuses,
children are a greater concern because they generally have higher exposures than fetuses. Fetal
exposures are lower because exposures to mothers are typically lower than exposures to children.
These and other health effects are described in further detail in Appendix C (Toxicity Profiles).

7.3     Method for Predicting Risks to Children

In contrast to risk assessment methodologies for predicting cancer or non-cancer risks, the lead
models  rely on central tendency exposure values to predict a central tendency (geometric mean)
blood lead level. The predicted geometric mean blood lead level is then used in conjunction with
an assumed geometric standard deviation to estimate the probability of exceeding a target blood
lead level of 10 |ig/dl established by the Centers for Disease Control (CDC, 1991).  In this way,
central tendency exposure estimates are used to estimate upper percentile blood lead levels.  An
example graph of an IEUBK Model run depicting the geometric mean and percent greater than 10
|ig/dl is shown in Figure 7-1. In the IEUBK model, a geometric mean blood lead level of 4.6
|ig/dl corresponds to a 5% chance of exceeding 10 jig/dl using the default geometric standard
deviation of 1.6 (USEPA, 1994b). Although lead risk assessment methods differ from that
employed for other chemicals, the goal of protecting  highly exposed individuals remains the
same.

The geometric standard deviation accounts for the variation in blood lead observed in children
exposed to similar environmental concentrations of lead.  The variation in observed blood lead
levels is attributed to differences in the children (behavior and metabolism); not the environment.
Because the geometric standard deviation accounts for behaviors that determine exposure levels
to lead,  applying the geometric standard deviation to high contact rate behaviors, including fish
ingestion, would over-estimate the variability and over-predict the probability of exceeding 10
                                             7-153

-------
              3  £
              1  1

           LEHD O.
e   10    12   14
LEHD CONCENTBRTION
0 to -7-2. Months
     Figure 7-1.  Sample IEUBK Model for Lead Output Graph.


Running the IEUBK Model with high-end fish consumption rates predicts the most likely blood
lead levels for people eating large amounts offish, although, the result does not correspond to the
geometric mean of a population consuming different amounts offish.  Blood lead predictions for
highly exposed  individuals facilitate comparison of lead risks to risks from other chemicals, but
results from high-end exposure inputs preclude application of the geometric standard deviation to
calculate risks of exceeding a 10 |ig/dl blood lead level. Risks to highly exposed individuals are
typically characterized by the 95th percentile of the blood lead distribution centered around the
predicted geometric mean blood lead rather than using the high-end fish ingestion values.

The IEUBK Model was run with all exposure parameters set to default levels with the addition of
dietary lead intake attributable to lead in fish tissue for the full range of lead concentrations
observed.  Default exposure parameters are based on national average levels of lead in air, water
food,  soil, and dirt (Table 7-1) and described in detail in  EPA guidance (USEPA, 1994b).
                                             7-154

-------
  Table 7-1. Default Input Parameters Used for the IEUBK Model Adapted from (USEPA,1994b)
  Input Parameter _ Value

  Soil lead concentration                                     200,000 |ig/kg

  House dust lead concentration (proportion of soil in dust = 0.7)     140,000 |ig/kg

  Combined soil and dust ingestion rate by age:

                          0-11 months                       85 mg/day
                          12-23 months                     135 mg/day
                          24-35 months                     135 mg/day
                          36-47 months                     135 mg/day
                          48-59 months                     100 mg/day
                          60-71 months                     90 mg/day

  Lead concentration in Air                                   0.10 • g/cubic meter
The default concentrations of lead in soil and house dust are representative of average, national
conditions. The default concentrations for lead in soil and house dust are 200,000 |ig/kg and
140,000 |ig/kg respectively (USEPA, 1994b). These values are appropriate for urban areas and
are likely to exceed the expected concentrations in rural areas surrounding the Columbia River
because lead levels increase with urbanization. A recent survey of 50 homes from small, rural
towns in Northern Idaho found soil lead concentrations less than 100,000 |ig/kg (Spalinger et al.,
2000).  These concentrations would not account for severe lead paint contamination. Lack of data
on specific soil and house dust concentrations remains a large source of uncertainty in this
evaluation because soil and dust in the home account for a large proportion of lead exposure in
young children (Manton et al., 2000) (Lanphear et al., 1998).

The IEUBK model has the capability to simulate exposures to locally grown vegetables,  game,
and fish.  The IEUBK default values for soil, house dust, air, diet, and water were used in
conjunction with an age-specific median fish ingestion rate of 16.2 g/day based on the fish
consumption survey of CRITFC's member tribes (CRITFC, 1994). Fish ingestion was specified
as the percentage of meat (Table 7-2) consisting of locally caught fish and the lead concentrations
in the fish.  There are other ways to simulate fish ingestion in the IEUBK Model (e.g.  by
specifying dietary lead intakes as |ig/day), but it was preferred to specify fish ingestion as a
percentage of meat to preserve the caloric and protein intake assumptions of the model.  This
approach substitutes fish for other protein sources rather than adding fish to the default diet.  This
approach conforms with IEUBK body weight and biokinetic assumptions and is described in EPA
guidance (USEPA, 1994b).

       Table 7-2.  Input Parameters Used in the IEUBK Model Meat Consumption Rate by Age
       in the IEUBK model Adapted from (USEPA. 1994b) _
     _ Age Range (months) _ Meat Consumption grams/day _
                      12-24                                     87
                      25-36                                     96
                      37-48                                     102
                      49-60                                     107
                      61-72                                     112
                    Average                                    101

                                             7-155

-------
The CRITFC study examined Columbia River fish consumption in young children as surveyed by
their parents. This study was selected as the most relevant study to assess the Columbia River
lead hazard for all children because it is specific to the place, CRITFC's member tribes, and the
age range specified by the IEUBK (CRITFC, 1994).  The tribal ingestion rates are likely to
overestimate fish consumption for non-tribal members.  Because the CRITFC study presents
consumption rates for children up to 72 months of age, the IEUBK Model was run for the same
age range.

To facilitate comparisons between risks from lead and other chemicals presented in Section 6, the
ingestion rates used for other chemicals are summarized in Table 7-3. Fish ingestion rates used to
estimate risks from chemicals other than lead are based on mean and 99th percentiles of both the
CRITFC survey and national data for the general public described in Section 4 of this report.

The distribution of child fish consumption rates from the CRITFC study is statistically skewed
because it included individuals with very high fish consumption rates relative to others. For
skewed data, the arithmetic mean is not an appropriate measure of central tendency because it is
highly influenced by the individuals with large fish consumption rates.  The median (50th
percentile) is a preferred central tendency measure of skewed data because it is less sensitive to
extreme values. The fish consumption data for CRITFC's member tribes (CRITFC, 1994) were
re-analyzed to omit children who did not consume fish from the data set (Kissinger and Beck,
2000). The re-analysis calculated a median  consumption rate occurred between 13 and  16.2
g/day, the 39th and 65th percentiles, respectively (see Table 7-4).  Rather than interpolate a median
value of 14.4 g/day between the 39th and 65th percentiles, the higher value was selected as a
protective central tendency consumption rate.
  Table 7-3.  Fish Ingestion Rates (grams/day) Used to Assess Risk for Lead and other Chemicals
              Target Population
Assessment
Population
Expo sure Level

Ingestion Rate
Basis
Age Range
Ingestion Rate
Basis
Lead
Native American
Central
High End
Mother and Fetus
39.2
50th CRITFC
389
99th CRITFC
Children < 72 Months
16
50th CRITFC
101
IEUBK MAX*
Non-lead
Native American
Central
High End
Adult
63.2
Mean CRITFC
389
99th CRITFC
Children < 72 Months
24.8
Mean CRITFC
162
99th CRITFC
Non-lead
General Public
Central
High
Adult
7.5
Mean EPA
142.4
99th
Children < 1 5 years
2.83
Mean
77.95
99th
* A fish ingestion rate of 101 g/day assumes that locally caught fish comprise 100% of all dietary protein sources and represents an upper
constraint of the IEUBK Lead Model for Children
                                             7-156

-------
Table 7-4. Percentages of Child Fish Consumption Rates for Consumers of Fish
From (Kissinger and Beck, 2000) analysis of (CRTTFC, 1994)
Grams/dav
0.4
0.8
1.6
2.4
3.2
4.1
4.9
6.5
Cumulative
Percent
1%
1%
5%
5%
9%
14%
16%
18%
Grams/dav
8.1
9.7
12.2
13.0
16.2
19.4
20.3
24.3
Cumulative
Percent
33%
35%
38%
39%
65%
66%
67%
70%
Grams/dav
32.4
48.6
64.8
72.9
81.0
97.2
162.0

Cumulative
Percent
84%
89%
93%
95%
97%
98%
100%

7.4    Risk Characterization for Children

Predicted blood lead levels spanning the full range of observed fish tissue concentrations are
shown in Figure 7-2.  Predicted geometric mean blood lead levels are plotted on the left axis with
a solid line. The corresponding probabilities of exceeding 10 |ig/dl are shown as percentages on
the right axis with a dashed line. Each of the 11 pairs of points represents a separate IEUBK
Model run at successively increasing concentrations of lead in fish.  These results indicate that for
fish containing lead up to 500 |ig/kg, the probability of achieving a blood lead level greater than
10 |ig/dl is no more than 5% and the predicted geometric mean blood lead level is 4.6 |ig/dl.  For
comparison, only the average concentration of whole body eulachon had a lead concentration of
500 |ig/kg. The next highest whole fish species is fall chinook, with an average lead
concentration of 220 |ig/kg. Average lead concentrations in all other whole fish and fillet
samples occur well below 500 ng/kg and concentrations in fillets averaged 200 ng/kg (Table 2-
14).
            6.00
            5.00
                                                                              Probability of
                                                                                 > 10 ug/dl
                                                                                12%
            0.00
                   0    100   200    300   400   500   600   700   800   900   1000
                                 Concentration of Lead in Fish (ug/kg)

       Figure 7-2. Predicted blood lead levels for children who consume offish collected from the
       Columbia River Basin assuming fish is 16% of dietary meat.
                                              7-157

-------
To explore the effect of an extremely high fish consumption rate in children, the IEUBK Model
was run assuming that fish replaced 100% meat in the diet (101 g/day) (Figure 7-3).  The IEUBK
Model was run repeatedly to determine the fish tissue concentration associated with a predicted
blood lead level of 10 |ig/dl. A lead concentration of 500 |ig/kg in fish tissue corresponded to a
predicted blood lead concentration of 10  |ig/dl. This is the same concentration associated with a
5% risk of exceeding 10 |ig/dl under the  16.2 g/day fish consumption scenario described in the
previous paragraph.
IO.U •

^)
D

0
_i
ro
0
_i
0
o
m 40 -

2.0 "

^^.L
~*lf'^
•'"Predicted ^^tf"^
Blood Lead Level ua/dl ^^^^
^^A
s*ir t

Z
t<

prf***




                     0   100  200   300   400   500   600   700  800  900  1000
                                      Fish Lead Concentration (ug/kg)
            Figure 7-3. Predicted blood lead levels for children (0-72 months) who consume
            101 g/day of fish collected from the Columbia River Basin, 1996-1998.

7.5    Uncertainties in risk estimates for Children

Lead risk assessment methods are unique because they use cumulative exposures to predict blood
lead levels in contrast to methods used for other chemicals which generally limit evaluation of
exposures to discreet sources.  Because lead risks are cumulative, uncertainties are compounded
by the many sources of exposure in addition to uncertainties arising from fish consumption. In
children, lead exposure occurs primarily from lead in soil and house dust rather than from typical
dietary sources (Manton et al, 2000).  Sources of lead exposure common to children and fetuses
include industrial or agricultural sources, occupational exposures, and environmental lead
originating from gasoline or leaded paint. Occupational exposures can track contaminants from
the workplace into the home, potentially spreading exposure among children and adults in a
household (Fenske et al., 2000). A major source of uncertainty in this risk assessment may be
attributable to sources of lead other than Columbia River fish. The magnitude of lead exposure
from fish consumption varies with selection offish parts eaten (e.g. whole versus fillet), species
offish, and the study site of the fish relative to sources of lead contamination.

The  IEUBK model is normally used to simulate blood lead levels for children up to 84 months of
age.  However, because the fish consumption data from the CRITFC study were reported for
children up to 72 months of age, IEUBK evaluation was limited to 72 months. A 72-month
                                             7-158

-------
model run predicts higher blood lead concentrations than an 84-month model run because blood
lead levels peak during the first 36 months. In the absence of data to estimate specific, concurrent
residential exposures, the default concentrations of lead in soil and house dust represent a large
source of uncertainty in the IEUBK evaluation because these sources are expected to account for
most of the lead exposure to young children. However, the default soil and dust concentrations
are unlikely to underestimate average levels of lead in the homes.

7.6    Method for Predicting Risks to Fetuses

The Adult Lead Model begins with a baseline blood lead level for adult women and then predicts
an incremental increase in blood lead levels associated with an increase in exposure that is not
included in the baseline blood lead levels (USEPA, 1996b and USEPA, 1999a). In the Adult
Lead Model, fetal blood lead  levels are set equal to 90% of the mother's blood lead level. If the
baseline blood lead reflects the modeled incremental exposure, then the exposure is counted twice
and the modeled blood lead level would be too high. In this study, the Adult Lead Model was
used to evaluate fish ingestion as the source of incremental exposure greater than the baseline
blood lead level.

The assumptions used in this approach include:

       1) Lead exposures from all sources except consuming fish from the Columbia River are
       captured in the baseline blood lead level, based on high end estimates from national blood
       lead surveys, and
       2) incremental ingestion offish is not included in the baseline blood lead level.

Selection of a high baseline blood lead level minimized the possibility of underestimating risk.
The lead ingested from fish is converted to a blood lead level by using a constant ratio of an
increase in blood lead concentration associated with a mass of absorbed lead. This ratio is the
Biokinetic Slope Factor (BKSF). The baseline blood lead level, the blood level in the absence of
lead exposure via Columbia River fish ingestion, is critical to this calculation.  A complete listing
of all the Adult Lead Model input values is included in Table 7.5.

The equations used in the Adult Lead Model are (USEPA 1999b):

Equation 7-1
Adult Blood Lead Level = Baseline Blood Lead Level + Increase in Blood Lead

Equation 7-2
Increase in Blood Lead =
             [(BKSF) * Fish Ingestion  Rate * Fish Concentration * Absorbed Fraction for Fish]

Equation 7-3
Fetal Blood Lead           = Adult Blood * 0.9
Equation 7-4

                                           7-159

-------
Probability that Fetal Blood Lead is greater or equal to 10 jig/dl using the z-value where:
       z = In (10)-ln (Fetal Blood Lead)/ln (Geometric Standard Deviation)


Analysis of the lead hazard associated with adult consumption of Columbia River fish was
conducted using the formula:

  Equation 7-5 PbBadul, central = PbBadult,0 + BKSF * (PBF * IRF * AFF * EFF) /AT


 Table 7-5.  Input Parameters Used for the EPA Adult Lead Model
    Variable    Description                                                    Value Used
    PbBa[|Ult 0    Adult blood lead concentration in the absence of other lead Central 1 .7 |ig/dl
               exposure.                                         High End 2.2 |ig/dl
      BKSF     Biokinetic slope factor relating the (quasi-steady state) increase in
               blood lead concent
      PbF      Fish lead concentration                              full range of values:  0-1000 jig/kg
      IRF      Intake rate of fish in g/day median of CRITFC Adult ConsumX9s2 g/day
      AFF      Absolute gastrointestinal absorption factor for ingested leadOiilO
               fish (dimensionless)
      EFF      Exposure frequency for ingestion offish (days of exposure dMfii^ays per year
               the averaging period); may be taken as days per year in
               continuing long term exposures.
      AT      Averaging time, the total period during which exposure may 365 days per year
Because study site-specific baseline blood lead levels and geometric standard deviations are not
available for consumers of Columbia River fish, the Adult Lead Model was run using both central
tendency and high-end estimates of the baseline blood lead level and the geometric standard
deviation described in (USEPA, 1996b). The larger baseline blood lead level increased the
predicted blood lead levels.  An increase in the Geometric Standard Deviation increased the
probability of exceeding 10 |ig/dl.  All input parameters are listed in Table 7.6.

   Table 7-6.  Adult Lead Model Baseline Blood Lead and Geometric Standard Deviations _
   Input Parameter         Baseline Blood Lead Level           Geometric Standard Deviation
   Central Values           1.7|ig/dl                          1.8
   High End Values _ 2.2 ug/dl _ 2.1 _

Fish ingestion rates for adult consumers of Columbia River fish are based on the median ingestion
rate of 39.2 g/day interpolated from Table 10 of the 1994 CRITFC consumption survey (CRITFC,
1994). Consumption rates were reported as  38.9 g/day and 40.5 g/day for the 49th and 53rd
percentiles respectively (CRITFC, 1994). For comparison, EPA provides a mean estimate of
national per capita fish consumption of 7.5 g/day (USEPA, 2000b). The Model was also run
using the 99th percentile ingestion rate from the CRITFC survey (389 g/day) to facilitate
comparison with the risks from chemicals other than lead (Table 7.1).
                                              7-160

-------
7.7    Risk Characterization for Fetuses

The Adult Lead Model was used to evaluate potential lead risks to the fetus following maternal
consumption of Columbia River fish. Predicted fetal geometric mean blood lead levels and
associated probabilities of exceeding the 10 jig/dl for a range of lead levels in fish are
summarized in Figures 7-4 and 7-5. Figure 7-4 shows results using the maximum recommended
exposure parameters for the baseline blood lead level of 2.2 jig/dl and geometric standard
deviation of 2.1 (USEPA, 1996b). Figure 7-5 is identical to Figure 7-4, but uses central tendency
estimates of baseline blood lead level of 1.7 |ig/dl and geometric standard deviation of 1.8.
Although, the predicted risks of exceeding 10 |ig/dl are substantially higher in Figure 7-4, the fish
concentration associated with a 5% risk of exceeding 10 |ig/dl is 700 |ig/kg.  Average fish
concentrations in whole fish and fillets were 0.12 and 0.02 respectively.  The highest lead
concentrations were found in whole-body samples of eulachon with an average fish tissue
concentration of 500 |ig/kg lead. For the fetus of an adult consuming 39.2 grams of whole fish
per day (129 |ig/kg), the Adult Lead Model predicts that fetal blood lead levels will exceed 10
|ig/dl less than 2% of the time using the high end values for baseline blood lead level and
geometric standard deviation.  Using high end values for baseline blood lead  level and geometric
standard deviation with the 389 g/day ingestion rate results in a predicted fetal blood lead level at
a fish concentration of 600 |ig/kg.
                     0   100   200  300   400   500   600   700   800  900   1000
                                  Concentration of Lead in Fish (ug/kg)

           Figure 7-4.  Predicted fetal blood lead levels with maternal fish ingestion rate
           of 39.2 g/day with baseline blood lead level at 2.2 ^g/dl and GSD = 2.1 ^g/dl.
                                              7-161

-------
                        3.5
                        3.0
                        2.5--
                       12.0
•-o15
"S°
                        1.0
                        0.5
                        0.0
                               FeHBcodLead
                               'Pn±elyoffeabtaodleaJ>10u9a
                                   .A'
                                      TlT
                                                                      20%
                                                 1.4% §>
                                                    O

                                                 1.2% A
                                                    T3
                                                    03
                                                 1.0% -1
                                                                        , m
                                                                         'E
                                                                      04%
                                                • 02%
                                                                      00%
                            0  1002003004005006007008009001CCO
                                      Corcertiefonaf Lead h Rsh (igkj)
                   Figure 7-5. Predicted fetal blood lead level with maternal fish
                   ingestion rate of 39.2 g/day with baseline blood lead level at 1.7 (ig/dl
                   andGSD=1.8ng/dl.

7.8    Uncertainty Analysis for Risk to Fetuses

Fetal risk estimates share common sources of uncertainties with the estimates for child risks
including the assumed fish lead concentrations and fish consumption rates.  Uncertainties unique
to the Adult Lead Model include the assumed baseline blood lead level and geometric standard
deviation parameters from the National Health and Nutrition Examination Survey
(USEPA,  1996b). The results are based on the highest recommend values for the baseline blood
lead levels and the geometric standard deviation. They are unlikely to underestimate risk.

7.9    Conclusions

Despite uncertainties in this assessment, lead levels in fish analyzed from the Columbia River
occur at levels unlikely to cause a blood level greater than 10 |ig/dl. Risks to children from fish
consumption are unlikely to exceed 5% at lead concentrations less than 500 |ig/kg
(Figure 7-2, 7-3). Similarly, fetal risks are unlikely to exceed 5% at concentrations less than
700 |ig/kg (Figure 7-4, 7-5).  These levels of concern occur at lead concentrations near the
maximum values of the samples.  This conclusion is supported by several analyses using health
protective exposure assumptions that are unlikely to underestimate risks from fish consumption.
The exposure assumptions are based on default and high end exposure parameters recommended
by EPA lead risk assessment guidance used in conjunction with fish ingestion rates from the
CRITFC fish consumption survey (CRITFC, 1994) .
                                             7-162

-------
8.0    Radionuclide Assessment

8.1    Radionuclide Data Reporting and Use

A unique characteristic of some radionuclide analytical data is the occurrence of numerically
negative results.  Radionuclide analyses usually require the subtraction of an instrument
background measurement from a gross sample measurement. Both results are positive, and when
sample activity is low (close to background), random variations in measurements can cause the
resulting net activity to be less than zero.  Although negative activities have no physical
significance, they do have statistical significance, as for example in the evaluation of trends or the
comparison of groups of samples. Good practice for laboratory reporting of radionuclide analysis
results therefore  dictates reporting results as generated: whether positive, negative, or zero,
together with associated uncertainties.

This is consistent with EPA guidance (USEPA, 1980a), which states: "When making
measurements near background levels, one can expect to frequently obtain values that are less
than the estimated lower limit of detection or minimum detectable concentration. If these values
are not recorded  and used in making average estimates, then these estimates are always going to
be greater than the "true" representation in the environment.  Therefore it is recommended that
every measurement result should be recorded and reported directly as found."

The general principles for evaluation of radionuclide data for this project were:

       a. It is generally best to use reported values plus the associated uncertainties.

       b. Reported values are better estimates of actual concentrations than are minimum
       detectable  concentrations.

       c. J-qualified (estimated) data should not be used for quantitative purposes where
       unqualified data is available to  substitute.

       d. All reported data (including U-qualified (nondetect) data, should be used in averages.

       e. Quantitative analyses should only be performed for those radionuclides which have at
       least one positive unqualified result reported.

       f For gamma data, the EPA's National Air and Radiation Exposure Laboratory (NAREL)
       reported  minimum detectable concentration values for certain radionuclides of interest
       even in cases where the radionuclide was not detected and no value was reported. If these
       minimum detectable concentrations are used for quantitative analyses, the results should
       clearly note the use of minimum detectable concentration-based input. If minimum
       detectable concentrations are to be used for quantitative purposes, the  minimum
       detectable concentrations may  need additional decay corrections where holding times
       exceeded 10 half lives.  This should not be an issue since no radionuclide with a half-life

                                             8-163

-------
       less than 10% of holding time was detected in any of the gamma analyses and therefore
       these short-lived radionuclides would not be used for analytical purposes.

8.2    General Information on Radiation Risk

Radiation is a known human carcinogen.  As such, the models used to estimate risk from
radiation exposure assume that at low levels of exposure, the probability of incurring cancer
increases linearly with dose, and without a threshold.

All of the epidemiological studies used in the development of radiation risk models involve high
radiation doses delivered over relatively short periods of time. Evidence indicates that the
response per unit dose at low doses and dose rates from low-linear energy transfer radiation
(primarily gamma rays) may be overestimated if extrapolations are made from high doses acutely
delivered. The degree of overestimation is often expressed in terms of a dose and dose rate
effectiveness factor that is  used to adjust risks observed from high doses and dose rates for the
purpose of estimating risks from exposures at environmental levels. EPA models for radiation
risk include a dose and dose rate effectiveness factor of 2 applicable to most low-linear energy
transfer radiation exposure. For high-linear energy transfer radiation (e.g. alpha particles), the
differences in relative biological effect are accounted for in weighting factors  applied in the
calculation of dose and risk.

In addition to cancer risk, radiation can also represent a risk for hereditary effects.  Radiation-
induced genetic effects have not been observed in human populations, however, and cancers
generally occur more frequently than genetic effects. The radiation-related risk of severe
hereditary effects in offspring is estimated to be smaller than that for cancer.  The risk of severe
mental retardation from radiation exposure to the fetus is estimated to be greater per unit dose
than the risk of cancer in the general  population, but the period of susceptibility is very much
shorter. Based on these considerations, EPA generally considers the risk of cancer to be limiting
and uses it as the sole basis for assessing radiation-related human health risks.

The risk coefficients used in this risk assessment are derived using age-specific models and are
age-averaged. This means that the risk coefficients are appropriate for use in estimating exposure
over a lifetime, since they are derived by taking into account the different sensitivities to radiation
as a function of age. The risk coefficients in this assessment may be used to assess the risk due to
chronic lifetime exposure of an average individual to a constant environmental concentration.
The risk estimates in this report are intended to be  prospective assessments  of estimated cancer
risks from long-term exposure to radionuclides in the environment.  The use  of the risk
coefficients listed for retrospective analyses of radiation exposures to populations should be
limited to estimation of total or average risks in large populations. The risk coefficients are not
intended for application to  specific individuals or to specific subgroups.

Estimates of lifetime risk of cancer to exposed individuals resulting from radiological and
chemical risk assessments may be summed to determine the overall potential human health
hazard.  It is standard practice, however, to tabulate the two sets of risk estimates separately.  This
                                             8-164

-------
is due to important differences in the two kinds of risk estimates.  For many chemical
carcinogens, laboratory experiments and animal data are the basis for estimates of risk. In the
case of radionuclides, however, the data come primarily from epidemiological studies of exposure
to humans. Another important difference is that the risk coefficients used for chemical
carcinogens generally represent an upper bound or 95th percent upper confidence level of risk,
while radionuclide risk coefficients are based on best estimate values.

8.3    Risk Calculations

Data qualifiers assigned during the data verification and validation process were used in making
decisions about numerical values for input into risk calculations. Reported values were used with
the following exceptions: zero was used where negative values were reported and one half of the
reported minimum detectable concentration was used where the result was reported as  minimum
detectable concentration.

The  naturally-occurring radionuclide potassium-40 (K-40) is a special case in the risk
calculations. Potassium is an essential nutrient which contains the naturally radioactive isotope
potassium-40, which has a half-life of more than one billion years.  K-40 constitutes 0.01% of
natural potassium which as a result has a specific activity of approximately 800 pCi/g of
potassium.  Variations in diet have little effect on the radiation dose received,  since the amount of
potassium in the body is under close hemostatic control. Although K-40 is the predominant
source of radiation exposure from food, calculation of dose or risk for specific food pathways is
not meaningful  since the biological control of potassium content in the body (and hence the
radiation dose due to potassium) means that the dose is independent of intake. Therefore, K-40
concentrations were not included in the calculations of cumulative risk from radionuclides in
samples. K-40 concentrations and risks are discussed separately for comparison.

Quantitative analyses were performed only for those radionuclides which had at least one positive
unqualified result reported.  Those radionuclides and their associated risk coefficients are:
Radionuclide Risk Coefficient (risk/Bo)
Uranium -234 (U-234)
Uranium-235+D (U-235+D)
Uranium-238+D (U-238+D
Strontium-90+D (Sr-90+D)
Plutonium-239 (Pu-239)
Bismuth-212 (Bi-212)
Bismuth-214 (Bi-212)
Cesium-137+D (CS-127+D)
Potassium-40 (K-40)
Lead-212(Pb-212)
Lead-214(Pb-214)
Raon-224(Ra-224)
Thorium-228+D (Th-228+D)
Radon-226+D (Ra-226+D)
Telllurim-208 (Tl-208)
2.58 x 10-9
2.63 x ID'9
3.36 x lO'9
2.58 x 10-9
4.70 x 10-9
included in Th-228+D coefficient
included in Ra-226+D coefficient
l.OlxlO-9
9.26 x lO'10
included in Th-228+D coefficient
included in Ra-226+D coefficient
included in Th-228+D coefficient
1.14xlO-8
1.39xlO-8
included in Th-228+D coefficient
                                                                                        Risks
                                             8-165

-------
for individual radionuclides were calculated using morbidity coefficients for dietary intake from
EPA guidance (USEPA 1999c). Many of the radionuclides detected are members of important
naturally-occurring decay chains (e.g. Ra-226 series, Th-228 series). For these radionuclides,
risks were calculated based on risk from the entire decay series in secular equilibrium. Risk
coefficients representing the entire decay series (identified with "+D" designation) were derived
by summing the risk coefficients for all decay chain members. For some decay series members
(e.g. Po-218) no data is available in EPA guidance and these radionuclides were not included in
the calculation of risk coefficients (USEPA, 1999d).  Based on data for these radionuclides
reported in FEAST the risks from radionuclides which are not included in EPA guidance are
insignificant in comparison to the risks from the other members of the decay series for which
EPA guidance provides data (USEPA, 1994c; USEPA, 1999d).

The general approach used in selecting data for input into decay series  calculations was to:
       1) use measured data wherever possible,
       2) prioritize measured data in accordance with assigned data qualifiers, and
       3) to use minimum detectable concentration values ( minimum detectable concentrations)
       for input only when other sources of data were not available.

In selecting the value to use for the concentration of the radionuclide at the head of the chain,
decay products were used as surrogates. This is consistent with the physical principles of
radioactive decay and secular equilibrium.  Where more than one decay product was available to
act as surrogate, positive values were selected over nondetect. The largest positive value was
used where two or more otherwise equally suitable results were available.

In cases where Tl-208 was used as a surrogate for the Th-228 decay series, the branching ratio of
the Bi-212 decay (36% decaying to Tl-208) was taken into account. If no decay chain member
data is available, one-half of the minimum detectable concentration value for Ra-226 was used
for input into the calculation for the Ra-226+D subchain.  Similarly, one-half the minimum
detectable concentration for Ra-228 was used as input into the Th-228+D subchain calculation
where necessary.  In the case of Cs-137, if no gamma peak was reported, one-half of the Cs-137
minimum detectable concentration was used as input for this radionuclide.

If there was a choice between uranium data from uranium alpha analyses and from gamma
analyses (e,g, U-235), the uranium alpha analysis data was used. Alpha analysis for uranium is a
more sensitive technique than gamma analysis. In particular, U-235 analysis by gamma
spectroscopy involves additional analytical uncertainty resulting from Ra-226 interference with
the spectral line used to quantify U-235. If only the gamma data was available, it was used with
appropriate consideration of data qualifiers.

Analytical results used for risk calculations included three samples which had a total of six "J"
qualified (estimated) results among them.  Five of these estimated values represented uranium
isotopes which are expected to be present, and for which the estimated values represent the best
available data for input into the risk calculation. In one case the estimated value used represented
a result for Pu-239. These estimated values were included in the calculations for completeness,


                                             8-166

-------
and their inclusion did not significantly alter the magnitude of the risks calculated.

8.4    Composite Study site Results

Plutonium, strontium and uranium analyses were not performed on all samples sent for
radionuclide analysis. For some of the composite groups of samples (composites 53 (study site
Columbia River 9U), 24 (study site Columbia River 7), and 25 (study site Columbia River 8),
only gamma analyses were performed. Risks were calculated based on the gamma component of
these samples only.  Risks were calculated based on a nominal consumption rate of 1 gram per
day and also for consumption rates of 7.5 g/day (average public consumption), 142.4 g/day (99th
percentile public consumption), 63.2 g/day (average CRITFC's member tribe consumption) and
389 g/day (99th percentile CRITFC's member tribe consumption).  These consumption rates are
the same as used for the nonradionuclide risk analysis. Risks were calculated for a 70 year
lifetime.  Composites of particular interest include Composite 54 (study site -K-Basin ponds) and
30 (study site Snake River 13). Table 8-1 presents a summary of the calculated risks for each
consumption rate.

8.4.1  Potassium-40 Results

As expected, the results for K-40 analyses are very consistent throughout the samples and
represent one of the most prominent sources of radioactivity in all samples analyzed.  The
concentrations in samples ranged between 1.7 pCi/g and 3.7 pCi/g with an average value of 2.8
pCi/g.  If this value were used to calculate risk in the same manner as the other radionuclides
detected, the resulting calculated average risk would be 1 x 10~3.  As noted previously, however,
although K-40 is the predominant source of radiation exposure from food, calculation of dose or
risk for specific food pathways is not meaningful since the biological control of potassium
content in the body (and hence the radiation dose due to potassium) means that the dose is
independent of intake. Therefore, K-40 concentrations were not included in the calculations of
cumulative risk from radionuclides in samples. K-40 concentrations and risks are presented
separately for the purposes of comparison.

8.5    Background

As anticipated, many of the radionuclides present in naturally-occurring background were also
present in the samples analyzed.  The sampling and analysis for radionuclides was not designed to
provide the statistical power necessary to quantitatively define background. The mobile nature of
the species sampled together with normal regional and local variations in concentrations of
naturally-occurring radionuclides in the environment make such an effort impractical in the
context of this project. However, an effort was made to obtain data that would provide  a
qualitative perspective on  background concentrations in fish. To this end, samples were taken
from the Snake River (composite group number 30; study site Snake River 13) to represent fish
that would not be affected by the operations of nuclear facilities in the Tri-Cities area.
Examination of the analytical results for the Snake River samples shows that in none of the
samples was there any Pu-239 or  Sr-90 detected.  Cs-137 was detected, as could be expected  from


                                            8-167

-------
the worldwide distribution of this radionuclide as a result of the atmospheric testing of nuclear
weapons during the 1950's and early 1960's.  In addition, naturally occurring radionuclides in the
uranium and thorium decay series were also detected.
                                            8-168

-------
  Table 8-1. Composite risks for consumption offish contaminated with radionuclides from the Columbia River Basin for the general public and
  CRITFC's member Tribes.
Fish Consumption Rates
Composite
number
(study sites)
52 (9E,9F)
53 (9F,9H)
54 (9K)
24 (7A)
25 (8F)
29 (8E,8B)
84 (8F)
85 (8F,8I)
86 (8C)
30(13E,13F)
87 (91)
88 (91)
78 (9Q,9P)
79 (9O,9N)
82 (9D,9B,9A)
83 (9A)
Species
Largescale sucker
Largescale sucker
White sturgeon
White sturgeon
White sturgeon
White sturgeon
Channel catfish
Largescale sucker
Channel catfish
White sturgeon
White sturgeon
White sturgeon
Mountain whitefish
Mountain whitefish
White sturgeon
White sturgeon
Unit
(Ig/d)
6 x 1O7
9 x 10-'*
6 x 1C'7
1 x lO6*
8 x 1C'7*
6 x 10-7
8 x 10-7
9 x 10-7
6 x 1O7
8 x lO'7
7xlO7
7xlO-7
8 x 1O7
6 x 1O7
8 x 1C'7
5 x 10-7
Average Public
(7.5 g/d)
5 x 10-"
7 x 106 *
5 x lO'6
8xl06*
6xlO-6*
5 x 10-6
6 x 10-6
7xlO-6
5 x 10-6
6 x 10-6
5 x 1O6
5 x 1C'6
6 x 10-6
5 x lO6
6 x 10-6
4 x 1O6
High Public
(142.4 g/d)
9 x lO5
1 x 10-4*
9 x 10-5
IxlO4*
1 x 10-4*
9 x 1O5
IxlO4
IxlO4
9 x 10-5
IxlO4
IxlO4
IxlO4
IxlO4
9 x 10-5
IxlO4
7xlO-5
Average CRITFC's
member tribe
(63.2 g/d)
4 x 10-5
6xlO-5*
4 x 10-5
6xlO-5*
5xlO-5*
4 x 10-5
5 x lO'5
6 x 10-5
4 x 10-5
5 x 10-5
4 x 10-5
4 x 10-5
5 x 10-5
4 x 10-5
5 x 10-5
3 x 10-5
High CRITFC's member
tribe
(389 g/d)
2xl04
4xl04*
2xl04
4xlO-4*
3xl04*
2xl04
3xl04
3xl04
3xl04
3xl04
3xl04
3xl04
3xl04
2xl04
3xl04
2xl04
*Composites 53, 24, and 25 did not have uranium, strontium or plutonium analyses performed, and the composite risks do not include contributions from those radionuclides.
                                                                    8-169

-------
8.6    Uncertainties

The uncertainty associated with cancer risk estimates for ingestion offish contaminated with
radionuclides includes contributions from the analytical uncertainties of the reported results, and
risk coefficients. The analytical uncertainties associated with the laboratory results are reported
at the two standard deviation level. For radionuclide analyses, uncertainties related to counting
statistics depend on the number of counts obtained, which varies with the analytical technique
used as well as the concentrations of radionuclide in the sample. As a percentage of the reported
result, their magnitude typically varies from a few percent in the case of gamma results which are
significantly greater than detection limits (e.g. K-40 results), to 20-40% for uranium results, to
more than 100% in cases of reported results which are classified as non-detect.

Some analytical results are qualified as estimated values due to interferences from other
radionuclides in the analysis. Additional uncertainty results from the use of some radionuclides
as surrogates for other radionuclides in decay series, the assumption of secular equilibrium, and
the use of minimum detectable concentration data in calculating risk. These uncertainties likely
result in overestimates of risk.

The uncertainties associated with the risk coefficients are likely to be larger than those due to
analytical uncertainties.  EPA guidance does not provide specific quantitative uncertainty
estimates of the cancer risk coefficients (USEPA 1999d).  National Council on Radiation
Protection and Measurements. (NCRP) Report 126 (NCRP, 1997), examined the question of
uncertainties in risk coefficients for the relatively simple case of external radiation exposure to
low linear energy transfer (primarily gamma) radiation. The conclusion was that the 90%
confidence interval encompassed a range approximately a factor of 2.5 to 3 higher and lower than
the value of the risk estimate.  Since estimates of risk from ingestion of food necessarily involve
the added complexity of modeling of physiological processes to determine dose and risk, the
uncertainties in this context are likely to be even greater.

The National Academy of Sciences Committee on the Biological Effects of Ionizing Radiation
(BEIR), in their report, addressed the issue of uncertainty in risk estimates for low doses from low
linear energy transfer radiation (NAS, 1990). BEIR V considered the assumptions inherent in
modeling such risks and concluded that at low doses and dose rates it must be acknowledged that
the lower limit of the range of uncertainty in the risk estimates extends to zero.

8.7    Discussion

Considering the number of samples, the mobility of the fish, and the range of results obtained, it
does not appear to be possible to attribute results to specific sources.  Most of the radionuclides
detected are known to be present naturally in the environment.  Cs-137 is also widespread in the
environment and was detected in many samples without apparent pattern.  There were three
samples in the vicinity of the Hanford Reach (Columbia River study site 9U) which showed
positive detection results for Sr-90.
                                             8-170

-------
Sr-90, like Cs-137, is a widespread radionuclide resulting from atomic testing in the atmosphere.
It is also associated with Hanford operations and is known from other environmental studies to be
present in Columbia River sediments near Hanford.

The estimated risks are similar across all composite groups (Table 8-1). This is consistent with
the observation that the majority of the estimated risk is generally due to radionuclides which are
members of naturally occurring decay chains.

8.8    Conclusions

The risks calculated for fish consumption (Table 8-1) are small relative to the estimated risks
associated with radiation from naturally-occurring background sources, to which everyone is
exposed. In the US, the average annual effective dose equivalent is approximately 300 millirem
including exposure to radon. The lifetime risk associated with this background dose can be
estimated to be approximately 1 x 10~2, or 1%.
                                             8-171

-------
9.0    Comparisons of Fish Tissue Chemical Concentrations

9.1    Comparison by Chemical Concentration

In this section the fish tissue residues from our study are compared to other food types and studies
of contaminants in fish reported in literature. This section also includes a comparison offish
tissue concentration data for smallmouth bass and channel catfish in addition to the 13 fish
species which were the main focus of this report.

9.1.1  Chlordane

Chlordane was used as a pesticide from the 1940's until the late 1980's. Until 1983 it was used on
corn and citrus fruits, lawns and gardens.  It was banned in 1988.

Like most of the other cylclodiene pesticides (heptachlor, heptachlor expoxide, aldrin, dieldrin,
endrin, and endosulfans I and n) chlordane degrades very slowly.  Various of its metabolites can
stay in the soil for over 20 years and can bioaccumulate in tissues of higher organisms.

Exposure to chlordane occurs largely from eating contaminated foods, such as root crops, meats,
fish, and shellfish, or from touching contaminated soil.  In the early 1980's chlordane was
detected in  4 of 324 food composites:  3 potato composites ranging from trace to 2 |ig/kg, and 1
garden fruit composite at a trace level (Gartrell et al., 1986). In the 1980 U.S. Food and Drug
Administration (USFDA) market basket survey of infant and toddler diet samples, chlordane was
detected at  5 |ig/kg in one of 143 toddler food composites (Gartrell et al., 1985).

Chlordane concentrations of 118 to 290 |ig/kg were measured in various estuarine fish in coastal
states surveyed (Butler and Schutzmann, 1978).  In a more recent  survey, Munn and Gruber
(1997) reported fish concentrations of 140 - 610 |ig/kg of the sum of chlordane in composite
samples of whole body fish from the Central Columbia Plateau.

The average concentrations of total chlordane found in anadromous fish tissue from our study
ranged from <4 |ig/kg in eulachon and coho salmon to 43 |ig/kg in Pacific lamprey (Table 2-3).
Egg samples from spring chinook sample had the highest average concentration (66 |ig/kg) in our
study (Table 2-3). The average concentrations of total chlordane in the resident fish species in
our study ranged from < 2.4 |ig/kg in rainbow trout and bridgelip sucker to 29 |ig/kg in white
sturgeon (Table 2-3).

9.1.2  Total  DDT

The legal use of DDT in agriculture has been banned in the United States since 1972. DDT and
its derivatives are persistent, bioaccumulative compounds which are ubiquitous in the organisms,
sediments, and soils.
                                            9-172

-------
Exposure to DDT and its structural analogs (DDE, ODD) occurs primarily from eating
contaminated foods, such as root and leafy vegetables, meat, fish, and poultry.  From 1967 to
1972 the concentrations of total DDT in meat, fish and poultry decreased from 3,200 |ig/kg to 900
jig/kg  (IARC, 1978).  From 1970 to 1973, DDE residues decreased only 27%, compared to a
decrease of 86% and 89% for DDT and ODD, respectively (USEPA, 1980).

Based on data from the US Fish and Wildlife Service National Pesticides Monitoring Program
(Schmitt et al, 1981), the DDT concentrations in fish ranged from 100 to 11,000 |ig/kg.

DDT was detected in meats (0.3 |ig/kg) and raw berries (2.0 |ig/kg) consumed by indigenous
residents of the Canadian Arctic (Berti et al.,  1998).

The maximum concentration of DDE in the fish from several USGS surveys was in a whole body
composite sample of carp (3,300 |ig/kg) from the Brownlee Reservoir on the Snake River, Idaho
(Table  9-1). The maximum concentration of DDE in our study was in the whole body composite
sample of white sturgeon (1400 |ig/kg) from the Hartford Reach of the Columbia River (study site
9U). The maximum concentrations of DDE in bridgelip sucker, rainbow trout, and largescale
sucker levels in our study were higher than levels found by Munn and Gruber (1997) in the
Central Columbia Plateau (Table 9-1). The largescale sucker levels in our study were similar to
the largescale sucker levels reported by Clark and Maret (1998) for the Snake River Basin.

 Table 9-1. Comparison of range concentrations of sum of DDE (o,p' & p.p') in whole body composite fish
 samples Columbia River Basin.
Fish
carp
bridgelip sucker
bridgelip sucker
bridgelip sucker
rainbow trout
rainbow trout
largescale sucker
largescale sucker
largescale sucker
US/kg
3300
87
120-340
347-612
9.5-32
5-89
33-1300
120-400
29-1312
Location
BrownleeReservoir, Snake River,Idaho
Palouse River, Central ColumbiaPlateau
Northern Desert, Central Columbia
Columbia River Basin
Northern Desert, Central Columbia
Columbia River Basin
Snake River Basin
PalouseRiver, Central ColumbiaPlateau
Columbia River Basin
Reference
Clark and Maret ,1998
Munn and Gruber, 1 997
Munn and Gruber ,1 997
Our study, 1996-1998
Munn and Gruber, 1 997
Our study, 1996-1998
Clark and Maret ,1998
Munn and Gruber, 1 997
Our study, 1996-1998
9.1.3  PCBs

PCBs, are stable, man-made chemicals that only degrade at very high temperatures. They do not
conduct electricity and most of the various types of PCBs and PCB mixtures take the form of
liquids.  For these reasons, PCBs have been used extensively in much of the world as electrical
insulating fluids, especially in capacitors and transformers which deliver high voltage in critical
devices and situations where fire prevention is of great concern.  PCBs have also been used
extensively as hydraulic fluids, as well as in the manufacture of carbonless copy paper, etc.
Environmental contamination with PCBs has resulted from industrial and domestic discharges,
landfills, and atmospheric transport of incompletely incinerated PCBs.

Under environmental conditions, PCBs are extremely stable and slow to chemically degrade

                                           9-173

-------
(Eisler, 1986b).  PCBs enter the environment as mixtures containing a variety of individual
components (congeners) and impurities that vary in toxicity.  The chlorinated nature of the
various PCB molecules also makes them more fat soluble, and thus capable of bioaccumulating in
aquatic food webs.  The lipid solubility of the PCBs increases with increased chlorine
substitution. This lipophilicity also tends to increase resistence to biodegradation.

Because of the relatively great environmental persistence and lipophilicity of this group of
pollutants, low-level PCB contamination is now a global phenomenon, with PCB residues
occurring almost universally in human milk, other human tissues, food, etc. For the general
population, likely routes of ongoing chronic exposure to PCBs are primarily from food
(Table 9-2).

             Table 9-2. PCB residues in raw agricultural commodities, 1970-76.
             (Source: Duggan et al, 1971)
Food Tvne
fish
eggs
milk
cheese
red meat
poultry
Number of
samples
2,901
2,302
4,638
784
15,200
11,340
Percent
Detected
46
9.6
4.1
0.9
0.4
0.6
Average
fue/ke)
892
72
67
11
8
6
The estimated PCB content of a typical teenage boy's diet was about 15 ng/day in 1971,
decreasing by 1975, to about 8.1 ng/day (IARC, 1978).  The levels of PCBs have declined in
ready-to-eat foods from 1978 to 1982 (Table 9-3). However, the human body burden remains
high.  The body burden of PCBs in human fat ranged between 500 and 1,500 |ig/kg in 1987
(USEPA, 1987).     	
                    Table 9-3.  The declining trends in PCBs in ready-to-eat foods collected
                    in markets of a number of US cities (Source: Duggan et al., 1971).

Year
1978
1979
1980
1981-82
Number of
samples
360
360
360
324
Percent
Detected
9
4
2
2
Average
( iie/ke
trace - 50

-------
                  Table 9^4. The 1976-80 ranges for PCB residues from 547 finfish from
                  the Chesapeake Bay and its tributaries ( Source: USEPA, 1987a).
                                     Year                     ug/kg
                                     1976                   ND-980
                                     1977                    30-510
                                     1978                   60-4,640
                                     1979                   10-1,600
                                      1980                    3-1,450
In later studies concentrations of total PCBs in a variety offish tissue types ranged from
10 |ig/kg in white sucker fillets in Saginaw Bay, Lake Huron, Michigan to 14,500 |ig/kg in fish
from the Spokane River, Washington (Table 9-5).  Measurements of Aroclor 1254 and 1260 in
white croaker muscle in California ranged from 1 |ig/kg to 713 |ig/kg (Table 9-6).


  Table 9-5.  Total PCB concentrations in fish tissue from studies reported in the literature from 1978-1994.

  Species & Tissue type        \ipfkp          Location/date of study                    Reference
  fish livers                 132-772         near the outfall for the Los  Angeles County Gossett et al., 1983.
                                           wastewater treatment plant 1980-81,

  750 fish samples            70-14,500       11 major lakes and rivers in Alberta, Canada  Chovelon et al., 1984

  25 white suckers fillets       10-180          Saginaw Bay, Lake Huron, 1979-1980       Kononen, 1989

  freshwater fish (whole body) mean = 36        Spokane River, WA, 1999                Johnson, 2001
                           maximum =930
               Table 9-6.  Concentrations Aroclor 1254 & 1260 in white croaker muscle
               tissue from California water bodies in the spring of 1994. (Source: Fairey et
               al., 1997)
                        ug/kg                                 Location
                       137 - 613                  13 locations throughout San Francisco Bay
                          1                         Southern California Dana Point,
             	757	Malibu	

The concentration of Aroclor 1254 ranged from 480 |ig/kg to 9,930 |ig/kg in lake trout from lakes
in Michigan (Table 9-7). The concentration of Aroclor 1254 in resident fresh water species from
our study ranged from 10 |ig/kg in rainbow trout to 930 |ig/kg in mountain whitefish.


               Table 9.7. Concentrations of Aroclor 1254 in lake trout from lakes in Michigan
               during 1978-82  (Devault et al., 1986).	
             	ug/kg	      	Location	
                          5630 - 9930                               Lakes Michigan
                          2100-3660                                 Lake Huron
                           480-1890                                 Lake Superior

The concentration of Aroclors in chinook salmon eggs from Lake Michigan were much higher


                                                9-175

-------
than the levels found in our study (Table 9-8).
       Table 9-8. Aroclor concentrations in chinook salmon eggs reported for Lake Michigan, Michigan,
       compared to our study of Aroclors in the chinook salmon eggs.
Mg/kg
Aroclor 1254
5,400
12
15-20
Aroclor 1260
1,100
<19
<18
N

1
6


1

salmon
chinook
fall chinook
spring chinook

chinook
fall chinook
spring chinook
Location/date of study
Lake Michigan, 1982 (Jaffet et al.,
Columbia River Basin, 1996-1998
Columbia River Basin, 1996-1998

Lake Michigan, 1982 (Jaffet et al.,
Columbia River Basin, 1996-1998
Columbia River Basin, 1996-1998

1985)


1985)


       < = detection limit

Concentrations of PCBs measured in fish from our study were compared to other fish surveys in
Lake Roosevelt on the upper Columbia River in Washington (Table 9-9). The maximum
concentration of Aroclors 1254 and 1260 in walleye and rainbow trout were lower in our study of
the Columbia River Basin than the EPA (USEPA, 1998c) and USGS (Munn, 2000) surveys of
Lake Roosevelt, Washington.  Concentrations of the Aroclors in white sturgeon were higher in
our study than the EPA study of Lake Roosevelt, Washington (Table 9-9).

     Table 9-9. Concentrations of Aroclors 1254 and 1260 in composite samples offish fillets from Lake
     Roosevelt, Washington compared concentrations measured in our study of the Columbia River
     Basin.
Fish Species

small walleye
large walleye
walleye
white sturgeon*
white sturgeon *
rainbow trout
rainbow trout
rainbow trout
smallmouth bass
smallmouth bass
kokanee
lake whitefish
small walleye
large walleye
walleye
white sturgeon*
white sturgeon *
rainbow trout
rainbow trout
smallmouth bass
smallmouth bass
kokanee
lake whitefish
Jig/kg
Aroclor 1254
30-10
35-89
12-14
15-77
10-190
13-45
3-49
10-20
ND-8
38-83
28-40
31-51
Aroclor 1260
4-13
23-32
13-102
13 - 200
5-72
3-6
68 - 220
10-14
16-29
N

9
2
7
2
16
10
16
7
9
3
4
3
9
2
7
2
16
10
7
9
3
4
3
Location

Lake Roosevelt, 1994
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Lake Roosevelt, 1998
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Reference

USEPA, 1998c
USEPA, 1998c
our study
USEPA, 1998c
our study
USEPA, 1998c
Munn, 2000
our study
USEPA, 1998c
our study
USEPA, 1998c
USEPA, 1998c
USEPA, 1998c
USEPA, 1998c
our study
USEPA, 1998c
our study
USEPA, 1998c
our study
USEPA, 1998c
our study
USEPA, 1998c
USEPA, 1998c
   N - number of samples  < = detection limit *White sturgeon were individual fillets without skin
                                              9-176

-------
9.1.4  Chlorinated Dioxins and Furans

Because of their chlorination and specific chemical structures, most chlorinated dioxins and
furans are highly fat soluble, and difficult for the body to quickly degrade and excrete. They are
similar to some of the other persistent chlorinated residues like DDT and PCBs. Also like PCBs
and DDTs, chlorinated dioxins and furans can bioaccumulate in fish.  The amount of furans in
fish can sometimes be tens of thousands times higher than the levels in the surrounding water.

The chlorinated dibenzodioxins and chlorinated dibenzofurans are not produced intentionally by
industrial processes.  Rather, most chlorinated dioxins and furans are generated in very small
amounts as unwanted impurities during the manufacture of several chlorinated chemicals and
consumer products, including certain wood treatment chemicals, some metals, and paper
products. When the waste water, sludge, or solids from these processes are released into
waterways or soil in dump sites, the sites may become contaminated with chlorinated dioxins and
furans. These unwanted contaminants also enter the environment from burning municipal and
industrial waste in incinerators, as well as from gasoline exhaust, and the burning of coal, wood,
or oil for home heating and production of electricity.  Other production chemicals which can
generate unwanted trace amounts of 2,3,7,8-TCDD have included the forestry herbicide 2,4,5-
trichlorophenoxy propionic acid (Silvex), and the industrial chemical
2,4,5-trichlorophenol.  Unwanted trace amounts of some of the higher-chlorinated dioxins,
especially the hexa and octa isomers, have also been associated with the production of the widely
used wood preservative, pentachlorophenol.

Many of the various chemicals and processes which significantly produce chlorinated dioxins and
furans in the environment  are either being  slowly phased out or are strictly controlled.  It is
currently believed that chlorinated dioxin and furan emissions associated with incineration and
combustion activities are the predominant environmental source of these contaminants (USEPA,
2000e). Chlorinated dioxins and furans also arise from natural processes in the environment such
as forest fires and volcanos.

TCDF is  often found in fish tissue because of its affinity for lipids and because of its formation as
a by-product in the industrial processes, especially pulp and paper mills (USEPA, 2000e).  The
concentration of 2,3,7,8-TCDF was measured in a variety offish species from Lake Roosevelt,
Washington by the USEPA in 1994 (Table  9-10). The concentrations of 2,3,7,8-TCDF in walleye
ranged from 0.0001 to 0.0063  |ig/kg (Table 9-10). The maximum concentration from our study
was lower than the maximum reported for Lake Roosevelt, Washington.  The white sturgeon
2,3,7,8-TCDF maximum concentration in our study was higher than the maximum from the 1994
Lake Roosevelt study (Table 9-10). The rainbow trout 2,3,7,8-TCDF concentrations were similar
in both studies.
                                            9-177

-------
Table 9-10. Concentrations of 2,3,7,8-TCDF in composite samples of fish fillets collected from Lake
Roosevelt, Washington in 1994 compared with our 1996-1998 survey of the Columbia River Basin.
Fish
small walleye
large walleye
walleye
white sturgeon
white sturgeon
small rainbow trout
large rainbow trout
rainbow trout
kokanee
smallmouth bass

lake whitefish
Jig/kg
0.0001 - 0.0016
0.0007- 0.0063
0.0006 - 0.00085
0.016- 0.025
0.0025- 0.054
0.000098-0.0015
0.0015-0.00188
0.0001- 0.0003
0.0028- 0.0031
0.00001 - 0.0041
0.0038 - 0.01610

N
9
2
3
2
16
6
10
7
4
9

3
Collection date
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Columbia River Basin, 1996-98
Lake Roosevelt, 1994
Columbia River Basin, 1996-98
Lake Roosevelt, 1994
Lake Roosevelt, 1994
Columbia River Basin, 1996-98
Lake Roosevelt, 1994
Lake Roosevelt, 1994

Lake Roosevelt, 1994
Reference
USEPA, 1998c
USEPAc 1998c
our study
USEPA, 1998c
our study
USEPA, 1998c
USEPA, 1998c
our study
USEPA, 1998c
USEPA, 1998c

USEPA, 1998c
N= number of samples
In the USEPA National Dioxin Survey (USEPA, 2000d) background levels of toxicity
equivalence concentrations for chlorinated dioxins, furans, and dioxin-like PCB congeners were
0.00116 ±0.00121 ng/kg in fish and 0.00046 ± 0.00099 |ig/kg in beef. In our study the average
toxicity equivalence concentrations ranged from a low of 0.0004 |ig/kg in fall chinook salmon to
the highest average concentration of 0.0063 |ig/kg in mountain whitefish.

9.1.5  Metals

The metals measured in our study are naturally occurring substances. Some of these metals are
essential at trace levels for survival of vertebrates. These chemicals may combine with other
chemicals to form compounds,(e.g. methylmercury, dimethyarsenic, arsenocholine, arsenosugars)
which alters their bioavailability and toxicity.  Most can become toxic if sufficiently high levels
are encountered in the environment. Many of the metals which are taken up by fish tend to
increase in concentration as the organisms age and increase in body size (Wiener and Spry, 1996,
reported in Clark and Maret, 1998).

Information about barium, beryllium, cobalt,  and  manganese and are not included in this section.
Background information on these chemicals is included in the Toxicity Profiles (Appendix C)

9.1.6  Aluminum

Aluminum is the most common and widely distributed metal in the earth's crust.  Concentrations
as high as 150,000 - 600,000 mg/kg have been reported in soil. The average ingestion of
aluminum by humans has been estimated at 30 - 50 mg/day (Bjorksten, 1982).  This estimate may
be low, in light of a 1997 United Kingdom (UK) total diet study involving 20 different food
groups from 20 representative towns, for the general UK population, where the highest mean
concentrations of aluminum were found in the bread (6,600 |ig/kg)  and fish (6,100 |ig/kg) (Ysart
et al., 2000). Aluminum is present in the natural diet, in amounts varying from very low in
animal products to relatively high in plants.
                                            9-178

-------
In our study the basin-wide average aluminum concentrations ranged from non-detect in coho
salmon (whole body and fillet) to 69,000 |ig/kg in whole body largescale sucker. The maximum
concentration was 190,000 |ig/kg in the largescale sucker composite sample from the main-stem
Columbia River (study site 8).

9.1.7   Arsenic

Arsenic is found widely in nature, and occurs most abundantly in sulfide ores. Arsenic levels in
the earth's crust average about 5,000 |ig/kg.  Arsenic is found in trace amounts in aquatic
environments. As was described in Section 5, arsenic exists in both organic and inorganic forms.
The most common combined form of arsenic is the inorganic compound, arsenopyrite (FeAsS).
The organic arsenic compounds are less toxic than the inorganic arsenic compounds.

Arsenic does not readily bioconcentrate in aquatic organisms.  It is typically water soluble and
does not combine with proteins. Since, aquatic invertebrates accumulate arsenic more readily
than fish biomagnification is unlikely (Spehar et al., 1980). Planktivorous fish are more likely to
concentrate arsenic than omnivorous or piscivorous fishes (Hunter et al., 1981). Eisler (1988a)
found no evidence that biomagnification occurs in aquatic food chains. In 1995, Robinson et al.,
found no evidence of arsenic uptake or accumulation from water in both rainbow and brown
trout.  The rainbow trout in our study had the lowest arsenic concentrations (<25 |ig/kg fillet; 120
jig/kg whole body) of the fish species sampled.

In a 1997 UK study, dietary exposures to arsenic were estimated to be about 65 jig /day (Ysart et
al., 2000). The "fish" food group had the highest mean arsenic concentration (400 |ig/kg; Ysart et
al., 2000).

Arsenic levels recorded for fish tissues seem to be quite variable. Fish taken from the Great lakes
contained 5.6-80 |ig/kg arsenic; primarily in the lipid fraction of the fish tissue
(Lunde, 1970).  In a study of African tilapia fish, muscle tissue contained arsenic levels ranging
froml 10 |ig/kg(Ikdu and Marget Lakes) to one specimen with 10,500 |ig/kg (Abu Quir Bay)
(El Nabawi et al., 1987). Ashraf and Jaffar (1988) measured arsenic levels of 2,880 jig/kg and
2510 |ig/kg in two tuna species from the Arabian Sea.  The authors noted that increased arsenic
content was proportional to increased weight in the tuna species.

The average arsenic levels in resident, fresh water fish species in our study ranged from not detect
in rainbow trout fillet to 490 |ig/kg in whole body walleye (Table 2-14).  The average
concentrations in anadromous species from our study ranged from 310 |ig/kg in Pacific lamprey
fillet to 890 |ig/kg in whole body eulachon.   There was no correlation between lipid and arsenic
in fish in our study, as was observed in the Great Lakes study (Lunde, 1970) or body weight and
arsenic as observed by Asraf and Jaffar (1988).
                                            9-179

-------
9.1.8   Cadmium

Cadmium naturally occurs in the aquatic environment, but is of no known biological use and is
considered one of the most toxic metals. While cadmium is released through natural processes,
anthropogenic cadmium emissions have greatly increased its presence in the environment. In
aquatic systems, cadmium quickly partitions to sediment, but is readily remobilized through a
variety of chemical and biological processes (Currie et al, 1997).  Cadmium does not
bioconcentrate significantly in fish species, but does tend to accumulate more readily in
invertebrates. Omnivorous and insectivorous predators tend to accumulate cadmium in their
tissues more than piscivorous predators (Scheuhammer, 1991).  Saiki et al., (1995) found no
evidence of biomagnification of cadmium in steelhead on the Upper Sacramento River. Eisler
(1985a) also maintains that evidence for cadmium biomagnification suggests that only the lower
trophic levels exhibit biomagnification.  Cadmium tends to form stable complexes with
metallothionein (a sulfhydryl-rich protein). The resulting cadmium complexes have long half-
lives and a tendency to accumulate with age in exposed organisms. As such, long lived species
tend to be at a higher risk from chronic low-level dietary cadmium exposure.

People who are smokers  are exposed to significant levels of inhaled cadmium. The major
exposure route for the non-smoking human population is via food. In a 1997 UK study, the
mean population dietary  exposures to cadmium was estimated to be about 12 jig/kg/day for the
general UK population (Ysart et al., 2000).  Cadmium concentrations were highest in the viscera
and trimmings of animals (77 |ig/kg), and nuts (59 |ig/kg), while the bread and potato food groups
made up the greatest contributions (both 25%) to dietary exposure of the general population.

Certain cruciferous vegetable crops are known to be able to sequester elevated cadmium levels if
grown in sufficiently contaminated soils.  Queiroloa et al. (2000) reported ranges of 0.2 to
40 |ig/kg for cadmium, with highest levels being found in potato skin in a study of vegetables
(broad beans, com, potato, alfalfa and onion) from farming villages in Northern Chile.

The WHO (1992) indicates that marine organisms generally contain higher cadmium residues
than their freshwater and  land-dwelling counterparts. In our study the highest cadmium levels
were in whole body samples of largescale sucker (250 |ig/kg ) followed by spring chinook salmon
(170 |ig/kg) and Pacific lamprey (150 |ig/kg).

Average cadmium concentrations ranged from non detect in fillet samples of walleye, coho
salmon, and fall chinook  salmon to 120 |ig/kg in whole body spring chinook salmon. The
maximum concentration (250 |ig/kg) was in the largescale sucker composite sample from the
Hanford Reach of the Columbia River (study site 9U).

9.1.9   Chromium

Chromium is widely distributed in the earth's crust, with an average concentration of about
125,000 |ig/kg. It is found in small amounts in all soils and plants.  Most of the chromium
present in food is in the trivalent form [Cr(m)], which is an essential nutrient. The hexavalent

                                            9-180

-------
form is more toxic, but is not normally found in food. In freshwater environments, hydrolysis and
precipitation are the most important processes in determining the environmental fate of
chromium, while absorption and bioaccumulation are considered minor. Chromium (VI) is highly
soluble in water and thus very mobile in aquatic systems (Ecological Analysts, 1981).

The mean daily dietary intake of chromium from air, water, and food, is estimated to be about
0.2 - 0.4 |ig, 2.0 |ig, and 60 |ig, respectively (ATSDR, 2000). The predicted  intakes from air
chromium are probably exceeded considerably in the case of smokers, and those who are
occupationally exposed.

In a 1997 UK study, meat products contained the highest mean chromium concentration
(230 |ig/kg), but beverages made the greatest dietary contribution (19%) to the population
exposure to chromium (Ysart et al., 2000). The US Food and Nutrition Board has recommended
a safe and adequate dietary intake of chromium of 0.05 - 0.20 |ig/day (Seller and Sigel, 1988).

Chromium was found in fish sampled from 167 lakes in the northeast United States at levels
ranging from  30-1,460 |ig/kg with a mean of 190 ug/kg (Yeardley et al., 1998). Seaweeds have
been shown to sequester total chromium by a bioaccumulation factor of about 100 times greater
than ambient levels in seawater (Boothe and Knauer, 1972).  Snails showed an accumulation
factor of 1 x 10 6 for total chromium (Levine, 1961).

In our study, basin-wide average chromium concentrations ranged from <100 |ig/kg in eulachon
to 360 |ig/kg in the whole body white sturgeon (Table 2-14).  The maximum concentration
(1000 |ig/kg) was measured in the whole body white sturgeon sample from the main-stem
Columbia River (study site 8)

9.1.10  Copper

Because of its ubiquitous occurrence in the environment, and its essentiality for life, copper is
found naturally at trace levels in aquatic and terrestrial organisms.  Copper is not strongly
bioconcentrated in vertebrates, but is more strongly bioconcentrated in invertebrates. In
salmonids the accumulation of copper in muscle, kidney, and spleen tissues occurred at copper
concentrations ranging from 0.52-3 |ig/L in both seawater and freshwater (freshwater
hardness=46-47 mg/L)(Camusso and Balestrini, 1995; Peterson et al.,  1991;  Saiki et al., 1995).
The concentrations of copper in fish tissues reflect the amount of bioavailable copper in  the
environment.  Baudo (1983, Wren et al. (1983), and Mance (1987) have all concluded that
copper, along with zinc and cadmium do not biomagnify in the aquatic environment.

Intake of copper from food tends to be about one order of magnitude greater than intake from
drinking water (USEPA, 1987). Exceptions to this are in relatively rare situations involving
consumption of "soft" drinking water sources supplied by copper pipes; which can result in daily
individual drinking water intakes of copper in excess of 2 mg/day. In a 1997 UK diet study,
copper was highest in viscera and trimmings (50,000 |ig/kg) and nuts (8,500 |ig/kg), with mean
concentrations in the other food groups ranging from 50 to 2,100 |ig/kg (Ysart et al., 2000).

                                            9-181

-------
In our study, the copper concentrations ranged from 250 |ig/kg in white sturgeon fillet sample to
4500 |ig/kg in whole body Pacific lamprey.  The maximum concentration (14,000 |ig/kg) was in
the whole body fall chinook salmon composite sample from the main-stem Columbia River
(study site 14).

9.1.11   Lead

Lead is a naturally occurring, ubiquitous compound that can be found in rocks, soils, water,
plants, animals, and air. Lead is the fifth most prevalent commercial metal in the US. Lead is
found naturally in all plants, with normal concentrations in leaves and twigs of woody plants of
about 2,500 |ig/kg, pasture grass 1,000 |ig/kg, and cereals from  100 -1,000 |ig/kg (IARC, 1980).

Absorption of lead by aquatic animals is affected by the age, gender and diet of the organism, as
well as the particle size, chemical species of lead, and presence of other compounds in the water
(Eisler, 1988b; Hamir et al, 1982). Although inorganic lead is poorly accumulated in fish, it has
been shown to bioconcentrate in aquatic species. Invertebrates tend to have higher lead
bioconcentration factors than vertebrates. A bioconcentration factor of 42 was observed in brook
trout embryos (Eisler,  1988b). Bioconcentration factors decrease as waterborne lead
concentrations increase, thus suggesting accelerated depuration or saturation of uptake
mechanisms (Hodson et al., 1984). Exposures of rainbow trout to 3.5-51  |ig/L tetramethyl lead
from 7-14 days resulted in rapid accumulation of lead. However, once the fish were removed to
clean water, lead decreased rapidly from organs, followed by a slower release from other body
components, until baseline levels were reached.  An increase in dietary calcium of 0-8400 |ig/kg
reduced the uptake of waterborne lead in coho salmon, possibly due to interactions with gill
membrane permeability (Hodson et al., 1984). In vertebrates, lead concentrations tend to increase
with age and localize in hard tissues such as bone or teeth.

The primary exposure route for lead is food (Table 9-11). Foods which are likely to have
elevated lead levels are dried foods, liver, canned food, and vegetables which have a high area-to-
mass ratio.  Historic use of soldered food cans greatly increased the lead content of prepared and
processed foods. Sherlock (1987) reported that while ravioli from welded (no lead) cans
contained 30 |ig/kg lead, ravioli from a 98% lead soldered can was found to contain a mean
content of 150 |ig/kg lead.
                                            9-182

-------
           Table 9-11. Lead concentrations in food purchased in five Canadian cities between
           1986 - 1988 (Source: Dabeka and McKenzie, 1995.
category % contribution to
dietary intake
fruits and fruit juice
miscellaneous
vegetables
meat and poultry
fish
sugar and candies
soups
bakery goods and cereals
beverages
fats and oils
milk and milk products
canned and raw cherries
canned citrus fruit
canned beans
canned luncheon meats
13.9
6.1
16.8
7.6
0.7
1.5
4.5
20.6
20.9
0.3
7.1




mean
Jig/kg
44.4
41.7
24.4
20.2
19.3
18.3
15.5
13.7
9.9
9.6
7.7




maximum
Jig/kg
372.7
178.9
331.7
523.4
72. 8
111.6
48.7
66.4
88.8
19.7
44.7
203
126
158
163
The basin-wide average lead concentrations in fish from our study of the Columbia River Basin
ranged from non detect in fillets of Pacific lamprey, walleye, and rainbow trout to 500 |ig/kg in
whole body eulachon (Table 2-14). The maximum concentration (1200 |ig/kg) in our study was
in the whole body fall chinook salmon from the main-stem Columbia River (study site 14).

9.1.12    Mercury

While mercury does occur naturally in small amounts in aquatic environments, the cycling of
mercury prolongs the influence of man-made mercury compounds (Hudson et al,  1995). Mercury
is cycled through the environment through an atmospheric-oceanic exchange.  This cycling is
facilitated by the volatility of the metallic form of mercury. Natural bacterial transformation of
mercury results in stable, lipid soluble, alkylated compounds such as methyl mercury (Beijer and
Jernelov,  1979.  In sediments, mercury is usually found in its inorganic forms, but aquatic
environments are a major source of methyl mercury (USEPA,  1985).  In background freshwater
systems, mercury occurs naturally at concentrations of 0.02-0.1 |ig/L (Moore and Ramamoorthy,
1984).

Mercury has been shown to bioconcentrate in a variety of aquatic organisms. Aquatic predators
face the greatest danger of bioconcentrating mercury, and thus their tissue concentrations best
reflect the amount of mercury available to aquatic organisms in the environment. Fish have been
shown to concentrate mercury as methyl mercury even when they are exposed to inorganic
mercury.  Fish, such as rainbow trout, have been found to accumulate mercury in the form of
methyl mercury at aquatic concentrations as low as 1.38 ng/L (Ponce and Bloom,  1991).

Some evidence supports the biomagnification of mercury in aquatic food chains. When
comparing benthic feeding fish, fish that feed on plankton, invertebrates, and vertebrates, the

                                            9-183

-------
greatest mercury concentrations were found in piscivorus fishes.  Thus, the authors of this study
concluded that mercury content in fish increased with higher trophic levels (Wren and
MacCrimmon, 1986).

Freshwater ecosystems historically associated with heavy gold mining activity have often been
impacted by elevated mercury levels in fish. This is in large part due to the use of liquid
elemental mercury, or quicksilver, as a means of separating out gold during the mining process,
especially during historic times.

Dietary sources greatly exceed other media like air and water as a source of human mercury
exposure and uptake.  In a 1997 UK diet study, fish contained the highest mean concentration (43
jig/kg), and made the greatest contribution (33%) to the population dietary exposure estimate
(Ysart et al., 2000). The World Health Organization, EPA, and others indicate that risk to
humans from mercury contamination via ocean fish is mainly through the consumption of
predator species like swordfish, king mackerel, and shark (WHO, 1976).

In a  monitoring study offish in British Columbia, Canada, mercury concentrations in muscle
tissue of various fish ranged from 40 |ig/kg in rainbow trout to 2,860 |ig/kg in lake trout
(Table 9-12). In our study, rainbow trout the average mercury concentrations ranged from
73 |ig/kg in whole body samples to 77 |ig/kg in the fillet samples (Table 2-14).

             Table 9-12. British Columbia monitoring study of mercury
             concentrations in fish fillet tissue. (Source:  Bligh and Armstrong 1971)
                      Fish Species (study location)                    Jig/kg
                      Rainbow trout (Tezzeron Lake)                     40
                               herring                              70
                   dolly varden or char (Carpenter Lake)              410-1,940
                      dogfish or shark (English Bay)                   1,080
            	lake trout (Pinchi Lake)	2,860	
A 1984 EPA national survey offish tissue found mercury ranging from 50 |ig/kg in salmon to 610
jig/kg in pike (Table 9-13).  In our study average mercury concentrations in fillet samples of
salmon was 84 |ig/kg in fall chinook, 100 |ig/kg in spring chinook, and 120 |ig/kg in coho.
(Table 2-14).
                                             9-184

-------
       Table 9-13. EPA 1984 survey of total mercury concentrations in edible fish tissue, shrimp,
       and prepared foods. (Source USEPA. 1984b)
Fish Species
salmon
whiting
sardines
flounder
snapper
bass
catfish
trout
pike
Hg/kg Invertebrates
50 shrimp
50
60
100
450
210
150
420
610
Hg/kg Prepared food MS/kg
460 fish sticks 210
canned tuna 240







In a more recent EPA national survey of mercury in fish tissue, median mercury levels ranged
from 1 ng/kg in largemouth bass, channel catfish, bluegill sunfish, and common carp to 8,940
Hg/kg in largemouth bass (Table 9-14).  The concentrations of mercury fillets offish tissue in our
study were 380 - 470  |ig/kg in smallmouth bass, 160 - 200  |ig/kg in walleye,  and
240 - 280 |ig/kg in channel catfish (Table 9-27). All of these fish species had lower
concentrations in our study than in the EPA 1990-1995 survey (USEPA, 1999e).


                 Table 9-14. Mercury concentrations from an EPA 1990 - 1995 national
                 survey offish fillets (Source : USEPA, 1999e).	
                         Species                               Jig/kg
                     largemouth bass                         1 - 8,940
                     Smallmouth bass                         8 - 3,340
                         walleye                             8 - 3,000
                       northern pike                          100 -4,400
                      channel catfish                          1 - 2,570
                      bluegill sunfish                          1 -1,680
                       common carp                           1 -1,800
                       white sucker                           2-1,710
                	yellow perch	10-2,140	

In 1999, May et al. (2000) collected 141 samples offish from reservoir and stream areas in the
Bear and South Yuba River watersheds in the Sierra Nevada of Northern California (Table 9-15).
Fish concentrations in the California survey ranged from 20 |ig/kg to 1,500 |ig/kg
(Table 9-15). Rainbow trout mercury concentrations in fillets ranged from 45 - 150 |ig/kg
(Table 9-27). Channel catfish mercury concentrations ranged from 240 - 280 |ig/kg
(Table 9-27).
                                             9-185

-------
                Table 9-15. TJSGS survey of mercury concentrations in fish tissue from
                reservoirs and streams in Northern California. (Source: May et al, 2000).
                Fish were fillets without skin
                            Reservoir
  us/kg
                         largemouth bass
                        Reservoir sunfish
                         channel catfish
                            Streams
20 - 1,500
< 100-410
 160 - 750
  us/kg
                           Brown trout
                          rainbow trout
  20 - 430
  60 - 380
Several recent surveys in Washington measured concentrations of mercury in resident fish species
(Table 9-16). The walleye samples from our study were within the range of the samples from
Munn and Short (1997) and Munn (2000). Smallmouth bass from our study were within the
range of the studies by Munn et al. (1995) and Sedar et al. (2001) although the maximum
concentrations in our smallmouth bass were lower than the levels found in Lake Roosevelt,
Washington (Munn et al.,1995) and Lake Whatcom (Serdar et al., 2001).  Serdar et al., (2001)
reported a mean concentration of (70 |ig/kg) in most fish species in Washington State.  The
authors found higher concentrations of mercury in 6 of 8 fillets with the skin off. In our study all
the fillets, except white sturgeon, were analyzed with skin.  There was also no consistent pattern
between fillets with skin or whole body. Rainbow trout concentrations from our study were also
within the range observed in rainbow trout from Lake Roosevelt, Washington, although the
maximum was lower than the maximum observed in Lake Roosevelt (Munn et al,  1995).

  Table 9-16. Mercury concentrations in fish fillets collected in Lake Whatcom and Lake Roosevelt,
  Washington compared to our study of the Columbia River Basin.
Fish species
walleye
walleye
walleye
smallmouth bass
smallmouth bass
smallmouth bass
rainbow trout
rainbow trout
perch
kokanee
pumpinkinseed
cutthroat trout
brown bullhead
Tissue Tvpe
composite
individual
composite
composite
individual
composite
individual
composite
individual
individual
individual
individual
individual
ug/kg
110-440
110-150
160 - 200
160 - 620
100 - 1840
380-470
110-240
45-150
120 - 290
100-130
70 -120
60-80
70 - 440
N
34
8
3
5
96
3
6
7
30
30
30
30
30
Location
Lake Roosevelt, 1994
Lake Roosevelt, 1998
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Lake Whatcom, 2000
Columbia River Basin, 1996-1998
Lake Roosevelt, 1994
Columbia River Basin, 1996-1998
Lake Whatcom, 2000
Lake Whatcom, 2000
Lake Whatcom, 2000
Lake Whatcom, 2000
Lake Whatcom, 2000

Munn and Short 1997
Munn 2000
our study
Munnetal., 1995
Serdar etal, 2001
our study
Munnetal., 1995
our study
Serdar etal., 2001
Serdar etal., 2001
Serdar etal., 2001
Serdar etal., 2001
Serdar etal., 2001
 N= Number of samples

9.1.13   Nickel

Nickel occurs naturally in rocks and soils and can leach into aquatic environments.  However,
weathering of nickel-containing substrates results in only small amounts of nickel entering into
aquatic systems. Manmade sources of nickel include mining, combustion of coal, petroleum and
tobacco, manufacture of cement and asbestos, food processing, textile and fur fabrication,
                                             9-186

-------
laundries, and car washes (USEPA, 1983).  The National Academy of Sciences reports that fish
contain nickel at a maximum of 1,700 |ig/kg (NAS, 1975).

Nickel concentrations the maximum nickel concentration was 17,000 |ig/kg in a whole body
steelhead sample from the Klickitat River (study site 56).  This sample was an anomaly since the
other samples from this site were 170 and 520 |ig/kg.  The average concentrations in fillet
samples ranged from 15 |ig/kg in Pacific lamprey to 260 |ig/kg in walleye; whole body ranged
from 50 |ig/kg in eulachon to 1200 |ig/kg in Coho salmon.

9.1.14   Selenium

While selenium is  ubiquitous in the earth's crust, only trace levels normally occur in aquatic
environments. Selenium enters aquatic habitats from a number of anthropogenic and natural
sources. Elevated levels in aquatic systems are found in regions where soil is selenium-rich or
where soils are extensively irrigated (Dobbs et al, 1996).  As an essential micronutrient, selenium
is used by animals for normal cell functions. However, the difference between useful amounts of
selenium and toxic amounts is small. Selenium at low levels in the diet is an essential element for
humans.  At elevated dose levels, it exhibits toxicity (selenosis).  Organic and reduced forms of
selenium (e.g. seleno-methionine and selenite) are generally more toxic and will bioaccumulate
(Besser et al., 1993; Kiffney and Knight, 1990).  Bioconcentration of selenium may be modified
by water temperature, age of receptor organism, organ and tissue specificity, and mode of
administration (Eisler, 1985a). Fish bioconcentrate selenium in their tissues with particularly
high concentrations observed in ovaries when compared to muscle tissues (Lemly, 1985;
Hamilton et al., 1990) and milt (Hamilton and Waddall, 1994). Selenium that is bioconcentrated
appears to occur in its most harmful concentrations in predator species such as chinook salmon
(Hamilton et al., 1990). Bioconcentration factors (BCFs) in rainbow trout range from 2-20 after
exposure to 220-410 jig/L selenium.  The magnitude of the BCFs appeared to be inversely related
to exposure concentrations (Adams and Johnson, 1977). Biomagnification of selenium has also
been well documented.  The magnitude of the biomagnification ranges from 2-6 times between
producers and lower consumers (Lemly and Smith, 1987). Piscivorous fish accumulate the
highest levels of selenium and are generally one of the first organisms affected by selenium
exposure, followed by planktivores and omnivores (Lemly,  1985).

Selenium has been frequently detected in a great variety of commonly consumed foods. In a
1997 UK diet study the mean selenium concentrations in the viscera and trimmings was estimated
to be 490 |ig/kg and 250 |ig/kg in nuts (Ysart et al., 2000).  Meat products (15%), fish (13%), and
bread (13%) groups make the greatest contributions to diet (Ysart et al., 2000).

In the US infant diet the average concentration of selenium was highest in grains and cereals
followed by fish (Table 9-17).
                                            9-187

-------
                   Table  9-17.    Selenium concentrations in  US  infant diet.  (Source:
                   Gartrell et al.. 1985 and 1986).
Food Group
other dairy products
potatoes
beverages
whole milk
vegetables
sugars and adjuncts
oils and fats
meat, fish and poultry
grain and cereals
1979 jig/kg
2
2
2
4
4
11
12
107
156
1981-1982 ng/kg
15
2

9
7

5
112
192
Selenium is well known to accumulate in living tissues.  Selenium has been found in marine fish
meal at levels of about 2,000 |ig/kg, which is about 50,000 times greater than the selenium levels
in seawater (Wilbur, 1980).  Table 9-18 is a list of selenium concentrations in a variety offish
tissue types.

   Table 9-18. Concentrations of selenium in fish reported in the literature.
   Fish type
              Location and date
                           Reference
   Razorback sucker eggs
   largemouth bass and bluegills
   gonads
   rainbow trout, edible portion
   northern pike, edible portion
   freshwater fish


   brown trout liver

   carp liver

   white sucker liver

   lake trout

   walleye and splake /backcross lake
    trout
   walleye and splake /backcross lake
    trout

   carp
Mean
3,700 -10,600
2,630 - 4,640

270
250
Geometric
mean
560
460
470
6,290

8,130

17,900

500 to 860

650 to 790

700 to 790

Maximum
3,650
Utah (1992)
power plant cooling reservoirs
(1994)
Toronto Harbor, Canada 1980
Toronto Harbor, Canada 1980
Hamilton and Waddell, 1994
Baumann and Gillespie, 1986

Davies, 1990
Davies, 1990
112 selected US monitoring
stations during from 1976-
1979
South Platte River Basin in
1992-93
South Platte River Basin in
1992-93
South Platte River Basin in
1992-93
Lake Huron from 1980 - 85

Lake Huron 1980- 85

Lake Huron 1979 and 1985,


Colorado River 1978-79,
Loweetal., 1985


Heiny and Tate, 1997

Heiny and Tate, 1997

Heiny and Tate, 1997

Great Lakes Water Quality
Board, 1989
Great Lakes Water Quality
Board, 1989
Great Lakes Water Quality
Board, 1989

Loweetal., 1985
The average concentrations of selenium in our study ranged from 220 |ig/kg in a rainbow trout
fillet to 1,100 |ig/kg in the white sturgeon fillet (Table 2-14). The maximum concentration
(2700 |ig/kg) was in a white sturgeon fillet sample from the Hanford Reach of the Columbia
River (study site 9U).
                                                9-188

-------
9.1.15   Vanadium

Vanadium is found in vegetables from about 0.5 to 2 |ig/kg, with an average of about 1 |ig/kg
(Beyerrum, 1991).  Veal and pork have been found to contain about 0.1 |ig/kg. According to
ATSDR (1992), foods containing the highest levels of vanadium include ground parsley, 1,800
jig/kg; freeze-dried spinach, 533 - 840 |ig/kg; wild mushrooms, 50 - 2,000 |ig/kg; and oysters,
455 |ig/kg.  Intermediate levels are found in certain cereals, like maize (0.7 |ig/kg), and
Macedonian rice 30 |ig/kg).  Also vanadium has been found in beef at 7.3 |ig/kg, and in chicken
at about 38 |ig/kg.  Seller and Sigel (1988) indicate that beverages, fats, oils, and fresh fruits and
vegetables contained the least vanadium, ranging from less than 1 to about 5 |ig/kg. Grains,
seafoods, meats, and dairy products were generally from about 5 to 30 |ig/kg. Prepared food
ranged from 11 to 93 |ig/kg, and dill seed and black pepper contained 431 and 987 |ig/kg
vanadium, respectively.  ATSDR (ATSDR, 1992) indicates that in general, seafoods have been
found to contain somewhat higher levels of vanadium than do tissues from terrestrial animals.

Mackeral has been found to contain about 3.5 |ig/kg of vanadium, with 28 |ig/kg in freeze-dried
tuna (ATSDR, 1992).  Konasewich et al. (1978) found vanadium in whole-fish samples of burbot
and bloater chub taken from Lake Huron at concentrations of 75 |ig/kg and 260 |ig/kg,
respectively.  The same authors also found vanadium in whole samples of lake trout from Lake
Superior, at 85 |ig/kg.  Nakamoto and Hassler (1992) found vanadium in the carcasses of male
and female bluegill taken from the Merced River and the Salt Slough, California, at mean
concentrations of 2,200 and 1,700 |ig/kg, respectively.

In our study the average vanadium concentrations ranged from 5 |ig/kg in fillet samples of spring
chinook salmon and walleye to 310 |ig/kg in whole body largescale sucker.  The maximum
concentration (770 |ig/kg) was in a whole body rainbow trout composite sample from the
Umatilla River (study site 101).

9.1.16   Zinc

Zinc occurs naturally in the earth's crust at an average concentrations of about 70,000 |ig/kg.  It is
introduced into aquatic systems via leaching from igneous rocks. Zinc is found in all living
organisms and is an essential element for growth, development and reproduction. However
aquatic animals tend to accumulate excess zinc which can result in growth retardation,
hyperchromic anemia, and defective bone mineralization.  Because zinc combines with
biomolecules in target species and most of these species accumulate more than they need for
normal metabolism, data showing bioconcentration factors for target receptors may be
misleading. Bioconcentration factors (BCFs) reported by EPA ranged from 51 in Atlantic
salmon (Salmo salar) to 1,130 for the mayfly (Ephemerella grandis) (USEPA, 1987c). Little to
no evidence exists indicating the successive biomagnification of zinc in tissues offish and avian
receptors (USEPA, 1987c).

In the ATSDR survey of food groups the levels for zinc ranged from 29,200 |ig/kg in
fish/meal/poultry to 2,300 |ig/kg in leafy vegetables (Table 9-19).
                                            9-189

-------
             Table 9- 19. Concentrations of zinc in food groups.  (Source: ATSDR, 1993)

             Food Group        Mg/kg              Food Group
             meat/fish/poultry   29,200              dairy products    4600
             grain/cereals       8,700               legumes         8300
             legumes          8,300               leafy vegetables   2300
The average concentrations of zinc in whole body fish tissue from our study ranged from
3800 |ig/kg in the white sturgeon fillet to 30,000 |ig/kg in the whole body coho salmon
(Table 2-14). The maximum concentration (40,000 |ig/kg) was in the whole body mountain
whitefish from the Deschutes River (study site 98).

9.2    Comparisons By Fish Species

This section includes general descriptions of each of the chemicals measured in this study
followed by brief comparisons of these chemicals with data reported in databases or other studies.
More information about each chemical is provided in Appendix C (Toxicity Profiles).  In addition
to chemical descriptions, this section includes a summary of the life history of the fish species.
This brief discussion of the habitat preferences and feeding habits is intended to provide some
understanding of how the fish may be exposed to pollutants. Appendix B (Fish Life Histories)
contains detailed information on each fish species.

The chemical levels measured in fish tissue from our study in largescale and bridgelip sucker,
mountain whitefish, rainbow trout, channel catfish, smallmouth bass, fall and spring chinook, and
coho were compared with levels reported in 4 databases and two other similar studies in the
Columbia River Basin.  Only those concentrations which had more than a 10 fold difference are
discussed.

Information on white sturgeon, walleye, steelhead, eulachon, and Pacific lamprey was not found
in these databases or reports. However their life histories and a synopsis of the literature
information described in Section 9.1 are added to this section to complete the summary for all
species from this study.

The 4 databases were developed by:

       1) the USGS, National Contaminant Biomonitoring Program (NCBP) database
        (Schmitt et al., 1999a),

       2) the USGS, Biomonitoring of Environmental Status and Trends (BEST) database
       (Schmitt et al., 1999b)

       3) the State of Washington, Puget Sound Ambient Monitoring Program (PSAMP) (West
       et al., 2001 and

       4) EPA's 1994 survey of literature reports on chemical data from the Columbia River


                                            9-190

-------
       Basin (USEPA 1994d)
The NCBP database includes data on persistent organochlorine insecticides, industrial chemicals,
herbicides, and potentially toxic contaminants that may threaten fish and wildlife resources
(Schmitt et al, 1999a). The NCBP database, from the early 1960's through 1986, contains
measured values of average whole-body composite fish samples where each composite sample
was comprised of five individual fish samples.

The  BEST database includes data from the smallmouth bass sampled from the Mississippi River
drainage during August-December 1995 (Schmitt et al., 1999b).  Fish tissue data consisted of
whole body composite samples, where, ideally, each composite sample consisted of 10 individual
fish samples.

The PSAMP database consists of measured chemical concentrations in fillet (without skin)
composites of adult chinook and coho salmon (West et al., 2001). Composite samples include 2-
5 individual fish, with five individual fish per composite being the most common.

EPA's 1994 database includes a compilation  of data from 1984 to 1994 on chemical
concentrations in fish tissue and sediments from the Columbia River Basin.  The information in
the database includes individuals and agencies contacted, data sources, abstracts for contaminant
studies, and an overview  of future or ongoing studies (USEPA, 1994d).

The data from  two surveys of chemicals in fish from the Columbia River Basin were also
compared to fish tissue residues from our study:

       1) The Lower Columbia River Bi-State Water Quality Program (Tetra Tech, 1996) and

        2) Willamette River Human Health Technical Study (EVS, 2000)

The Lower Columbia River Bi-State Water Quality Program (Tetra Tech, 1996) characterized
potential human health risks associated with consuming  fish from the lower Columbia River,
below the Bonneville Dam. The Bi-State study was conducted during two periods: 1991-1993
and 1995.  Data from 1991-1993 consisted of data that measured chemical contaminant
concentrations in fillet tissues of five different  resident target fish species (largescale sucker, carp,
peamouth, white sturgeon, and crayfish). Five individual fish were composited to form single
composite samples. Data from 1995 included measured chemical concentrations in fillet fish
tissue from  largescale sucker, smallmouth bass, chinook salmon, and coho salmon. Fish tissue
data for these species consists of range and mean data from three composite samples where each
sample was made up of eight fish.

The Willamette River Human Health Technical Study (EVS, 2000) included data from four fish
species of which smallmouth bass and largescale sucker were used for comparisons with our
study. Data were compared for both fillet with skin and whole body tissue. All samples from the

                                           9-191

-------
Willamette study were composite samples formed by homogenizing tissue from five to eight
individual fish.
9.2.1  Largescale Sucker (Catostomus macrocheilus) and Bridgelip Sucker (C. columbianus)

The largescale sucker is native to the Pacific Northwest in tributaries to the Pacific Ocean from
the Skeena River in British Columbia to the Sixes River in Oregon (Scott and Grossman 1973).
Largescale suckers are abundant throughout the Columbia River and are the most common
resident fish species collected in the Hanford Reach (Gray  and Dauble 1977).

Dauble (1986) found that algal periphyton was the major food item for fry, juvenile, and adult
largescale suckers in the Columbia River.  The stomachs of adults may also contain crustaceans,
aquatic insect larvae, snails, fish eggs, sand, and bottom debris (Dauble  1986, Scott and Grossman
1973). Stream fish appear to feed upon more algae, diatoms, and aquatic insect larvae other than
Chironomidae, whereas lake fish include Amphipoda and Mollusca (Carl 1936).

The bridgelip  sucker is found in the Fraser and Columbia river basins from British Columbia to
southeastern Oregon, including the Harney basin, below Shoshone Falls in the Snake River,  and
in northern Nevada (Scott and Grossman 1973, Lee et al.  1980). Throughout its range in coexists
and hybridizes with the largescale sucker (C. macrocheilus) (Dauble and Buschbom 1981).

The life history and behavior of the bridgelip sucker are poorly understood. According to Scott
and Grossman (1973), this fish usually inhabits small, swift, cold-water rivers with gravel to
rocky substrates, whereas Wydoski and Whitney (1979) report it inhabits quiet backwater areas or
the edges of the main current of rivers with sand or mud bottoms. In the Yakima River, Patten et
al. (1970) found this fish in warm flowing waters. In the mid Columbia River during the day,
Dauble (1980) found that subadult and adult bridgelip suckers were common in the tailouts of
pools, at the end of riffles, and above boulders in the main current.  At night, these fish were more
abundant near shore in flowing water 0.6 to 1.5 m deep.

The diet of C. columbianus is almost entirely periphyton during all seasons. This fish has an
expanded cartilaginous lower lip on its mouth that enables it to efficiently crop algae attached to
the bottom. However, like almost all other suckers, this species also feeds to some extent on
aquatic insect larvae and crustaceans (Dauble 1978, Wydoski and Whitney 1979). Mammals and
some birds prey on this species (Scott and Grossman 1973).

Chemical concentrations in largescale sucker fish tissue were compared for arsenic, cadmium
copper, mercury, lead, selenium, zinc, p,p'-DDE, p,p'-DDT, Aroclor 1254,  and Aroclor 1260
were compared data in the NCBP databases and the Bi-State and Willamette River studies  (Table
9-20a).

While the metal concentrations in largescale sucker from our study were within the range of the
other studies and databases examined, the maximum concentrations of metals were higher or

                                            9-192

-------
lower depending on the chemical (Table 9-20a).  Cadmium concentrations were 25 times higher
in our study than in the Willamette River study and National NCBP database.  Lead in largescale
sucker from our study was 9 times higher than in largescale sucker from the NCBP National
database.

The organic chemical comparisons in largescale sucker were also quite variable (Table 9-20a).
With exception of the Aroclors the organic chemical concentrations in our study were all within
the range of the other databases and studies.  However, the maximum concentrations were
different. The maximum concentration of p,pDDE in largescale sucker was 9 times higher in our
study than in the Bi-State study, and 14 times higher than in the NCBP Columbia River station
98.

The maximum Aroclor 1254 concentrations in largescale sucker were higher in the Columbia
River NCBP stations (from 8x to 46x) than in our study. The detection limits  were too high in
the National NCBP database to discern a difference in Aroclor 1254 and our  study.

With the exception of cadmium, the Willamette River study results for metals and organic
chemicals were similar to our study.

 The concentrations of chemicals in bridgelip  sucker were within the range found in largescale
sucker, except the largescale sucker had higher maximum concentrations (Table 9-20a,b).
                                           9-193

-------
Table 9-20a. Comparison of chemical concentrations in composites samples of whole body
USGS- NCBP- Columbia River
Station


Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
Columbia
(46)


Vg/kg
<50 - 870
<50 - 160
850-1340
90 - 390
50 - 320
60 - 430
20 - 2000
10 -270
100-2100
100-700
Columbia
(47)

range
Hg/kg
130-290
<50 - 600
1070 - 1283
100 - 520
<10-160
60 - 386
20-1100
10-430
5 - 3000
<5 - 100
Columbia
(98)

range
Vg/kg
111-333
50-410
720-1150
160-2570
20-130
190-250
10-90
10-70
100 - 600
100 - 300
Basin
Snake
(41,42,96)

range
vg/kg
<50 - 260
<50 - 260
490- 4318
10-290
10-230
170-450
50 - 560
10 - 440
<5 - 500
<5 - 300
USGS- NCBP
National

range
vg/kg
40 - 270
<5-9
600-1010
20 - 120
10 - 370
80 - 340
10 - 970
10-190
<100
<100-300
largescale sucker.




EPA
Willamette
single
comnosite
vg/kg
120
10
1780
37
121
ND
835
190
53
36
Bi-State

mean
[ig/kg
8
37
912
171
122
132
59
10
176
35

max
Vg/kg
385
66
1230
860
264
260
150
56
270
1300
Our study

ave
Hg/kg
160
55
1400
170
130
310
370
33
30
38

range
vg/kg
74- 320
13-250
800-5600
27-1100
<58-250
<1 80-500
28-1300
<1-180
<14-65
<12-100
Min= minimum; Max. = maximum, Ave = average < = detection limit
NCBP = USGS National Contaminant Biomonitoring Program 1969-1986. Range of average whole body composites. Station numbers are in parentheses.
Willamette = composites without replication, EVS, 2000.
Bi-State = whole body concentrations of fish collected during 1991-1993 from the lower Columbia River, below Bonneville Dam. Mean and maximum (max) TetraTech, 1996
EPA- Our study = range of composite fish samples from sites in the Columbia River Basin.  See table 1-1 and 1-2 for description of sites.
                                                                                            9-194

-------
Table 9-20b . Comparison of ranges of chemical concentration in composite
Station
Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p"-DDE
p,p"-DDT
PCB1254
PCB1260
USGS - NCBP- Columbia River Basin
Salmon (43) Snake (96) Columbia (98)
Hg/kg ng/kg ng/kg
160-330
20-50
680 - 1900
100 - 220
40-80
200 - 470
10-30
<10-20
<100
<100
No Data
No Data
No Data
No Data
120
No Data
340 - 440
190-200
<100-500
<100
180-270
70 - 280
No Data
530 - 1000
20-70
200 - 260
<10-40
<10-40
<100
<100-4800
samples of whole body bridgelip sucker.
NCBP
National
US/kg
60
<50 - 60
No Data
<100-110
80 - 160
No Data
200 - 350
180 - 380
1000-2800
No Data
EPA
Our Study
Vg/kg
260 - 300
22-32
880 - 1800
37-78
<40 - 53
280
310-560
37-52
18-32
27-49
       < = detection limit
       NCBP = USGS National Contaminant Biomonitoring Program 1969-1986 Range of average whole body composites. Station numbers
       are in parentheses.
       EPA- Our Study = range of composites from the Yakima River (study site 48).

9.2.2  Mountain Whitefish (Prosopium williamsoni)

The mountain whitefish is native to cold water rivers and lakes in western North America, both
east and west of the Continental Divide (Scott and Grossman 1973).  Seven-year old fish range in
length and weight from 307 to 387 mm and from 475 to 890 g, respectively, while the ranges for
8-year old fish are 330 to 410 mm and 501 to 944 g (Scott 1960, Pettit and Wallace 1975,
Thompson and Davies 1976). Mountain whitefish feed primarily on immature forms of bottom-
dwelling aquatic insects such as Diptera (true flies and midges), Trichoptera (caddisflies),
Ephemeroptera (mayflies), and Plecoptera (stoneflies) (Wydoski and Whitney 1979, Cirone et al.
2002).

The ranges of chemical concentrations in the whole body mountain whitefish, from the present
study were compared with mountain whitefish data from the NCBP database (Table 9-21). There
was no consistent pattern between the metal concentrations in our study of mountain whitefish
and NCBP database (Table 9-21).  The maximum arsenic and cadmium levels were similar in our
study and the NCBP database. The maximum copper concentrations in mountain whitefish in our
study were 6 to 9 times higher than the concentrations in the NCBP database. Lead
concentrations were higher in the NCBP database. The maximum mercury levels measured in the
Salmon River in NCBP database were higher than the levels measured in our study; the levels in
the NCBP Snake River mountain whitefish were lower. The maximum selenium concentrations
were lower in the NCBP database than in our study.

The maximum p,p' DDE concentrations in mountain whitefish in our study were 700 times higher
than the concentrations in mountain whitefish from the NCBP Salmon River station.  The Aroclor
concentrations were not comparable because of the higher detection limits in the NCBP database.
                                            9-195

-------
Table 9-21. Comparison of ranges chemical concentrations in composite
samples of whole body mountain whitefish.
USGS -NCBP - Columbia River Basin
Station
Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
Salmon f43)
US/kg
120
40
840
100
290
680
<10
20
<100
<100
Snake (96)
US/kg
No data
No data
590
103
65
472
590
30
100
100
Columbia (97)
US/kg
No data
No data
No data
No data
190
No data
1410
350
<100
100
EPA
Our Studv
fig/kg
120-180
<4-54
620 - 5000
10-72
<47-130
590 - 1800
13-770
<2-49
<21 - 140
<18-130
      < = detection limit
    NCBP = USGS National Contaminant Biomonitoring Program 1969-1986. Range of average whole body composites. Station numbers
       are in parentheses.
    EPA- Our Study = range of composite fish samples from sites in the Columbia River Basin. See table 1-1 and 1-2 for description of sites

9.2.3  White Sturgeon ( Acipenser transmontanus)

White sturgeon is native to the Pacific Northwest where it has evolved life history characteristics
that have allowed them to thrive for centuries in large, dynamic river systems containing diverse
habitats.  These characteristics include opportunistic food habits, delayed maturation, longevity,
high fecundity, and mobility (Beamesderfer and Fair 1997).  White sturgeon may attain lengths
and weights of more than 6 m and 580 kg, respectively, during a life span of over 100 years (Scott
and Grossman 1973).  White sturgeon body weight ranged from 9 to 34 kg.

White sturgeon take advantage of scattered and seasonal food sources by moving between
different riverine habitats. They feed on a wide range of food items including zooplankton,
molluscs, amphipods, aquatic larvae, benthic invertebrates, and fish (McCabe et al. 1993). White
sturgeon are more predaceous than any other North American sturgeon (Semakula and Larkin
1968) and can capture and consume large prey (Beamesderfer and Fair 1997). Seasonal
migrations occur in the Lower Columbia River where sturgeon move to feed on eulachon
(Thaleichthys pacificus\ northern anchovy (Engraulis mordax\ American shad (Alosa
sapidissima), moribund salmonids, amphipods, and other invertebrates (DeVore et al. 1995).

Concentrations of the Aroclors and 2,3,7,8-TCDF and in white sturgeon from our study of the
Columbia River Basin were higher than the EPA 1994 (USEPA,  1998c) studies of Lake
Roosevelt, Washington (Tables 9-9 and 9-10).

9.2.4  Walleye (Stizostedion vitreum)

The original range of the walleye generally east of the Rocky Mountains was expanded when it
was introduced to the Columbia River below Roosevelt Dam in the 1940's or 50's (Wydoski and
Whitney 1979).  This species shows a preference for large,  semi-turbid waters, but is capable of
inhabiting a large range of physical  and chemical conditions (Colby et al. 1979).
                                             9-196

-------
Feeding usually occurs near or at the bottom, and walleye may move into shallow water to feed.
Walleye fry feed on rotifers, copepods, and cladocerans.  Juvenile and adult walleye are largely
piscivorus, but invertebrates (e.g., mayfly nymphs and amphipods) may be a large part of their
diet in the late spring and early summer. Cannibalism is common with this species (Colby et al.
1979, Eschmeyer 1950). Prey for this species in the Columbia River includes mainly  cottids,
cyprinids, catostomids, and percopsids; out migrating juvenile salmonids were a smaller part of
their diet (Zimmerman 1999).

Adult walleye are not usually preyed upon by other fish. However, in its native range  northern
pike and muskellunge do prey on this fish (Colby et al. 1979). They are also probably preyed
upon by fish eating birds and mammals (Sigler and  Sigler 1987).

The maximum concentration of Aroclors 1254 and 1260 and 2,3,7,8-TCDF  in walleye were lower
in our study of the Columbia River Basin than levels found in surveys of Lake Roosevelt,
Washington, (USEPA, 1998c; Munn, 2000) (Tables 9-9 and 9-10).

9.2.5  Channel catfish (Ictalurus punctatus)

The original  range of the channel catfish, east of the Rock Mountains was expanded when it was
introduced to Idaho waters in 1893, but the date of its introduction to Washington waters is
unknown (Wydoski and Whitney 1979, Simpson and Wallace 1982).

Young channel catfish tend to feed primarily on aquatic insects and bottom arthropods, but after
attaining about 100 mm in length they are usually omnivorous or piscivorus (Carlander 1969).
Adult channel catfish consume a wide variety of plant and animal material including clams,
snails, crayfish, pondweed, and small terrestrial vertebrates (Eddy and Underbill 1976, Moyle
1976).

Young channel catfish are prey to a variety of fishes and piscivorus birds but the adults, due to
their size and bottom occurrence, are probably free of predation (Scott and Grossman 1973,
Schrammetal. 1984).

The concentrations of chemicals measured in channel catfish our study were compared to levels
reported in the NCBP database (Table 9-22). The concentrations of metals were higher in the
National and Columbia Basin NCBP databases with two exceptions. The maximum
concentrations of arsenic and selenium concentrations in channel catfish were 10 times higher in
our study than the NCBP Willamette station.  The concentrations of the following metals were
higher in the NCBP national database:  cadmium  29x , lead 60x, mercury 14x, and selenium 4
times higher.

The concentrations of organic chemicals were higher in the NCBP National database than in our
study. The maximum concentrations of the following chemicals in channel catfish from the
National NCBP database were higher than the levels in channel catfish in our study: p,p'DDE
47x, p,p'DDT 166x, Aroclor 1260 672x, and Aroclor 1260 42 times higher.  The concentrations


                                            9-197

-------
of p,p' DDT in the NCBP Columbia Basin stations were 5-23 times higher than in our study.
The maximum concentrations of Aroclor 1254 in channel catfish was from the NCBP Columbia
Basin Stations were 24 to 76 times higher than in our study.
Table 9-22. Comparison of ranges of chemical concentrations in whole body channel
catfish tissue from our studv with the TJSGS-NCBP database.
Station

Chemical
Arsenic
Cadmium
copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260

Willamette (45)
fig/kg
<50
<50
no data
100
290
60
570
<10-1050
4400
No Data
USGS - NCBP
Snake (96)
fig/kg
<50-610
<50
no data
<100-210
80 - 900
70-180
<10-1050
<10-220
<10 - 1400
<100-500

National
fig/kg
10-630
3-760
no data
30 - 2000
<10-4500
<50 - 2500
10-42300
<5 - 7500
<50 - 39000
<50 - 5900
EPA
Our Study
ave

230
17
510
21
210
500
570
21
38
77

fig/kg
110
13
410
12
140
410
280
0.8
25
32-


-430
-26
-590
-33
-320
-630
-900
-45
-58
-140
       *Samples are fillet with skin;     Ave= average
       NCBP = USGS National Contaminant Biomonitoring Program 1969-1986. Range of average whole body composites. Station numbers
       are in parentheses.
       EPA-Our Study = whole body composite samples from the Columbia River (study site 8) and the Yakima River (study site 48)
9.2.6  Smallmouth Bass (Micropterus dolomieu)

The range of the smallmouth bass, originally restricted to freshwaters of eastern-central North
American, was expanded by plantings in the Pacific Northwest in the late 1800s and early 1900s.
In Washington, smallmouth bass are most numerous in the Columbia and Snake rivers (Wydoski
and Whitney 1979, Simpson and Wallace 1982).

Smallmouth bass fry initially eat copepods and cladocerans and at lengths of 2 to 5 cm change to
a diet of insects and small fish (Hubbs and Bailey, 1938).  Tabor et al. (1993) found that
salmonids made up from 4 to 59% (by weight) and from 19 to 30% (by volume) of the diet of
samllmouth bass in the Columbia River Basin. The authors concluded that predation rates on
salmonids were high during the spring and early summer when subyearling salmon were
abundant and of suitable forage size and shared habitat with the smallmouth bass.

Smallmouth bass in the Columbia River grow at a rate equal to or better than that of bass from
other locations in the United States.  In a 1952 study, the weights and total lengths of the
Columbia River fish at age four were 510 g and 32 cm; age six, 794 g and 38 cm; age eight, 1,304
g and 43  cm; and at age ten, 1,814 g and 47 cm, respectively (Henderson and Foster 1957,
Wydoski and Whitney 1979).  The body weight of smallmouth  bass in our study ranged from
1300 to 1400 g.

Smallmouth bass from our study were compared to data reported in the BEST and NCBP
databases (Table 9-23).  The concentrations of all chemicals in smallmouth bass from the NCBP
National database were higher than in our study. In particular, Aroclor 1254 was higher (68x) in

                                            9-198

-------
the NCBP National database.  The Aroclor concentrations in Columbia River Basin NCBP
stations had higher detection limits than in our study.
Table 9-23. Comparison of ranges of chemical concentrations in whole body smaUmouth bass.
Chemical
Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
Yakima (44)
fig/kg
No data
No data
No data
No data
140 - 270
No data
940 - 1660
200 - 420
100 - 600
200
Snake (42)
fig/kg
50-60
10-50
380
<100
150-280
440
80 - 2540
80 - 170
<100
<100 - 800
USGS-NCBP
Salmon f43)
fig/kg
<30 - 50
6-60
1182
100-170
210-360
606 - 830
280 - 690
80 - 170
<50 - 400
<50 - 100
Willamette(45)
fig/kg
250
50
No data
120
130
No data
60
20
<400
<200
National
fig/kg
40 - 670
2-50
257-1950
10 - 320
60 - 1200
80 - 1260
10-950
<5 - 590
<50 - 6400
<50 - 1300
uses
BEST
fig/kg
<178-263
<36-43
445-591
8-100
80 - 280
203-491
10-65
10-84
No data
No data
EPA
Our Studv
fig/kg
160 - 170
5-19
500 - 560
10-140
220 - 360
480-710
970 - 1700
44-80
46-94
80-190
NCBP = USGS National Contaminant Biomonitoring Program 1969-1986. Range of average whole body composites. Station
numbers are in parentheses.
BEST = USGS Biomonitoring of Environmental Status and Trends Program - 1995 Fish Samples from the Mississippi Delta.
EPA- Our Study = whole body composite samples from the Yakima River (study site 48)

9.2.7  Rainbow and Steelhead (Oncorhynchus mykiss)

Oncorhynchus mykiss are native to the Pacific Northwest and appear in two forms: the resident
rainbow trout and the anadromous Steelhead, both of which occur in the Columbia River Bbasin.
It also has the greatest diversity of life history patterns of any Pacific salmonid species (Wydoski
and Whitney 1979, Pauley et al. 1986). This diversity includes degrees of anadromy, differences
in reproductive biology, and plasticity of life history between generations (Peven 1990, Busby et
al. 1996).

The diet of rainbow trout and juvenile Steelhead changes seasonally, depending on food
availability.  They may feed on aquatic insects, amphipods, leaches, snails, and fish eggs.  The
Steelhead's diet in the ocean includes crustaceans, squid, herring, and  other fish (Withler, 1966;
Wydoski and Whitney, 1979).  Adult non-migratory rainbow trout average 0.9 to 1.8 kg in weight
and usually have a life span of 5 to 6 years (Simpson and Wallace, 1982; Sigler and Sigler, 1987).
 Steelhead can achieve 9 years of age, weights of 16 kg, and lengths to 122 cm (Scott and
Grossman, 1973; Wydoski, and Whitney, 1979). The average body weight of rainbow trout in
our study ranged from 47 - 571g. The Steelhead average body weight ranged from 1633 to 6440g.

The chemical residues in rainbow trout measured in our study were compared to the NCBP
databases (Table 9-24).  The maximum concentration of p,p' DDE in rainbow trout was 300
times higher in the NCBP Columbia River Basin station  (Snake River) than in our study.

Steelhead concentrations of metals in fish tissue were within the range of rainbow trout (Table 9-
24).  The maximum concentrations of arsenic and lead were higher (4x and 2x respectively) in the
Steelhead, while p,p'DDE was lower in the Steelhead than the rainbow trout.
                                            9-199

-------
    Table 9-24. Comparison of ranges of chemical concentrations in composite samples of whole body
    rainbow trout

Station
Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
USGS-
Snake (41)
us/ks
<50 - 145
5-50
680-3130
9-100
30-130
220 - 540
80 - 25400
5-70
100-600
<50
NCBP
National
us/ks
<50 - 260
10-70
1130-4620
10-650
10-270
170 - 3000
10-140
5-40
<50 - 300
<50 - 100
EPA ( Our Study)
rainbow trout
us/ks
<50 - 560
<4-58
900 - 5000
<10-88
<33 - 380
230 - 790
3-84
<2-12
<10-20
<6-22
steelhead

290 - 1200
29-88
1900-6800
<10-360
<50 - 420
460 - 940
5-33

-------
The chemical concentrations in fall and spring chinook salmon from our study were similar to
each other with the exception of cadmium, lead, and mercury which were higher in spring
chinook (15x, 8x, and 5x, respectively; Table 9-25).
Table 9-25. Comoarison of chemical concentrations in chinook salmon fillet with skin.
Station
EPA
1994
Database
PSAMP
Bi-State
EPA
Our Study
fall chinook salmon
Chemical
Arsenic

Cadmium
Copper

Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
2,3,7,8-TCDD
2,3,7,8-TCDF
range
US/kg
20-1110

20-50
240-1900

20-40
62 - 164
360 - 370
no data
3
18-20
16-30
0.00014
0.0009
range
US/kg
570-
1600
No data
370-
1200
no data
58-160
no data
4-48
0.5-4
5-88
1-72
no data
no data
ave
Vg/kg
13

2
860

7
100
280
8.5
1.5
0.9
10
0.0002
0.0016
max
vg/kg
23

2.5
1010

10
130
340
11
3
0.9
15
0.0006
0.00027
ave
Vg/kg
810

<2
640

7
84
330
12
2.5
17
9.9
0.00002
0.00068
range
Vg/kg
530-1100

<4
540 - 760

<10-16
<50-150
280 - 380
4-26
<2-8
9-35
<19
<0.00001-0.00005
0.00003-0.0014
spring
ave
Vg/kg
850

2
790

14
100
350
12
4
16
11
0.00002
0.0006
chinook salmon
range
Vg/kg
560 - 1200

<4-15
240 - 1000

<10-140
<83-510
290-430
6-18
3-8
9-24
<18
<0.00001-0. 00005
0.0004-0.00074
Ave = average; max = maximum < = detection limit
EPA 1994 database = EPA survey of data from the Columbia River Basin from 1983-1994. Does not differentiate between spring and fall chinook
salmon
Bi-State = 1995 concentrations in fillets offish from the lower Columbia River, below Bonneville Dam. Does not differentiate between fall and
spring chinook salmon (Tetra Tech, 1996).
PSAMP = 1992-1995, data is for fillet without skin. Does not differentiate between fall and spring chinook salmon
EPA- Our study = range of composite fish samples from sites in the Columbia River Basin. See table 1-1 and 1-2 for description of sites

9.2.9   Coho Salmon (Oncorhynchus kisutch)

Coho salmon are one of the five Pacific salmon species in North America. The life span of most
coho is three years, during which they attain average weights ranging from about 3,000 to 6,000g
(Wydoski and Whitney 1979). The average body weight of the coho salmon in  our study was
2,855g to 3,960g.

The coho salmon fish typically spend up to 21 months in freshwater followed by  approximately
16 months in the ocean before returning to freshwater where they will spawn and die.  These fish
rarely feed on non-moving food or off the bottom in streams (Sandercock 1991). Juveniles
consume insects (larvae, pupae, and adults), worms, small fish, and fish eggs. In reservoirs, coho
juveniles feed primarily on zooplankton and emerging insects (Wydoski and Whitney 1979).

Samples of coho salmon from our study were compared to data from PSAMP and the Bi-State
study (Table 9-26).  The maximum concentrations of several chemicals were higher in coho
salmon from our study than the coho salmon from the Bi-State study: arsenic (85x), lead (25x),
and Aroclor 1254 (19x).
                                               9-201

-------
Table 9-26. Comparison of chemical concentrations in coho salmon fillet with skin.
Station

Chemical
Arsenic
Cadmium
Copper
Lead
Mercury
Selenium
p,p'-DDE
p,p'-DDT
Aroclor 1254
Aroclor 1260
2,3,7,8-TCDD
2,3,7,8-TCDF
PSAMP
range
tig/ks
570 - 1600
No data
410-1010
No data
58-160
No data
1.3-26
0.52-1.4
2-66
1-32
No data
No data
Bi-State
mean
l*g/kg
2.7
3
810
4
44
168
3
0.8
0.6
3
0.0003
0.0007
max
US/kg
1
5
850
9
48
188
5
1
0.9
4
0.0009
0.0009
EPA
ave
tig/ks
540

1700
81
120
290
33
2
16

0.000017
0.0005
- Our study
range
US/kg
450 - 600
<4
680 - 3600
<10-230
110-120
270-310
29-35
<2-4
12-19
<18
<0.00001 - 0.00004
0.0004 - 0.0005
 Ave = average; max = maximum; < = detection limit
 PSAMP = 1992-1995, data is for fillet without skin
 Bi-State = 1995 whole body concentrations offish from the lower Columbia River, below Bonneville Dam. (TetraTech, 1996)
 EPA - Our study = range of composite fish samples from sites in the Columbia River Basin. See table 1-1 for site descriptions.

9.2.10   Pacific Lamprey (Lampetra tridentatd)

The Pacific lamprey is a native anadromous fish with a widespread distribution in the Columbia
River Basin (Wydoski and Whitney 1979).

The adults overwinter in freshwater, do not feed during this time, and spawn the following spring
(Beamish 1980).  Larvae (ammocoetes) leave the gravel approximately 2 to 3 weeks after
hatching, drift down current, settle in slow back water areas, burrow in soft substrates with
organic debris, and take up a filter feeding existence (Fletcher 1963, Kan 1975). The ammocoete
life stage may range from 4 to 7 years, during which time they remain buried in the sediment
(Beamish and Levings 1991, Close et al. 1995). Ammocoetes are reported to feed on vegetative
material (Clemens and Wilby 1967), diatoms and desmids (Fletcher 1963), and detritus and algae
suspended above and within the substrate (Moore and Mallatt  1980).  Juvenile lampreys play an
important role in the diets of many freshwater fishes, including channel catfish, northern pike
minnow, and several species of cyprinids and cottids. Salmonid fry prey upon lamprey eggs, but
do not feed on the ammocoetes.  The larvae are also taken by several species of gulls and terns
(Fletcher 1963, Close et al.  1995).

Metamorphosis occurs from July to October.  Shortly thereafter, the downstream migration of
young adult lampreys begins usually at night and with an abrupt increase in river flow. Pacific
lampreys migrate to salt water where they take up a parasitic life, but feeding may start in
freshwater (Fletcher 1963, Beamish 1980,  Beamish and Levings 1991).

The ocean phase of the adult life cycle may last 3.5 years (Beamish 1980).  In ocean and estuarine
areas, adults are important prey for several  pinniped species. After entering the Columbia River
they become a prey item for white sturgeon (Wydoski and Whitney 1979, Roffe and Mate 1984,
Close et al.  1995).
                                             9-202

-------
There were no comparable studies of Pacific lamprey in the literature.

9.2.11 Eulachon (Thaleichthys paciftcus)

The eulachon occurs only on the west coast of North America, including the Columbia River
Basin (Scott and Grossman 1973).  This anadromous species spawns in the main channel of the
Columbia River and periodically in the Grays, Cowlitz, Kalama, Lewis, and Sandy Rivers (Smith
and Saafeld 1955).

It is believed that developing larvae do not to feed in freshwater, but rely on their yolk sac for
nourishment until they reach the ocean (Smith and Sallfeld 1955, Scott and Grossman 1973). At
sea, post-larval eulachon move into deeper water as they grow. They feed on plankton, mysids,
ostracods, copepods and their eggs, and barnacle, cladoceran, and polychaete larvae (Hart 1973).
Juvenile and adult fish feed primarily on euphausid shrimp, crustaceans, and cumaceans.  Adults
do not feed after they return to freshwater (Barraclough 1964).

As are other smelts, T. pacificus is a very important food item for a wide variety of predators.
Adults are fed on by many piscivorus fishes including Pacific salmon and white sturgeon, marine
mammals ranging from the harbor seal to the finback whale, seabirds, waterfowls, and gulls
(Scott and Grossman  1973). The larval and post larval stages contribute modestly to the diet of
small salmon off the Fraser River (Hart 1973).

There were no comparable studies of eulachon in the literature.

9.3 Comparisons across all species

9.3.1  Resident Fish

White sturgeon, mountain whitefish, whole body walleye, largescale sucker,  smallmouth bass,
and channel catfish had the highest concentrations of organic chemicals of all the species tested in
this study (Table 9-27a,b). Bridgelip sucker and walleye fillet samples had much lower chemical
residues,  similar to the salmonids and eulachon.

The largescale sucker was the fish species with the most frequent detection  of PAHs (Table 2-la).
The phenols were detected in only one white sturgeon sample from the main-stem Columbia
River (study site 8) (Table 2-la).

The basin-wide average concentrations of total DDT (Table 2-4) in the salmonids (chinook, coho,
rainbow trout, and steelhead ) and eulachon were much lower than, white sturgeon, mountain
whitefish, largescale sucker, and smallmouth bass. The maximum concentrations p,p'DDE was
found in whole body smallmouth bass followed by white sturgeon fillet, channel catfish fillet, and
whole body largescale sucker (Table 9-27a).

The white sturgeon, mountain whitefish, whole body walleye, and smallmouth bass had the

                                            9-203

-------
highest concentrations of Aroclors. The maximum concentration of TCDF was in the white
sturgeon (Table 9-27a,b). The next highest average concentration was in the mountain whitefish.

The maximum concentrations of metals (arsenic, cadmium, copper, lead, mercury, selenium)
were lower in the resident species than in the anadromous species, except for largescale sucker
which had the highest concentration of cadmium  (Table 9-27a,b). When doing a comparison of
fish tissue across all species it is important to not only consider the maximum concentrations but
also some measure of the variability. In this study, the average concentration is a measure of
variability.  While the maximum mercury and selenium concentrations were in the spring
chinook salmon, the basin-wide average concentrations of mercury were highest in the largescale
sucker, walleye,  and white sturgeon.

The higher concentration of organic chemicals may be  attributed to size in some species or lipid
content. The white sturgeon were some of the largest fish measured in the study. The samples
included only single fish.  It is also known to have  a very long life span. Thus, it is not clear
whether the high levels of organic chemicals in this fish may be due to an anomaly in the few fish
that were sampled, their size, or their age.

The association of organic chemical concentrations in the tissues of resident species and percent
lipid was not particularly evident in this study. There was an association with lipid in the white
sturgeon samples from one study site (study  site 6).  The difference in chemical content between
the whole body walleye and the fillet was also associated with lipid. However, there were no
other clear associations of whole body and fillet with lipid and organic chemicals in fish tissue.

There was an indication of high concentrations of organic chemicals in the resident fish collected
from the Hartford Reach of the Columbia River (study  site 9U).  However, there is no
information in this study to explain the levels in fish from this study site.

9.3.2   Pacific lamprey and eulachon

Of the anadromous fish species, Pacific lamprey had  maximum concentration of organic
chemicals (DDE and Aroclor 1254; Table 9-27b).  The high concentration of organic chemicals in
the Pacific lamprey may have been due to its high  lipid content.
The metals content of the Pacific lamprey was not consistent across different metals. For
example when compared to the other anadromous species, the arsenic concentrations were low
for Pacific lamprey while concentrations of copper, lead, mercury, and selenium were within the
range of the range of these other fish species.

While eulachon also had a high lipid content, they  had some of the lowest levels of organic
chemicals of all the species test. Aroclors and chlordane were not detected in the eulachon.
Eulachon had the highest average concentration of arsenic and lead.
                                             9-204

-------
9.3.3  Salmonids

The salmonids had the lowest concentrations of organic chemicals with a few exceptions.  There
were no semi-volatile chemicals detected in the fall chinook salmon or coho salmon tissue
samples. Pyrene was found at the highest concentrations of all the PAHs in a rainbow trout
collected from the upper Yakima River (study site 49).  The fillet or whole body samples of
rainbow trout, eulachon, and coho salmon had no detectable concentrations of any of the
chlordane compounds.

The concentrations of metals in the chinook salmon and steelhead were higher than the other
resident or anadromous  fish species.  Steelhead had the maximum concentration of arsenic.
When doing a comparison offish tissue across all species it is important to not only consider the
maximum concentrations but also some measure of the variability. In this study, the average
concentration is a measure of variability. Thus, while steelhead had the maximum concentration
of arsenic, the average concentrations were higher in eulachon, and chinook salmon (Table 2-14).
From this study, the salmon, steelhead, and eulachon had higher concentrations  of arsenic than
the resident species and  Pacific lamprey.  Fall chinook salmon had the maximum  concentration
of lead (Table 9-27b). The average concentrations of lead were highest in eulachon, fall chinook
salmon, and whole body walleye (Table 2-14).

Although the egg samples from the salmon and steelhead had high percent lipid, the concentration
of organic compounds was generally lower than the fish tissue of the anadromous  or resident fish
with a few exceptions. The highest concentrations of total chlordane were in egg samples from
the spring chinook salmon. The maximum  concentrations of copper and selenium  were in egg
samples from the salmon and steelhead (Table 9-27b).  The basin -wide average concentrations of
copper were highest in the egg samples from the salmon and steelhead followed by the whole
body Pacific lamprey. The basin-wide average concentrations for selenium were highest in
spring chinook salmon egg samples followed by white sturgeon and mountain whitefish.  The
high concentration of selenium may also be associated with the high percent lipid in the egg
samples.
                                            9-205

-------
Table 9-27a. Range of chemical concentrations in resident fish tissue samples from our study of the Columbia River Basin, 1996-1998.
Chemical
N-FS
N-WB
Arsenic

Cadmium

Copper

Lead

Mercury

Selenium

p,p'-DDE

p,p'-DDT

Aroclor 1254

Aroclor 1260

2,3,7,8-TCDD

2,3,7,8-TCDF

T


FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
FS
WB
largescale
sucker
US/kg
19
23
50 - 100
74- 320
<4 - 24*
13-250
430 -870
800 - 5600
10-140
27-1100
71 - 370
<58-250
130-400
<180-500
14 - 740
28-1300
<2 - 92*
<1 - 180
10-46
<14-65

-------
Table 9-27b. Range of chemical concentrations ( ug/kg) in anadromous fish tissue samples from our study of the Columbia River Basin.
N-Egg
N-FS
N-WB
Arsenic


Cadmium


Copper


Lead


Mercury


Selenium


p,p'-DDE


p,p'-DDT


Aroclor 1254


Aroclor 1260


2,3,7,8-TCDD


2,3,7,8-TCDF


T

E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
E
FS
WB
steelhead
1
21
21
ND
280-1500
290 - 1200
34
<4-9
29-88
18,000
540 - 940
1900 - 6800
41
<10-23*
<10 - 360
<43
70-210
<50 - 420
4500
<250 - 500
460 -940
7
5-28
5-33
<2

-------
10.0    Uncertainty Evaluation

There are many uncertainties in completing a survey of contaminants in fish tissue and in
estimating risks from consumption of these fish.  This section provides a summary of the
assumptions and uncertainties in evaluating the fish contaminant data and preparing the risk
assessment.  Some of the types of uncertainty which were encountered in this study include:

        1) errors in sampling, fish preparation, and chemical analysis,

        2) variability in fish tissue concentrations within fish, across species and tissue types, and
        among stations,

        7) lack of comparable data-sets for comparisons, and

        3) lack of knowledge regarding human exposure and toxicity.

10.1    Fish Tissue Collection

Uncertainty in toxic chemical levels is primarily associated with variability in fish tissue
concentrations over space and time as well as errors in chemical analytical methods.  The
temporal (seasonal, annual) range of chemical concentrations in fish species was not known.

There was some measure of spatial variability in certain fish species which were collected at a
number of sites (largescale sucker, white sturgeon, mountain whitefish, rainbow trout, chinook
salmon, steelhead, Pacific lamprey).  Coho salmon, bridgelip sucker, and eulachon were each
only collected at one location, therefore there was no measure of spatial variability in these
species. Pacific lamprey and walleye were only collected at two locations. Therefore, there were
gaps in our information on contaminant levels in these species from other sections of the
Columbia River Basin.  In addition to a limited number of sampling locations, some of the sites
included large stream reaches (Table 1-1). Therefore, the average concentrations from these sites
represent sampling areas of several miles.

Individual fish tissue were composited to obtain a representative sample of the mean
concentrations offish tissue. However, by compositing the  fish there is a loss of certainty in the
variance among individual fish samples. To  reduce some of the uncertainty associated with
composites,  an attempt was made to collect fish: 1) at the same time and 2) of the same size.

To maintain uniformity in sample size within composites the  smallest individual within a
composite was supposed to be no less than 75% of the total length of the largest individual.
Seventy-nine percent of the composites were within this guideline.  Of the composite samples not
meeting the guideline, roughly one-half were within 70% of the total length of the largest
individual. The compositing goals were not fully met in all samples because:
                                            10-208

-------
       1) larger fish (rainbow trout and mountain whitefish) were added to some composites to
       gain enough fish tissue for analyses,
       2) tribal members requested that small fall chinook salmon (jacks) be added to samples of
       larger adults, or
       3) spatial and temporal variability in fish species limited the number offish available for
       sampling.

To maintain uniformity across composites the relative difference between the average length of
the individuals in the smallest-sized composite (i.e., the one with the smallest average body
lengths) was to be within 10% of the average length of the largest-sized composite. Eighty-nine
percent of the composites were within the 10% guideline.  Of the 11% not meeting the guideline,
5 composites were steelhead, and one each were walleye, largescale  sucker, rainbow trout, and
spring chinook salmon.

In addition to collecting composites of the same size an attempt was made to collect replicate
samples at each study  site to provide a more accurate estimate of the variance in tissue analyses.
The goal of collecting  at least three replicate composite samples for each sample type from each
study site was met at 92% of the study sites.  Only two replicates or less were collected at 8% of
the study sites. Replication was limited at study site 30 on the Umatilla River because the
electro-fishing boat broke down, which prohibited additional collections of walleye and
largescale sucker.  There were a low number of rainbow trout available from study site 98 in the
Deschutes River.

The uncertainty in the tissue concentrations is also associated with the  sampling design.  The fish
type, tissue type, and sample location were all predetermined during the planning conference.
This type of sampling is biased with unequal sample sizes and predetermined sample locations
rather a random design. This bias is to be expected when attempting to provide information for
individuals or groups based on their preferences.  The results of this survey should not be
extrapolated to any other fish or fish from other locations.

EPA's guidance for preparing fish tissue for chemical analysis recommends scaling fish (USEPA,
2000f). However, CRITFC's member tribes do not typically scale their fish (CRITFC tribes,
personal communication). The results of some of the chemical analyses in this report may be
affected by the amount of certain chemicals (e.g. metals) which may be concentrated in the fish
scales.

The homogeneity of ground fish tissue can vary considerably, depending upon the nature of the
tissue sample and the grinding procedures. In this project we attempted to minimize variability of
chemical measurements by specifying the fish grinding procedure (See Volume 5) and by
monitoring the homogeneity of composite samples.

With the  exception of white sturgeon, fish tissue chemical residues were measured in fillet with
skin and whole body.  White sturgeon were the only species which were analyzed as fillet without
skin. As discussed in Section 2, whole body fish  tissue samples tend to be somewhat higher in


                                            10-209

-------
lipids than fillet with skin samples for some fish species.  This difference in lipids between whole
body and fillet fish samples was not consistent across species. This was not surprising since the
preparation of fillets with skin usually left a thin layer of subcutaneous fat remaining under the
skin.

The fillet and whole body samples were not from the same fish. Therefore, any comparisons
between them will be affected by the natural variability in fish samples as well as the tissue type.

10.2   Chemical Analyses

All data quality objectives established for this project were met.  However, there were
uncertainties in the chemical analysis due to interferences, detection limits, and method
development.

A number of problems were encountered in the measurement of target  compounds. For
dioxins/furans, dioxin-like PCBs,  non-acid labile chlorinated pesticides, and Aroclors, the
primary analytical problem encountered by the laboratories was the interference of chlorinated
and brominated non-target compounds in extracts of project fish samples.  For dioxin-like PCBs,
many sample extracts had to be diluted and re-measured because of high levels of dioxin-like
PCB target compounds in some samples.

The metallic equipment used to grind fish samples was tested prior to sample analysis for
possible interferences.  The results indicated that lead, manganese, nickel, copper, aluminum,
zinc, and PCB 105 were found in the rinsate blanks from the fish grinder.  The levels of
manganese, nickel, copper, aluminum, zinc, and PCB 105 were in negligible quantities and
should not affect the study results. However, the lead levels (77 jig/1) in the rinsate were higher;
therefore, the results reported in this study for lead may  be increased over levels that would be
found in tissue samples.

Modifications to digestion procedures for high levels of lipids in some project samples improved
measurements of metals and mercury using EPA methods 200.8 and 251.6. The chemical
analysis of chlorinated phenolics (EPA Method 1653) and neutral semi-volatiles (EPA Method
8270) had the largest number of data which were not acceptable due to  high quantitation limits.

For this project, analytical methods were chosen to provide detection or quantitation limits which
were as low as possible given available analytical methods and resources. The true value of
chemicals which were "not detected" is actually somewhere between the reported detection limit
and zero. For this study /^ the detection limit was used to estimate chemical concentrations.
Appendix E lists each chemical concentration as equal to:  1) the detection limit, 2) zero, and 3)
one-half the detection limit. The use of 1A the detection limit may have over or underestimated
the true fish tissue concentration.

In the quality assurance review of the chemical data, certain chemical concentrations were
qualified with a "J". The "J" qualifier designates a concentration which  is estimated. EPA

                                            10-210

-------
recommends that the J-qualified concentrations be treated in the same way as data without this
qualifier with acknowledgment that there is more uncertainty associated with "estimated" data
(USEPA, 1989). We chose to use these data in this assessment without conditions.  Use of this
data to calculate fish tissue concentrations may overestimate the true concentration since these
levels may be incorrect.  The data qualifiers are listed with each data point in Appendix D of
Volume 1 and in Volume 4.
The percent difference in field duplicates was estimated for all chemicals analyzed. There was
less than 10% difference between most of the duplicate samples. The samples with greater than
10% difference are shown in Table 10-1.  The maximum difference was 157% in cobalt
concentrations in fall chinook from study site 48 (Table 10-1). There was no consistent pattern of
error in field duplicate by study site, chemical, or fish species.

The difference in duplicate fillets from the same fish is an indication of the variability of
chemicals within fish tissue, since the fillets were from the opposite sides of the same fish.  In this
study, the duplicate values were averaged. By averaging the concentration of the duplicate
samples fish tissue concentrations and risk estimates may be lower than the  actual exposure that
would occur if the higher fish tissue concentration was used.
   Table 10-1. Percent difference in field duplicate samples from the Columbia River Basin.  Fish are
   listed with study site ID in parentheses. The maximum percent difference is given for the chemical
   within a chemical group.
Percent difference for analytes (greater than 10%)
Species (study sites)
steelhead (96)
spring chinook (94)
fall chinook (8)
fall chinook (48)

mountain whitefish (98)

white sturgeon (13)

white sturgeon (6)

white sturgeon (9)
Dioxins & Furans
46 (OCDD)
13 (HXCDF)

18 (TCDF)

29 (TCDD)

29 (HxCD)

57 (TCDF & HxCDF)

50 (OCDD)
Metals
68 (Ba)
62 (Cd)
29 (Hg)
107 (Cr);
157 (Co)
70 (Pb)

54 (Hg)

42 (Co)

144 (Co)
PCBs
56 (PCB 123)
17 (PCB 189)
14 (PCB 157)
28 (PCB 126);
18(Aroclorl254)
32 (PCB 167);
32 (Aroclor 1254)
15 (PCB 11 8);
11 (Aroclor 1260)
39 (PCB 105);
109 (Aroclor 1254)
27 (PCB 169)
Pesticides
67 (DDT)
15 (DDT)
11 (ODD)


35 (DDE)

124 (nonaclor)

119 (DDT)

59 (oxychlordane)
10.2.1 Lip id analyses

All samples were measured for percent lipids according to the procedure described in EPA
Method 1613B.  Other percent lipid procedures such as the three extraction methods described in
EPA Method 8290 would have produced different percent lipid results because of the different
extraction solvents used and different extraction conditions. While the lipid values reported in
our study were consistent because the analyses were all done within one laboratory using one

                                             10-211

-------
method, there would be considerable uncertainty in comparing the lipid levels measured in this
study with other data generated by different methods or different laboratories.

10.3   Comparing Chemical Data Across Fish Species and with Other Studies

The comparison of this study with other studies is confounded by the methods that were used to
collect the samples, the tissue type, number of samples, and species as well as the inconsistency
in chemical methods.  In particular, methods for analyzing fish tissue for dioxins, furans, and
PCB congeners have changed recently. Thus, chemical analysis offish tissue data for these
particular chemicals from the 1970's through the early 1990's will not necessarily give the same
results as were seen in this study.

10.4   Risk Assessment

Uncertainties can occur in all parts of the risk assessment-exposure assessment, toxicity
assessment, and risk characterization. An uncertainty evaluation has been done as a part of this
risk assessment to show how the risk characterization could be affected if alternative assumptions
had been made and/or different parameters had been used to calculate the cancer  risks and non-
cancer hazard indices.

10.4.1 Exposure Assessment

10.4.1.1  Contaminant Concentrations in Fish Tissue

As  discussed earlier in this report, the fish species collected and the sampling study sites selected
were based primarily on data from CRITFC's Fish Consumption Report (CRITFC, 1994) and
discussions with tribal staff. Although samples were taken from the study sites used most
frequently by the tribes, many other study sites used for fishing were not sampled.  In addition, as
discussed in Section 4.5, there were limited data on the species collected and fishing locations
used  by non-tribal populations in the Columbia River Basin.  Therefore, while the concentrations
of chemicals in fish tissue have been used to characterize risk for the general public in this study,
this characterization was uncertain due to the lack of data on fishing practices for the general
public.

Another source of uncertainty for this risk assessment involves the use of the average chemical
concentrations for fish collected over a short period of time to estimate human exposure over 30
and 70-year durations.  If average chemical concentrations in fish tissue have changed over time,
or were likely to change in the future, the risk estimates presented in this report may either
underestimate or overestimate the risk to individuals.  The relatively small  amount of existing
historical data on chemical contaminants in fish within the Columbia River Basin was insufficient
to reliably evaluate trends in chemical concentrations. The seasonal range of chemical
concentrations in the target species evaluated in this risk  assessment is also not known.

Thus, the risk estimates presented  in this report could increase or decrease depending upon how

                                            10-212

-------
concentrations vary over location and time.

As discussed in Section 1.7.5, to calculate average contaminant levels in fish, a value of one-half
the detection limit was used in some cases for non-detected chemicals.  Risk characterization
based upon one-half the detection limit could be either an overestimate or an underestimate of the
actual risks.

10.4.1.2 Tissue Type

For this study, both whole fish and fillets were analyzed when possible.  The fillet and whole
body sample types were chosen based on the fish consumption survey for CRITFC's member
tribes (CRITFC, 1994).  In this study, respondents were asked to identify the fish parts they
consume for each species.  For most of the fish species sampled as a part of this study, 50% or
more of the respondents said that they consume fish skin.  A smaller proportion of the tribal
members consumed other fish parts (head, eggs, bones and organs). In addition to the question of
people consuming fish parts, some chemicals preferentially accumulate in fat or internal organs,
thus  having both whole body and fillet fish tissue samples provides a more comprehensive picture
of the amount of chemical accumulated throughout the fish tissue.  Fillets were analyzed with
skin  because most tribal members consumed the skin with the muscle tissue.

Information  on the portions offish that are consumed most frequently by the general public were
not available. However, respondents to the qualitative fish consumption survey of people from
Wheatland Ferry to Willamette Falls Reach of the Willamette River, Oregon indicated that they
consume primarily fish fillets as well as other fish parts and the whole body (EVS, 1998).

In Section 6.2.4, the ratios of the estimated hazard indices and cancer risks for whole body to
filleted fish samples were calculated to determine the possible impact of tissue type on the risk
characterization. These results were calculated for those species that had both fillet and whole
body samples analyzed at a given site.  For non-cancer effects, whole body to fillet ratios were
calculated for the total hazard index as well as for the endpoints of immunotoxicity and
reproduction. The number of whole body to fillet ratios that were greater than 1 compared to the
total  number of samples was also shown. These calculations (Table 6-23) did not show a
consistent pattern in whole body to fillet ratios for the total hazard indices,  the immunotoxicity
hazard indices, or cancer risks at a given site for a species. The whole body to fillet ratios ranged
from 0.2 to greater than 1 for a  few species/sites (e.g.  high of a ratio 6.6 for fall chinook,
immunotoxicity hazard index).  For reproductive effects, the ratios of the hazard indices for
reproductive effects in whole body to fillet samples appear to be less than 1  more frequently than
those for the other hazard indices or cancer risks. This may be because the  hazard index for
reproductive effects is based largely upon the contaminant mercury which  is not lipophilic and
binds strongly to protein (e.g., muscle tissue).

Any  conclusions, however, on the results of whole body to fillet samples are limited by the small
sample sizes (usually 3 or less)  at each site and by the fact that whole body samples were always
from a composite offish different than those used for the whole body samples (i.e., fillet and
                                            10-213

-------
whole body samples are not from the same fish).

10.4.1.3  Exposure Duration

Exposure duration is defined as the time period over which an individual is exposed to one or
more contaminants.  For adults, two different exposure durations were used for the risk
assessment: 70 years, which represents the approximate average life expectancy of all individuals
born in the United States in the late 1960s; and 30 years, which represents the 90th percentile
length of time that an individual stays at one residence (USEPA, 1997b).

The value of 70 years was assumed for lifetime exposure in this risk assessment because it is the
value commonly assumed for the general population in most EPA risk assessments. Also, 70
years is the primary assumption used in the derivation of many of the cancer slope factors found
in IRIS (USEPA, 2000c).

As was discussed in Section 4, changes in exposure duration do not impact the exposures
estimated for calculating non-cancer health impacts. This is because the product of the exposure
frequency (EF) times exposure duration (ED) is always equivalent to the averaging time (AT)
(see Equation 4-1 in Section 4.3).

However,  since the averaging time for estimating exposure for cancer risks is always a person's
lifetime, changing exposure duration does impact the estimated risk.  The cancer risk estimates
for an individual who consumes fish over an exposure duration that differs from the exposure
durations used in this report (ED new) can be determined using the following equation:


(Equation 10-1)       ECRnew  = ECR?o x EDnew/ED?o

       where:

       ECRnew  = Excess cancer risk for the new exposure duration
       ECRvo   = Excess cancer risk estimate for a lifetime exposure duration of 70 years
       ED new  =  Individual exposure duration in years
       EDvo     = Default lifetime exposure duration of 70 years

Equation 10-1 shows that the excess cancer risk will change in direct proportion to the ratio of the
new and default exposure  durations. For example, if an exposure duration of 9 years was
selected, which is the median length of time an individual stays at one residence, the lifetime
exposure cancer risk estimates would be multiplied by a factor of 0.13 (9 years + 70 years = 0.13)
to obtain revised cancer risk estimates for a 9-year exposure duration. Thus, all total excess
cancer risk estimates for 70 years exposure duration for the fish species and tissue types evaluated
in this report would decrease by approximately an order of magnitude (i.e. ten-fold) for an
exposure duration of 9 years.

10.4.1.4  Consumption Rate

                                           10-214

-------
In this risk assessment, exposures were estimated for both the general public and for members of
CRITFC's member tribes. For the general public, adequate quantitative information on fish
consumption rates for those areas of the Columbia River Basin sampled in this study was not
available. Therefore, the ingestion rates assumed for those individuals in this risk assessment

were based on a national report offish consumption (USEPA, 2000b). For CRITFC's member
tribes, ingestion rates were taken from CRITFC's fish consumption study (CRITFC, 1994). For
both the general population and the tribes, mean and a 99th percentile ingestion rates for children
and adults were selected to evaluate potential risks over a range of possible ingestion rates.

It is not known if the ingestion rates selected for this risk  assessment are representative of the
actual consumption practices of individuals consuming fish from the study area.  The exposures
estimated in this report are likely to be higher than those expected for a recreational fisherman
who infrequently fishes at any of the study sites. On the other hand, as discussed in Section 4,
Harris and Harper (1997) suggest that an ingestion rate of 540 g/day is more  appropriate for a
tribal member who pursues a traditional lifestyle. This is higher than the 99th percentile CRITFC
member tribal fish consumption rate of 389 g/day used in this report.

10.4.1.5 Multiple-Species Consumption Patterns

The hazard indices and cancer risk estimates in this report were primarily based upon the
consumption of individual fish species and tissue types. However, these estimates which  are
based upon individual fish species may not be an adequate representation of risk for most
individuals  since most people likely eat a diet composed of multiple fish species. Therefore, as a
part of the risk characterization,  a hypothetical multiple-species diet was also evaluated using
tribal fish consumption data from CRITFC's fish consumption study.  For this hypothetical
multiple-species diet, information from Table 17 of the CRITFC fish consumption study
(CRITFC, 1994) was used. This table from the CRITFC consumption survey  provides
information on the percentage of adults that consumed 10 fish species evaluated in the study
(CRITFC, 1994).  As was shown in Table 6-24 and Figures 6-35 and 6-36 the resultant cancer
risk and non-cancer hazards of the multiple species diet reflect the proportion of the different
types offish in the diet and the contaminant levels in those fish.  Therefore, the estimated  cancer
risks and non-cancer hazards from consuming fish from the Columbia River Basin for any one
individual depend upon the types and amounts offish they eat and may be very different from
those estimated in this report for individual species.

As part of this uncertainty analyses, an estimate of the total cancer risks and non-cancer hazards
from a multiple species diet using data from Table  18 in the CRITFC fish consumption study in
addition to that in Table 17 was calculated (CRITFC, 1994). Table 18 provides average
consumption rates (grams per day) for each species for those adult respondents  in the survey who
consume fish. These  rates were determined by combining the average consumption rate  for each
individual who consumed a particular species with the average serving size in ounces for that
individual and then calculating the mean of all of the individual consumption rates. The
differences in the consumption rates for the hypothetical multiple diet using the two  CRITFC
tables (Table 17 versus Table 18) are shown in Table 10-2.  As can be seen from Table  10-2, the

                                           10-215

-------
consumption rates, cancer risks and total hazards for each individual fish species differ using the
results from the two different tables in the CRITFC consumption study (CRITFC, 1994).
However, the total estimated cancer risks and total non-cancer hazard indices from consuming all
species are approximately the same using either table.

Table 10. 2. Comparison of estimated total cancer risks and hazard indices for a hypothetical multiple
species diet using data from Table 17 and Table 18 in the CRITFC fish consumption report (Source:
CRITFC, 1994).
Results using Table 17 in the CRITFC fish consumption
study(1)


Fish Species T
salmon FS
trout FS
whitefish FS
smelt WB
lamprey FS
walleye FS
sturgeon FW
sucker FS
Totals

Percentage of
Hypothetical Diet
27.7%
21.0%
6.8%
15.6%
16.3%
2.8%
7.4%
2.3%
100.0%
Consumption
Rate
(grams/day)
17.5
13.3
4.3
9.9
10.3
1.8
4.7
1.5
63.2
Results using Table 18 in the CRITFC
fish consumption study
Total Non- Cancer Consumption
Cancer Effects (total Rate
Risk
6E-05
3E-05
9E-05
3E-05
1E-04
4E-06
7E-05
9E-06
4E-04
HI)
0.6
0.3
0.7
0.1
0.7
0.1
0.6
0.1
3.2
(1) These results are those presented in Section 6.2.5 and Table 6-24
FS= fillet with skin FW
= fillet without skin WB
= whole body


(grams/day)
25.7
9.6
8.9
4.8
4.7
3.8
3.3
2.8
63.6
T= tissue type
Ffl = hazard index
Total
Cancer
Risk
8E-05
2E-05
2E-04
2E-05
5E-05
9E-06
5E-05
2E-05
4E-04


Non Cancer
Effects
(total HI)
0.9
0.2
1.5
0.0
0.3
0.2
0.4
0.2
3.8


10.4.1.6  Effects of Cooking

It was assumed for this risk assessment, that (with the exception of skinless white sturgeon fillets)
the skin and fatty areas of the fish are not removed during preparation, and that there is no net
reduction in contaminant concentrations during cooking.  Anglers who prepare fillets by skinning
and trimming away the fatty area may reduce their exposure to chemicals (such as
organochlorines) that accumulate in fatty areas.  It has also been shown that cooking the fish may
affect exposure concentrations of such chemicals, depending on the cooking method.

EPA's guidance (USEPA, 2000a) provides a summary of the effects on organochlorine (e.g.,
PCBs, DDT, chlordane, dioxins/furans) contaminant levels in fish as a result offish preparation
and cooking. This summary shows that the reductions in chemical concentrations vary
considerably among the different studies because of different fish species, contaminants, cooking
methods, etc. In these studies most of the percent reductions in chemical concentrations ranged
from about 10 to 60%.  However, much higher losses were also seen as were net gains of one
contaminant (PCBs). Overall, these studies support the conclusion that organochlorines can be
lost during cooking.  But,  based on the available information, it is difficult to quantify these
losses for use in a risk assessment since the actual losses from cooking depend upon the cooking
method (i.e., baking, frying, broiling, etc.), the cooking duration, the temperature during cooking,
preparation techniques (i.e., trimmed or untrimmed, with or without skin), the lipid content of the
fish, the fish species, and the contaminant levels in the raw fish.
                                             10-216

-------
Also as discussed in EPA guidance (USEPA, 2000a), several studies indicate that some organo-
metal compounds bind to different fish tissues than the tissue which bind organochlorines.
Mercury, for example, binds strongly to protein, thereby concentrating in the muscle tissue of
fish. Mercury also concentrates in liver and kidney, though at generally lower rates. Thus,
preparations such as trimming and gutting, can actually result in a greater average concentration
of mercury in the remaining tissues compared with the concentration in the whole fish
(Gutenmann and Lisk, 1991).  As discussed previously in the discussion on effects of sample type
on the risk characterization (Section 6.2.4 and Table 6-23), the ratios of the hazard indices for
reproductive effects in whole body to fillet samples appear to be less than 1 more frequently than
the ratios for the total hazard index, hazard index for immunotoxicity, and cancer risks.  This may
be because the hazard index for reproductive effects is based largely upon the contaminant
mercury which is not lipophilic and binds strongly to protein (e.g., muscle tissue). However, any
conclusions based on the ratios of whole body to fillet samples are limited by the small sample
sizes (usually 3  or less) at each site and by the fact that whole body samples were always from a
composite offish different than those used for the whole body analysis (i.e., fillet and whole body
samples are not from the same fish).

The impact of cooking on mercury levels was studied by Morgan et al., 1997. They found that
mercury concentrations (wet weight basis) in pan-fried, baked and boiled walleye fillet ranged
from 1.1 to 1.5 times higher than in the corresponding raw portions; in lake trout the range was
1.5 to 2.0 times higher.

10.4.2 Toxicity Assessment

There are also uncertainties in the toxicity assessment. These include uncertainties (1) in the
toxicity values (i.e., reference  doses and cancer slope factors) used; (2) in the toxicity equivalence
factors developed for dioxins/furans and dioxin-like PCBs and in the relative potency factors used
for PAHs; (3) in the lack of toxicity data for some of the chemicals that were detected in fish,
and; (4) in the manner in which certain chemicals (Aroclors, dioxin-like PCBs, DDT/DDE/DDD,
and arsenic) were evaluated.

10.4.2.1  Toxicity Values

As discussed in Section 5.0, the majority of the toxicity factors used in estimating hazard indices
and cancer risks were taken from EPA's IRIS database which is a database of human health
effects that may result from exposure to various  substances found in the environment. For a
small number of chemicals whose toxicity factors were not available in IRIS, toxicity factors
developed by NCEA were used.  Although the development of the IRIS toxicity factors has been
reviewed by a group of EPA health scientists using consistent chemical hazard identification and
dose-response assessment methods, there are still several sources of uncertainty in these factors
and their relevance to the populations for which the risk assessment is being conducted.  As
discussed in EPA's guidance (USEPA, 1989), some of these uncertainties may include:

•      using dose-response information from effects observed at high doses to predict the


                                            10-217

-------
       adverse effects that may occur in humans following exposure to the lower levels expected
       from human exposure in the environment;

•      using dose-response information from short-term studies to predict the effects of long-
       term exposures;

       using dose-response information from animal studies to predict effects in humans; and

       using dose-response information from homogenous populations or healthy human
       populations to predict the effects likely to be observed in the general population consisting
       of individuals with a wide range of sensitivities.

In addition to the uncertainties in developing reference doses and cancer slope factors based upon
the data that are available, there are also uncertainties in the fact that specific types of effects data
are often not available for a given chemical. Some examples include the lack of data on a
chemical's cancer and non-cancer impact on vulnerable populations (e.g., children) and a lack of
information for some chemicals on non-cancer endpoints such  as reproductive, developmental,
and endocrine disruption.  However, the lack of data on non-cancer effects is usually considered
when determining what uncertainty factors and modifying factors should be used to develop a
reference dose for a given chemical. The lack of data on cancer is partially addressed by using
conservative assumptions (e.g., upper confidence levels, the most sensitive species) in estimating
cancer slope factors.  All of these assumptions are intended to provide a margin  of safety to
ensure that the health impacts for an individual chemical are not likely to be underestimated.

To better understand the uncertainties associated with the toxicity factors for each of the
chemicals evaluated in this risk assessment, refer to the Toxicity Profiles in Appendix C. These
profiles review the data upon which the reference doses and cancer slope factors were developed.

10.4.2.2 Toxicity Equivalence Factors for Dioxins, Furans, and Dioxin-like PCB Congeners
and Relative Potency Factors for PAHs

Toxicity equivalence factors  were used for the chlorinated dioxins and furans and the dioxin-like
PCBs measured in this study to calculate toxicity equivalence concentration.  These toxicity
equivalence factors were calculated using all of the available data and were selected to account
for uncertainties in the available data and to avoid underestimating risk (Van den Berg et al.,
1998). Alternative approaches, including the assumption that all dioxin-like PCBs carry the
toxicity equivalence of 2,3,7,8-TCDD, or that all chlorinated dioxins, furans, and dioxin-like PCB
congeners other than 2,3,7,8-TCDD can be ignored, have been generally rejected as inadequate
for risk assessment purposes  by EPA and many other countries and international organizations.
These toxicity equivalence factors are order-of-magnitude estimates relative to the toxicity of
2,3,7,8-TCDD. Therefore, their use creates uncertainty in the  risk assessment, especially since
chlorinated dioxins/furans and dioxin-like PCBs contribute significantly to the cancer risks
estimated in this risk assessment.
                                             10-218

-------
Also, it should be noted that the cancer slope factor for 2,3,7,8-TCDD is being re-evaluated as
part of a current review by EPA (USEPA, 2000e). A review of the most current draft document
suggests that this cancer slope factor may increase. This change would affect both the cancer risk
estimates associated with 2,3,7,8-TCDD as well as those risk estimates calculated for the other
chlorinated dioxins, furans, and dioxin-like PCB congeners having toxicity equivalence factors.
If the slope factor increases, cancer risks estimated for these classes of compounds would also
increase.

As discussed in Section 5, EPA has developed provisional guidance on estimating risk from
exposure to PAHs (USEPA, 1993).  A cancer slope factor is available for only one PAH,
benzo(a)pyrene. In this provisional guidance, relative potency factors have been developed for
six PAHs relative to benzo(a)pyrene. These relative potency factors were used to estimate cancer
risk from PAHs in this risk assessment.  As with the toxicity equivalence factors these relative
potency factors are order-of-magnitude estimates and, therefore, have inherent uncertainties.
However, unlike the toxicity equivalence factors, these relative potency factors for the PAHs are
considered to be more uncertain because they do not meet all of the criteria for the application of
toxicity equivalence factors to mixtures.

In our study, with the exception of one composite sample of largescale sucker taken at study  site
13 (see discussion in Section 6.2), PAHs do not contribute significantly to the levels of
contaminants in fish or to  cancer risk estimates from consuming fish.  Therefore, the uncertainties
in the use of relative potency factors for PAHs should not greatly impact the overall risks
characterized in this report.

10.4.2.3  Chemicals Without Quantitative Toxicity Factors

As shown in Table 5-1, there were 23 chemicals that were analyzed for in fish tissue that do not
have a cancer slope factor or reference dose. Of the 23 chemicals without toxicity values, the
following 14 chemicals were not detected in any fish species: delta-BHC, dibenzofuran, gamma-
chlordene, tetrachloroguaiacol, 4-bromophenyl-phenylether,  4-chloroguaiacol, 4-chlorophenyl-
phenylether, 3,4-dichloroguaiacol, 4-chloro-3-methylphenol, 4,5-dichloroguaiacol, 4,6-
dichloroguaiacol, 3,4,5-trichloroguaiacol, 3,4,6-trichloroguaiacol, and 3,5,6-trichloroguaiacol.
Six additional chemicals were detected in less than 3% of the samples: acenaphthylene, alpha-
chlordene, benzo(ghi)perylene, phenanthrene, retene, and 1-methyl-naphthalene. Of the
remaining 3 chemicals, DDMU was detected less than 10%;  2- methyl-naphthalene and
pentachloroanisole were detected greater than 10% of the time.

As discussed in the Toxicity Profiles (Appendix C), the toxicity and mechanism(s) of action(s) of
pentachloroanisole are similar to those of its parent chemical, pentachorophenol. However,
methylation of the chlorophenols makes them more polar, and thus likely to be somewhat less
reactive in biological systems. Thus the extent of both acute and chronic toxicity of
pentachloroanisole can be reasonably anticipated to be somewhat less than its chlorinated parent,
PCP.  DDMU is a breakdown product of the DDT.  Little information is available on DDMU or
2-methyl-naphthalene.
                                            10-219

-------
It is impossible to predict how the lack of toxicity information on these 23 chemicals might
impact the characterization of risk in this report.  However, given the fact that only 2 of these
chemicals (2- methyl-naphthalene and pentachloroanisole) were detected in greater than 10% of
the samples, any under estimation of cancer risk and non-cancer hazards is unlikely to be great.

There are no EPA consensus reference doses available for the chlorinated dioxins and furans and
the dioxin-like PCB congeners, therefore, the possible non-cancer health effects from exposure to
these chemicals from fish consumption could not be estimated in this report.  From the most
recent draft of EPA's reassessment of the toxicity of these compounds (USEPA, 2000e), it is
clear that these compounds can cause non-cancer effects at very low levels of exposure. The
inability to characterize the non-cancer hazards from these compounds may result in an
underestimate of the non-cancer hazards calculated in this report.

10.4.2.4  Risk Characterization for PCBs

As discussed in Section 1, two different measurements were used in this study to determine PCB
concentrations in fish tissue: 1) analysis of Aroclors which are commercial mixtures of both
dioxin-like and non-dioxin-like PCB congeners, and 2) analysis of individual dioxin-like PCB
congeners.  The Aroclor methodology included the analysis of 7 Aroclors: Aroclor 1016, Aroclor
1221, Aroclor 1232, Aroclor 1242, Aroclor 1248, Aroclor  1254, and Aroclor 1260. Only
Aroclors 1242, 1254, and 1260 were detected. Eleven  dioxin-like PCB congeners that exert
toxicity similar to 2,3,7,8 -TCDD were also measured.  PCB 170 and PCB 180, though measured,
were not considered in the risk assessment as dioxin-like PCB congeners because they do not
currently have associated toxicity equivalence factors.

Cancer Risks for PCBs

Because Aroclors are a mixture of both dioxin-like and non-dioxin-like PCB congeners,
calculating and summing the risk associated with both Aroclors and with individual dioxin-like
PCB congeners would likely overestimate cancer risk by accounting for the dioxin-like PCB
congener risk both individually and within the risk estimates for Aroclors. Therefore, before
using the Aroclor fish concentrations to calculate cancer risk, an adjustment was made to the
Aroclor concentrations by subtracting the concentration of dioxin-like PCB congeners from the
total Aroclor concentrations for each sample. This resulted in what is called the "adjusted
Aroclor" value.

To estimate the impact of using this method on the cancer risk, a comparison was made for
estimates of cancer risk from PCBs using different methods.  The excess cancer risks calculated
with these methods (using basin averages) for each fish species are shown in Table 10-3. The
risk from dioxin-like PCB congeners alone ranged from 0.5 (coho salmon) to 3.5 (rainbow trout)
times (column B/A) the risk calculated for total unadjusted Aroclors alone. Because the mass of
dioxin-like PCB congeners is so small compared to that of the Aroclors, the risk estimated for
adjusted Aroclors (subtracting the concentration of dioxin-like PCB congeners from the total
Aroclor concentrations) (column C)  is only slightly lower than that for total unadjusted Aroclors
                                            10-220

-------
(Column A).  Characterizing PCB risks by combining either total Aroclors plus dioxin-like PCB
congeners (A + B) or adjusted Aroclors plus dioxin-like PCB congeners (B + C) is approximately
the same. The PCB risks estimated from using "adjusted Aroclors plus dioxin-like PCB
congeners" is from 1.5 to 4.3 times that estimated from using total unadjusted Aroclors alone
(Column B+C /A).

Table 10-3. Estimated Cancer Risks for PCBs Using Different Methods of Calculation.  CRTTFC's member
tribal adult, average fish consumption, 70 years exposure using average Columbia River Basin-wide
chemical concentrations.










bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
coho
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey
A






Total
unadjusted
Aroclors
1.1E-04
7.6E-05
3.5E-04
2.0E-04
2.3E-05
2.5E-05
4.6E-05
3.1E-05
2.9E-05
4.4E-05
ND
1.6E-04
B






Dioxin-
like PCB
congeners
1.2E-04
1.1E-04
7.7E-04
1.7E-04
2.6E-05
8.7E-05
2.5E-05
3.6E-05
4.8E-05
7.5E-05
9.5E-06
3.3E-04
B/A







Risk
Ratio
1.1
1.4
2.2
0.8
1.1
3.5
0.5
1.2
1.7
1.7
NA
2.1
C






Adjusted
Aroclors
only
l.OE-04
7.1E-05
3.0E-04
1.9E-04
2.1E-05
2.2E-05
4.5E-05
3.0E-05
2.8E-05
4.2E-05
ND
1.5E-04
A+B




Total
Aroclors
plus dioxin-
like PCB
congeners
2.3E-04
1.8E-04
1.1E-03
3.7E-04
4.9E-05
1.1E-04
7.0E-05
6.8E-05
7.7E-05
1.2E-04
9.5E-06
4.8E-04
B+C




Adjusted
Aroclors plus
dioxin-like
PCB
congeners
2.3E-04
1.8E-04
1.1E-03
3.6E-04
4.6E-05
1.1E-04
7.0E-05
6.6E-05
7.6E-05
1.2E-04
9.5E-06
4.7E-04
(B+C)/
(A+B)






Risk
Ratio
0.98
0.97
0.96
0.97
0.95
0.97
0.99
0.98
0.98
0.99
1.00
0.98
(B+C)/A

Adjusted
Aroclors
plus dioxin-
like PCB
congeners /
total
unadjusted
Aroclors
2.1
2.4
3.1
1.8
2.0
4.3
1.5
2.1
2.6
2.7
NA
3.0
 ND = not detected  NA = not applicable

Non-Cancer Effects from Aroclors

The immunological endpoint was based upon the toxicity of Aroclors. However, only one of the
three Aroclors detected in the fish samples has a reference dose - Aroclor 1254. Therefore, two
possible methods were available to estimate the non-cancer hazard for the immunotoxicity
endpoint.

•      (A) - estimate the hazard index using the concentration of Aroclor 1254 only and the
       reference dose for Aroclor 1254, or

       (B) - assume that the reference dose for Aroclor 1242 and 1260 are equivalent to that for
       Aroclor 1254; estimate the hazard index by summing all three Aroclor concentrations and
       use this sum with the reference dose for Aroclor 1254.

Method B was used in this risk assessment. To show the potential uncertainties with using
Method B, the hazard indices calculated with both methods (using basin averages) for each fish
species are shown in Table 10-4.
                                            10-221

-------
 Table 10-4 Comparison of Hazard Indices for the Immunological End point Based on Alternative
 Treatments of Aroclor Data. CRTTFC's member tribal adult, average fish consumption, using average
 Columbia River Basin-wide chemical concentrations.	
                        Endpoint specific hazard index for
                               immunotoxicitv




bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
coho salmon
fall chinook salmon
spring chinook salmon
steelhead
eulachon
Pacific lamprey


(A)
Aroclor 1254
1.1
0.8
5.1
2.6
0.6
0.6
0.7
0.8
0.7
0.7
ND
3.9

(B)
sum of Aroclors 1242, 1254,
and 1260
2.7
1.9
8.7
5
0.6
0.6
1.1
0.8
0.7
1.1
ND
3.9
(B/A)
Ratio of the hazard index for the
of Aroclors to the hazard index
Aroclor 1254 only
2.5
2.4
1.7
1.9
1.0
1.0
1.6
1.0
1.0
1.6
ND
1.0

sum
for













ND = Not Detected
Table 10-4 also shows the ratio of the hazard index calculated using (A) Aroclor 1254
concentrations only or (B) the sum of all three Aroclors. For walleye, rainbow trout, spring
chinook, fall chinook, and Pacific lamprey, the method used has no impact on the hazard index
calculated for the immunotoxicity endpoint. This is because for these five species, only Aroclor
1254 was detected in the fish sampled. For the other species, the hazard index based on Method
B (using the sum of all Aroclor concentrations) is from 1.6 to 2.5 times higher than the hazard
index based upon Aroclor 1254 alone (column B/A).

10.4.2.5 Non-Cancer Effects from DDT,  ODD, and DDE

DDT and its derivatives, ODD and DDE, were measured in fish tissue samples; however, only
DDT has a reference dose. The reference dose for DDT is based upon its toxic effects on the
liver (hepatotoxicity).  For the non-cancer hazard assessment done in this report, two possible
methods for the estimation of the hazard quotient and hazard index from these chemicals were
possible:

       (A) - estimate the hazard quotient using the concentrations of DDT only and the reference
       dose for DDT, or

•      (B) - assume that the reference doses for DDD and DDE are equivalent to that for DDT.
       Therefore, first sum the concentrations of all of the DDD, DDE and DDT species in each
       sample and utilize the reference dose for DDT to estimate the hazard quotient from the
       summed concentrations of DDD, DDE, and DDD
                                            10-222

-------
Table 10-5. Comparison of Hazard Quotients and Hazard Indices for the Hepatic Health Endpoint Based on
Alternative Treatments of DDT, ODD, and DDE Data.  CRTTFC's member tribal adult, average fish
consumption, using average Columbia River Basin-wide chemical concentrations.	
                   Hazard quotient
Hazard Index for hepatic
      endpoint

A
Snecies DDT only
bridgelip sucker
largescale sucker
mountain whitefish
white sturgeon
walleye
rainbow trout
coho salmon
fall chinook
spring chinook
steelhead
eulachon
Pacific lamprey
0.08
0.04
0.03
0.02
0.00
0.01
0.00
0.00
0.01
0.00
ND
0.06
B

C
(B/A)
HQ (Total DDT)/
Total DDT HO (DDT) DDT only
0.95
0.44
0.76
1.04
0.10
0.05
0.01
0.03
0.04
0.03
0.02
0.17
11
11
27
52
28
8
4
7
4
8
NA
3
0.13
0.10
0.19
0.36
0.47
0.04
0.06
0.08
0.08
0.07
0.05
0.22
D
sum of DDT,
DDE. and ODD
1.00
0.50
0.93
1.38
0.57
0.09
0.07
0.10
0.11
0.10
0.07
0.33
(D/C)
HI (Total DDT)/
HI (DDT)
7.5
5.0
4.8
3.9
1.2
2.1
1.2
1.4
1.3
1.4
1.4
1.5
ND = not detected; NA = not applicable
HS = hazard quotient
HI = Hazard index
Total DDT = sum of DDT, ODD, DDE
Method B was used to characterize non-cancer health effects in this study. Because DDT has
been identified as having a hepatic (liver) toxicity endpoint, the treatment of DDT and its
derivatives will affect not only the hazard quotient for the these species, but also the hazard index
for the hepatic (liver) toxicity endpoint.

Table 10-5 compares the hazard quotients for DDT and its derivatives (in columns A and B) as
well as the hazard indices for the hepatic endpoint (in columns C and D) using the two methods.
As can be seen from Table 10-5, the hazard quotient increased from about 3 times for Pacific
lamprey to 52 times for white sturgeon when all three species (DDT, DDE, DDD) are summed to
calculate the hazard quotient compared to calculating the hazard quotient using DDT data alone.
The impact on the hepatic endpoint is less because for some fish species other chemicals in
addition to DDT and its derivatives are included in the calculation of the hazard index for
hepatotoxicity. The ratio between the hepatic hazard index using DDT, DDE, and DDD to the
hepatic hazard index using DDT alone ranges from between 1.2 for coho salmon to 7.5 for
bridgelip sucker, with the highest ratios seen in some of the resident fish species.  Thus, the
endpoint specific hazard indices for hepatotoxicity that are discussed in Section 6 may be an
overestimate if DDE and DDD are less toxic to the liver than DDT. This is primarily true for
several of the resident species.

10.4.2.6  Risk Characterization for Arsenic
As discussed in Section 5.3.3, total arsenic was measured in fish tissue samples in this study.
Because a reference dose and cancer slope factor are available for only inorganic arsenic, an

                                             10-223

-------
assumption about the percent of inorganic arsenic in fish had to be made to estimate the non-
cancer hazards and cancer risks. The non-cancer hazards and cancer risks discussed in Section
6.2.1 and 6.2.2, respectively, assumed that for all fish species (resident fish and anadromous fish)
caught in this study, 10% of the total arsenic was inorganic arsenic.  The data in Section 5.3.3
also suggests that an alternative assumption for anadromous fish species should be considered -
the assumption that 1% of the total arsenic is inorganic.  Therefore in Section 6.2.6, the non-
cancer hazards and cancer risks were recalculated for anadromous fish species using basin data
assuming that 1% of the total arsenic was inorganic.

This comparison of the results from using the two different assumptions (1% versus 10%) for
arsenic in fish shows that the reduction of the non-cancer hazards is less than 12% for all
anadromous fish species, except eulachon which had about a 50% reduction. However, the
impact is greater on the estimates of cancer risk. With the exception of lamprey for which cancer
risks were reduced by  only 6%, the reductions in cancer risks for steelhead were about 29%. The
cancer risks for the other anadromous fish species were reduced from about 40% to 50%. Thus,
the assumptions used for percent inorganic arsenic have the most impact on the cancer risks
estimated for salmon, steelhead and eulachon and on the non-cancer hazards for eulachon.

10.4.3 Risk Characterization

10.4.3.1  Cancer Risk Estimates

As recommended by EPA's guidance on mixtures (USEPA, 2000g), the total cancer risk  from a
sample is calculated by summing the risk of individual carcinogenic compounds in that sample.
This approach for carcinogens (response addition)  assumes independence of action by the
components in a mixture (i.e., that there are no synergistic or antagonistic interactions among the
carcinogens in fish and that all chemicals produce the same effect, cancer). If these assumptions
are incorrect, over- or under-estimation of the actual risks could result.  The underlying biological
basis for assuming synergism is that cancer is a multistage process where a series of events
transforms a normal cell into a malignant tumor.  If two carcinogens act at different stages,  their
combined effect can be greater than either acting alone.  For example, initiation-promotion
studies have demonstrated synergistic effects for some pairs of carcinogens.  On the other hand,
similar-acting carcinogens can  compete with each other to result in antagonism. For example, the
presence of one metal  can decrease the absorption  or effectiveness of a similar metal.
Interactions can be quite complex and can depend  on dose or  other factors, including background
exposures to other carcinogens. In general, available information seldom allows quantitative
inferences to be made  about potential interactions among carcinogens. In the absence of such
information, the practice is to assume additivity, particularly at low doses for mixtures.

Summation of carcinogenic risks for substances with different weights-of-evidence for human
carcinogenicity is also an uncertainty. The cancer risk equation for multiple substances sums all
carcinogens equally, giving as much weight to class B or C as to class A carcinogens.  Using the
assumption of additivity gives equal weight to all slope factors without regard to their basis from
human data. In this assessment, only arsenic is in the class A carcinogen group (human
carcinogen based on human data) and all of the other major contributors to cancer risk (e.g., DDT

                                            10-224

-------
and DDE, DDD, Aroclors, dioxin-like PCB congeners and chlorinated dioxins and furans) are in
the class B2 group (probable human carcinogen based on sufficient evidence in animals and
inadequate or no evidence in humans).  It should be noted, however, that EPA's most recent draft
document on the toxicity of 2,3,7,8-TCDD and related compounds (USEPA, 2000e) characterizes
the complex mixtures of dioxins to which humans are exposed as "likely human carcinogens".

The cancer slope factors used in this risk characterization are primarily from EPA's database,
IRIS. Most of the IRIS cancer slope factors are considered to be plausible upper bounds to the
actual lifetime excess cancer risk for a given chemical. Concern has often been raised that adding
multiple carcinogens, whose slope factor are upper bound estimates, will lead to unreasonably
high estimates of the actual risk.  Statistical examination of this issue suggests that the error in the
simple addition of component upper bounds is small compared to other uncertainties, and that as
the number of mixture components increases, summing their upper bounds yields an inflated but
not misleading estimate of the overall risk (Cogliano, 1997). In fact, division by a factor of two
can be sufficient to convert a sum of upper bounds into a plausible upper bound for the overall
risk.  If one or two carcinogens predominate the risk, however, this is not of concern.

10.4.3.2  Non-Cancer Health Effects

In Section 6, non-cancer health impacts were evaluated in several ways. First, the hazard quotient
was calculated. The hazard quotient, which is the ratio between an individual's estimated
exposure to a chemical compared to the reference dose for that chemical,  assumes that there is a
level of exposure (i.e., the reference dose) below which it is unlikely for even sensitive
populations to experience adverse health effects.  As a rule, the greater the value of the hazard
quotient, the greater the level of concern. However, it is important to emphasize that the level of
concern does not increase linearly as the reference dose is approached or exceeded for each
chemical because reference doses for different chemicals do not have equal accuracy or precision
and are not based on the same severity  of toxic effects.  Therefore, the possible health impacts
resulting from exposures greater than the reference dose can vary widely depending upon the
chemical.

Based on EPA guidance (USEPA,  1986a; USEPA, 1989; USEPA, 2000g), the hazard quotients
calculated for each chemical in a sample were then summed to give a hazard index.  This
approach of adding all of the hazard quotients regardless of endpoint (dose addition)  has several
uncertainties because it assumes that all compounds in a mixture have similar uptake and
pharmacokinetics (absorption, distribution, and elimination in the body) and it results in
combining chemicals with reference doses that are based upon very different critical effects,
levels of confidence, uncertainty/modifying factors, and dose-response curves. Since the
assumption of dose additivity is most properly applied to compounds that induce the same effect
by the same mechanism of action, EPA guidance recommends that when the total hazard index
for a mixture exceeds 1, the chemicals in that mixture  should be segregated by effect and
mechanism to derive endpoint-specific  hazard indices (USEPA, 1986a).

Although deriving endpoint specific hazard indices, as was done for this risk assessment, likely
reduces the uncertainty in the non-cancer hazard evaluation in this risk assessment, these

                                            10-225

-------
uncertainties are not eliminated. For example, calculation of endpoint specific hazard indices
may still be incorrect estimates of non-cancer health impacts. Although two chemicals may
affect the same organ (e.g. the liver), they may not necessarily do so by the same specific
lexicological process.

However, it should be noted that in this assessment the majority of the estimated non-cancer
hazards resulted from a limited number of chemicals: Aroclors, mercury, total DDTs, and arsenic.
The highest endpoint specific hazard indices were for immunotoxicity (due to Aroclors), central
nervous system and reproduction/developmental (due to mercury), liver (due primarily to DDT,
DDE and ODD), and hyperpigmentation/cardiovascular (due to arsenic).  These endpoint specific
hazard indices are based in large part on a single chemical or class of chemical (e.g. total DDTs).
Therefore, the many uncertainties regarding calculation of endpoint specific hazard indices using
a mixture of chemicals should not play a major role in the characterization of non-cancer hazards.

10.4.3.3 Cumulative Risk from Chemical and Radionuclide Exposure

Risks were combined for all carcinogens to equal a total cancer risk. However, radionuclides
were not included in this estimate because radionuclide analyses were not completed for all
species in this assessment.

10.5   Risk Characterization for Consumption of Fish Eggs

As discussed in Section 4.5, a small number of egg samples were collected for some of the
anadromous fish species. Although the fish consumption studies discussed in this report suggest
that both CRITFC's member tribes and some of the general public consume eggs, none of these
studies provided information on the amount of eggs consumed.  Therefore, a risk characterization
of eggs was not included in Section 6. However, to provide information on the potential risks
from consuming eggs, the average fish ingestion rates for adults and children (general public and
CRITFC's member tribes)  were used for estimating cancer risk (adults only) and non-cancer
hazards (adults and children) for eggs. These estimates for eggs, which are shown in Appendix P,
are very uncertain but they  serve as a useful comparison to the results for fish consumption.

Three samples of eggs were collected from coho salmon (Umatilla), fall chinook (Columbia, site
8), and steelhead (Columbia, site 8) and six egg samples were collected from spring chinook (3 at
the Umatilla and 3 at Looking Glass Creek).

Endpoint specific and total  hazard indices for eggs were calculated using the average fish
ingestion rates for each population (adult and child, general public and; adult and child,
CRITFC's member tribes )(Tables 1.1 and 1.2 (coho salmon), 2.1 and 2.2 (fall chinook salmon),
3.1 and 3.2 (spring chinook salmon), 4.1 and 4.2 (steelhead)).  This provides estimates of the non-
cancer hazards for two ingestion rates for adults (7.5 and 63.2 g/day) and children (2.83 g/day, up
to age 6; and 24.8 g/day, up to age 15). No endpoint specific hazard indices and no total hazard
indices greater than 1 were  found using the average fish consumption rate for the general public,
adult or child. At  the average consumption rate for CRITFC's member tribal adults and children,


                                            10-226

-------
some of the total hazard indices were greater than 1 for eggs, the highest being approximately 4
for steelhead eggs at the average fish consumption rate for CRITFC's member tribal children.
Endpoint specific hazard indices greater than 1 (high of 2) for liver, immunotoxicity, and
selenosis were seen for CRITFC's member tribal child, average ingestion rate for spring chinook
and steelhead; an immunotoxicity endpoint specific hazard index of approximately 1 was seen for
coho. Endpoint specific hazard indices greater than 1 were due to exposures greater than the
reference dose for total Aroclors (immunotoxicity) and selenium (selenosis and liver).

Cancer risks for eggs were calculated using the average fish ingestion rates for both adult
populations (general public adult and CRITFC's member tribal adult) for both 30 and 70 years of
exposure. These results are found in the tables in Appendix P (Tables 1.3 (coho salmon), 2.3 (fall
chinook salmon), 3.3 (spring chinook salmon), and 4.3 (steelhead). As can be seen from these
tables, cancer risks from consumption of eggs ranged from 4 X 10~6 for both fall chinook and
steelhead at the lowest exposures (general public adult, average fish ingestion rate, 30 years
exposure) to a high of 8 X 10"5 for the highest exposure calculated (average fish consumption rate,
CRITFC's  member tribal adult, 70 years of exposure).  For these same exposures, coho salmon
eggs ranged from 7 X 10'6 to 1 X 10'4 and spring chinook eggs from 9 X 10'6 to 2 X 10'4.
                                            10-227

-------
11.0   Conclusions

The goals of this study were to determine:

               1)     if fish were contaminated with toxic chemicals,

               2)      the difference in chemical concentrations among fish species and study
                      sites, and

               3)      the potential human health risk due to consumption offish from the
                      Columbia River Basin.

The results of the study showed that all species offish had some levels of toxic chemicals in their
tissues and in the eggs of chinook and coho salmon and steelhead.   The concentration of organic
chemicals in the egg samples was lower than expected, given the high lipid content of the egg
samples.  The fish tissue chemical concentrations were quite variable within fish (duplicate
fillets), across tissue type (whole body and fillet), across species, and study sites. However, the
chemical residues exhibited some trends in distribution.  The concentrations of organic chemicals
in the salmonids (chinook and coho salmon, rainbow and steelhead trout) were lower than any
other species. The concentrations of organic chemicals in three fish species (white sturgeon,
mountain whitefish, largescale sucker) were higher than any other species.  Pacific lamprey had
higher organic chemical concentrations than anadromous species but lower than resident species.
The concentrations of metals were variable with maximum levels of different metals occurring in
a variety  of species. The distribution across stations was variable although fish collected from the
Hanford Reach of the Columbia River and the Yakima River tended to have higher
concentrations of organic chemicals than  other study sites.

 The concentrations of toxic chemicals found in fish from the Columbia River Basin may be a risk
to the health of people who eat them depending on:

       A.      the toxicity of the chemicals,

       2)      the concentration of chemicals in the fish,

       3)      fish ingestion rates

       4)      fish species, and tissue type

The chemicals which contributed the most to the hazard indices and cancer risks were the
persistent bioaccumulative chemicals (PCB, DDE, chlorinated dioxins and furans) as well as
some naturally occurring metals (arsenic,  mercury). Some pollutants persist in the food chain
largely due to past practices in the United States and global dispersion from outside North
America.  Although some of these chemicals are no longer allowed to be used in the United
States, a survey of the literature indicates  that these chemical residues continue to accumulate in a

                                            11-228

-------
variety of foods including fish. Human activities can alter the distribution of the naturally
occurring metals (e.g. mining, fuel combustion) and thus increase the likelihood of exposure to
toxic levels of these chemicals through inhalation or ingestion of food and water.

Many of the chemical residues in fish identified in this study were not unlike levels found in fish
from other studies in comparable aquatic environments in North America.  The results of this
study, therefore, have implications not only for tribal members but also the general public.

While contaminants remain in fish, it is useful for people to consider ways to still derive
beneficial effects of eating fish, while at the same time reducing exposure to these chemicals.
Fish are a good source of protein, low in saturated fats, and contain oils which may prevent
coronary heart disease.  Risks can be reduced by decreasing the amount offish consumed, by
preparing and cooking fish to reduce contaminant levels, or by selecting fish species which tend
to have lower concentrations of contaminants.

Reducing dietary exposure through cooking or by eating a variety of fish will decrease the
consumer's exposure, but not eliminate these chemicals from the environment. Reduction of
many of the man-made chemicals from the environment will take decades to centuries.
Regulatory limits for new waste streams and clean up of existing sources of chemical wastes can
help to reduce exposure. The exposure to naturally occurring chemicals can be reduced through
better management of our natural resources. The results of this study confirm the need for
regulatory agencies to continue to pursue rigorous controls on environmental pollutants and to
remove those pollutants which have been dispersed into our ecosystems.

There are many uncertainties in this risk assessment which could result in alternate estimates of
risk. These uncertainties include our limited knowledge of the mechanisms which cause disease,
the variability of contaminants in fish, changes in fish tissue concentrations over time, ingestion
rates, and the effects of food preparation. The uncertainties in our estimates  may increase or
decrease the risk estimates reported in this  study.

The chemicals which were estimated to contribute the most to potential health effects (PCB,
DDE, chlorinated dioxins and furans, arsenic, mercury) are the chemicals for which regulatory
strategies need to be defined to eliminate or reduce these chemicals in our environment.
                                            11-229

-------
12.0   References

Adams, WJ; Johnson, HE. (1977) Survey of the Selenium Content in the Aquatic Biota of
Western Lake Erie. J Great Lakes Res 3(1/2): 10-14.

Adolfson Associates, Inc. (1996) Technical Memorandum on the Results of the 1995 Fish
Consumption and Recreational Use Surveys - Amendment No. 1. Prepared for the City of
Portland. Adolfson Associates, Inc., Portland, OR.

Agency for Toxic Substances and Disease Registry (ATSDR) (1992) Toxicological Profile for
Vanadium. ATSDR/TP-91-2a.

ATSDR (1993) Toxicological Profile for Zinc. ATSDR/TP-93-15.

ATSDR (2000) Toxicological Profile for Chromium. ATSDR, Atlanta, GA.

American Cancer Society. (2002) What is Cancer? www.cancer.org

Ashraf, M; Jaffar, M. (1988) Weight Dependence of Arsenic Concentration in the Arabian Sea
Tuna Fish.  Bull Env Contam Toxicol 40(2): 219-225.

Ay, O; Kalay, M; Tamer, L; Canli, M. (1999) Copper and Lead Accumulation in Tissues of a
Freshwater Fish Tilapia Zillii and its Effects on the Branchial Na,K-ATPase Activity. Bull
Environ Contam Toxicol 62: 160-168.

Barraclough, WE.  (1964) Contributions to the Marine Life History of the Eulachon,
Thaleichthyspacificus. J. Fish.  Res. Board Can. 21(5): 1333-1337.

Baudo, R. (1983) Is Analytically-defined Chemical Speciation the Answer We Need to
Understand Trace Element Transfer Along a Trophic Chain?  In: Leppard GC ,ed. Trace Element
Speciation in Surface Waters and  Its Ecological Implications. Plenum Press, New York.

Baumann, PC;  Gillespie, RB. (1986) Selenium Bioaccumulation in Gonads of Largemouth Bass
and Bluegill From Three Power Plant Cooling Reservoirs. Environ. Toxicol. Chem 5:695-701.

Beamesderfer,  RCP; Fair, RA. (1997) Alternatives for the Protection and Restoration of
Sturgeons and their Habitat. Environ. Biol. Fish. 48:407-417.

Beamesderfer, RCP; Rien, TA; Nigro, AA. (1995) Differences in the Dynamics and Potential
Production of Impounded and Unimpounded White Sturgeon Populations in the Lower Columbia
River.  Trans. Am. Fish.  Soc  124(6): 857-872.

Beijer, K; Jennelov, A. (1979) Methylation of Mercury in Natural Waters.  In: Nriagu, JO (ed.).
The Biogeochemistry of Mercury in the Environment. Elsevier/North Holland  Biomedical Press,

                                          12-230

-------
New York, NY.

Beamish, RJ. (1980) Adult Biology of the River Lamprey (Lampetra ayesf) and the Pacific
Lamprey (Lampetra tridentata) from the Pacific Coast of Canada. Can. J. Fish. Aquat. Sci.
37:1906-1923.

Beamish, RJ; Levings, CD (1991) Abundance and Freshwater Migrations of the Anadromous
Parasitic Lamprey, Lampetra tridentata, in a Tributary of the Fraser River, British Columbia.
Can. J. Fish. Aquat. Sci. 48(7): 1250-1263.

Berti, PR; Receveur, O; Chan, HM; Kuhnlein, HV.  (1998)  Dietary Exposure to Chemical
Contaminants from Traditional Food Among Adult Deneetis in the Western Northwest
Territories, Canada. Environ Res 76(2): 131 -142.

Bessser, JM; Canfield TJ: LaPoint TW. (1993) Bioaccumulation of Organic and Inorganic
Selenium in a Laboratory Food Chain.  Environ Toxicol Chem 12(1): 57-72.

Bjorksten, JA. (1982)  Aluminum as a Cause of Senile Dementia.  Comprehensive Therapy 8(5):
73-76.

Bligh, EG; Armstrong, FAJ. (1971) Marine mercury pollution in Canada.  International Council
for Exploration of the  Sea.  Rep No. CM 1971/E34. .  Cited in National Research Council
Canada (1979) Effects of Mercury in the Canadian Environment. NRCC Report No. 16739.

Boothe, PN; Knauer, GA. (1972) The Possible Importance of Fecal Material in the Biological
Amplication of Trace and Heavy Metals. Limnology & Oceanography  17: 270-274.

Butler, PA; Schutzmann, RL. (1978)  Fish, Wildlife, and Estuaries: Residues of Pesticides and
PCBs in Estuarine Fish, 1972 - 76-National Pesticide Monitoring Program. Pestic Monit J 12(2):
51.

Beyerrum, RU. (1991) Vanadium.  In: Merian E, ed. Metals and Their compounds in the
Environment.  VCH Weinheim, New York-Basel-Cambridge ISBN 3-527-26521.

Busby, PJ; Wainwright, TC; Bryant, GJ; Lierheimer, LJ; Waples, RS; Waknitz, FW;
Lagomarsino. IV. (1996) Status Review of West Coast Steelhead from  Washington, Idaho,
Oregon, and California. U.S. Dept. Com., NOAA Tech. Mem. NMFS-NWFSC-27. 195 p.  + 6
appendices.

Camusso, M; Vigano, L; Balestrini, R. (1995) Bioconcentration of Trace Metals in Rainbow
Trout: a Field Study. Ecotoxicol Environ Saf 31(2):  133-141.

Carl, GC.  (1936) Food of the Coarse-Scaled Sucker (Catostomus macrocheilus Girard). J. Biol.
Board Can. 3(l):20-25.

                                          12-231

-------
Carlander, KD. (1969) Handbook of freshwater fishery biology.  Vol. 1, Iowa State Univ. Press,
Ames.

Center for Disease Control (CDC) (1991) Preventing Lead Poisoning in Young Children: A
Statement on Preventing Lead Poisoning in Young Children. Atlanta, GA.

Chovelon, A; George, L; Gulayets, C; Hoyano, Y; McGuinness, E; Moore, J; Ramamoorthy,  S;
Singer, P.  (1984) Pesticide and Poly chlorinated Biphenyl Levels in Fish from Alberta (Canada).
Chemosphere 13(1): 19.

Cirone, PA;  Kama, DW; Yearsley, JR; Falter, CM; Royer, TV. (2001) Ecological Risk
Assessment for the Middle Snake River, Volume 1. EPA/600/R-01017.

Clark, GM; Maret, TR. (1998) Organochlorine Compounds and Trace Elements in Fish Tissue
and Bed Sediments in the Lower Snake River Basin, Idaho and Oregon. USGS Water Resources
Investigations Report 98-4103.

Clemens, WA; Wilby. GV. (1967)  Fishes of the Pacific Coast of Canada.  Fish. Res. Board of
Can. Bull. No. 68.

Close, DA; Fitzpatrick, M; Li, H; Parker, B; Hatch, D; James, G. (1995) Status of the Pacific
Lamprey (Lampetra tridentata) in the Columbia River Basin. U.S. Department of Energy,
Bonneville Power Administration, Project No. 94-026, Contract No. 95B139067, Portland, OR,

Cogliano, VJ (1997) Plausible Upper Bounds: Are There Sums Plausible? Risk Analysis 17 (1):
77-84.

Colby, P.J., R.E. McNicol, and R.A. Ryder.  1979.  Synopsis of Biological Data on the Walleye
Stizostedion v. vitreum (Mitchell 1818). FAO United Nations, Fish. Synopsis No. 119, Ontario
Min. Natur. Resources, Fish. Res. Sect., Contrib. No. 77-13.  123 p. + 2 appendices.

Columbia River Intertribal Fish Commission (CRITFC) (1994) A Fish Consumption Survey of
the Umatilla, Nez Perce, Yakama, and Warm Springs Tribes of the Columbia River Basin.
Technical report 94-3. Columbia River Inter-Tribal Fish Commission, Portland, OR.

Coon, JC. (1978) Movement, Distribution, Abundance and Growth of White Sturgeon in the
Mid-Snake River. Doctoral Dissertation, University of Idaho, Moscow.

Currie, RS; Fairchild, WL; Muir, DCG. (1997) Remobilization and Export of Cadmium from
Lake Sediments by Emerging Insects. Environ Toxicol Chem 16(11):2333-2338.

Dabeka, RW; McKenzie, AD. (1995) Survey of Lead, Cadmium, Fluoride, Nickel and Cobalt in
Food Composites and Estimation of Dietary Intakes of these Elements by Canadians in 1986-
1988. JAOACInternational 78(4):897-909.

                                          12-232

-------
Dauble, DD. (1978) Comparative Ecology of Two Sympatric Catostomids, Catostomus
macrocheilus. and Catostomus columbianus. in the Middle Columbia River. M.S. Thesis,
Washington State Univ., Pullman, WA.

Dauble, DD. (1980) Life History of the Bridgelip Sucker in the Central Columbia River. Trans.
Amer. Fish. Soc. 109:92-98.

Dauble, DD (1986) Life History and Ecology of the Largescale Sucker (Catostomus
macrocheilus) in the Columbia River. Amer. Midland Natur.  116(2): 3 5 6-3 67.

Dauble, DD; Buschbom, RL. (1981)  Estimates of Hybridization Between Two Species of
Catostomids in the Columbia River.  Copeia (4):802-810.

Davies, K. (1990) Human Exposure Pathways to Selected Organochlorines and PCBs in Toronto
and Southern Ontario. In:  Nriagu JO and Simmons MS, eds. Food Contamination from
Environmental  Sources.  Adv Environ Sci Technol 23:525-540.

Devault, DS; Willford, WA; Hesselberg, RJ; Nortrupt, DA; Rundberg, EGS; Alwan, AK;
Bautista, C. (1986) Contaminant Trends in Lake Trout Salvelinus Namaycush from the Upper
Great Lakes: USA, Canada Arch Environ Contain Toxicol 15(4):349-356.

DeVore, JD; James, BW; Tracy, CA; Hale, DA. (1995) Dynamics and Potential Production of
White Sturgeon in the Unimpounded Lower Columbia River.  Trans. Am. Fish. Soc  124(6):845-
856.

Dietrich, W. (1995) Northwest Passage: the Great Columbia River.  Simon &  Schuster, NY.

Dobbs, MG; Cherry, DS; and Cairns, J. (1996)  Toxicity and bioaccumulation of selenium to a
three-trophic level food chain. Environ. Toxicol. Chem 15:340-347.

Duggan, RE; Lipscomb, GQ; Cox, EL; Heatwole, RE; Kling, RC. (1971) Pesticide Residue
Levels in the United States from July 1, 1963 to June 30, 1969. PesticMonitJ 5(2):73-113.

Ecological Analysts, Inc. (1981) The Sources, Chemistry, Fate and Effects of Chromium in
Aquatic Environments. American Petroleum Institute, 2101 L St. NW, Washington, D. C. 20037.

Eddy, S; Underbill, JC. (1976) Northern Fishes. Third Ed., Univ. Minnesota Press, Minneapolis,
MN.

Eisler, R. (1985a).  Cadmium Hazards to Fish, Wildlife and Invertebrates: A Synoptic Review.
U. S. Fish and Wildlife Service Biological Report 85(1.2).

Eisler, R. (1985b)  Selenium Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review.
U. S. Fish and Wildlife Service, Biological Report 85(1.5).

                                           12-233

-------
Eisler, R. (1986) Polychlorinated biphenyl hazards to Fish, Wildlife, and Invertebrates: A
Synoptic Review.  U. S. Fish and Wildlife Service Biological Report 85(1.7).

Eisler, R. (1988a) Arsenic Hazards to Fish, Wildlife and Invertebrates: A Synoptic Review. U.
S. Fish and Wildlife Service Biological Report 85(1.12).

Eisler, R. (1988b)  Lead Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review. US
Fish and Wildlife Service Biological Report 85(1.14).

El Nabawi, A; Heinzow, B; Kruse, H. (1987) Arsenic, Cadmium, Copper, Lead, Mercury and
Zinc in Fish from the Alexandria Region, Egypt. BullEnv Contam Toxicol 39(5): 889-897.

Eschmeyer, PH. (1950)  The Life History of the Walleye in Michigan. Michigan Dep. Conserv.,
Inst. Fish. Res., Bull. 3.

EVS (1998) Willamette River Basin Studies, Human Health Technical Study, Willamette River
Qualitative Fish Consumption Survey. Prepared for Oregon Department of Environmental
Quality, Salem, OR. EVS Environment Consultants, Seattle, WA.

EVS (2000)  Human Health Risk Assessment of Chemical Contaminants in Four Fish Species
from the Middle Willamette River, Oregon. Prepared for Oregon Department of Environmental
Quality, Portland, OR. EVS Environment Consultants, Seattle, WA.

Fairey, R; Taberski, K; Lamerdin, S; Johnson, E; Clark, RP;  Downing, JW; Newman, J; Petreas,
M. (1997) Organochlorines and other Environmental Contaminants in Muscle Tissues of
Sportfish Collected from San Francisco Bay. Mar Pollut Bull 34(12): 1058-1071.

Fenske, RA; Lu, C; Simcox, NJ; Loewenherz, C; Touchstone, J. (2000) Strategies for Assessing
Children's Organophosphorus Pesticide Exposures in Agricultural Communities. J Expo Anal
Environ Epidemiol 10:662-71.

Gartrell, MJ; Craun,  JC; Podrebarac, DS; Gunderson, EL.  (1985)  Pesticides, Selected Elements
and other Chemicals in Infant and Toddler Total Diet Samples: October 1978 - September 1979.
JAOAC 68(5): 123-144.

Gartrell, MJ; Craun,  JC; Podrebarac, DS; Gunderson, EL.  (1986)  Pesticides, Selected Elements
and other Chemicals in Infant and Toddler Total Diet Samples: October 1980 - March 1982.
JAOAC 69(1): 146-169.

Goldstein, GW. (1990)  Lead Poisoning and Brain Cell Function. Environ Health Perspect 89:
91-4.

Gossett, RW; Brown, DA; Young, DR. (1983) Predicting the Bioaccumulation of Organic
Compounds in Marine Organisms Using Octanol-Water Partition Coefficients.

                                          12-234

-------
Mar Pollut Bull 14(10):387-392.

Gray, RH; Dauble,DD. (1977) Checklist and Relative Abundance of Fish Species from the
Hanford Reach of the Columbia River. NWSci.  51:208-215.

Great Lakes Water Quality Board (1989) Report to the International Joint Commission.

Gutenmann, WH; Lisk, DJ. (1991) Higher Average Mercury Concentration in Fish Fillets After
Skinning and Fat Removal. J. Food Safety 11:99-103.

Hamilton,  SJ; Buhl, KJ; Faerber, NL; Wiedmeyer, RH; Bollard, FA.  (1990)  Toxicity of Organic
Selenium in the Diet to Chinook Salmon.  Environ Toxicol Chem 9(3):347-358.

Hamilton,  SJ; Waddell, B. (1994) Selenium in Eggs and Milt of Razorback Sucker (Xyrauchen
texanus) in the Middle Green River, Utah. Arch Environ Contam Toxicol 27(2): 195-201.

Hamir, AN; Sullivan, ND; Handson, PD.  (1982) The effects of Age and Diet on the Absorption
of Lead from the Gastrointestinal Tract of Dogs. Australian Veterinary Journal 58:266-268.

Harris, SG; Harper, BL. (1997)  A Native American Exposure Scenario. Risk Analysis 17(6):789-
795.

Hart, J.L.  1973.  Pacific Fishes of Canada.  Bull. Fish. Res. Board Can. No. 180. 740 p.

Healey, M. C.  1991. Life History of Chinook Salmon (Oncorhynchus tshawytscha\ p. 311-393.
In C. Groot and L. Margolis (editors) Pacific Salmon Life Histories, Univ. British Columbia
Press, Vancouver, BC.

Heiny, JS; and Tate, CM. (1997)  Concentration, Distribution and Comparison of Selected Trace
Elements in Bed Sediment and Fish Tissue in the South Platte River Basin, USA, 1992- 1993.
Arch Environ Contam Toxicol 32(3): 246-259.

Henderson, C; Foster, RF. (1957) Studies of Smallmouth Black Bass (Micropterus dolomieu) in
the Columbia River near Richland, Washington. Trans. Amer. Fish. Soc. 86:112-127.

Hodson, PV; Whittle, DM; Wong, PTS; Borgman, U; Thomas, RL; Chau, YK; Nriagu, JO;
Hallet. DJ. (1984)  Lead Contamination of the Great Lakes and its Potential Effects on Aquatic
Biota. In: Nriagu JO; Simmons MS eds., Toxic Contaminants in the Great Lakes, John Wiley &
Sons, New York, NY.

Hubbs, C.L., and R.M. Bailey.  1938. The Small-mouthed Bass.  Cranbrook Inst. Sci. Bull.
No. 10. 89 p.

Hudson, RJM; Gherini, SA; Fitzgerald, WF; Porcella, DB. (1995) Anthropogenic Influences on

                                          12-235

-------
the Global Mercury Cycle: A Model-Based Analysis. Water Air SoilPollut 80(l-4):265-272.

Hunter, RG; Carroll, JH; Butler, JS.  (1981)  The Relationship of Trophic Level to Arsenic
Burden in Fish of a Southern Great Plains Lake, USA. JFreshwater Ecol 1(1): 121-127.

ICF Kaiser (1996) Toxicity and Exposure Concerns Related to Arsenic in Seafood: An Arsenic
Literature Review for Risk Assessments. Prepared for U.S. Environmental Protection Agency,
Seattle, WA. ICF Kaiser, Seattle, WA.

International Agency for Research on Cancer (IARC) (1978)  Monographs on the Evaluation of
Carcinogenic Risk of Chemicals to Man. Poly chlorinated Biphenyls and Polybrominated
Biphenyls. Vol 18:62.  World Health Organization, Lyons, France

IARC (1980)  Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Man.
Some Metals and Metallic Compounds. Vol. 23: 325-415. World Health Organization, Lyons,
France

Jaffe, R; Stemmler, EA; Eitzer, BD; Kites, RA. (1985) Anthropogenic Polyhalogenated Organic
Compounds in Sedentary Fish from Lake Huron and Lake Superior: USA-Canada Tributaries and
Embayments.  Great Lakes Res 11 (2): 156-162.

Johnson,  A. (2001). An Ecological Hazard Assessment for PCBs in the Spokane River.
Washington Department of Ecology, No. 01-03-015.

Kan, TT.  (1975) Systematics, Variation, Distribution, and Biology of Lampreys of the Genus
Lampetra in Oregon. Ph.D. Thesis,  Oregon State Univ., Corvallis, OR.

Kiffney, P; Knight, A.  (1990) The Toxicity and Bioaccumulation of Selenate, Selenite and
Seleno-L-Methionine in the Cyanobacterium Anabaena Flos-Aquae. Arch Environ Contain
Toxicol  19(4):488-494.

Kissinger, L; Beck, N.  (2000) Evaluation of Cadmium, Lead and Zinc Contamination in
Spokane River Spokane, Spokane County, Washington pp. 1-40. Washington State Department of
Health, Washington Department of Ecology,. Olympia, WA.

Konasewich, D; Traversy, W; Zar, H. (1978) Status Report on Organic and HeavyMmetal
Contaminants in the Lakes Erie, Michigan, Huron and Superior Basins.  Great Lakes Water
Quality Board, p. 195-257.

Kononen, DW. (1989) PCBs and DDT in Saginaw Bay, Michigan, USA White Suckers.
Chemosphere 18(9-10):2065-2068.

Lang, WL; Carriker, RC.  (1999)  A Resurgent Columbia River: An Introduction. In Lang, WL;
Carrider, RC, eds. Great River of the West: Essays on the Columbia River. Univ. Washington

                                          12-236

-------
Press, Seattle, WA.

Lanphear, BP; Matte, TD; Rogers, J; Clickner, RP; Dietz, B; Bornschein RL, Succop P, Mahaffey
KR, Dixon S, Galke W, Rabinowitz M, Farfel M, Rohde C, Schwartz J, Ashley P, Jacobs DE.
(1998)  The Contribution of Lead-Contaminated House Dust and Residential Soil to Children's
Blood Lead Levels. A Pooled Analysis of 12 Epidemiologic Studies. Environ Res 79:51-68.

Lanphear, BP; Dietrich, K; Auinger, P; and Cox, C. (2000) Cognitive Deficits Associated with
Blood Lead Concentrations <10 microg/dL in US Children and Adolescents. Public Health Rep,
115: 521-9.

Lee, DS; Gilbert, CR; Hocutt, CH; Jenkins, RE; McAllister, DE; Stauffer Jr.,JR. (1980) Atlas of
North American Freshwater Fishes. Pub. No. 1980-12 North Carolina Biol. Surv., North Carolina
State Mus. Natur. Flist., Raleigh, NC.

Lemly, AD.  (1985)  Toxicology of Selenium in a Freshwater Reservoir:  Implications for
Environmental Hazard Evaluation and Safety. Ecotoxicol Environ Saf 10(3):314-338.

Lemly, AD; Smith, GJ.  (1987)  Aquatic Cycling of Selenium: Implications for Fish and Wildlife.
US Fish and Wildlife Service Leaflet (12): 1-10.

Lepla, JB. (1994) White Sturgeon Abundance and Associated Habitat in Lower Granite
Reservoir, Washington. M.S. Thesis, University of Idaho, Moscow, ID 77p.

Levine, EP. (1961) Occurrence of Titanium, Vanadium, Chormium, and Sulfuric Acid in the
Asddian Eudistoma ritteri.  Science 133:1352-1353.

Lowe, TP; May, RW; Brumbaugh, WE; Kane, DA. (1985) National Contaminant Monitoring
Program: Concentrations of 7 Elements in Freshwater Fish, 1978- 1981. Arch Environ Contain
Toxicol 14(3):363-388.

Lunde, G (1970)  Analysis of Arsenic and Selenium in Marine Raw Materials. J Sci Food
Agric 21-.242-247.

Ma, M;  Le, XC.  (1998) Effect of Arsenosugar Ingestion on Urinary Arsenic Speciation.
Clinical Chemistry 44 (3):539 -  550.

Mance, G. (1987) Pollution Threat of Heavy Metals in Aquatic Environments. Elsevier Applied
Science. New York.

Mann, DL. (1988)  Analysis of Boron in Contaminated Shrimp by Inductively Coupled Plasma
Spectroscopy.  Spectroscopy 3(3):37-39.

Manton, WI; Angle, CR; Stanek,  KL; Reese, YR; Kuehnemann, TJ. (2000) Acquisition and

                                          12-237

-------
Retention of Lead by Young Cildren. Environ Res 82:60-80.

May, JT; Hothem, RL; Alpers, CN; Law, MA. (2000) Mercury Bioaccumulation in Fish in a
Region Affected by Historic Hold Mining: the South Yuba River, Deer Creek, and Bear River
Watershed, California, 1999. US Geological Survey, Open-file Report 00-367. USGS,
Sacramento, Calif.

McCabe Jr., GT; Emmett, RL; Hinton,SA . (1993) Feeding ecology of juvenile white sturgeon
(Acipenser transmontanus) in the Lower Columbia River.  NW Sci. 67(3): 170-180.

Meehan, WR; Bjornn, TC .  (1991) Salmonid Distributions and Life Histories, p. 47-82. In W.R.
Meehan (editor) Influences of Forest and Rangeland Management on Salmonid Fishes and their
Habitats. Amer. Fish. Soc. Spec. Pub.  19.

Miles, RL.  (1977) Commercial Fishery Investigations; Completion Report, National Marine
Fisheries Service, Washington, DC.

Moore, JW; Ramamoorthy S. (1984)  Heavy Metals in Natural Waters: Applied Monitoring and
Impact Assessment.  Spring-Verlag, New York, NY. Cited  in: SAIC (1993) Draft Biological
Evaluation for Reissuance of a National Pollutant Discharge Elimination System Permit for the
Potiatch Corporation, Lewiston, Idaho.  Science Applications International Corporation. San
Diego, CA.

Morgan,!;  Berry, M.; Graves, R. (1997) Effects of Commonly Used Cooking Practices on Total
Mercury Concentration in Fish and their Impact on Exposure Assessments. Journal of Exposure
Analysis and Environmental Epidemiology 7(1):119-133.

Moore, J.W., and J.M. Mallat. 1980. Feeding of Larval Lamprey. Can. J. Fish. Aquat. Sci.
37:1658-1664.

Moyle, PB (1976) Inland Fishes of California. Univ. Calif.  Press, Berkeley.

Munn, MD; Cox, SE; Dean, CJ. (1995)  Concentrations of Mercury and Other Trace Elements in
Walleye, Smallmouth Bass, and Rainbow Trout in Franklin D. Roosevelt Lake and the Upper
Columbia River, Washington, 1994.  USGS Open File Report 95-195. USGS, 12101 Pacific Ave.,
Ste 600, Tacoma, WA 98402.

Munn, MD; Short, TM.  (1997)  Spatial Heterogeneity of Mercury Bioaccumulation by Walleye
in Franklin D. Roosevelt Lake and the Upper Columbia River, Washington. Transactions of the
American Fisheries Society 126:477-487.

Munn, MD; and Gruber, SJ. (1997) The Relationship Between Land Use and Organochlorine
Compounds in Streambed Sediment and Fish in the Central Columbia Plateau, Washington and
Idaho, USA. Env Tox andChem 16:1877-1887.

                                          12-238

-------
Munn, MD.  (2000)  Contaminant Trends in Sport Fish from Lake Roosevelt and the Upper
Columbia River, Washington, 1994-1998. USGS Water Resources Investigations Report 00-
4024.

Musick, JA; Harbin, MM; Berkeley, SA; Burgess, GH; Eklund, AM; Findley, L; Gilmore, RG;
Golden, JT; Ha, DS; Huntsman, GR; McGovern, JC; Parker, SJ; Poss, SG; Sala ,E; Schmidt, TW;
Sedberry, GR; Weeks, H; Wright, SG. (2000) Marine, Estuarine, and Diadromous Fish Stocks at
Risk of Extinction in North America (Exclusive of Pacific Salmonids). Fisheries 25(11):6-30.

Nakamoto, RJ; Hassler, TJ. (1992)  Selenium and Other Trace Elements in Bluegills from
Agricultural Return Flows in the San Joaquin Valley, California. Arch Environ Contam Toxicol
22(l):88-98.

National Academy of Sciences (NAS) (1975) Report on the Committee on Medical and
Environmental Pollution: Nickel.  Cited in: National Toxicology Program (NTP) (1979)
Chemical Selection Profile: Nickel Chloride.

NAS (1983) Risk Assessment in the Federal Government: Managing the Process. National
Academy Press. Washington, D.C.

NAS (1990) Committee on the Biological Effects of Ionizing Radiation (BEIR), Health Effects
of Exposure to Low Levels of Ionizing Radiation (BEIR V, 1990).

National Research Council (NRC) (1977) Drinking Water and Health. Vol 1. National
Academy Press, Washington, DC.

NRC (1993) Measuring Lead Exposure in Infants, Children, and Other Sensitive Populations.
National Academy Press: Washington, D.C.

National Council on Radiation Protection and Measurements (NCRP) (1997) Report 126.
Uncertainties in Fatal Cancer Risk Estimates Used in Radiation Protection, October 1997.

Netboy, A. (1980)  The Columbia River Salmon and Steelhead Trout, their Fight for Survival.
Univ. Washington Press,  Seattle, WA.

Patten, BG; Thompson, RB; Gronlund, WD. (1970) Distribution and Abundance of Fish in the
Yakima River, Wash., April 1957 to May 1958.  U.S. Fish Wildl. Serv., Spec. Sci. Rep., Fish. No.
603.

Pauley, GB; Bortz, BM; Shepard, MF. (1986) Species Profiles: Life Histories and Environmental
Requirements of Coastal  Fishes and Invertebrates (Pacific Northwest) - Steelhead Trout. U.S.
Fish Wildl. Serv. Biol. Rep. 82(11.62). U.S. Army Corps of Engineers, TREL-82-4.

Peterson, LK; D'Auria, JM; McKeown, BA; Moore, K;  Shum, M. (1991) Copper Levels in the

                                          12-239

-------
Muscle and Liver of Farmed Chinook Salmon, Oncorhynchus tshawytscha. Aquaculture 9:105-
115.

Pettit, SW; Wallace, RL. (1975)  Age, Growth, and Movement of Mountain Whitefish,
Prosopium williamsoni (Girard), in the North Fork Clearwater River, Idaho.  Trans. Am. Fish.
Soc. 104(l):68-76.

Peven, CM.  (1990) The Life History of Naturally Produced Steelhead Trout from the Mid-
Columbia River Basin.  M.S. Thesis, Univ. Washington, Seattle, WA.  .

Fletcher, TF (1963) The Life Flistory and Distribution of Lampreys in the Salmon and Certain
Other Rivers in British Columbia. M.S. Thesis, Univ. British Columbia., Vancouver, BC

Ponce, R; Bloom, NS. (1991) Effect of pH on the Bioaccumulation of Low-level, Dissolved
Methyl Mercury in Rainbow Trout (Oncorhynchus mykiss). Water Air Soil Pollut 56(0):631-640.

Queiroloa, F; Stegen, S; Restovica, M;  Paza, M; Ostapczuk, P; Schwugerb, MJ; Munozc, L.
(2000) Total Arsenic, Lead, and Cadmium Levels in Vegetables Cultivated at the Andean
Villages of Northern Chile.  Sci Total Environ 255(l-3):75-84.

Robinson, BH; Brooks, RR; Outred, HA; Kirkman, JH. (1995) Mercury and Arsenic in Trout
from the Taupo Volcanic Zone and Waikato River, North Island, New Zealand.  Chemical
Speciation andBioavailability  7(l):27-32.

Rodier, PM. (1995) Developing Brain as a Target of Toxicity. Environ Health Perspect 103(6):
73-6.

Roff, TJ; Mate, BR. (1984) Abundances and Feeding Habits of Pinnipeds in the Rogue River,
Oregon. J. Wildlife. Management. 48(4): 1262-1274.

Rogan, WJ; Dietrich, KN; Ware, JH; Dockery, DW; Salganik, M;  Radcliffe, J; Jones, RL; Ragan,
NB; Chisolm, JJ; Rhoads, GG. (2001)  The Effect of Chelation Therapy with Succimer on
Neuropsychological Development in Children Exposed to Lead. NEnglJMed: 344: (19) 1421-60

Rosen, JF; Mushak, P.  (2001) Primary Prevention of Childhood Lead Poisoning — The Only
Solution.  N EnglJMed 344:1470-1471.

Saiki, MK; Castleberry, DT; May, RW; Martin, BA; Bollard, WN. (1995) Copper, Cadmium
and Zinc Concentrations in Aquatic Food Chains from the Upper Sacramento River (Calif) and
Selected Tributaries.  Arch Env Contam  Toxicol 29(4):484-491.

Sandercock, FK.  (1991) Life History of Coho Salmon (Oncorhynchus kisutch)., p. 395-445. In
C. Groot and L. Margolis (editors) Pacific Salmon Life Histories. Univ. British Columbia Press,
Vancouver.

                                          12-240

-------
Scheuhammer, M.  (1991) Effects of Acidification on the Availability of Toxic Metals and
Calcium to Wild Birds and Mammals. Environ Pollut 71 (2-4): 3 29-3 76.

Schmitt, CJ; Ludke, JL; Walsh, DF. (1981) Organochlorine Residues in Fish: National Pesticide
Monitoring Program, 1970 - 74. PesticMonitJ 14(4):36-155.

Schmitt, CJ; Zajicek, JL; Ribeck, MA. (1985) National Pesticide Monitoring Program: Residues
of Organochlorine Chemicals in Freshwater Fish: 1980-1981. Arch Environ Contam Toxicol
14(2):225-260.

Schmitt, CJ; Zajicek, JL; May, TM; Cowman, CF; (1999a) Organochlorine Residues and
Elemental Contaminants in U.S. Freshwater Fish, 1976-1986: National Contaminant
Biomonitoring Program:  Reviews of Environmental Contamination and Toxicology 162:43-104

Schmitt, CJ; Bartish, TM; Blazer, VS; Gross, TS; Tillitt, DE; Bryant,  WL; DeWeese, LR.
(1999b) Biomonitoring of Environmental Status and Trends (BEST) Program: Contaminants and
their Effects in Fish from the Mississippi, Columbia, and Rio Grande Basins. Cited in:
Morganwalp, D.W. and H.T. Buxton, eds.  U.S. Geological Survey Toxic Substances Hydrology
Proceedings of the Technical Meeting.  Charleston, S.C., March 8-12, 1999. Volume 2 of  3?
Contamination of Hydrologic Systems and Related Ecosystems. U.S. Geological Survey-Water
Resources Investigations Report 99-4018B.

Schramm, FfL; Ednoff, M; French, B. (1984) Depredation of Channel Catfish by Florida
Double-crested Cormorants.  Prog. Fish-Cult. 46(l):41-43.

Scott, WB. (1960)  Summaries of Current Information on Round Whitefish and Mountain
Whitefish. Ontario Dep. Lands Forests, Res. Inform. Paper (Fish.) No. 8.

Scott, WB; Grossman, EJ. (1973)  Freshwater Fishes of Canada. Fish. Res. Board Can.
Bull. No. 184.

Seller, HG; Sigel, H. (1988) eds.,  Handbook on the Toxicity of Inorganic Compounds. Marcel
Dekker, Inc. New York.

Semakula, SN; Larkin, PA.  (1968) Age, growth, food, and yield of the white sturgeon
(Acipenser transmontanus) of the Fraser River, British Columbia. J. Fish. Res. Board    Can. 25:
2589-2602.

Serdar, D; Johnston, J; Mueller, K; Patrick, G. (2001) Mercury Concentrations in Edible Muscle
of Lake Whatcom Fish.  Washington Department of Ecology Publication No. 01-03-012

Sherlock, JC.  (1987)  Lead in Food and the Diet.  Environ Geochem Health 9(2):A3-77.
                                          12-241

-------
Sigler, WF; Sigler, JW. (1987) Fishes of the Great Basin, a Natural History. Univ. Nevada
Press, Reno, NV.

Simpson, JC; Wallace RL. (1982) Fishes of Idaho. Univ. of Idaho Press, Moscow, ID

Smith, WE; Saalfeld, RW.. (1955) Studies on Columbia River Smelt, Thaleichthys pacificus
(Richardson).  Washington Dep. Fish., Fish. Res. Papers l(3):3-26.

Sokal, R; Rohlf, JF.  (1981) Biometry, 2nd Edition. W.H. Freeman and Company.

Spalinger, SM; von Braun, MC; Petrosyan, V; von Lindern, IH.  (2000)  A Comparison of House
Dust and Soil Lead Levels in Northern Idaho to the Bunker Hill Superfund Site. Unpublished
Manuscript.

Spehar, RL; Fiandt, JT; Anderson, RL; DeFoe, D. (1980) Comparative Toxicology of Arsenic
Compounds and their Accumulation in Invertebrates and Fish. Arch Environ Contam Toxicol
9(l):53-63.

Tabor, R.A., R.S.  Shively, and T.P. Poe.  1993. Predation of Juvenile Salmonids by Smallmouth
Bass and Northern Squawfish in the Columbia River near Richland, Washington. N. Amer. J.
Fish. Manage. 13:831-838.

Tetra Tech (1996) Assessing Human Health Risks from Chemically Contaminated Fish in the
Lower Columbia River. Prepared for Lower Columbia River Bi-State Water Quality Program,
Portland, OR and Olympia, WA. Tetra Tech, Inc., Redmond, WA.

Thompson, G.E., and R.W. Davies.  1976. Observations on the Age, Growth, Reproduction, and
Feeding of Mountain Whitefish (Prosopium williamsonf) in the Sheep River, Alberta. Trans. Am.
Fish. Soc. 105(2):208-219.

USEPA (1980a) Upgrading Environmental Radiation Data. EPA 520/1-80-012.

USEPA (1980b) Ambient Water Quality Criteria Document; DDT. EPA 440/5-80-038.

USEPA (1980c) Ambient Water Quality Criteria Document: Beryllium.  EPA 440/5-80-024.

USEPA (1980d) Biological Monitoring of Toxic Trace Elements Report. EPA 600/3-80-090.

USEPA (1983) Health Assessment Document: Nickel. EPA 600/8-83-012.

USEPA (1984a) Risk Assessment and Management: Framework for Decision Making.  EPA
600/9-85-002

USEPA (1984b) Mercury Health Effects Update. EPA 600/8-84-019f

                                         12-242

-------
USEPA (1985)  Ambient Water Quality Criteria  for Mercury - 1984. EPA 440/5-84-026.

USEPA (1986a) Guidelines for Health Risk Assessment of Chemical Mixtures. Federal Register
51(185):34014-34025.

USEPA (1986b) Guidelines for Carcinogen Risk Assessment. Federal Register 51(185): 33992-
34003.

USEPA (1987a) Drinking Water Criteria Document for Polychlorinated Biphenyls (PCBs). EPA
ECAO-CIN-414.

USEPA (1987b) Health Issue Assessment: Copper.  EPA 600/8-87/001

USEPA (1987c) Ambient Water Quality Criteria for Zinc. EPA 440/5-87-003.

USEPA (1989) Risk Assessment Guidance for Superfund. Volume I. Human Health Evaluation
Manual (Part A). Interim Final. EPA/549/1-89/002.

USEPA (1990) National Oil and Hazardous Substances Pollution Contingency Plan. Federal
Register 40(55):8666-8865.

USEPA (1991) Workshop Report on Toxicity Equivalency Factors for Polychlorinated Bipenyl
Congeners. EPA/625/3-91/020.

USEPA (1992a) National Study of Chemical Residues in Fish. EPA 823-R-92-008a.

USEPA (1992b) Guidance on Risk Characterization for Risk Managers and Risk Assessors. F.
Henry Habicht, Deputy Administrator, U.S.  Environmental Protection Agency, Washington, DC.

USEPA (1993) Provisional Guidance for Quantitative Risk Assessment of Polycyclic Aromatic
Hydrocarbons. ECAO-CIN-842.

USEPA, (1994a) Integrated Exposure Uptake Biokinetic Model for Lead in Children (IEUBK)
version. 99d. U.S. EPA. Washington, DC.  www.epa.gov/superfund/programs/lead/adult.htm

USEPA,  (1994b) Guidance Manual for the Integrated Exposure Uptake Biokinetic Model for
Lead in Children. U.S. EPA. Washington, DC.  www.epa.gov/superfund/programs/lead/adult.htm

USEPA (1994c). Health Effects Summary Tables.  Superfund Slope Factors for Radionuclide
Carcinogenicity. EPA/540/R-94/059.

USEPA (1994d).  Columbia River Basin Contaminant Database: Data Abstract Report.  USEPA,
Office of Water, Wasington, D.C.,  internal  report.
                                         12-243

-------
USEPA (1995) Policy for Risk Characterization at the U.S. Environmental Protection Agency.
Carol M. Browner, Administrator, USEPA, Washington, DC.

USEPA (1996a) PCBs: Cancer Dose-Response Assessment and Application to Environmental
Mixtures. EPA 600-P-96-001F. U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment, Washington, DC.

USEPA (1996b)  Recommendations of the Technical Review Workgroup for Lead for an Interim
Approach to Assessing Risks Associated with Adult Exposures to Lead in Soil, internal EPA
report,  www.epa.gov/superfund/programs/lead/adult.htm

USEPA (1997a) Method 1668,  Toxic Polychlorinated Biphenyls (PCBs) by Isotope Dilution
HRGC/HRMS, Draft Revision, March.

USEPA (1997b) Exposure Factors Handbook. Volume HI. Activity Factors. EPA/600/P-
95/002Fc.

USEPA (1997c) Exposure Factors Handbook. Volume I. General Factors. EPA/600/P-95/002Fa.

USEPA (1997d) Health Effects Assessment Summary Tables - FY 1997 Update. EPA 540-R-97-
036.

USEPA (1998a) Guidance for Conducting Fish and Wildlife Consumption Surveys. EPA 823-B-
98-007.

USEPA (1998b)  Clarification to the 1994 Revised Interim Soil Lead Guidance for CERCLA
Sites and RCRA Corrective Action Facilities. EPA-OSWER Directive #9200:4-27

USEPA (1998c) Assessment of Dioxins, Furans, and PCBs in Fish Tissue From Lake Roosevelt,
Washington,  1994. USEPA Region 10, internal  report.

USEPA (1999a)  USEPA, Frequently Asked Questions (FAQs) on the Adult Lead Model,
Technical Review Workgroup for Lead, Internal EPA Report.
www.epa.gov/superfund/programs/lead/adult.htm

USEPA. 1999b. Use of the Technical Review Workgroup for Lead, Interim Adult Lead
Methodology in Risk Assessment, Internal EPA Report.
www.epa.gov/superfund/programs/lead/adult.htm

USEPA (1999c) Cancer Risk Coefficients for Environmental Exposure to Radionuclides.  EPA
402-R-99-001.

USEPA (1999d) Federal Guidance 13. Cancer Risk Coefficients for Environmental Exposure to
Radionuclides. EPA402-R-99-001.

                                        12-244

-------
USEPA (1999e) The National Survey of Mercury Concentrations in Fish. Database Summary
1990-1995. EPA 823-R-99-014.

USEPA (2000a) Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories.
Volume 2. Risk assessment and fish consumption limits. Third Edition. EPA 823-B-00-008.

USEPA (2000b) Estimated Per Capita Fish Consumption in the United  States. EPA-821-R-00-
025.

USEPA (2000c) Integrated Risk Information System (IRIS), www.epa.gov/iris

USEPA (2000d) The National Dioxin Study of 1990.   EPA/600/3-90/022.

USEPA (2000e) Exposure and Human Health Reassessment of 2,3,7,8 Tetrachlorodibenzo-/?-
dioxin (TCDD) and Related Compounds (Draft Final) EPA/600/p-00/001B..

USEPA (2000f) Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories.
Volume 1. Fish Sampling and Analyses, Third Edition.  EPA 823-B-00-007.

USEPA (2000g) Supplementary Guidance for Conducting Health Risk Assessment of Mixtures.
EPA/630/R-00/002

USEPA (2001) Cacodylic Acid -Re- evaluation- Report of the Hazard Identification Assessment
Review Committee. HED Document No. 014468.

US Department of Health and Human Services (USDHHS). (1999) Toxicological Profile for
Lead. U.S. Department of Health and Human Services: Atlanta, GA.

US Geological Survey (USGS) (1992) Surface Water Quality Assessment of the Yakima River
Basin, Washington. Water Supply Paper #2354-B

Van den Berg, M; Birnbaum, L; Bosveld, ATC; Brunstrom, B; Cook, P; Feeley, M; Giesy, JP;
Hanberg,  A; Hasegawa, R; Kennedy, SW;  Kubiak, T; Larsen, JC; van Leeuwen, FXR; Djien
Liem, AK; Nolt, C; Peterson, RE;  Poellinger, L; Safe, S; Schrenk, D; Tillitt, D; Tysklind, M;
Younes, M; Waern, F; Zacharewski, T.  (1998) Toxic Equivalency Factors (TEFs) for PCBs,
PCDDs, PCDFs for Humans and Wildlife.  Environ. Health Perspect 106(12):775-792.

Washington Department of Health (WDOH)  (1997) Consumption Patterns of Anglers Who
Frequently Fish Lake Roosevelt. Office of Environmental Health Assessment Services, Olympia,
WAGartrell, MJ et al.  1986. JAOAC 69:146-169.

Weast, RC.  (1988) CRC Handbook of Chemistry and Physics. 69th Ed. CRC Press, Inc. Boca
Raton, FL.
                                         12-245

-------
West, J; O'Neill S; Lippert, G; Quinnell, S. (2001) Toxic Contaminants in marine and
Anadromous Fishes from Puget Sound, Washington, Washington Dept of Fish and Wildlife,
Olympia, WA.

Wiener, JG;  Spry, DJ.  (1996) lexicological Significance of Mercury in Freshwater Fish.  In:
Beyer, WN; Heinz, GH; Redmon-Norwood, AW, eds. Environmental Contaminants in Wildlife:
Interpreting Tissue Concentrations. Special Publication of the Society of Environmental
Toxicology and Chemistry.  Less Publishers, Boca Raton, Florida

Wilbur, CG  (1980)  Toxicology of Selenium: A Review,  din. Toxicol 17:171-230.

Withler, I.L.  1966. Variability in Life History Characteristics of Steelhead Trout (Salmo
gairdnerf) along the Pacific Coast of North America. J. Fish. Res. Board Can. 23(3):365-393.

World Health Organization (WHO) (1976)  Environmental Health Criteria: Mercury. No. 1,
Geneva, Switzerland.

WHO (1992) Environmental Health Criteria: Cadmium: Environmental Aspects. No. 135,
Geneva, Switzerland

Wren, CD; MacCrimmon, HR; Loescher, BR. (1983) Examination of Bioaccumulation and
Biomagnification of Metals in a Precambrian Shield Lake.  Water Air Soil Pollution 19(3):277-
291.

Wren, CD; MacCrimmon, HR. (1986) Comparative Bioaccumulation of Mercury in Two
Adjacent Freshwater Ecosystems.  Water Res 20(6):763-770.

Wydoski, RS; Whitney, RR. (1979) Inland Fishes of Washington. Univ. Washington Press.

Yeardley, RB; Lazorchak, JM; Paulsen, SG. (1998) Elemental Fish Tissue Contamination in
Northeastern US Lakes: Evaluation of an Approach to Regional Assessment. Environ  Toxicol
Chem 17(9): 1875-1884.

Ysart, G; Miller, P; Croasdale, M; Crews, H; Robb, P; Baxter, M; De L'Argy, C; Harrison, N.
(2000) 1997 UK Total Diet Study-Dietary Exposures to Aluminum, Arsenic, Cadmium,
Chromium, Copper, Lead, Mercury, Nickel, Selenium, Tin and Zinc.  FoodAddit Contain
17(9):775-786.

Zimmerman, MP.  (1999) Food Habits of Smallmouth Bass, Walleye, and Northern Pikeminnow
in the Lower Columbia River Basin during Outmigration of Juvenile Anadromous Salmonids.
Trans. Amer. Fish. Soc. 128:1036-1054.
                                          12-246

-------