UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Environmental impacts and economic costs : A study of pulp mill effluent in British Columbia Lempriere, Tony Christopher 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1995-0182.pdf [ 8.1MB ]
JSON: 831-1.0086813.json
JSON-LD: 831-1.0086813-ld.json
RDF/XML (Pretty): 831-1.0086813-rdf.xml
RDF/JSON: 831-1.0086813-rdf.json
Turtle: 831-1.0086813-turtle.txt
N-Triples: 831-1.0086813-rdf-ntriples.txt
Original Record: 831-1.0086813-source.json
Full Text

Full Text

E N V I R O N M E N T A L I M P A C T S A N D E C O N O M I C COSTS: A STUDY O F P U L P M I L L E F F L U E N T IN BRITISH C O L U M B I A  By TONY CHRISTOPHER LEMPRIERE B . A . , University of Toronto, 1985 M . A . , Queen's University, 1988  A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF M A S T E R O F SCIENCE in T H E F A C U L T Y O F G R A D U A T E STUDIES (Resource Management and Environmental Studies)  We accept this thesis as conforming to the required standard  T H E UNIVERSITY O F BRITISH C O L U M B I A April 1995 ® Tony Lempriere, 1995  In presenting this jthesis in partial fulfillment of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t j t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d | by the head of my department or by h i s or her representatives. I t i s understood t h a t copying o r p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be allowed without my written|permission. j  Department of  ^p^mi/Jol /^QftctjX<frWf c^^^^nu^on/T^^ticJ-  The U n i v e r s i t y of B r i t i s h Columbia Vancouver/ Canada Date  Do  A p r s i  (  WwdcQ  11  •  ABSTRACT  Lempriere, T . C . [1995]. Environmental impacts and economic costs: A study of pulp mill effluent in British Columbia.  Master of Science Thesis, Resource Management and Environmental Studies,  University of British Columbia, Vancouver.  In this thesis I study the economic cost of the environmental impacts of pulp and paper mill effluent in British Columbia. The thesis is not a benefit-cost analysis of the industry. Rather, my primary objective is to assess what the process of trying to estimate the costs reveals about the difficulties and limitations in the economic analysis of environmental damages.  I review the environmental impacts of pulp mill effluent in British Columbia.  There is a great deal of  uncertainty in many respects about these impacts, but it is clear that in some cases they have been significant. A considerable body of literature addresses the theory and methodology of how to measure such impacts in economic terms. I draw upon this literature in four case studies of certain environmental impacts and specific types of economic damage. The case studies examine 1) dissolved oxygen reductions in Alberni Inlet and the impact on sport salmon fishing; 2) chlorinated dioxin contamination in Howe Sound and commercial shellfishing ground closures; 3) chlorinated dioxin contamination in the Columbia River and the impact on sport fishing; and 4) fish tainting at Kitimat and its effect on the Haisla people. In each case study, I estimate a range of values for the economic cost of the specific environmental impact in question. In each, there are significant and difficult questions which must be addressed.  I conclude by noting that there are many instances where credible economic value estimates of environmental damages can be derived.  Even partial and incomplete estimates can be helpful in demonstrating that the  environment is an important source of economic value. However, the case studies suggest that five potentially severe sets of problems will invariably accompany economic analysis of environmental impacts. Thefirstfour  ;  U  — complexity, uncertainty, site-specificity  J  and limited information — are typical of most studies of  environmental issues, but economic analysis adds a new dimension to each. unique to economic analysis.  The fifth set of problems is  These difficulties limit the ability of economists to undertake comprehensive  analyses of the environment or environmental degradation.  :  1Y_  TABLE OF CONTENTS ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF TABLES  vi  LIST OF FIGURES  vii  ACKNOWLEDGEMENTS  viii  1. INTRODUCTION  1  1.1 Goal and Motivation 1.2 The British Columbia Pulp and Paper Industry  1 5  1.3 Thesis Outline  7  2. THE ENVIRONMENTAL IMPACTS OF PULP MILL EFFLUENT IN BRITISH COLUMBIA  11  2.1 De-Oxygenating Substances 2.2 Total Suspended Solids (TSS) 2.3 Chemical Contaminants 2.4 Summary 3. THE ECONOMIC COST OF WATER POLLUTION  13 19 22 36 37  3.1 3.2 3.3 3.4 3.5  Types of Economic Value The Theory of Economic Value Measuring Economic Value Measurement Issues in Assessing The Economic Cost of Water Pollution Summary  41 43 48 53 65  4. DISSOLVED OXYGEN REDUCTIONS AND SPORT FISHING - ALBERNI INLET  67  4.1 Background 4.2 Economic Valuation 4.3 Summary 5. DIOXIN CONTAMINATION AND SHELLFISHERY CLOSURES - HOWE SOUND 5.1 Background 5.2 Economic Valuation 5.3 Summary 6. DIOXIN CONTAMINATION AND CONSUMPTION ADVISORIES - COLUMBIA RIVER 6.1 Background 6.2 Economic Valuation 6.3 Summary  67 78 89 90 90 96 103 105 105 110 118  ,  ; ,  Y.  ;  7. FISH TAINTING A N D FOOD FISHING - KTTIMAT 7.1 7.2 7.3  Background Economic Valuation Summary  8. DISCUSSION  __ 119 119 124 129 130  8.1  Case Study Results  130  8.2  Difficulties in the Economic Analysis of Pulp Mill Pollution  138  9. CONCLUSION  144  REFERENCES CITED  147  APPENDIX A . FISHERY CLOSURES A N D FISH CONSUMPTION ADVISORIES, BRITISH COLUMBIA, 1988 T O FEBRUARY 1995  161  •  ;  Vi  ;  LIST OF TABLES Table 1.1  British Columbia Pulp and Paper Mills, 1994  6  Table 2.1  British Columbia Pulp Mills: Production and B O D and TSS Discharges, 5  1980, 1990 and 1994  15  Table 2.2 British Columbia Pulp Mills: A O X Discharges, 1989, 1991 and 1994  24  Table 2.3  British Columbia Pulp Mills: Average Acute Toxicity, 1987, 1990 and 1994  28  Table 3.1  Assessing the Economic Cost of Water Pollution  38  Table 3.2  Components of the Economic Value of Aquatic Environmental Services  42  Table 3.3  Summary of Common Environmental Damage Valuation Techniques  50  Table 3.4 Examples of Economic Costs Due to Water Pollution  54  Table 4.1  70  Chinook and Sockeye Run, Alberni Inlet, 1956-93  Table 4.2 Dissolved Oxygen Criteria for Salmonids  72  Table 4.3  79  Sport Catch and Effort, and Components of Sockeye Catch, Alberni Inlet, 1956-94  Table 4.4 Factors, Alternative Assumptions and Estimated Economic Cost to Sport Fisheries of 1990 Sockeye Mortality in Alberni Inlet Table 4.5 Estimates of WTP Values Relevant to Alberni Inlet Sport Salmon Fishing  81 86  Table 5.1  Shrimp, Prawn and Crab Commercial Harvests in British Columbia and D F O Management Area 28, 1981-94 Table 5.2 Factors, Alternative Assumptions and Estimated Economic Cost to Commercial Shellfishers of Chlorinated Dioxin-Related Closures, Howe Sound  98  Table 5.3  Estimated Geographical Distribution of Area 28 Shellfish Catch Prior to Closures  99  Table 6.1  Factors, Alternative Assumptions and Estimated Economic Cost of Reduced Sport Fishing Due to Chlorinated Dioxin-Related Consumption Advisories, Lower Columbia River  Table 6.2 The Effects of Pollution on Lower Columbia Sport Fishing, 1990 Table 7.1  Table 8.1  93  111 116  Factors, Alternative Assumptions and Estimated Economic Cost to the Haisla of Eulachon Tainting at Kitimat, 1973-1995  125  Summary of Case Study Cost Estimates  131  Table 8.2 Potential Economic Costs of Pulp Mill Effluent at Case Study Sites  134  .  •.  •.  yjj  :  LIST OF TABLES (Continued)  Table A . 1 Fishery Closures and Consumptions Advisories Due to Chlorinated Dioxin Contamination, Coastal British Columbia Locations, 1988-1995  162  Table A.2 Fish Consumption Advisories Due to Chlorinated Dioxin Contamination, Interior British Columbia Locations, 1989-1994  164  LIST OF FIGURES  Figure 3.1  Supply, Demand and Economic Value  44  Figure 4.1  Port Alberni and Area  68  Figure 4.2 .  Dissolved Oxygen Saturation by Depth, Port Alberni, 1977-88  74  Figure 5.1  Howe Sound  91  Figure 6.1  Lower Columbia River and Area  106  Figure 7.1  Kitimat and Area  120  ACKNOWLEDGEMENTS I especially wish to thank my supervisor, Ilan Vertinsky, for his support arid financial assistance. I also wish to thank the other members of my Supervisory Committee ~ Ken Hall, Les Lavkulich and Peter Nemetz ~ for their confidence and helpful suggestions. Numerous officials from the Federal Department of Fisheries and Oceans, Environment Canada, the British Columbia Ministry of Environment, Lands and Parks, and the British Columbia Ministry of Agriculture, Fisheries and Food provided a great deal of helpful information without which much of my research could not have been completed. I would like to express my deepest appreciation to Jerry Carter for his unflagging support and confidence throughout the preparation of this thesis.  1 1. INTRODUCTION  1.1.  Goal and Motivation  In this thesis I ask what are some of the economic costs of the environmental impact of pulp and paper mill effluent in British Columbia?  I intend not to provide a definitive answer to this question but rather to  examine certain specific impacts and, in doing so, to investigate some of the difficulties that arise in the economic analysis of the industry's effluent.  By environmental impact I refer to obvious effects such as the destruction or alteration of natural habitats, mortality of aquatic organisms and changes in species mix, but also to less obvious chronically toxic effects which kill or impair over a longer period. Often, concerns about the environmental impacts of wastewaters focus on human health rather than the health of the aquatic environment.  By economic cost I refer to the economic value lost to British Columbia as a result of pulp and paper mill environmental impacts. Applied welfare economics suggests that the conceptually appropriate measures of such costs are reductions in consumer and producer surplus.  For example, an activity such as salmon  harvesting or boating along a river provides a value or benefit to its participants and society. However, the actual benefit derived, the net economic value or surplus, is the total value less the cost of undertaking the activity.  It is the reduction in these economic values due to mill-related environmental damage that I  investigate.  The importance and size of the pulp and paper industry in Canada, and the extent of its impact on the environment, have meant that it has been of particular regulatory and public concern. Regulatory activity during this decade has substantially reduced pulp mill pollution by requiring full effluent treatment and by forcing various production process changes. A number of economic analyses have sought to assess this and  2 earlier regulatory activity. To cite recent examples, Sinclair [1990] reviewed the economic and environmental performance of mills in the context of the 1971 Federal Pulp and Paper Effluent regulations; the process of regulation leading to the 1992 Federal regulations has been described by Stanbury [1993] and Harrison [1993]; and Inch [1993] studied the process of organochlorine regulation in Canada.  The studies by Sinclair and  Harrison also looked at the extent of compliance by mills with the 1971 effluent regulations regarding the three "conventional" pollutant parameters, biochemical oxygen demand (BOD), total suspended solids (TSS) and acute toxicity.  In terms of measuring the cost of compliance, McCubbin et al. [1990] prepared mill-by-mill estimates associated with the proposed Federal regulations and Sonnen and Laurence [1990] estimated the macroeconomic impact.  H.A.Simons Ltd. [1992] estimated the cost-of-compliance with British Columbia's  Adsorbable Organic halogen (AOX, a measure of organochlorine in effluent) regulations.  Thus there has  been considerable examination of the way in which the pulp and paper industry is regulated, and of the costs imposed on the industry by regulation.  What is striking is that little attempt has been made to examine the economic cost to society of the environmental damage associated with mill effluent (or, the economic benefit associated with varying degrees of regulation).  The problem is that the environment is seldom explicitly included in markets, making the  economic analysis of environmental damage costs extremely difficult. For example, during the preparation of the 1992 Federal Pulp and Paper Regulations, Environment Canada contracted a consultant to produce national estimates of the economic benefits of the regulations, but the task was found to be too difficult [Stanbury 1993].  Similarly, the Federal Department of Fisheries and Oceans (DFO) was asked by  Environment Canada to estimate the value for recreational and commercial fisheries of more stringent effluent regulations, but was unable to do so. In the end, the Regulatory Impact Analysis Statement prepared for the regulations noted only that the benefits to users and non-users of water courses were significant but nonquantifiable [Canada Gazette, Part II, 126(11), 1941-2006, 7 May 1992]. Stanbury [1993] severely criticized  3 the lack of any detailed study of the economic benefits of the regulations, and indeed efforts have been made outside of Canada. For example, in 1993, the United States Environmental Protection Agency (EPA) [1993] published an assessment of the economic benefits of reduced pollution associated with proposed U.S. federal regulations to limit pulp and paper industry air and water pollution. In the assessment the EPA attempted an analysis of several benefit categories but also noted that it was impossible to derive a full benefit study. Moreover, derivation of some of the benefits relied on very significant assumptions, yielding what arguably may be considered rather meaningless estimates. Rick Freeman, a respected environmental economist who has spent much of his career studying the measurement of environmental economic value, has noted that (t)he difficulty in tracing out the effects of the discharge of a pollutant on the many parameters of environmental quality and, in turn, their effects on man's uses of the environment, substantially limit our ability to do careful benefit-cost analyses of environmental quality and improvements. This is not fully appreciated by many advocates of greater use of benefit-cost analysis in this field [1993, p.33].  Clearly, pulp mill effluent has the potential to cause a wide variety of environmental impacts, especially at effluent outfall locations, but also at considerable distances from discharge sites. Reductions in the dissolved oxygen content of water can create lethally hypoxic conditions. Suspended solids increase turbidity and reduce primary productivity. Settling solids destroy benthic habitat and smother bottom-dwelling organisms. An immense array of chemical compounds contaminate water, sediments and the biota.  These contaminants  include the relatively persistent, toxic and bioaccumulative synthetic chlorinated hydrocarbons such as the chlorinated dioxins, furans and phenolics.  Biological uptake of persistent compounds can lead to  bioaccumulation up the food chain. Some of the compounds are known or suspected to be carcinogenic in humans and other organisms at minute doses. Lesions and other physiological damage to fish, and elevated biochemical processes, provide evidence of the stress associated with mill effluent.  Not all of these effects  are observed at all mills, but all mills cause some degree of environmental degradation.  Environmental impacts impose economic costs on society.  Impaired fish stocks affect the quantity that  commercial, recreational and First Nations anglers can harvest. Consumption advisories affect the quality of the harvest, the value of recreational activity and the cultural and food value of fish for First Nations  4 communities.  Problems of taste and odour in fish and drinking water occur.  Effluent reduces aesthetic  appreciation of the environment. It compromises the benefit that people derive from simply knowing that a particular environment is pristine.  The goal of this thesis is not to demonstrate that these economic costs do or do not justify a certain level of regulation in British Columbia. Nor do I wish to demonstrate that pulp mill effluent has imposed either an unacceptably high or reasonably low level of cost compared to the benefits of pulp mill production, benefits that are especially important in the many communities which rely heavily on a pulp mill and the associated activities. This thesis is not a benefit-cost assessment. My objectives are more modest ~ they are to present estimates of specific costs and to examine what the process of deriving these results tells us about the issues that arise in attempting to translate environmental impacts into economic costs.  My motivation lies in a recognition that environmental problems involve a complexly inter-related web of ecological, economic, social and political factors. No environmental problem can be fully resolved without recognizing this and recognizing the limits of any single disciplinary approach.  Full solutions require  interdisciplinary or multidisciplinary approaches. With respect to economic analysis this means, first, that analysis must be grounded in a sound understanding of the non-economic dimensions of the problem at hand. It means, second, that economic analysis is at least as limited and problematic as the analysis of any other single discipline.  Concerns with the economic analysis of environmental impacts focus on the fact that such analyses cannot account for the complexity and multifunctionality of the environment, nor for our incomplete understanding of how we affect the environment. Economic valuation is by nature reductionist and directed toward analysis of efficiency — the optimization of net economic benefits ~ but other societal goals, such as social, environmental and economic sustainability often may be deemed more important.  Moreover, economic  analysis usually cannot fully measure the economic effects of environmental impacts; inevitably, substantial  5 uncertainty exists.  Two solutions exist to counter these concerns. The first is to attempt to make better use of economic analysis, to understand the limitations and problems in its use. A second and complementary solution is to ensure that the decision-making process itself is open, and explicitly considers a variety of criteria and interests, including economic ones, but others as well. The work in this thesis is meant to be part of the first solution.  1.2.  The British Columbia Pulp and Paper Industry  Pulp and paper production is one of the largest manufacturing industries both in Canada and in British Columbia.  In 1993, the industry employed about 19,000 people in the province (11% of manufacturing  employment) and shipped $4.1 billion of product (15% of manufacturing output), much of it exported [B.C. Ministry of Finance and Corporate Relations 1994]. The indirect and induced effects of this activity are significant and affect all areas of the province — in the manufacturing sector, only the wood industry is more important. Pulp and paper production also plays a crucial role in many communities: three "single-company" towns (Port Alice, Gold River, Powell River) rely on the local pulp mill for over one-quarter of their direct employment [DREE 1979, Pharand 1988]. Many other towns also rely to a significant extent on employment from the local pulp and paper mill and on employment in the forestry industry which supplies the mill.  Table 1.1 lists the 23 pulp and paper mills which currently operate in the province outside the Vancouver area . Ten mills discharge to marine or estuarine environments and 13 operate in the interior of the province. 1  The first pulp mill in the province was built in 1894 near Port Alberni. It had a relatively short life but the  Four mills operate in the Vancouver area. In New Westminster, Scott Paper produces hardwood mechanical pulp and paper products and NewsTech Recycling produces de-inked newsprint pulp. Paperboard Industries operates in Burnaby and Island Paper Mills operates on Annacis Island; neither produce pulp. These mills are not included in the discussion in this thesis because they are much smaller than most other mills in the province, they produced qualitatively different products (paper only, hardwood pulp, recycled newsprint pulp), and the impact of their effluent can not be isolated from that of the other very substantial municipal and industrial discharges in the Vancouver area. 1  6  I  CN -H CN Os os os OS OS OS  n n n « & & & £  3 85 PH CO  -I  m « vi OS OS 00 Os Os os  00 -H JN r- J CJ Os Os Os  PH  VI  CO  8  O N  O  SO 00 SO CN 00 00 SO SO SO t"» 00 so oo Os OS OS OS Os Os Os  •3 a 2 00  o  00 SO CN -H OO 00 so so t— 00 SO 00 Os OS Os Os Os Os  vs O CN O vo r— r- t— Os OS Os Os  'S >  o n •a •O 60 a e s o s co —. 2 o S "5 § S> •a 3 •3 J S «  11  IS  ii  On a  1  l l  5?3  3 ' OH  T3  ll  >  e 3 o co  u  i fra55 P i l s  II  l l  2  m  a  PH S OH  2  1111-  HF 8 5  ,1 O  1  I PH  8 0 0 0 0 0 0  mm-HOTi-Ood '-' TJ- v> CN m CN v» 0  •O OH  V  t~©oosomcN-HcnsocN OCAHMTf\OnoO>0^ M O v i f i c i H i n ^ ^ M  1-3 H m n « H ^ S i n H H O » c VI W W H TJ- -H TJ- Tf CN — CN CN -- •- -  e  '1  e 0  a ti U  I  J2  1  S3 PH  s 3 e  a.  e  o O  «*  S  51  •a > PH -O0 —  U  s  « « « ~ ePQ CQ  £  0  ON ON ON ON ON OS ON  PH  2 <"  U O  i^  PH 5 TJ HH •a •a § OS " c •S 8 g £• O" O-PH tA - 1 S .9 O PH PH PH 3 a '3 u S ^ o 3 O J ^ PH OH U  iI  1  o CQ r-; •* Os  3 00 3 \o 3 rCQ H 00 00 ) H o vo <o ^ <o h> 00 CO  i  1  II W  3 e a, ^ '•8 3 PH U v. IHo g PH CH l « S O S o 2o  I 2 > ^  ^ o 2 C S .2 •3 PH V ,03 OU ^ C l l a  S  s  •a  a!  I I •§ §  S'gz •S PH O "3 fi  S 2 .-2  z u <y o PH 1-0  1  E E  PS  3  PH •o §_  60 60 W O H H O O PH O O c a - o £••0 W 5 •3 3 O U ra .-a P co 3 iH | H o o J= J3 1 1 1 1 1 1 o o > i— — -S! O *H  •c •S H  ti s CQ" CQ" 5 <S " . . . a a J2 G  I*  03 3 o  IH  11  S  03 CQ 2  vn o s t a g *  PH CO Q  u  PH  42 £2  ti t : a] a) J2 JB CQ PQ  43 ti Is lis OJ 03 • J2 5 PQ CQ PQ co ca  ,s a if  o  g  o | . 6  M . .a =5  £ *5 b  6  P.  6 6  W O BC PH OH OH tstt ti CH J  0 0 0 ,y 60 60 6« 3 U U fc- ^  2I  in  oa £  S I S -B -8 1 8 8 8 8 8 -g 2 £ £ £ a a S h  O to  a  7 first two decades of the twentieth century brought the construction of a series of mills on Vancouver Island and the southern mainland coast. In the 1950s several more mills were built on the coast including two of the largest at Crofton and Elk Falls. It was not until 1960 that the first interior mill was built, at Castlegar on the Columbia River.  This was followed in quick succession over the following 12 years by the  construction of ten new mills, nine at inland locations and one at Gold River on Vancouver Island. Two inland mills were built in the 1980s, at Quesnel and Taylor, and in 1991 a mill was completed at Chetwynd. This last mill produces a specific type of pulp which allows it to be effluent-free.  In 1992, conditional  governmental approval was given for the construction of a zero-effluent integrated pulp and paper mill between Kitimat and Terrace, but the project is on hold.  1.3. Thesis Outline  The key questions of this thesis are:  1)  What are the environmental impacts of pulp and paper mill effluent in British Columbia?  2)  What can be said about the economic costs of the impacts?  3)  What issues arise in attempting to meaningfully analyze the impacts in economic terms?  In Chapter 2,1 answer the first of the key questions by summarizing current knowledge regarding the aquatic environmental impacts of the mills shown in Table 1.1. sound understanding of these impacts.  The valuation of economic costs must start with a  I focus on the effects of dissolved oxygen reductions, TSS and  chemical contamination. The chapter shows the great diversity and site-specificity of the impacts and the significant uncertainty that exists with respect to their exact nature, especially at levels of greatest relevance to many economic considerations (i.e., fish populations).  8 Chapter 3 turns to the economic side of the issue. Within the literature on environmental valuation there are a great many issues of theory, method, data use and data interpretation. A full review of these issues lies beyond the scope of this thesis and I present only the basic concepts relevant for the later discussion. The purpose of this thesis is not to evaluate the theory itself but to examine what its application in the "real world" tells us about the difficulties encountered in the economic valuation of water pollution.  The next four chapters present case studies ~ retrospective analyses of some recent economic costs associated with mill effluent in British Columbia.  The case studies directed the course of the overall research and  provide a means of using applied economic analysis of specific environmental impacts to illustrate the issues and difficulties that arise in this sort of work.  The first case study (Chapter 4) examines the effect of DO reductions on sport fishing value in Alberni Inlet. Here the pronounced effect of effluent on water quality has demonstrable consequences for the biota and these consequences can be discussed in terms of the economic value of recreational fishing, among other things. The other three case studies are somewhat simpler since economic valuation starts with the known impact of effluent on human uses of the aquatic environment.  In addition, the consequences of effluent for aquatic  organisms themselves are not an issue. In Chapter 5, the second case study examines commercial shellfishing ground closures in Howe Sound. The economic effects arise because one particular environmental impact (accumulation of chlorinated dioxins) has potential human health consequences, and this has led to closures. In Chapter 6, the third study considers the impact of chlorinated dioxin-related fish consumption advisories on sport fishing on the Columbia River between Castlegar and the Canada-U.S. border. Here again the concern is with human health but the consumption advisories also raise the issue of the importance of perceptions with respect to certain types of economic value.  In Chapter 7, the final study looks at the  economic consequence of tainting of the First Nations eulachon harvest at Kitimat.  Here, the economic  damage from tainting arises not from human health concerns (as in the previous two case studies), nor from demonstrable effects of effluent on the biota (as in the first case study), but from the effect of effluent on the  9 quality of fish for human uses.  In each case study I first present background information which describes the environmental and economic context of the economic cost estimate. Each study uses general approaches and results found in the existing theoretical, methodological and applied economic literature to guide the analysis.  Each requires that  assumptions, sometimes very significant ones, must be made about various factors in the analysis, and the combination of these assumptions yields the cost estimate. Often each factor can not be determined with any great degree of certainty ~ there are usually reasons to think that a particular assumption may create an upward bias in the resulting value estimate, as well as reasons to think that the assumption may create a downward bias. Thus, in each case study I present what I called intermediate assumptions; these generate a point estimate of the economic cost. Given the wide degree of uncertainty about the various considerations, as well as the need for sensitivity analysis, I also present low and high assumptions for each factor, and corresponding low and high economic cost estimates. These extreme values are meant to capture at least some of the possible biases in the point estimates. They represent a "reasonable" range for the costs, and sensitivity analysis of any particular assumption probably will yield cost estimates that fall inside this range. In some cases, an intermediate assumption reflects merely an intermediate point between the assumed low and high possibilities.  In a few cases, most notably with respect to a critical relationship in the Howe Sound case  study, I use an ad hoc assumption that is meant to be illustrative of a possibility rather than a foundation for a highly credible estimate.  Note that each of these case studies addresses only the economic value of fishing activity (commercial, recreational and First Nations food fisheries). There are, of course, many other sources of value associated with aquatic environments, and many other potential environmental consequences of pulp and paper mill effluent.  M y choice of these particular case studies is, in itself, instructive as to the limits and problems  confronting the analyst who attempts to bridge the gap between environmental impacts and economic analysis. None of these case studies is meant to be definitive. Time restrictions prevented extensions of the research  10 and analysis which would have yielded more complete and credible results; nevertheless, the case studies are indicative of the difficult and sometimes intractable problems that arise in this sort of analysis, even when the potential economic impact appears well-defined.  The case study results should not be taken out of the context in which they were derived. They are estimates of very specific economic costs resulting from specific environmental impacts. They depend heavily on the information available to me, and on the assumptions made. I did not examine other potential economic costs associated with these environmental impacts and others. Indeed, often such costs may not be calculable, for one reason or another.  In Chapter 8,1 comment on the case study results and on the difficulties and limits they reveal in the economic analysis of environmental impacts. Many of these difficulties — those related to uncertainty, complexity, lack of data and the site-specific nature of many environmental problems ~ are found in most environmental studies, although economic analysis adds new dimensions to each. As well, there are a number of important issues and difficulties unique to economic analysis. In Chapter 9,1 conclude the thesis with general comments on environmental valuation.  11 2. T H E E N V I R O N M E N T A L I M P A C T S O F P U L P M I L L E F F L U E N T IN BRITISH C O L U M B I A  In this chapter I survey pulp mill effluent environmental impacts in British Columbia in terms of 1) deoxygenating substances, primarily biochemical oxygen demand (BOD ); 2) total suspended solids (TSS); and 5  3) chemical contaminants.  Several factors influence the environmental impact of pulp mill discharges.  Pollutant loading and character depend upon the site-specific characteristics of the pulping, bleaching and treatment processes. In the environment, the persistence, fate and ultimate impact of pollutants is mediated by other sources of anthropogenic stress as well as by the biological, chemical and hydrological characteristics of the receiving environment. Thus pulp production and pollutant loading do not necessarily bear any simple relationship to environmental impact ~ the effluent character and the assimilative capacity of the receiving environment can vary considerably. Also, loadings vary greatly from mill to mill and over time ~ this means that average loadings data are indicative only of typical discharges, and can hide large variations.  The production of pulp is a complex process yielding wastewaters at many points, and mills discharge very substantial effluent volumes.  On British Columbia's coast, only the municipal discharges of Victoria and  Vancouver match the permit effluent volumes of each mill [Kay 1989]. Final mill effluent reflects the type of wood used and the wood preparation, pulping, bleaching and treatment processes. A range of pulping processes exist to separate wood fibres and break down the lignin that binds them together. These are broadly classified as mechanical or chemical.  The simplest process involves feeding logs or chips through a  mechanical grinder with water to create mechanical pulp (MP). The thermomechanical pulp (TMP) process involves first softening the wood chips with steam at high pressure.  In the chemothermomechanical pulp  (CTMP) process, the softening of the chips is achieved by cooking briefly with a dilute chemical solution. The T M P and C T M P processes have become increasingly common over the last 15 years ~ in British Columbia, the three mills constructed since 1980 produce only these types of pulp.  However, the most  common pulping process remains the kraft or sulphate chemical process. Kraft pulping separates wood fibres by dissolving lignin in an alkaline solution at a high temperature. A second and increasingly less common  12 chemical pulping process is sulphite pulping which uses an acid cooking solution. In British Columbia only the Port Alice mill currently produces pulp of this type. Eighteen of the 23 mills shown in Table 1.1 produce kraft pulp and 11 of the 23 mills produce mechanically derived pulps.  The colour of pulp ranges from the dark brown of kraft pulps to the near-white of some mechanical pulps. For many purposes, these colours are satisfactory, but for other purposes the mill bleaches the pulp, traditionally using gaseous molecular chlorine (Clj). Wood fibres are inherently white — it is primarily the lignin which creates colour as a result of transformations undergone in chemical pulping. Thus bleaching of chemical pulps is the norm, while bleaching of mechanical pulp occurs less frequently.  As concern has  increased regarding the environmental impacts of the chlorinated organic compounds discharged by mills using C l , a variety of modifications to the traditional bleaching processes have been made, and new processes have 2  been developed.  The most widely practised adjustment to bleach plant operations in British Columbia has  been the substitution of chlorine dioxide (C10 ) for molecular chlorine. The use of C10 reduces the overall 2  2  production of chlorinated organics and decreases the degree of chlorination of the chlorinated organic compounds that are produced, thus reducing overall toxicity [Solomon et al. 1993]. Only two of the 23 mills shown in Table 1.1 do no bleaching and 17 of the remaining 21 use chlorine of some sort in their bleaching operations. 50-100% C10 substitution is now standard practice at all mills except Port Alice [Government 2  of Canada 1995].  The quality and quantity of pulp mill effluent can be altered by a wide variety of process adjustments within the mill, of which the substitution of C10 for C l is just one. 2  2  Once the effluent leaves the pulp mill it is  usually treated to varying degrees and this too significantly alters its quality. Table 1.1 shows the year in which full primary and secondary treatment were first installed at mills in British Columbia. All interior mills with the exception of the Castlegar mill have had full primary and secondary treatment since start-up. In contrast, at the coastal mills, full treatment has come on-line only since 1990, although a number of mills have had partial treatment for some time. The Skookumchuk mill differs from all others in the province in that  13 effluent is only discharged during periods of high river flow. At low flow, effluent is discharged to a ground disposal system [C. Johnson, B . C . Environment, Nelson, pers. comm., 29 March 1994].  Primary treatment clarifies effluent by removing about 80-95% of the TSS, with a resulting fall in B O D of 5  about 10% [McCubbin 1990]. From the perspective of chemical contamination, primary-treated effluent can be considered untreated since the short retention time in primary clarifiers prohibits significant contaminant reduction [McLeay 1987]. Microorganisms (bacteria, algae, protozoa, fungi) which biodegrade effluent form the basis of secondary treatment systems, of which there are several types. The efficiency of such systems varies greatly from mill to mill and over time, but well-designed and operated secondary treatment substantially reduces deoxygenating substances and reduces toxicity by 50-90% [McCubbin 1990, CEPA 1991]. After full treatment only residual amounts of B O D and TSS remain, along with nutrients, colour and 5  a host of inorganic and organic contaminants. The placement of effluent diffusers also can significantly reduce effluent impact. Relative to a submerged diffuser, a surface diffuser generally results in much slower effluent dilution and more severe effects on near-shore life.  2.1.  De-Oxygenating Substances  The dissolved oxygen (DO) balance in water is determined by a set of processes which govern the supply of and demand for oxygen.  Photosynthetic inputs and diffusion from the atmosphere represent the principal  sources of oxygen with the former usually being the major source.  The supply of oxygen, and its  hydromechanical distribution in the aquatic environment, is counter-balanced by chemical demands related to oxidation of organic and inorganic matter and by the consumptive use of aquatic organisms. Pulp mill effluent can reduce oxygen supply through direct toxicity to the photosynthetic biota or through effluent colour which decreases light transmission, and hence reduces photosynthesis. For example, reductions in phytoplanktonic activity attributable to effluent have been observed near the outfalls at Crofton, Powell River, Port Alberni,  14 Port Mellon and Woodfibre [Colodey and Wells 1992]. Effluent colour is known to have an impact in some of these cases.  However, more important than reductions in the supply of DO are increases in D O demand. Mill effluent increases D O demand primarily in one of two ways: 1) from dissolved organic matter which exerts a shortterm biochemical oxygen demand, usually measured over 5 days (hence BOD ); and 2) from the slow 5  decomposition of wood fibres which settle out and exert a long-term demand . The long-term effects, related 2  to TSS deposition, are described in the next section. In British Columbia, only the municipal discharges of the Lower Mainland and Victoria generate B O D loadings greater than those of individual pulp and paper 5  mills.  Chemical pulping processes tend to create a significantly greater volume of waste organics than  mechanically-based processes. Among chemical processes, sulphite mills traditionally discharge much greater amounts of B O D than do kraft mills (roughly 10 times as much per unit of effluent) because sulphite waste 5  liquor is much less amenable to recycling than is the case with kraft liquors [Waldichuk 1983].  Table 2.1 shows B O D loading and pulp production data between 1980 and 1994 for the 21 direct discharge 3  mills outside the Vancouver area (the two Canfor mills at Prince George use the same effluent treatment system and are treated as one mill). The data show that BOD discharges have fallen over time for most mills, 5  both in absolute terms and per unit of pulp produced. Substantial reductions in loading can occur when mill processes change or with the institution of new treatment facilities.  For example, the recent installation of  full secondary treatment facilities at most of the coastal mills reduced B O D considerably between 1990 and 5  1994. In 1990, total loading from all mills amounted to an average of 370 tonnes per day (t/d), over 80%  A third type of oxygen demand induced by effluent can occur if the nutrients released in effluent stimulate eutrophic processes. Nutrients are sometimes added to effluent treatment facilities to promote biological degradation of organic matter, but usually this is not considered to pose a significant hazard to the receiving environment. One exception in British Columbia occurred in 1972 in Kamloops when a massive algal bloom in the Thompson River below Kamloops Lake was coincident with a five-fold expansion of the bleached kraft mill [Servizi 1989, Bothwell 1992]. Fish kills were observed. Bothwell [1992] attributed the bloom to mill effluent, although municipal effluent is also discharged to the Thompson River. He has shown that a release of a small amount of phosphorous from the mill would have been sufficient to cause the bloom. 2  15 cn  cnWvoc<ovnodr~>nQ  "?r-.vqoot-;OO q'' :r-;Ov"^ v  uivri<Nrr'vni^-^ov--<cn _*_HrH*-4*-HrTCS*-*CSOV  60  cs o\ © vo <N cn so ^ <e r i « - •* n »  co H  ,  i  '—•  P  «  ° . " J OV  cs  S n o V h n O i d o  CN  O  m  •—•  (  1 ' r n *-<  I  8  oo »  oo ov oo Tt; vo cn <"":  vq ov  viovcjOTrcnov^J^t-^csP,  S ov  vo n  8  *  »  o  n  IF,  -  0\  ^  © © cn 1  vr; ON  vq  rvo  cncsp^t^vrjovpoocn-^  •"tovpvqvq^iovqt-;  1  o o O V T t - T T ^ H ^ O v ^ r o ^ * N n N ^ Tf VO Tf rCN CN CN CN ov -H — m cs  o 00 OS  ee S3  •s a  " O C J N o i S ^ o ' v .  o 60  j  ^vqvq  vo Ov  n;cni>cScn  Q O  - " n n t N d * v o - < n c > «  pvqvqovvocnoowicn c^vdod-- 'c>od<~- o T f - t t - H C N o o c N r o  v^o\i^oOTfcnooiriTti>-p Scncn-HrtTrvdcn-^vdiN  :  -s  o  a, 50  n u u a e o  v  ©  0 0  o 00 ov  vncncni^cnovncSr^Ov o o i ^ n o o ' t N ' j t S v o m w • ^ • c n ^ - H F - H r ^ ^ H t s c n ^  |  OvrlvOOvoOrtvOc^OVVO  cn - O e e 2 o •S o o *-> T3  S o f- Zl ~* ov Si 22 ov r- r! oo S f- » •—' CN '  8  "S  8 "C a  o  00  3 •* «  ov <N  8  2 ™ M r- o 3  VO  ^ t n fv r o 5 « vj t O •» cn i— cs £ O~ VO ~5 5 rgr 8 cn <Nt> oo »n * ?  * oo TI-  s  ca  e 0 CA  5 el i  B O . 23 oi 5 .a = •a «  53  2  cS  o  tt ti ti ji J? & & J? O H Oil ( X C L ,  <s CN OV  2  i  i  s  '-3  c -a  6§ .is S IS S  "1 cs  EE P  3 g-  §1 3 o H  3  -o o o i- >  3  I  S  |-o W c vi^oJtNinfjvBf-  5  00  ^ • * ^ T j - v o O V O V t ~ t ~ O i N ovri-Ti-^^oor^  u oo e  PQ  !  g 00 5 3  SSSNvonrtootJvnoo  1  .1 cs vo Ov  ^ 0 v m ^ v 9 - < c N i ; f j v O t ^ - < * t ^ m < < i  a,  t  •a  •S •§ •a S  h  Ov  Si  —  ,* °  r - - v O ' n v o : ^ - c - r i C N ' t O — ' O g c N r i — •  9  •8  Ov CN  :  3  H  a  :  o ca o TT ° it co O ^*  rr  ,a  oa  i  q q o \ v 5 i - . q v o . « Ti Ti cS(n>n'*wi^vo'  Icfl ^ O  8  CQ 03 00  SP  S u  a  = 3o u « * S S £ £ ao'1 o. H 3  3 <^ a  o ,-,  oa oo '3 t-< t-i CB  i  c .§  co H  2  ll ») Ov  16 of it from coastal mills, and up slightly from 1980.  By 1994, this had fallen to 127 t/d, with over 60%  accounted for by the Port Alice sulphite mill.  Distinct differences exist among the mills in the province with respect to the impact of effluent on ambient DO conditions. The characteristics of the receiving environment and the naturally occurring DO concentration determine the extent to which reduced DO supply and increased D O demand associated with effluent can be absorbed while still leaving sufficient oxygen for aerobic biota. The capacity of receiving water to assimilate effluent depends on such hydrological conditions as mixing and dispersion ability, seasonal flows, tides and natural aeration capacity. Turbulence and large volumes of water ensure rapid dilution, while naturally high turbidity reduces the significance of extra colour.  At British Columbia's freshwater mills there have been no published reports of significant long-term reductions in DO resulting from pulp mill discharges. These mills discharge to rivers or lakes which generally have high capacity for dilution and continual oxygenation by aeration, even at fairly low flows. Most importantly, all of the mills with one exception have used biological treatment since their start-up (see Table 1.1) because of concerns about impacts on fish stocks, particularly the valuable Fraser River system anadramous Pacific salmon stocks. The exception was the mill at Castlegar which had only partial treatment for much of its life. Despite this, effects of the mill's effluent on D O have been minimal due in part to the discharge of welloxygenated water through a dam 3 km upstream of the mill [Butcher 1992].  In contrast to the freshwater mills, secondary treatment facilities at all but one of the coastal mills are of very recent origin (see Table 1.1). This likely reflects that fact that many were built at a time when there was less concern and knowledge about effluent impacts, and because of a perception that the assimilative capacity of marine environments is generally sufficient to dispose of wastewaters adequately.  In fact, the mixing and  dilution characteristics of coastal discharge sites vary considerably. The degree of salinity stratification is also an important mediating factor since pulp mill effluent tends to be less dense than brackish or salt water. Most  17 coastal mills discharge to marine environments characterized by good to excellent mixing and dispersion abilities. Effluent at Harmac and Elk Falls is rapidly mixed and dispersed [Packman 1979b, Sullivan 1980, Sullivan 1981, Colodey and Wells 1992, Lim 1992]. The Elk Falls mill has historically discharged quite high B O D loadings because of its large size and lack of secondary treatment, yet because of the nature of the 5  receiving environment little change in D O has been observed, even in the immediate vicinity of the outfall.  Conditions are somewhat less favourable at Prince Rupert, Port Mellon, Woodfibre, Crofton and Powell River, although no D O reductions, or only localized and relatively minor reductions, have been observed at the latter four sites [Sullivan 1980, Sullivan 1981, Colodey and Tyers 1987, Colodey et al. 1990]. Some reductions in primary productivity have been noted near the outfall areas at the four sites [Anderson 1983, Colodey and Wells 1992] but, overall, D O reductions have not posed an environmental threat. However, problems at Prince Rupert have been serious in the past.  From the start-up of the mill until 1976, when  sulphite pulping was discontinued, the sheer quantity of discharged B O D resulted in a steady decline in DO 5  concentrations near the mill, despite the relatively favourable mixing and dispersion characteristics of the receiving environment. Hypoxic conditions led to fish kills. Once kraft pulping was substituted for sulphite pulping, D O concentrations returned to levels approaching normal [Packman 1979b].  Today, the only serious D O reductions known to result from currently operating mills occur at the Port Alberni and Port Alice mills, and probably to a lesser extent at the Gold River mill. These mills are all on the west coast of Vancouver Island. They discharge effluent to poorly mixed and vertically highly stratified (by salinity) estuarine inlets which trap effluent in the upper layer near the surface.  The situation at Port  Alberni is described in detail in the case study in Chapter 4. In Neroutsos Inlet, to which the Port Alice mill discharges, a combination of factors have created very significant DO reductions for many years and continues to do so [Colodey and Pomeroy 1986, Birtwell 1989, Stucchi 1990, Colodey and Wells 1992]. These include poor flushing capacity, a relatively small freshwater in-flow, TSS deposition (which increases sediment oxygen demand), very high B O D loading (see Table 2.1) and perhaps reduced photosynthetic activity due to the light5  18 masking effect of effluent colour.  Conversion in 1977 to a new pulping process which allowed greater  sulphite liquor recovery, and other changes, such as the installation of an effluent clarifier in 1985, do not appear to have had a major positive impact on DO concentrations [Colodey and Wells 1992]. At Gold River, no measurable impact on DO had been observed in Muchalat Inlet in the mid-1970s after about 8 years of operation of the mill [Sullivan 1979], but more recent observations imply that D O concentrations have been lowered in the intermediate and lower layers of the inlet [Lim and Colodey 1991]. At Ocean Falls (where a mill existed from 1912 to 1980 without any treatment whatsoever), the receiving environment of Cousins Inlet is similar in character to Alberni Inlet and Neroutsos Inlet, and D O at lower depths was quite impaired when the mill was in operation [Packman 1979a]. Here the primary cause has been very substantial TSS deposition, since effluent was discharged into the well-oxygenated tailrace of an hydroelectric dam.  The seriousness of changes in the DO balance depend on the extent of the change, the degree of use of a particular receiving environment by aquatic organisms, and on their oxygen requirements. Inadequate DO levels reduce livable habitat and can have pronounced impacts on behaviourial, swimming, and respiration patterns of organisms as well as causing decreased fecundity, disease resistance and growth [Davis 1975, Birtwell 1989]. Some degree of acclimation and compensatory adjustment of bodily processes in response to lowered ambient DO can occur, but this expends energy, reducing that available for other activities. Severely hypoxic conditions are lethal. In addition, hypoxic conditions increase the toxicity of chemical contaminants [CCREM 1987] including pulp mill effluent [Alderdice and Brett 1957, Hicks and DeWitt 1971].  The effects of changes in the DO balance have been fairly extensively studied at Port Alberni and at Port Alice. At both mills, Pacific salmon migrations have been obstructed, fish kills have occurred and the benthic and intertidal communities have been altered [Colodey and Pomeroy 1986, Birtwell 1989, Stucchi 1990, Stucchi et al. 1990, Colodey and Wells 1992]. At Gold River, fish habitat has been severely constricted [Colodey et al. 1990].  19  2.2. Total Suspended Solids (TSS)  Wood and bark particles of various sizes form the bulk of TSS found in pulp mill effluent [Poole et al. 1978, Waldichuk 1988]. Various inorganic substances are also found in small proportions and biotreatment adds microscopic biosolids of living and dead bacteria, and partially decomposed organic matter.  Suspended  dissolved solids increase turbidity but a large portion of TSS eventually settles out. Significant deposition typically yields a jelly-like mat of black anaerobic material on the water bottom [Colodey and Wells 1992]. Wood debris and bark from log storage and sorting activities at pulp mills also result in deposits.  Table 2.1 shows TSS loadings at B.C.'s 21 direct discharge mills. Here, as with B O D discharges, the level 5  of effluent treatment plays a significant role in loading. Primary treatment reduces most of the TSS while secondary treatment reduces toxicity, much of which is associated with the residual TSS. During the 1970s and 1980s many of the coastal mills installed at least partial primary treatment or undertook process changes to reduce TSS so that by 1990 there was little difference between the coastal and interior mills in fibre loss per tonne of pulp produced. For example, installation of primary treatment at Powell River in 1978 reduced TSS loading from 39 kg/ADt in 1977 [Nelson 1979b] to 19 kg/ADt in 1980 (see Table 2.1).  The addition  of up-to-date full secondary treatment since 1990 has meant that the coastal mills now are relatively more efficient that the interior mills with respect to fibre loss (see Table 2.1). With secondary treatment the effects of TSS generally are localized to only the immediate outfall area.  The extent and loading of the TSS deposits depend upon the water velocity and circulation patterns in the receiving environment.  In marine environments, TSS rapidly flocculates on contact with sea water,  contributing to increased weight and deposition of the material [Poole et al. 1978, Waldichuk 1988]. High natural sedimentation rate reduces the effect of TSS deposition.  TSS deposition reduces invertebrate community diversity and species abundance in heavily impacted areas.  20 This can be due to direct toxicity to pollution intolerant species or to the physical unsuitability of the degraded habitat. Changes to local fish communities, or complete loss, can also occur. Deposition creates a long-term reduction in benthic feeding and reproductive habitat and smothers the benthic biota. Biological degradation of the organic matter, and chemical oxidation of reduced compounds, elevates sediment oxygen demand, leading to hypoxic or even anoxic (no oxygen) conditions in severe cases.  This represents a significant  problem at Port Alberni and Port Alice. The action of bacteria in the sediment also generates methane and hydrogen sulphide (H S) gas. The up-welling of such toxic gases from benthic deposits can result in fish kills, 2  as has been observed at Powell River and perhaps at Port Alice [Colodey and Wells 1992]. The sediment itself may be toxic to benthic invertebrates, as observed in laboratory bioassays using sediment samples from Port Alberni and Port Mellon [Colodey and Wells 1992]. TSS deposits serve as repositories of chemical contaminants ~ it has been estimated that more than 90% of chlorinated organics, such as the chlorinated dioxins and phenolics, are associated with TSS [Butcher 1992]. The uptake of contaminants through ingestion of settling or suspended particulate matter represents an important mechanism by which contaminants enter the food chain.  In British Columbia, the most significant loss of habitat and the most severe effects on community structure have occurred at the coastal mills and at the Castlegar mill. Varying degrees of sediment degradation and changes in the invertebrate community have been recorded at all coastal mills [Sullivan 1979, Sullivan 1981, Sullivan 1982, Pomeroy 1983, Waldichuk 1988, Colodey and Wells 1992]. Sediment samples have often contained noticeable H S. Colodey and Wells [1992] surveyed studies of B . C . coastal mills and concluded 2  that benthic degradation extended from 0.5-5.0 km from mill sites and covered areas ranging from 1-8 km . 2  Pomeroy [1983] concluded that the extent and seriousness of fibre deposits at coastal mills tended to be greater than those resulting from the two other major sources of TSS on the British Columbia coast, log storage/handling activities and domestic sewage outfalls.  The single clearest example of TSS-related habitat degradation occurred at the Ocean Falls mill on Cousins  21 Inlet [Packman 1979a, Pomeroy 1983, Waldichuk 1988]. The lack of primary treatment, coupled with the inlet's poor dispersion capability, led to a very significant deposit.  Environment Canada researchers in the  mid-1970s found deposits of fibre up to 4.5 km from the mill and up to 15 m thick. HjS gas was released frequently, the lower waters were anoxic and life was non-existent in a large area. Since the closure of the mill in 1980, some recovery in the fish community has occurred [Pomeroy 1983], but benthic recolonization will likely be slow due to low natural sedimentation rates [Waldichuk 1988].  At interior mills the observed impacts appear to have been much less pronounced.  Decreases in benthic  pollution-intolerant species have been noted at Skookumchuk [Derksen and Lashmar 1981b] and at Kitimat [Derksen 1981a], although the extent and nature of TSS deposits were not assessed. Early studies in the vicinity of the outfalls from the Prince George mills showed no apparent effect on the invertebrate community structure [Derksen 1981b]. Schreier et al. [1991] reported that studies in the late 1980s found a shift in structure toward pollution-tolerant species downstream of the mills at Prince George and Quesnel, although a 1989 study found no evidence of such degradation [Dwernychuk 1990]. At Kamloops, little change has been seen [Schreier et al. 1991].  Although based on limited data, it appears that the primary and secondary  treatment systems of interior mills have minimized the impact of TSS on the benthic community relative to the coastal mills. In the Fraser River system, and perhaps elsewhere in the interior, seasonally high natural sedimentation rates have also played a role in limiting impacts.  The effects at the Castlegar mill have been more pronounced. Surveys of the Columbia River bed in the few kilometres downstream of the mill outfall between 1975 and 1990 documented serious degradation of bottom sediments [Celgar Expansion Review Panel 1991, Butcher 1992].  It should be noted that a lumber mill  situated adjacent to and just upstream of the pulp mill contributes to sediment degradation, although the areas of impact can be differentiated.  During this period, the fibre mat resulting from pulp mill TSS ranged in  thickness from 0.15-0.90 m at the diffuser, stretched downstream several kilometres and covered much of the width of the river at one location. The richness, density and pollution sensitivity of the benthic community  22 increased with distance away from the mill.  In 1992 (i.e., before secondary treatment), Butcher [1992]  concluded that the fibre deposition represented a significant withdrawal of fish feeding and spawning habitat.  2.3. Chemical Contaminants  Pulp mill effluent contains a substantial array of chemical contaminants ~ compounds that normally are not found in aquatic environments or that, although naturally occurring, are found only in relatively small amounts compared to quantities discharged in pulp mill effluent. These compounds form through the reaction of wood constituents with pulping and bleaching compounds and vary greatly in concentration between mills and for any given mill over time.  Most of the following discussion focuses on the chlorinated organics, reflecting the predominant interest of researchers since Swedish work in the early 1980s highlighted their presence in effluent from pulp mills using chlorine bleaching. In 1991, effluent from mills using chlorine in bleaching was classified as toxic under the 1988 Canadian Environmental Protection Act (CEPA) as having immediate and long-term harmful effects on the environment [CEPA 1991]. It is important to note, however, that the environmental hazard posed by effluent contaminants derive from a wide variety of sources, not just chlorinated organics. While interest at any given point in time tends to focus on specific groups of compounds, there have invariably been other compounds which prove to be of particular concern as measurement techniques improve and knowledge expands.  2.3.1. Loading and Fate  The recent concern with chlorinated organics derives from the fact that, in comparison to analogous non-  23 chlorinated compounds, chlorinated organics tend to be more persistent, toxic and bioaccumulative. properties increase as the degree of chlorination increases.  These  Most of the chlorinated organics are hydrophilic  and are too large to permeate cell walls, or, if small, can be readily metabolized [CEPA 1991, McKinnon 1992]  They are therefore thought to pose relatively little risk to the environment, although biodegradation  or biotransformation can result in more hazardous contaminants.  It is the remaining small fraction (on the  order of 1%) which represents the most significant hazard. However, for practical purposes, measurement and characterization of all of the individual chlorinated compounds in effluent remains impossible. Adsorbable Organic halogen (AOX) measurements provide one indication of the total organochlorine content (the chlorine part of chlorinated organic compounds) of effluent.  Reduction of A O X means that less organochlorine is  being discharged and for that reason A O X has been introduced as a regulatory parameter in British Columbia (late 1990) and many other jurisdictions [McKinnon 1992]. However, it must be noted that low levels of A O X in the environment correlate poorly with the toxicity of effluent or with the observed persistence, toxicity or bioaccumulation of specific contaminants of concern [Neilson et al. 1991, CEPA 1991, McKinnon 1992, Solomon et al. 1993].  Table 2.2 summarizes A O X discharges from B . C . mills in 1989, 1991 and 1994. In 1989, before secondary treatment at coastal mills, and before high or complete C10 substitution became the norm, A O X loading 2  amounted to about 101 t/d, an amount that was probably roughly representative of loadings in previous years. C10 substitution significantly reduced organochlorine production so that by 1994 A O X loading had fallen to 2  about 14 t/d. These loadings correspond to approximately 1,300 t/d and 180 t/d of chlorinated organics in 1989 and 1994, respectively, using a rough factor of 13 to convert organochlorine loading to chlorinated organic loading [Colodey and Wells 1992].  Among the chlorinated organics, the chlorinated phenolics (especially the chlorophenols, chlorocatechols and chloroguaiacols) have received considerable attention.  The use of C10 in place of C l reduces the 2  2  concentration of chlorinated phenolics and results in a shift towards compounds with fewer chlorine  24  Table 2.2. British Columbia Pulp Mills: AOX Discharges, 1989, 1991 and 1994 tonnes/day Mill Location  kg/Air Dried tonne  1989  1991  1994  1989  1991  1994  Crofton Elk Falls Gold River Harmac Port Alberni Port Alice Port Mellon Powell River Prince Rupert Wood fibre  13.1 12.2 4.3 7.4 7.5 2.4 2.8 26.1 3.5 2.3  3.1 2.4 1.6 2.3 1.2 3.5 0.7 1.6 2.9 1.1  0.7 1.0 1.0 1.0 0.3 3.0 0.6 0.4 1.4 1.2  7.1 6.0 6.2 7.1 5.8 5.2 5.8 14.9 3.1 3.7  3.0 2.8 2.1 1.9 3.4 6.5 0.6 6.0 1.9 1.9  0.7 1.4 1.3 1.0 0.8 6.5 0.5 1.3 1.6 1.6  Total  81.6  20.4  10.6  7.2  2.5  1.4  Castlegar Kamloops Mackenzie (Fletcher Challenge) Prince George (Canfor) Prince George (Northwood) Quesnel (Caribou) Skookumchuk  2.2 3.4 1.2 4.9 4.2 2.6 0.5  1.8 3.9 0.6 2.3 2.8 3.9 0.7  0.3 0.8 0.2 0.8 0.6 0.7 0.1  4.2 3.0 2.7 3.6 3.0 3.2 0.8  3.8 2.7 1.0 1.5 1.8 3.8 1.4  0.3 0.6 0.4 0.6 0.5 0.9 0.3  Total  19.0  16.0  3.5  3.0  2.2  0.5  A.  Coastal Mills  1  B. Interior Mills  3  2  2  2  2  Note: Annual averages may be calculated in different ways for different mills and years, and data are not always consistent. 1  2  3  1993. Derived using 1990 production. Effluent discharged to a ground disposal system during periods of low flow.  Source: based on data from G. Leu, B.C. Ministry of Environment, Lands and Parks for 1991 and 1994; and Colodey and Wells [1992], Dwernychuk [1990] and West Coast Environmental Law Association [1989] for 1989.  25 substituents: the result is a reduction in toxicity and persistence [Kringstad and Lindstrom 1984, Dahlman et al. 1993, O'Connor et al. 1994]. In recent years, the greatest concern about chlorinated organics has been with respect to the polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran family of compounds (henceforth referred to simply as chlorinated dioxins and furans) which are known to be extremely toxic to test mammals and fish at very low doses. There are 75 chlorinated dioxin congeners and 135 chlorinated furan congeners, each differing by the number and position of the chlorine atoms on the molecule [CEPA 1990]. 2,3,7,8-tetrachlorodibenzo-dioxin (TCDD) and 2,3,7,8-tetrachlorodibenzofuran (TCDF) are consider the most toxic members of the family, and analyses of chlorinated dioxins and furans usually convert concentrations to units of 2,3,7,8-TCDD based on relative toxicity factors [CEPA 1990].  Chemists and toxicologists have been aware of dioxins for several decades but concern about production in pulp and paper mills first arose in the mid-1980s. A 1986 study by the United States EPA demonstrated the existence of dioxins in effluent, sludge and pulps from 5 bleached kraft pulp mills in the United States [DFO 1988a]. In Canada, T C D D was reported in sludge from an Ontario kraft pulp mill in 1986 [Nassichuk 1992]. Over the 1982-87 period, chlorinated dioxins were found in great blue heron eggs collected near Crofton and other Strait of Georgia locations, leading to suggestions that pulp mill effluent was one likely source [Nassichuk 1992]. A Federal government program of fish and sediment sampling in the vicinity of pulp mills and other sites began in 1987. Researchers found concentrations in effluent to be quite small, at the parts per trillion level, but still sufficient to cause appreciable bioaccumulation in some aquatic organisms and to pose a potential threat to human health.  Lack of data preclude fate and persistence characterization of most chlorinated organic contaminants but some general properties are known [Suntio 1988, Bonsor et al. 1988, CEPA 1990, CEPA 1991, McKinnon 1992, Solomon et al. 1993]. In addition, the fate and persistence of the chlorinated phenolics and the most toxic chlorinated dioxin and furan isomers are fairly well understood. Half-lives in the receiving environment can range up to several weeks for some of the chlorinated phenolics and several years for the chlorinated dioxins.  26 Partitioning to particles and sedimentation are key processes but the contaminants may still be available to fish and other organisms via the benthic invertebrates that feed in the sediment.  In addition, they can become  bioavailable after discharge ceases if desorption from contaminated sediments occurs.  Aquatic organisms can detoxify and/or eliminate many contaminants. In some cases bioaccumulation occurs, increasing the possibility of toxic effects both for the organisms and for organisms which feed upon them. The liver and hepatopancreas, organs involved in detoxification, display the greatest bioaccumulation. Skeletal muscle tissue, of greatest concern for human health, displays lower biaccumulative potential. Some effluent contaminants, notably the chlorinated phenolics and dioxins, display very high bioconcentration factors, with T C D D possessing the highest known bioaccumulative ability of all contaminants discharged by pulp mills [CEPA 1991, Solomon et al. 1993, U.S. E P A 1993]. Long-term exposure rather than high bioaccumulative potential can also result in high contaminant body-burdens [CEPA 1991].  Highly mobile organisms like  migrating fish might be expected to be able to eliminate some compounds when they leave contaminated waters.  Other contaminants, such as the chlorinated dioxins, can persist for long periods in fish tissue,  exhibiting half lives of at least several months [Owens 1991, CEPA 1991].  Biodegradation and biotransformation of chlorinated organics have been observed to occur, primarily in laboratory studies [Neilson et al. 1991].  However, many toxic compounds, especially chlorinated organics,  biodegrade only slowly and exhibit considerable persistence.  Chloroguaiacols have been found in water 650  km downstream of a bleached kraft mill on the Athabasca River and in fish in the Slave River over 1,500 km downstream of the nearest mill [CEPA 1991]. In the Fraser River system, chloroguaiacols were found in overwintering juvenile chinook over 600 km from the nearest mill [Rogers et al. 1988]. At some coastal mills, chlorinated phenolics have been found in sediments up to 10 km from the mill. Some of the metabolites of biodegradation processes, such as chloroveratrole, are more hazardous than the original contaminants in terms of toxicity and bioaccumulative potential [McLeay 1987, CEPA 1991].  27 2.3.2. Environmental Impacts  Effects of chemical contaminants on the biota include acute toxicity, sublethal toxicity, alterations in community structure and tainting. Acute toxicity is usually described in terms of the concentration (lethal concentration, LC) of effluent at which a certain proportion, often 50%, of the test organisms are killed over a 96 hour period, hence 96-h L C ^  (low values indicate high effluent toxicity).  L C tests are a standard  regulatory parameter in Canada. It should be noted that these tests provide only a crude measure of effluent toxicity since they do not reproduce environmental conditions, especially as related to temperature, DO content and pH, and they obviously do not account for longer term effects. Lethality of effluent as measured by the 96-h LCSQ varies from mill to mill and over time but generally is attributed to non-chlorinated and chlorinated resin and fatty acids and to chlorinated phenolics [McLeay 1987, Owens 1991]. Good secondary treatment can eliminate 90% of acute toxicity [CEPA 1991].  Kovacs [1986] suggested that under normal operating conditions the 96-h LCJO of non-treated bleached kraft mill effluent typically occurs around 40% effluent concentration with a low value of about 10%.  For  biotreated effluent, his survey suggested a typical range of 80-100% with a low value of about 40%. These ranges are roughly consistent with the average 96-h LCJQ values for British Columbia mills in 1987 as shown in Table 2.3.  The use of C10 and the installation of secondary treatment has clearly raised the average  concentration of the 96-h LCJQ  2  Currently, the dilution capacity of British Columbia's aquatic environments  should quickly remove the possibility of acute toxicity in most cases. For example, at Castlegar, acute toxicity has not been considered a problem even at low flow and despite the absence of secondary treatment until recently [Butcher 1992]. Similarly, acute toxicity should not be a problem downstream of the Fraser and Thompson River mills [Schreier et al. 1991].  Few  reported instances of fish kills in British Columbia can be linked specifically to effluent chemical  contaminants from pulp and paper mills, although kills resulting from spills of chemicals at mill sites have  28  Table 2.3. British Columbia Pulp Mills: Average Acute Toxicity 1987, 1990 and 1994  1  Mill Location  A.  1987  1990  1994  30 23 54 75 64 33 24 59 11 11  42 32 71 47 100 24 100 na 78 13  100 100 100 98 100 27 100 100 100 100  30 100 95 100 100 100 100 98 76 na 100  100 100 97 100 100 97 100 100 100 100 99  Coastal Mills  Crofton Elk Falls Gold River Harmac Port Alberni Port Alice Port Mellon Powell River Prince Rupert Woodfibre B. Interior Mills Castlegar Kamloops Kitimat Mackenzie (Fletcher Challenge) Mackenzie (Finlay Forest) Prince George (Canfor) Prince George (Northwood) Quesnel (Caribou) Quesnel (Quesnel River) Skookumchuk Taylor 4  2-19 na na na na 96 98 67 86 na  2  3  3  3  3  -  96-h LCJO = effluent concentration (volume effluent / total solution volume) at which 50% of the test organisms (usually rainbow trout) die in a 96-hour period. 1  2  3  4  Data are the average of all tests conducted during the year. Averages may be calculated in different ways for different years and over time. A value of 100 is used in calculating the averages when no organisms or fewer than 50% die in effluent concentrations of 100%. Range for January 1987 to February 1989. 1989. Effluent discharged to a ground disposal system during periods of low flow.  Source: based on data from G. Leu, B.C. Ministry of Environment, Lands and Parks for 1990 and 1994; and Colodey and Wells [1992], Dwemychuk [1990] and Butcher [1992] for 1987.  29 been reported [Colodey and Wells 1992]. Before the shutdown of the sulphite mill at Prince Rupert in the mid-1970s, highly toxic effluent, in combination with low D O , destroyed an area of what have been very productive herring spawning grounds. More recently, effluent toxicity was likely partially responsible for some of the 14 reported lethal incidents near Port Alice in 1986. These ranged from kills of 26 ratfish, to 10,000 herring, to 90% of all intertidal life over a 10 km distance from the mill [Colodey and Wells 1992]. As already described in Section 2.1, low D O has also plagued the area.  Note that high typical or average 96-h L C ^ values can obscure instances of serious concern. For example, 1989 effluent monitoring records reproduced by Dwernychuk [1990] for the Prince George and Quesnel mills on the Fraser River show that acute toxicity of effluent was measured at very low concentrations (under 10%) on quite a few occasions.  These instances suggest the risk of lethality at low river flow conditions, and  especially in effluent outfall areas.  They tended to be bunched together over periods of several weeks,  indicating the possibility of relatively toxic effluent discharges for extended periods.  The point is that  concentrations can vary significantly, but that is difficult to know the extent of acute toxicity, except in average terms.  Concentrations of specific chlorinated phenolics and chlorinated resin acids in effluent are generally well below levels demonstrated to cause 96-h LC^, acute lethality [CEPA 1991]. For example, observations of various chlorinated phenolics in 1981 and 1989 in effluent from the Prince George, Quesnel and Castlegar pulp mills [Dwernychuck 1990, Butcher 1992, Voss and Yunker 1983 as cited in Schreier et al. 1991] show concentrations (0-100 ppb range for the various compounds) generally well below the estimated acutely toxic concentrations (100-3000 ppb range), as reported in CEPA [1991] and U.S. E P A [1993]. Similarly, water concentrations of some chlorinated phenolics observed near coastal mills in 1989 (0-60 ppb range) [CEPA 1991] were below the estimated acutely toxic thresholds. However, some of the compounds have very high bioconcentration factors [CEPA 1991, U.S. EPA 1993], and concentrations in fish exposed for relatively long periods downstream of the Prince George and Quesnel mills have been observed to enter the range of  30 estimates for acute toxicity [Rogers et al. 1988, Rogers et al. 1989, Dwernychuk 1990]. The simultaneous accumulation of a number of contaminants might increase lethality (synergistic or additive effects). Aside from specific chemical compounds related to chlorine bleaching, acute lethality may also be caused by exposure to other compounds and by rapid variations or extremes in temperature, p H and D O content.  While acutely lethal effluent contaminant concentrations likely are not pervasive in British Columbia, the health of individual organisms, and potentially of populations, will be adversely affected if effluent concentrations reach the threshold at which chronic or sublethal effects occur. The concern of toxicologists currently focuses on such effects, and the potential for damage is clear. Sublethal effects include deleterious anatomical, behaviourial, physiological and biochemical changes which do not cause immediate death. Damage to the individual could be manifested in developmental, growth and reproductive abnormalities, or by increased susceptibility to other stressors such as parasitism. Long-term survival and reproductive ability of individuals could be affected, perhaps creating effects at the population level.  Extensive reviews of sublethal effects report biochemical, physiological and developmental changes under a wide range of effluent characteristics and concentrations in Canada and elsewhere [Kovacs 1986, McLeay 1987, CEPA 1991, Owens 1991, Colodey and Wells 1992, McKinnon 1992, Axegard et al. 1993]. Much less is known about such effects in the marine environment than in freshwater environments. Sublethal effects have been observed to occur at 0.5-5% concentrations of bleached pulp mill effluent in the Canadian environment [CEPA 1991].  Bleached kraft mill effluent sublethal toxicity thresholds for fish have been  suggested as 5-10% of the 96-h L C ^ [Kovacs 1986, Owens 1991]. If untreated effluent is assumed to be lethal at effluent concentrations in the 10-40% range (the range suggested by Kovacs [1986]), then the threshold for sublethal effects might be reached when effluent represents 0.5-4% of water volume.  If  biotreated effluent is lethal at concentrations in the 40-100% range then sublethal effects might be observed when effluent represents 2-10% of water volume. Generally, such concentrations occur in the immediate outfall zone of mill effluent in British Columbia but the extent of occurrence elsewhere depends on the mixing  31 and dispersion characteristics of each receiving environment. For example, under low flow conditions on the Fraser River, and assuming complete mixing and dilution of effluent, the discharge from the five mills could result in sublethal effects on fish. At Castlegar, sublethal stress on aquatic organisms was assumed to have occurred based on the dilution capacity of the Columbia River [Derksen and Lashmar 1981a].  The toxicity of individual contaminants should also be considered but knowledge regarding sublethal toxicity thresholds and effects of specific contaminants are limited. However, EPA ambient water quality criteria for sublethal effects of various chlorinated phenolics (3-20 ppb range) [U.S. E P A 1993] suggest that observed concentrations in the British Columbia environment can cause sublethal effects. For T C D D and T C D F , the EPA criteria are under 1 ppt.  Fish which spend considerable time in the vicinity of mills are at particular risk. Biochemical changes in the form of increased liver enzyme activity have been found in juvenile chinook which overwinter near mills on the Fraser and Thompson Rivers in conditions of reduced D O and low flow (high effluent concentration) [Rogers et al. 1988, 1989]. The response in the fish was likely related to bioaccumulation of chlorinated dioxins or other chlorinated organics.  The enzyme in question plays a role in steroid metabolism and  reproduction, as well as contaminant detoxification. Studies of white sucker in the vicinity of a bleached kraft mill on Lake Superior revealed similar increases in enzyme production along with biochemical and developmental changes including lower steroid levels [Munkittrick et al. 1991, 1992]. Observed abnormalities in growth rates, reproductive development, age-fecundity relationships, age to maturation, gonadal size and egg size appeared to be correlated with the biochemical changes.  Carey et al. [1993] have reviewed recent  Canadian field and laboratory studies regarding physiological changes in fish induced by effluent.  Their  assessment serves to highlight the fact that the threshold effluent concentration for sublethal effects remains uncertain, as does the source of such effects. They are worth quoting at length: The physiological disruptions in fish near pulp mills are associated with significant effects on their reproductive development. Canadian studies have clearly shown physiological effects in fish when they are exposed to effluent, whether from mills using chlorine bleaching or not—These effects are present at very low effluent concentrations ( < 1 %), are not removed  32 during normal secondary treatment of effluent, and existing toxicity tests can not predict whether or not they will be seen in wild fish. These effects do not correlate with A O X or levels of dioxins and furans, and represent a previously unidentified category of impact, distinct from the issue of chlorine use and the potential discharge of persistent, bioaccumulating toxic substances. However, the effects are transient and reversible, and once appropriate changes in process or effluent treatment have been designed, recovery of wild fish populations should happen quickly [p.8].  Liver lesions in sole taken at Port Mellon and Woodfibre, and gill lesions in sole from Port Mellon and Crofton have been observed and are consistent with long term sublethal exposure to carcinogens in sediment [Brand et al. 1992]. Compared to a reference site, an elevated incidence and severity of internal and external physiological abnormalities (discolourations, lesions, cysts etc.) have been observed in whitefish caught downstream of the Castlegar mill, although differences in age among the fish samples prevented clear attribution to pulp mill contaminants [Nener et al. 1994]  Some sublethal changes in response to effluent have been observed in organisms other than fish [Colodey and Wells 1992]. Reproductive impairment and slower growth occurred in blue mussels and barnacles transplanted to locations near Port Mellon. A laboratory study found that crabs fed clams contaminated with chlorinated dioxins experienced fewer molts and slower limb regeneration compared to controls. Clams environmentally contaminated with chlorinated dioxins displayed reduced size, internal lesions and thinner shells. Blue heron eggs near Crofton exhibited various morphological and physiological changes when adults were contaminated with chlorinated dioxins.  Tainting of fish flesh represents another form of sublethal response.  Compounds known to occur in pulp  effluent have been implicated in incidents of fish tainting although it has been difficult to make conclusive identifications because of the highly complex matrix of compounds present. Chlorophenolic compounds and their microbial metabolites (chloroveratroles and chloroanisoles) cause tainting [CEPA 1991].  Other  compounds suggested include resin acids, organosulphur compounds and volatile organics such as terpenes [Kovacs 1986, McLeay 1987, McKinnon 1992]. Untreated bleached pulp mill effluent has been observed to cause tainting at 0.5% effluent concentration [CEPA 1991]. Biological treatment of effluent can cause a 2-to-  33 10-fold decrease in tainting [Sprague 1988 as reported in Butcher 1992], suggesting that the threshold might lie in the 1-5% range. Now that C10 substitution is widespread in B . C . the average tainting threshold, what 2  ever it is, may have moved higher.  Reported incidents of fish tainting in B . C . appear to be fairly limited. A 1973-75 study found that treated effluent (pulp and municipal combined) seriously reduced the palatability of salmonids in the Lower Thompson (Servizi 1989). Complaints were also expressed in the 1970s with respect to the Skookumchuck and Castlegar mills [Derksen and Lashmar 1981a, 1981b]. However, the most serious prolonged instance of tainting has occurred at Kitimat where complaints have been continuously registered since 1972.  The case study in  Chapter 7 examines this problem in more detail.  Acute lethality and sublethal impacts, whether caused by chemical contamination or other effluent components, have with-in organism and whole organism effects.  Moving from these effects to evaluating consequences  for populations and communities poses significant difficulties [Shuter and Regier 1989]. Usually, the only clear effect at these levels is on benthic invertebrate communities or other relatively immobile populations. For example, some highly productive commercial oyster beds near Crofton were extinguished for commercial purposes within 7 years after start-up of the mill in 1957 [Lim 1992a]. The two ecological end-points relevant for assessing the impact of effluent on populations are mortality and reproductive success. An unquantifed risk of acute lethality has existed at pulp mills in British Columbia.  Sublethal effects (which might affect  reproductive success or increase long-term mortality) seem likely at many British Columbia locations, but are difficult to comprehensively assess. On average in the province, the risk of acute lethality and the extent of sublethal effects has likely decreased in recent years due to process changes and improved treatment. However, connections do appear to exist between observed biochemical/physiological changes in fish and reproductive dysfunctions and decreased reproductive effort, connections observed at very low concentrations of biotreated effluent.  Thus the degree and quickness of dilution which occurs, and the time spent by  organisms in effluent-containing water are key.  34 Any attempt to assess population effects of effluent must account for natural population variability and buffering capacity and the effect of other human interventions.  This can prove extremely difficult. For  example, two systematic attempts to assess the impact of wastewaters on Pacific salmon in the Fraser River system have been done. Brox [1976] studied the Chilko River sockeye stock for which relatively high quality data existed. He modeled various life stages from adult spawning escapement (total migration less catch) to adult recruitment using estimated survival rates at each stage. He found that knowledge about the sublethal effects of Lower Fraser River wastewaters on these rates was insufficient to do more that forecast future potential effects under various scenarios. Any past effects were captured in variations of the observed survival rates. Servizi [1989], in a much simpler study, assessed historical stock records but found that he could give only a limited assessment of the effect of wastewaters in the Fraser River above Hope on Pacific salmon (the pulp mills account for most of the wastewater). Simply put, the data were insufficient to adequately account for the various influences on the stocks. He was able to conclude only that there was no indication of any obvious effects on sockeye and pink stocks and that nothing could be said about effects on chum, coho and chinook stocks.  Concerns about the effects of pulp mill effluent on human health have been related primarily to the contamination of finfish and shellfish by chlorinated dioxins and furans. Food sources have been estimated to account for over 90% of human intake of these substances in Canada [CEPA 1990]. In B . C . , prior to efforts to reduce discharges, just over half of chlorinated dioxins emitted to the environment were estimated to come from pulp mills, as measured in terms of toxicity equivalents to 2,3,7,8-TCDD, the most toxic isomer [Radloff 1988 as reported in Schreier et al. 1991]. Data in CEPA [1990] support this calculation.  T C D D and T C D F are considered probable human carcinogens and T C D D is known to have adverse effects on human reproductive capacity and liver function, and to cause chloracne, a usually temporary skin condition [CEPA 1990, U.S. EPA 1993]. T C D D is also suspected to cause miscarriages, birth defects and behaviourial and neurological problems. Studies of long-term exposure to T C D D in mammals other than humans reveal  35 immune system damage, liver dysfunction, impaired reproduction, birth defects and cancer. It is also known to be acutely toxic to animals at relatively small doses (in the ppb to ppm range). As a result, in 1990 the chlorinated dioxins were classified as toxic under CEPA as representing a danger to human health, for their toxicity and because of their potential for short and long-term harmful impacts on the environment [CEPA 1990]. Health Canada has set a maximum concentration in fish flesh for human consumption of 20 ng/kg (parts per trillion). Federal regulations gazetted in May 1992 required that mills built before 1 June 1990 were to achieve a limit of 20 ppq (parts per quadrillion) for 2,3,7,8-TCDD and 50 ppq for 2,3,7,8-TCDF by January 1994 and by July 1992 for mills built after 1 June 1990 [Canada Gazette, Part II, 126(11), pp. 1940-1946, 7 May 1992].  Contamination of some species of shellfish and finfish has led to a series of consumption advisories and fishing ground closures in British Columbia since 1988.  The area of marine habitat closed to shellfishing due to  chlorinated dioxin contamination from pulp mills currently amounts to 841 km  2  [B.C. Ministry of  Environment, Lands and Parks 1994]. This is small compared to the open area of 76,367 km but note that 2  what matters more than a comparison of geographical areas is a comparison of shellfish productivity across locations.  The fish closures and consumption advisories are described further in the Howe Sound and  Castlegar case studies, and summarized in Appendix A . Contamination of ducks and seals has also been observed, leading to consumption advisories with regard to seabird livers at some locations.  Measurable  quantities of chlorinated dioxins in effluent were reduced substantially by 1991, and all mills had complied with the T C D D target by the end of 1993 through the use of C10 substitution and the elimination of dioxin 2  precursors on pentachlorophenol-treated wood chips and in defoaming agents [Government of Canada 1995]. A few mills have yet to consistently achieve the T C D F target. Concentrations observed in sediments and biota have also fallen to the point that some re-opening of closed fishing grounds occurred in February 1995, although new closures were also enacted as the geographical extent of contamination was further delineated [Government of Canada 1995].  36  2.4. Summary  All but one of British Columbia's pulp and paper mills now have secondary treatment, and B O D and TSS 3  loadings have been cut quite considerably. De-oxygenation due to B O D and TSS are serious at Port Alberni 5  and Port Alice. With good secondary treatment it is quite likely that the extent of fibre deposits at various mills will decrease and that some recolonization and rehabilitation of the benthic habitat will occur gradually. Of course, in cases of severe degradation and low natural sedimentation rates, such as at Ocean Falls, full rehabilitation may take decades.  Acute lethality as measured by the 96-h L C ^ is not a pervasive problem near British Columbia mills. A large number of sublethal effects have been observed in fairly recent studies although relatively few studies have examined B . C . environments specifically. However, it is clear that pulp mill effluent has affected the health of parts of British Columbia's aquatic systems. Past conditions, especially at those mills without secondary and/or primary treatment, were considerably worse than at present. The population and community effects of effluent have proven impossible to assess except for effects on benthic and relatively stationary organisms near outfalls. The human health concerns related to chlorinated dioxin and furan contamination have led to closures and consumption advisories.  37  3. THE ECONOMIC COST OF WATER POLLUTION  The preceding chapter summarized the diverse environmental impacts of pulp and paper mill effluent in the British Columbia environment. In this chapter I review how economists analyze these impacts. In economic terms, the aquatic environment can be considered as natural capital analogous to human-produced capital [Pearce and Turner 1990, E l Serafy 1991, Freeman 1993]. As such, aquatic environmental assets produce a stream of service flows including 1) commercially exploitable natural resource flows such as fish, and water for industrial and agricultural purposes; 2) amenity and cultural services which support recreational opportunities, scenic values, and subsistence activities; 3) services not related to any direct use; 4) life-support services; and 5) waste dispersion and storage services. All of these services provide economic value to society but deriving the value from one often comes at the expense of the value from another. In particular, the use of aquatic environments for pollution dispersion and storage services impinges on the economic values provided by the other flows.  As shown in Table 3.1, a rigorous assessment of the economic cost of effluent would relate the impact of discharges to water quality parameters and then relate these effects to the economic value people derive from the environmental service flows. This is an extremely difficult task. Pollution changes water quality in ways which often are complex and not well understood.  Sometimes physical and chemical measures of water  quality affect economic value and sometimes it is the impacts on the biota which affect economic value. In either case, the problem is that water quality or biotic health cannot be represented by a single index which can be meaningfully related to economic value. Moreover, environmental impacts suggest potential economic consequences, but just as environmental impacts depend on site-specific process and treatment parameters and on the characteristics of the receiving environment, so too are economic consequences highly site-specific. They are determined by the scope and type of economic activity or value associated with the particular location.  38  Table 3.1. Assessing the Economic Cost of Water Pollution Polluting Activities JJ U Polluting Substances (Large-scale wastes, BOD, chemical and biological contaminants, suspended solids)  JJ Receiving environment  JJ  Environmental Impacts (Contamination of sediments, water and biota; changes in dissolved oxygen; turbidity; reduced habitat; mortality of biota; sublethal effects; tainting; populations reduced; community structure changed)  JJ Economic activities and values  JJ  Economic Costs (Reductions in services provided by aquatic environments)  JJ JJ Valuation (Costs measured in terms of use and non-use values by consumer and producer surplus)  Source: adapted from Freeman [1993]  Two points must be kept in mind in the following discussion. First, it is important to distinguish between "economic value" and "economic impact". Total economic value is a measure of the benefit derived from a good or service. Usually, obtaining the good or service involves some expense so that net economic value is total value less the expenditure. In this thesis I examine net economic value as measured by producer and consumer surplus. When I use the term "economic value" I refer to these surplus measures. I do not discuss economic impacts, which refer to the taxation revenue, employment and income that result from the direct,  39 indirect and induced expenditures related to the activity of interest, for example, boating. In the absence of boating opportunities, the expenditures would be made on something else — thus while the exact distribution of the impacts would differ, economists usually assume that the total impact would not differ greatly. Under this assumption, economic impacts have great relevance for consideration of distributional questions, employment and government finances, but not for questions of economic value. Assessment of economic impacts provides information about how expenditures flow through the economy in market transactions, but their use in measuring value can give quite misleading results. For example, if the expenditure associated with boating were to fall because the cost of gas fell, this would not mean that boaters attach less value to the activity. At best, expenditures provide a lower bound to total value, and say nothing about net value.  The second and related important point is that measures of net economic value are contingent upon the scale of analysis chosen — the calculation of net value as it benefits a given area differs according to whether the value is obtained by residents or non-residents. In this thesis I focus on economic costs as they affect British Colombians.  An analysis which looked at regions within the province or at Canada as a whole would have to be undertaken in a slightly different manner. For example, I assume that expenditures on recreational activity by a British Columbia resident do not represent new economic activity in the province since he or she would spend the same amount on other things in the province if not on recreation. The only exception to this occurs where the resident has dedicated a certain amount to recreation, either within or outside the province, and would spend the same amount outside British Columbia if he or she could not recreate in the province. Thus the expenditures of British Columbia residents are not relevant for measuring net economic value — as noted above they indicate economic impact.  However, economic value does accrue to British Columbia from the  expenditures of non-residents on recreation since the expenditures represent new spending in the province. The cost of providing the recreational activity to non-residents must be accounted for so that the appropriate measure of net economic value to the province in the case of a non-residents is the sum of the net returns to  40 capital, labour and resource owners resulting from the activity (i.e. profits, wages and tax revenues).  The crucial assumption implicit in the above reasoning is that labour and capital are perfectly mobile and fully employed. Expenditures in a sector result in increased returns to labour and capital employed in the sector. The assumption that these expenditures do not contribute to the net economic value of the sector rests in turn on the assumption that, in the absence of expenditures in the sector, the labour and capital could readily move to other sectors and receive the same returns. Obviously, in an economic climate characterized by persistent and high unemployment this assumption is questionable, at least for labour. Thus expenditures in a given sector likely do contribute net economic value to labour and capital in the sector and should be accounted for. The addition to net economic value contributed by the sector would be measured as the difference between the total returns they receive (wages and profits) and the returns they would otherwise receive in the next best alternative (the opportunity cost).  I do not mean this chapter to be a comprehensive review of the environmental valuation literature. Rather, it serves as a background to the case studies presented in the following chapters. It also suggests some of the issues and difficulties that arise in the applied economic analysis of environmental value and the impact of pollution. Just et al. [1982] provide a detailed exposition of the standard theory of welfare economics as it applies to individuals and firms.  Freeman [1993] presents a recent comprehensive discussion of the  application of this theory to environmental issues while McCollum et al. [1992] provide a summary for noneconomists.  Other recent reviews include OECD [1989], Bergstrom [1990], Pearce and Turner [1990],  Cropper and Oates [1992] and OECD [1992]. Here I consider only some of most important points that arise in applied analysis.  In the first three sections of this chapter I discuss the theory and general methodology of economic valuation of aquatic environments.  In Section 3.1, I present a taxonomy of economic values.  In Section 3.2, I  summarize the basic theory of economic value in terms of consumer and producer surplus. I then briefly  41 describe measurement techniques in Section 3.3. Finally, in Section 3.4,1 look in more detail at some of the issues that arise in the applied analysis of the economic value of aquatic environmental services and the cost of reductions in those services.  3.1. Types of Economic Value  In economic terms, the impact of water pollution depends on what the water is valued for. Economists have yet to reach a consensus on the details of how to describe these economic values [Randall 1992, Freeman 1993]. Table 3.2 provides one possible taxonomy for the economic value of aquatic environmental services. An obvious distinction is between use values and non-use values. Use values can be consumptive, as in the extraction of fish, disposal of wastes (water quality is used) or the diversion of water for drinking, irrigation or industrial processes.  Alternatively, use values may be non-consumptive and related to in-water use such  as boating, or related to near-water use such as hiking along a river edge or beach. Aesthetic services also can be considered a source of non-consumptive use value in that they enhance activities involving the water or the ownership of property near water.  For both consumptive and non-consumptive uses some aspect of the environment is explicitly used or experienced, and this provides something of economic value. In contrast, non-use values arise when benefit results from the environment even when no use has occurred. Humans value the existence of landscapes, ecosystems and other species independent of use, or indeed of any intent to use.  Such values are certainly  real, but from an economic perspective pose singular measurement difficulties in that they do not result in observable behaviour in the way that use values do. The concept of existence value, as originally defined by John Krutilla in 1967, emphasized the uniqueness of the valued thing [Kahneman and Knetsch 1992]. It is recognized now that such values are also relevant for non-unique aspects of the environment as well as for temporary damages, although they are likely most important for unique, irreplaceable and long-lasting aspects  42  Table 3.2. Components of the Economic Value of Aquatic Environmental Services Value Class  Value Category  Use  Consumptive (in-water use) - commercial (fishing) - recreational (fishing) - subsistence - waste disposal Consumptive (diversionary use) - municipal (drinking water) - industrial (process/cooling water) - agricultural (irrigation) - hydroelectric Non-consumptive (in-water use) - recreational (boating, swimming) - aesthetic Non-consumptive (near-water use) - near-water recreational (hiking, picnicking) - aesthetic  Non-use  Existence - stewardship (preservation) - vicarious consumption Bequest Option (in face of uncertainty) Quasi-option (in face of uncertainty, irreversibility)  Source: adapted from Fisher and Raucher [1984], Mitchell and Carson [1989], OECD [1989] and U.S. EPA [1993].  [Bishop and Welsh 1992,  Arrow et al. 1993]. The motivation behind existence values may lie in altruism  towards others who are users of the environment or in a desire to preserve other species and their habitats, yielding what have been termed vicarious consumption and stewardship values [Fisher and Raucher  1984,  Mitchell and Carson 1989].  Three other categories of non-use value are sometimes recognized — bequest value, option value and quasioption value [Brookshire et al. 1987, O E C D 1989, Mitchell and Carson 1989, Randall 1992, Freeman 1993].  43 Each has an important temporal element.  Bequest value represents the value an individual derives from  passing on the environment to successive generations. Some authors consider it a motive underlying existence value rather than a separate category of value. Option values arise from the desire to retain the ability to make choices about how to benefit from the environment in the future: having that choice is of value today in the context of uncertainty about the impact of current decisions.  Quasi-option value measures the value  of delaying a choice between various alternatives when the net benefit of each is uncertain and at least one involves an irreversible effect on the environment.  The value arises because of information which may  become available in the future and which could help to inform a better choice. The definition of both option and quasi-option value, and what they may actually measure, has been controversial. Freeman [1993] has gone so far as to argue that neither is meaningful except perhaps as theoretical constructs.  Water pollution can reduce any of the economic values shown in Table 3.2.  Although not explicitly shown  in the table, the life-support services of aquatic environments can also be cast in terms of economic value. Clean water and healthy biota support human life ~ water pollution reduces the quality of drinking water and fish or other food from the aquatic environment, resulting in increased risks of mortality or sickness. These increased risks reduce the welfare of individuals and society.  3.2. The Theory of Economic Value  The essential concepts necessary to an understanding of economic value can be seen by considering the demand and supply for a good or service, as shown in Figure 3.1. Economists assume that the satisfaction of an individual's wants or preferences can be measured conceptually in terms of utility, a measure of welfare. Goods and services have value to an individual because they provide utility, although utility cannot readily be measured empirically. However, the individual is assumed to attempt to maximize his or her welfare and the result of this maximization can be observed in the form of demand relationships which indicate how much  45 the individual values varying levels of goods and services given his or her income level and preferences. Demand curves slope downward because as consumption of the good increases the marginal benefit of any additional amount decreases.  If the individual purchases amount q in Figure 3.1, the total benefit of having that quantity, and hence the maximum amount that the individual would be willing to pay (WTF) for it, is given by area A+B+C.  What  he or she actually pays is the market price (equals marginal WTP for the corresponding quantity) times the quantity, area B+C.  Since the amount paid reduces welfare, the individual only benefits from the net  economic value or consumer surplus, area A. As changes in the amount demanded occur because of price changes, quantity restrictions or quality changes, the welfare impact on the individual is measured as the change in consumer surplus. Thus, in estimating the economic value of environmental damages as they affect individuals, economists seek to derive demand relationships and measure changes in consumer surplus . 3  In fact, economists use several measures of consumer surplus. Marshallian demand curves — the type which relate values and quantities given a fixed income ~ yield Marshallian consumer surplus (MS). The problem is that, theoretically, MS can rise when in fact welfare has actually fallen [Just et al. 1982]. A theoretically more appropriate measure is given by Hicksian demand curves which reflect demands where an individual's utility rather than income is held constant. For example, suppose an individual who owns a riverside house (or enjoys fishing or knowing that a certain water-based ecosystem exists) initially benefits from environmental quality qO which provides welfare or utility at a level of UO. Now suppose that quality rises to ql, raising utility to Ul (welfare improves). The value of this increase, the most that she would be WTP to achieve ql, would be that amount which returns her to the initial utility level UO — at that point she is no worse off than before.  However, if instead ql was a lower level of environmental quality, then utility would fall and  theoretically she should be willing to accept compensation (WTA) for the change to ql provided it was at least  The term "consumer" surplus is often somewhat of a misnomer in the context of the environment where economic value derived from non-consumptive and non-use activity can be very important. However, I continue to use the term for the sake of brevity. 3  46 sufficient to ensure that UO was maintained. The WTP or W T A measures represent the net economic value of the change in individual welfare when faced with a change in environmental quality.  These two examples suggest an important point of special relevance in environmental valuation, the question of the assignment of rights [Bergstrom 1990, Cropper and Oates 1992]. In a typical market situation, the seller has the rights to the good or service in question which the buyer can then obtain by paying for it. The term "rights" is problematic when applied to the environment since it carries with it a sense of legally defined property rights and ownership, although often these are not well-defined with respect to environmental quality and quantity. Moreover, they change over time. A preferable term is "entitlements" by which is meant both property rights and the conditions which may be considered ethically or historically to belong to an individual or society.  The presumption as to where these entitlements lie is key.  In the above examples, the individual was  implicitly assumed to have entitlement to the initial river quality qO and the corresponding level of utility UO. Thus she would be WTP to acquire a greater quality, but had to be compensated if quality was restricted. However, if the entitlement applies not to the initial quality qO but to the subsequent level ql and the corresponding utility UI, then the move to ql from qO must be treated differently.  If ql > qO then the  individual would be W T A compensation which would at least raise her to UI in exchange for foregoing the increase to the environmental quality ql to which she is entitled. If ql < qO then she would be WTP to prevent the decrease because her utility is higher at current environmental quality qO, a level which exceeds the entitlement.  Thus the choice of WTP or W T A depends on whether environmental quality is rising or  falling, and on the placement of entitlements.  The choice of WTP or W T A is important because the measurements are not theoretically equivalent: W T A is expected to be larger than WTP for a given change in q [Just et al. 1982, Mitchell and Carson 1989, Cropper and Oates 1992]. The magnitude of the difference should be small when the change in q is small or  47 when WTP or W T A comprise only a small part of income. When q is large the difference could theoretically be infinite. The difference between WTP, W T A and MS should also be small when small changes in q or small budget items are involved.  In many instances, an examination of entitlements suggests that W T A represents the appropriate measure of the cost of environmental damage imposed on individuals by environmental damage.  Nevertheless, the  generally accepted practice has been to use WTP or MS based on the theoretical justification that the amounts involved often represent only a small portion of the individual's income. However, numerous studies have found that W T A estimates can be substantially greater than corresponding WTP estimates, with differences that appear too large to be ascribed to income effects and which do not seem to reflect methodological flaws in the estimation procedures [Mitchell and Carson 1989, OECD 1989, Knetsch 1990, Adamowicz et al. 1993]. People may attach a greater value to a loss than they do to a gain of the same magnitude [Kahneman and Tversky 1979, 1984]. Moreover, they may be unwilling even to consider the value of something they already have.  Thus other reasons for the common use of WTP are that W T A can be difficult to assess and an  argument that conservative assessments are most appropriate (for example, see Arrow et al. 1993). Why conservative assessments should be most appropriate is unclear ~ they could introduce considerable bias toward excessive environmental degradation [Mitchell and Carson 1989, Knetsch 1990]. In fact, derivation of W T A may not always be as problematic as assumed — Edwards and Carlson [1989] have summarized four methods by which data used to derive MS or WTP can be used to derive W T A .  The concepts of total and net economic value apply to producers just as they do to individuals [Just et al. 1982]. Environmental services serve as intermediate production inputs of producers. Thus reductions in the quantity or quality of the services used as inputs result in a decline in producer welfare. The estimation of producer welfare and how it changes is a much simpler process than that required for estimating an individual's welfare. The relative simplicity here lies in the fact that WTP and W T A are always the same and equivalent to producer surplus.  48 Economists generally assume that a firm's production is characterized by relatively fixed costs, such as debt servicing, and variable costs which rise as output rises. This implies that the marginal cost of producing each successive unit usually rises -- thus the producer's supply curve slopes upward, as shown in Figure 3.1. If the producer is assumed to face exogenously determined input and output prices, and behaves so as to maximize profits, then it will produce and sell a quantity of output such that its marginal cost equals the market price for its output . At the market price it produces an amount q in Figure 3.1, and the total value 4  of the production or the revenue that it receives is area B+C.  The variable cost of that production is area  C so that the net economic value or producer surplus obtained by the producer is area B. Thus in estimating the net economic value of environmental damages as they affect producers, economists seek to measure changes in producer surplus.  In some cases, and in the long run, profits and producer surplus are equivalent, but in general changes in profit do not accurately reflect producer welfare changes. Similarly, changes in gross revenue generally bear no obvious relationship to changes in producer surplus and the two measures may in fact move in opposite directions. However, when market prices are unchanged by the factors that affect producer surplus, changes in gross revenue do provide an upper bound to changes in producer surplus.  3.3. Measuring Economic Value  At this point I have described types of economic value and the theoretical constructs of economic surplus with which they should be measured.  Several methods have been used to actually assess the economic cost of  environmental damage, and there is a considerable literature on each. Only a brief discussion is possible here.  The assumption that the firm exists in a perfectly competitive world in which it has no influence on prices is often unrealistic. Nevertheless, it has always represented the benchmark for the economic analysis of supply on the basis that it provides at least a first approximation to what goes on in the real world. Imperfect competition can add considerably to the complexity of any issue. 4  49 Table 3.3 summarizes the key features of common approaches, showing the application, advantages and problems of each. The character of the damage in question, and the type of value of interest, determine which method is appropriate in a given situation.  The value of goods and services produced using aquatic environmental services as an input of some sort can be relatively easily determined when they are exchanged in markets. When pollution reduces the quality or quantity of the environmental input the relationship between it and market production can be assessed and changes in market consumer and producer surplus determined. In the absence of a market, producer surplus is not an issue and valuation approaches focus on environmental services as they relate to individual welfare. Grounded in the theory of demand, these approaches look for ways in which observed or hypothetical individual behaviour can be used to determined preferences, demands and economic value estimates. Two general approaches exist. The first assumes that estimates of economic value can be derived based on the individual's revealed preferences for the environmental service, as indicated by his or her behaviour in markets related to the good in question. The travel cost method (TCM) and hedonic price method (HPM) belong to this category. The T C M assumes that the behaviour of people with respect to their choice and use of recreational sites reveals their preferences, and that the cost associated with this recreational activity is an indicator of WTP for recreational sites and their attributes (see Smith 1989, OECD 1989, Freeman 1993). The hedonic price concept assumes that the value of an object can be decomposed into the values of its constituent parts or attributes (see OECD 1989, Cropper and Oates 1992, Freeman 1993). For example, lake water quality and the natural vista are constituent elements of the value of a lake-side home. The analyst assumes that an individual's choice of a home and its particular features reveals his or her preferences with respect to the features. Econometric analysis is used to isolate the value of the constituents of interest.  There can be no explicit accounting for environmental entitlements in either of these revealed preference approaches. In contrast, the contingent valuation method (CVM) ~ can theoretically capture either WTP or W T A and so can be used in conjunction with a consideration of entitlements. It can be used to value anything  50  Table 3.3.  Summary of Common Environmental Damage Valuation Techniques* Consumer Surplus Estimation in the Absence of a Direct Market Revealed Preference Hedonic Price Method (HPM)  Expressed Preferences Travel Cost Method (TCM)  Contingent Valuation Method (CVM)  Market Producer and Consumer Surplus Estimation  Behaviourial basis  Observed behaviour with respect to market goods; environment is a component of certain market goods  Observed behaviour with respect to market goods; environment and market goods/services are weak complements  Hypothetical behaviour; market for environmental good/service constructed in a survey  Behaviour of producers and consumers in actual markets  Key Assumption  Individuals purchase market goods with exact bundle of attributes they want, including enviro. attributes  Individuals' choices of recreational sites, and the cost of using them, reflect how they value sites  Survey respondents accurately indicate behaviour as if the market were real  Based on standard economic theory  Demand/supply function estimated  Marshallian demand  Marshallian demand  Hicksian demand, WTP or WTA compensation  Market supply and demand  Ease of Use  Complex  Complex  Complex  Medium complex.  Type of Value  Non-consumptive use  Consumptive use; non-consumptive use  All types including non-use and option  Consumptive use  Type of Service  Some amenity services  Some amenity services  Amenity, non-use and life support services  Natural resource service flows  Application:  Advantages  Relates property prices to environmental damage  Established method; very useful for study of recreation  Very versatile and only approach for non-use value; Hicksian measures  Based on market data  Problems  Key assumption often questionable; data problems likely; statistical problems  Measuring travel cost; considers existing users only; accounting for substitute sites  Respondent's inexperience with valuing environmental goods; WTP/WTA disparity  How to relate supply to the environment  Several other less common methods are not shown here. These include the HPM as applied to wages, hedonic travel cost methods, the averting behaviour method in which expenses associated with avoiding or mitigating pollution are used for valuation purposes, and a variety of ad hoc methods which emphasize physical damage functions and are not grounded in economic theory. 1  Source: adapted from OECD [1989, 1992].  51 but it is especially useful when an environmental service is completely unrelated to any market good so that there is no observable market behaviour (see Mitchell and Carson 1989, Arrow et al. 1993, Freeman 1993, Hausman 1993). The C V M uses structured surveys of individuals to ascertain how they would value the environmental service in question.  The survey instrument creates for the individual a hypothetical or  contingent market including a description of the service to be valued and how the respondent wOuld pay for it or would receive compensation. The individual's response to a WTP or W T A question is then assumed to accurately reflect what would actually occur if the situation or market were real. In other words, it is assumed that the survey respondents adequately express the preferences which would determine behaviour in an actual market.  Each of these techniques has been used extensively for several decades. Table 3.3 indicates that each presents problems and is likely to be complex. Overall, the difficulties encountered in any approach suggest that these methods can be expected to give, at best, only order-of-magnitude value estimates for environmental damage. Each approach has its uses and faults and the validity and reliability of specific studies depend upon the care with which they have been undertaken, the attention give to methodological issues, and the quality of the data available.  This is no less true of market producer and consumer surplus studies, although such studies  generally can be expected to yield more precise estimates.  Original research using any of the techniques represents a costly and time-consuming exercise.  Thus great  interest has arisen with respect to the possibility of value transfers — transferring a per unit value estimate derived in one case, such as the net economic value a recreationalist derives from a day spent fishing or swimming, to the case of interest [Boyle and Bergstrom 1992, Brookshire and Neill 1992, Desvousges et al. 1992]. The methodological and theoretical problems in each of the valuation techniques suggest that such transfers must be approached cautiously. Moreover, very great differences may exist from study to study in most of the important influences on value, as well as in methodological choices which affect value estimates. If the original study is well done in terms of adherence to acceptable theory and valid empirical techniques,  52 and focuses on a situation, environmental service, and economic damage that share qualitatively similar characteristics with the problem at hand, then transfer would be easy. Such similarities seldom exist.  Every value estimate is derived within a certain economic context including time, location, general economic conditions and the economic and social characteristics of the individuals examined.  In addition, the  expenditures associated with the measured activity may vary from study to study, and these affect the measurement of net economic value. For example, a 1975 study may have assessed the economic value of expensive open-ocean salmonfishingundertaken by non-local recreationalists with above average income. Value estimates from such a study are not of obvious relevance for river salmonfishingundertaken primarily by local middle-income people in 1994. In general, the context may not be transferrable or only transferable by explicitly accounting for some of the economic and environmental differences [Loomis 1992, Desvousges et al. 1992]. Thus, although the analyst does not undertake a full valuation exercise, even a value transfer ideally requires a significant amount of primary data to undertake a systematic adjustment to the original study. At the very least, the analyst should qualitatively assess the biases that the transfer estimate may introduce.  The role of judgement in determining the outcome of value transfer studies can not be over-emphasized. For example, Smith [1992] compared two studies of the economic benefit of imposing stricter effluent controls on U.S. pulp and paper mills. He found that one of the large number of mills studied by Luken et al. [1992] and Desvousges et al. [1992] was the same. For the mill in question, Luken et al. concluded that the cost of controls exceeded the benefits in terms of environmental improvement, as measured by transfers derived from three CVM studies. However, the transfer-based estimates derived by Desvousges et al. led to the opposite conclusion.  53  3.4. Measurement Issues in Assessing The Economic Cost of Water Pollution  Table 3.4 shows the variety of economic damages that arise when water pollution reduces the quality or quantity of the services provided to humans by aquatic environments.  In this section I discuss certain  measurement issues in the evaluation of the economic cost of the damages to natural resource service flows, amenity and cultural services, non-use services and life-support services.  One primary economic interest with respect to aquatic systems is fish. Fish are used non-consumptively as part of certain recreational activities, such as hiking along a river edge, visiting hatcheries or watching spawning runs.  They are used consumptively in commercial fishing.  More complicated to consider in  economic terms are the consumptive uses involved in sport fishing or subsistence food fishing which are inextricably mixed with experiential and non-consumptive uses of the broader habitat, or with social and cultural functions. Finally, fish have an element of non-consumptive non-use value in their own right and as part of their habitat.  Most of the following discussion concerns fish, reflecting their importance for the  economic value of aquatic environments and the focus of the case studies in the following four chapters. It should be kept in mind, however, that a great many other economic values can be reduced by water pollution.  3.4.1. Natural Resource Service Flows  Water pollution can affect any industrial, commercial or agricultural activity which requires water of good quality. In general, it appears that relatively little attention has been paid to such damages, at least in the published literature.  Interest has focused primarily on how degradation of aquatic environments affects  commercial fisheries, with the cost to fishers measured in terms of lost producer surplus due to higher fishing costs and/or lower catches. Most obvious are the damages when a fishery in a particular area is completely destroyed either because of physical extinction of the stock, commercial extinction when an event such as toxic  54  •o  §  §  C  •i  I  ll  2  o  OH  BC  1  § H  2  33 •S §  11 o-  2  e  1 •a g  1  §g CO  §  PH  3  1  I  >  <• e  1  E  a -a  11 OH  i  §  -8  I If 3  e o O  I 3  c o O  8 I3 1  i e o O  I o  a ©  c V c  CO  GO  g  o W  c o O  1*  1  •I  2  I  | II OH  1  I  1  IS  I  t5  o cOH o  1  !  > c  •8  1 o  .1 in a  1.1  •a  60 _  Si  o 8  <» -2 S S3  H3 3  •18 2  "S  8 | co-  e  | -a »  a  8 „ 1|  •I .a  3  1  •tj §  "a 3  e  •8  1  S3  c  1  >  c ©  I  0  >  -3  Q  11 CO  c o O  Ia  s 3 C O  z  a u .a  I  J -SI 11  55 contamination forces a closure, or severe quality reductions (due to tainting or a shift in the species mix toward commercially less valuable species). The economic damage to fishers of such a closure is the entire foregone producer surplus unless they can use their equipment and skills in a switch to another fishery or another location. When that occurs, it is the change in producer surplus between the old and new activities which represents the economic cost to commercial fishing.  Cropper and Oates [1992] suggest that, where a fish stock is partially depleted in terms of quality or quantity, one way to think about the economic cost is in terms of averting behaviour in which intensified fishing effort (increased use of boats, gear and labour) is used to offset the reduced fish population (the supply curve shifts in). However, if over-exploitation occurs, then the effect of intensified effort will be simply to drive the stock to commercial (and perhaps biological) extinction sooner. McConnell and Strand [1989] have emphasized the importance of considering the level of fishing in relation to the sustainable yield when determining the net economic value of water quality or habitat changes.  The fisheries management regime and market structure must also be carefully considered. For example, if area licensing or restrictions on entry in other fisheries or locations exist, then relocation will not be an option. In addition, in most commercial fisheries several marketing levels exist including fishers, processors, wholesalers and retailers: each seller may obtain some amount of consumer and producer surplus as the fish product moves along the chain to the final consumer. Estimates of the total economic cost of pollution must include changes in the net economic value at each point in the chain ~ the foregone producer surplus of fishers provides a lower limit. Note that so far I have implicitly assumed that the effect of a depleted stock in a particular area is not sufficient to significantly affect market supply or change the market price of the fish ~ in other words, the depleted catch represents only a small portion of the total amount available in the market. When this is not true and the market price does change, then the impact on the consumer surplus of final consumers must also be considered. Overall, the analysis may become quite complex. One rough rule-ofthumb used by the U.S. EPA [1993] in applied analysis is that the producer and consumer surplus associated  56 with a commercial fishery likely represents 50-90% of the gross value of the landed catch.  A significant problem the analyst often faces is establishing a linkage between environmental impacts of water pollution, changes in fish stock and changes in harvests. Population fluctuations occur naturally: as already noted in Section 2.3, the difficulty lies in separating these natural variations from the effects of variable harvesting rates, stock enhancement programs and a variety of habitat quantity and quality changes caused by humans. Directly incorporating environmental variables in the economic production function can be useful. Ellis and Fisher [1987] use the example of blue crab on the Florida Gulf coast. These crab require wetland habitat and the inclusion of exogenously determined fixed wetland area as an input in the estimated commercial crab production function brings environmental quality into explicit consideration.  If the wetland area is  reduced by water pollution, the supply curve shifts and the resultant change in producer surplus can be estimated, given knowledge about market demand conditions.  In the rare case where a single type of water pollution is known to be the only human influence on the fish stock then it could be relatively easy to disentangle naturally induced fluctuations from those induced by the human influence, especially when the latter are large. For example, Sieler [1989] used a comparable and adjacent relatively pristine habitat as a control to assess how a long history of pulp effluent discharges had affected adult coho salmon production in Grays Harbour/Chehalis River in Washington state. He found that coho smolt survival to adulthood recruitment was half of the rate in the control system, with the impact of attributable to pulp mill effluent — the effluent is by far the major source of habitat disruption. He was then able to assess what this meant in lost annual catch but incorrectly estimated the economic cost to commercial fishers as the foregone gross revenue. If it can be assumed that all of the additional catch would be sold at the prevailing market price then the lost gross revenue provides an upper bound to the lost producer surplus.  57  3.4.2. Amenity and Cultural Services  Most empirical economic analyses of water pollution focus on amenity services to recreationalists, especially anglers.  In part this reflects that fact that these damages often seem most susceptible to economic  quantification. As well, many analysts consider damage to recreational sites to be the most significant economic cost of water pollution, a belief that may reflect an unpublished 1979 study by Rick Freeman for the U.S. EPA (as cited by Vaughan and Russell 1982). In the study, he synthesized the existing material to derive a rough estimate of the economic benefit to the United States of attaining the 1985 water quality objective mandated by Federal regulation. He found that recreational activity (largely fishing) accounted for over half of the benefits.  The economic value of most recreational activity derives from the subjective experiential elements of the activity and the recreational site (aesthetic enjoyment of the setting, being outdoors, interacting with friends/family etc.). To a very large extent this is equally true of sport fishing, and usually the size and the number of the fish caught are of relatively little consequence.  Thus the benefit that anglers or other  recreationalists derive from their activity must be measured in units which capture the experiential element - common measures are time-related units which, in the case of anglers, are angling-day or boating-day. The total economic value of this benefit is the entire area under the demand curve for the recreational experience. Subtracting the expenditures undertaken to participate in the activity then yields consumer surplus. Just as with commercial fishing, the possibility of moving to alternative sites or switching activities must be kept in mind. The loss in consumer surplus to an individual resulting from a shift to a less preferred alternative recreational site or activity is the valid measure of the cost of water pollution at a particular location. Only if all alternatives are lost is the entire consumer surplus lost.  As already emphasized, the value of fishing is not the same as the value of fish. However, often the interest lies in determining the consumer surplus value that can be attributed to increases or decreases in fish stocks.  58 Estimates of this value are usually derived using average fish values, calculated as average angling-day value over average catch per day. Clearly, such an approach overestimates the value of fish to anglers since much of the angling-day value derives from elements other than the catch. It also assumes that changes in the total harvest result from changes in the number of angling-days. An alternative is that the total harvest changes due to changed catch rates, which alter the value of existing fishing days. Changes in the catch rate are less valuable than additional fishing days so that the usual method generally yields an overestimate, perhaps a substantial one. For example, a C V M study of British Columbia south coast salmon anglers by Cameron and James [1987] found that each angler caught an average of 0.5 chinook per angling-day and had WTP value for the current day of fishing of about $44. The common approach to valuing an increase or decrease in the total sport harvest would use a figure of about $88 (44/0.5) per chinook, based on this study. However, Cameron and James also found the marginal WTP value for an additional (marginal) chinook caught and kept to be about $6, a much lower amount.  In general, the few available estimates suggest that the marginal  economic value of a sport fish is an order magnitude lower than the economic value of a sport fishing day [Norton et al. 1983, Johnson and Adams 1989]. Of course the relationship will vary according to the catch rate and the time spent fishing. Also, at the extreme of no fish or a low or zero catch rate, the difference between fish value and fishing day value becomes meaningless since the activity only occurs because fish exist.  However, these results do reinforce the fact that fish catch is only one factor which generates  recreational fishing value.  A key issue in valuation of damage to any aquatic environmental service flow, but especially with respect to amenity services, is how to relate scientifically meaningful measures of environmental damage to economic values. One approach is to explicitly include pollution indicators in estimated recreational demand functions. For example, in a T C M study Bockstael et al. [1987] examined how specific water quality variables, along with other variables, affected the choice and intensity of use of Boston-area beaches. Such an approach seems to provide a link between pollution and economic value provided that people actually do use water quality measures in making their choice of recreational site, or that such measures are highly positively correlated  59 with what really determines their choice. This might be questionable since perceptions of water quality are likely more important than an understanding of actual water quality parameters in the choice of recreational sites.  Another approach has been the use of variants of the water quality ladder developed in 1981 by William Vaughan of Resources for the Future and published in an EPA draft report by Robert Mitchell and Richard Carson (cited in Smith et al. 1983, Luken 1990). The ladder defines various possible recreational uses of water in terms of the dissolved oxygen (DO) content assumed to be required for the activity. For example, < 3.5mg/L DO is considered unusable for any recreational activity, 3.5-4.0 mg/L can be used for boating but not fishing or swimming, and so on. Changes in DO likely would be of little meaning to most people but recreational users of a particular site can be asked in a C V M study about the value they attached to changes in the level of water quality such that feasible recreational uses change. Alternatively, value transfers from other studies can be used to value changes in water quality. The attraction of this approach is that the analyst can use it to relate economic value to changes in a quantifiable and scientifically meaningful water quality parameter.  It also provides a connection between recreational uses and pollutants since the relationship  between deoxygenating substances and DO in a particular instance can be modelled given hydrological information about the water body. These advantages account for the use of the approach in quite a few studies in the United States, including a number which examined the benefit of greater control of pulp mill pollution [Vaughan and Russell 1982, Naughton and Desvousges 1986, Smith and Desvousges 1986, Luken 1990]. One serious problem with the method is that while DO is certainly relevant for fishing, its relevance for other types of recreational activity is less clear, and it may have little or no correlation with some of the water quality parameters which do affect these activities [Russell and Smith 1990].  Value estimates are unlikely to be independent of the methodological choices and judgement of analysts. The very substantial number of estimates of the economic value of environmental amenity services has led to a number of meta-studies (studies of studies) to try to explain the observed variation in estimates in the context  60 of the need for a better understanding of the influences which must be considered in value transfers. For example, Walsh et al. [1992] reviewed 20 years of recreational demand studies in the United States and found 287 T C M and C V M studies which estimated net economic values for a total of 19 recreational activities (camping, boating, hiking, various types of fishing and hunting, and so on). The mean value across all studies was $34 1978 U.S. dollars per activity day, with a range of $4-200. The analysis was only of limited success as the cross-study econometric analysis was not robust, but it did reveal a number of apparently systematic differences in value estimates, related, for example, to the type of C V M or T C M study performed. In a similar vein, Russell and Smith [1992] reported a meta-analysis of 77 T C M studies which yielded 734 observations of net economic value for recreational activity at seven types of sites (river, lake, wetland and so on). This study also found systematic influences reflecting researcher's varying approaches.  Although most analysis of the economic value of amenity services provided by aquatic environments has focused on recreation activities, there are at least two other important sources of economic value here. One is the aesthetic value provided to those who live beside water bodies. This can be measured by the H P M and Steinnes [1992] has reviewed the few attempts to do this. He argued that all of these studies contain serious problems which make their results suspect. His own study of Minnesota lake-side lots illustrates the difficulty of linking scientific assessments of water quality to economic value.  He found that higher water quality  contributed to higher economic value, but the measure of water quality used, clarity, is one that influences individual's perceptions of quality while potentially having little relationship to scientific assessments of water quality.  Subsistence harvesting activity represents another important source of economic value based on aquatic environmental services.  Subsistence harvesting of fish and other aquatic life represents an important use of  water bodies in some locations, especially remote areas. This activity is of primary concern with respect to First Nations peoples for whom it provides food and other products as well as incorporating significant cultural, social and recreational roles [Muth et al. 1987].  These roles cannot be measured readily by  61 economic analysis although attempts have been made to do so.  For example, the U.S. E P A [1993] used  examples of compensation settlements for the loss of native subsistence uses to derive per capita compensation estimates as a proxy for the value of the services provided to natives by the environment.  An important  problem with such an approach is that there may be few relevant settlements. More seriously, potentially relevant settlements may be imposed or contain some other element of coercion, and thus cannot be considered as valid indicators of the economic value of subsistence activity. This was a problem in the EPA study.  A variety of methods have been used in the valuation of the food component of subsistence harvests. Brown and Burch [1992] discuss the economic theory behind some of these approaches.  One is the use of  commercial landed fish prices, on the basis that individuals would likely consume less if they had to purchase the fish at retail prices and because many indigenous people desire access to commercial markets for the fish they catch [Pearse Bowden Economic Consultants 1975, Pearse 1994]. In some cases, access to commercial markets is indeed desired for additional fish caught above and beyond cultural and food needs. In such cases, using landed prices would be valid for valuing those additional fish. However, if native people are considered to have an entitlement to a certain catch, and water pollution reduces the catch below that level, then this approach is not satisfactory for assessing the associated economic cost of the lost food source. It does not recognize that the correct measure of the reduced welfare should be based on a valuation consistent with a return to the original (pre-pollution) utility level. Thus a second method involves the use of replacement costs. Usher [1976, 1987] advocated the valuation of a subsistence harvest using local retail prices for the same item or a nutritionally close substitute.  If the cost of engaging in subsistence harvesting is subtracted from the  replacement or landed price then a net economic value for the food component of the harvest can be derived. Alternatively, use of the unadjusted replacement or landed price goes some (unknown) way toward accounting for the full cultural value of the harvest. A third approach has been the use of "local exchange value," the price one member of a group would charge another [Usher 1976]. Berger [1977] used both this method and retail replacement prices in his evaluation of the effects of the proposed Mackenzie Valley pipeline. Finally, a fourth method values subsistence harvests using economic values derived in studies of recreational activity,  62 with the idea that these values will at least partially account for cultural value [Scott et al. 1987].  3.4.3. Non-Use Services  Economic damage to the non-use values of aquatic environmental services can be estimated only via the C V M . The difficulty is that survey respondents have no experience in such valuations, creating considerable concern as to the validity of the resulting non-use value estimates. The most thorough review of the issue to date has concluded that well-designed and implemented C V M studies do provide reasonable estimates of these values [Arrow et al. 1993]. There in no doubt that non-use values can be quite substantial and more important than use values in some cases [Bishop et al. 1987, OECD 1989, Green and Tunstall 1991, OECD 1992]. The greatest conceptual difficulty lies in the determination of non-use value for non-unique individual sites or areas.  Fisher and Raucher [1984] reviewed a number of studies which attempted to estimate current  recreational use, non-use and option values for specific recreational sites. They concluded that while most of the studies are problematic in various respects, collectively they support the idea that combined non-use and option values reflect a very important component of the total value of recreational sites. Overall, they suggested that using 50% of recreational use value will not grossly overstate the existence and option values of a site for recreational users, and may even substantially understate them if the site possesses an element of uniqueness.  In contrast to the thinking of Fisher and Raucher, some writers do not believe that it is possible to attach nonuse values to particular recreational sites, but only to the environment as defined broadly.  For example,  Green and Tunstall [1991] believe that people may hold both specific and generalized preferences. It seems implausible, they argue, that an individual values each and every stream (specific preferences) but they very likely value all streams (generalized preferences). The problem is then how to determine what fraction of the total value applies to the stream of interest.  63 A number of studies in the United States have broached the question of the non-use value of aquatic environments at the national level. For example, Carson and Mitchell [1993] reported the results of a 1983 U.S. national C V M study to determine the economic value attached to various levels of clean water. Survey respondents were told what proportion of freshwater bodies were boatable, fishable and swimmable, where these proportions were based on the water quality ladder described above and D O concentrations estimated in a national water quality model for freshwater sites. Survey households indicated an average annual WTP of $106 1983 U.S. dollars for national water quality defined as boatable, an additional $80 for water defined as fishable, and finally another $89 for water defined as swimmable (median values were about 40% of these average values).  In an earlier 1981 study, Mitchell and Carson found that about two-thirds of the value  households were WTP for national fishable water could be attributed to the desire to have fishable water in their own state (as cited in U.S. E P A 1993).  Part of the economic value expressed in these studies  represented the use value that some households derived from aquatic environments but some of it represented non-use value — clearly, such values can be quite large at the national level, if the results of these studies are valid. The U.S. EPA [1993] has used both these types of non-use value estimates and the Fisher and Raucher rule-of-thumb described above to assess the non-use benefits derived from stricter control of pulp and paper mill chemical contaminants.  3.4.4. Life Support Services  The most basic service provided to humans by the environment is life support. It is ethically difficult to attach economic value to these services ~ this would involve establishing the value of human lives supported by the environment. However, changes in life-support services which lead to changes in mortality and morbidity risks can be more readily places in ah economic context.  Estimating the economic value of these changes  involves two highly complex and potentially controversial methodological issues.  64 The first issue is how to measure the incidence of health impacts associated with the pollutants of interest. Essentially this involves determining what health impacts may occur, identifying dose-response relationships, and an assessment of the extent, rate and nature of human exposure [U.S. EPA 1993]. The result would be a measure of the change in the risk of death or disease due to a particular contaminant. Obtaining such results is by no means an easy task and great uncertainty is likely to exist in each component.  The second complex issue involves how to attach economic values to these risk estimates [Cropper and Oates 1992, Frankel and Linke 1992, Freeman 1993, Viscusi 1993].  The valuation of changes in the risk of  mortality or morbidity requires estimates of individual's WTP for reduced risk or W T A compensation for increased risks, and these can be derived using HPM or C V M studies. However, even more so than in the case of the amenity and non-use services of the environment, valuation of changes in risk can vary considerably according to the context of the situation analyzed — income, age, the characteristics and magnitude of the risk, and other factors, all matter. In addition, there can be significant differences between technical risk estimates and public perceptions of risk. Overall, the transfer of estimates from one study to another is extremely problematic. With respect to valuing changes in mortality risks, economists typically translate value estimates into estimates of the "value of a statistical life." This translation is essentially one of analytical convenience and the estimates refer not to the economic value of an actual life but to the value of changes in the risk that a death will occur. Nevertheless, serious ethical concerns about such an approach persist.  Two recent studies of pulp and paper mill effluent have broached the question of the economic damage associated with increased human health risks due to chemical contamination. Luken [1990] used a range of 2,3,7,8-TCDD concentrations observed in tissues of fish downstream of mills, combined with estimates of daily fish consumption, the population estimated to fish near the mills, and an EPA dose-response relationship, to derive a range of estimates for the annual incidence of cancer in terms of the number of expected cancer cases. He did not estimate the economic cost of the increased incidence of cancer but the U.S. EPA [1993]  65 did take the analysis this extra step in a similar study. The EPA evaluated the national economic value of reductions in the risk of cancer from six probable carcinogens found in pulp and paper mill effluent, including 2,3,7,8-TCDD and 2,3,7,8-TCDF. They then used a range of estimates reported in the literature for the value of a statistical life to derive a total value for the reduced risk of cancer. The potential for controversy with respect to these incidence and value estimates is clear since they involve very considerable uncertainty and many assumptions, not to mention the ethical concerns alluded to above.  The public perception of the safety of water and fish from pollution sites may be as important in economic terms as the actual health risks. Perceived as well as actual risks influence the use of aquatic environments and hence help to determine economic value. For example, health advisories since 1989 regarding chlorinated dioxin contamination of various fish species near four interior British Columbia mills may have reduced the economic value of sport fishing for all species near these mills. A consumption advisory can reduce economic value by reducing participation in a sport fishery or by reducing the economic value derived by those who continue to fish. Based on an assessment of several studies of the effect of fish consumption advisories on angler behaviour in the United States, the EPA has concluded that the common action of those anglers who are aware of an advisory is a change in fish preparation and consumption habits, rather than changes in either fishing locations or participation rates [U.S. EPA 1993]. Overall, however, the EPA found that the literature suggested that consumption advisories reduce angler participation by about 20%. This result is suggestive of the potential magnitude of the effect of health concerns on sport fishing activities.  3.5 Summary  Aquatic environments provide a variety of services which yield a benefit to society. Net economic value is the total value of these benefits less the expenditures made to obtain them. Valuation of economic damage arising from water pollution means assessing reductions in net economic value as the quality or quantity of  66 aquatic environmental services is reduced. Economists measure these reductions in terms of consumer surplus and producer surplus.  Valuation of each type of aquatic environmental service, and of damages to it, presents a variety of problems and issues which the analyst must consider in deriving theoretically sound and empirically valid estimates. Economists have developed a variety of measurement techniques but each is problematic in certain ways. In practice, the transfer of value estimates from another study to the problem at hand is common, but must be done cautiously if the results are be meaningful.  67  4. DISSOLVED OXYGEN REDUCTIONS AND SPORT FISHING - ALBERNI INLET  4.1. Background  MacMillan Bloedel constructed a pulp and paper mill at Port Alberni on Vancouver Island in 1947 [Koning et al. 1993]. Current production capacity stands at 175 ADt/d of market unbleached softwood kraft pulp and 950 ADt/d of newsprint derived from unbleached kraft and several types of mechanically derived pulp [Pulp and Paper Canada 1994]. The mill stopped bleaching in 1993 after several years of using peroxide rather than chlorine as a bleaching agent. It discharges effluent to the head of Alberni Inlet at the mouth of the Somass River and is by far the principal source of pollution in the system (see Figure 4.1). Alberni Inlet is a 45 km estuarine inlet extending from Barkley Sound on the west coast of Vancouver Island to the mouth of the Somass River.  The river has a profound impact on the Oceanography of the inlet.  A pronounced and  permanent halocline (salinity gradient) and pycnocline (density gradient) over the length of the inlet separate the upper seaward flow of river freshwater from the deeper inward flow of denser saltwater [Tully 1949]. The thickness of the upper layer varies according to winds, tide and river flow. As it flows seaward, fed by the Somass River, it entrains cooler more saline water from beneath the halocline, which averages 1-4 metres thick. As a result, the lower level has a net flow toward the river mouth. Effluent is discharged to the upper level to take advantage of its seaward flow. Because the liquid portion of the effluent is lighter than the deep saline water there is virtually no downward mixing through the halocline. The distinct stratification of the inlet, and its quiescent nature, mean that mixing and dispersion of the effluent occur slowly so that severely impaired dissolved oxygen conditions result from effluent BOD and colour, and a long history of TSS deposition.  The creeks and rivers feeding into Alberni Inlet, especially the Somass River, provide spawning ground for one of the largest populations of Pacific salmon on Vancouver Island. Barkley Sound, Alberni Inlet and the Somass River support major commercial, recreational and First Nations food salmon fisheries dependent on  69 natural stocks as well on Department of Fisheries and Oceans (DFO) enhancement programs at Henderson Lake and on the Somass River system.  In order of importance, the Pacific salmon using Alberni Inlet are  sockeye, chinook, coho and chum. Within DFO Management Area 23 (Alberni Inlet and Barkley Sound), the waters feeding Alberni Inlet have historically accounted for the spawning escapement of all of the sockeye, roughly 95% of the chinook, 80% of the coho and 30% of the chum [Brown et al. 1979, D F O 1990]. Sockeye spend the first year or two of their life in freshwater lakes before migrating to the sea ~ in recent years, Sproat Lake and Great Central Lake on the Somass River system have accounted for roughly 90-95% of the sockeye, with the remainder associated with Henderson Lake. Similarly, probably 90% or more of the chinook and coho which use the inlet migrate to the Somass River. Thus a substantial portion of Area 23 stocks experiences direct exposure to pulp mill effluent in the upper inlet and the Somass River estuary. The focus here is on chinook and sockeye.  Table 4.1 presents estimates for 1956-1994 of the total sockeye and chinook run in Alberni Inlet. The effect of enhancement programs on the stocks since the early 1970s can be seen clearly. Salmon habitat, passage and stock enhancement efforts have had a century of history in Alberni Inlet and Somass River system, as described for sockeye by Hyatt and Steer [1987]. There is no evidence that the recent efforts (since 1970) have been undertaken in response to the perceived or expected effects of effluent on the salmon stocks. The earliest direct salmon enhancement effort was the creation of a sockeye hatchery at Henderson Lake in 1910. It was closed in 1935 for financial reasons but, in the 1960s, hatchery experiences in the U.S. Pacific Northwest suggested that with new knowledge and techniques, such operations could be economically justified [Department of Fisheries and Forestry 1970]. The Federal Department of Fisheries developed a renewed interest in a hatchery program and the existing structures at Robertson Creek in the Somass River system (from an unsuccessful spawning channel in the early 1960s) made it an obvious location to establish a hatchery. Thus in 1972 a hatchery was built there to enhance chinook and steelhead stocks, and the first chinook returns occurred in 1976.  In 1977, D F O committed significant resources to the Salmonid  Enhancement Program (SEP), a large scale effort with the objective of doubling British Columbia salmon  70  Table 4.1. Chinook and Sockeye Run , Alberni Inlet, 1956-93 (thousands) 1  Sockeye  1956-60 1961-65 1966-70 1971-75 1976-80 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993  1  Chinook  64,600 75,900 175,300 332,800 953,700  18,500 15,800 19,200 33,200 59,700  1,392,100 940,300 1,604,600 1,198,400 724,700 375,500 686,400 794,000 435,600 338,700 1,863,200 996,800 1,281,100  81,800 129,800 117,100 149,300 123,700 73,700 90,000 138,700 161,500 229,400 298,900 216,200 201,000  Total Alberni Inlet run = (escapement to Alberni Inlet + commercial net fishing in Barkley Sound + inlet sport fishing + First Nations fishing + DFO test fishing). 1956-75 data refer to total Management Area 23: all sockeye and most chinook use the inlet. Troll commercial fishing is not included since it is usually conducted offshore on stocks not native to Area 23. Barkley Sound sport fishing not included. First Nations food fishing for all of Alberni Inlet after 1990, up to 1990 for Somass River only. Creel coverage of the sport fisheries varies from year to year. Chinook escapement for 1976-1993 estimated as Somass River escapement plus 1,500 to account for other rivers.  Source: based on data from Knapp and Cairns [1978] for chinook, 1956-75 and sockeye, 1956-70; Starr et al. [1984] for sockeye 1971-80; and W. Luedke and T. Gjernes, DFO Nanaimo, for other years.  stocks [Pearse 1994].  The Robertson Creek hatchery has consistently represented the most important SEP  facility on the west coast of Vancouver Island.  It is now the largest chinook hatchery in British Columbia,  with a production capacity of over 8 million chinook juveniles and over 1 million coho juveniles [Cross et al. 1991].  71 DFO also has devoted considerable attention to sockeye enhancement over the past 25 years [Hyatt and Steer 1987]. In the summers of 1970-73, DFO personnel conducted controlled nutrient additions to Great Central Lake to increase primary productivity [Lebrasseur et al. 1978]. The experiments confirmed that this would increase production at succeeding trophic levels, eventually resulting in a significant increase in juvenile sockeye production. The success of the experiments led to fertilization at a number of B . C . coastal lakes. Nutrient addition began again at Great Central Lake in 1977 and began at Henderson Lake in 1976, and continues to this day. At Sproat Lake, residents of the area disliked the reduction in water clarity that resulted so that fertilization was undertaken only in 1985 and 1986 [K. Hyatt, DFO Nanaimo, pers. comm., 7 Dec. 1994].  D F O scientists believe that roughly 50% of juvenile sockeye production in treated lakes can be  attributed to fertilization [Hyatt and Steer 1987].  The migration patterns of the salmon mean that exposure to pulp mill effluent is maximized just as the fish are moving through the system.  Juveniles pass through the inlet in the spring and early summer. In the  summer and fall, adults migrate up the inlet and congregate at the estuary, before continuing up the Somass River to their spawning grounds [Knapp and Cairns 1978, Brown et al. 1979, Stucchi et al. 1990]. After spending 2-3 years at sea, Somass system sockeye begin to arrive in Barkley Sound in May and arrive in the area of Port Alberni in July and August. Spawning starts in late October and ends a month later. Chinook spend 3-5 years at sea and first arrive at the Somass River in August and September. After moving upstream they spawn from October through December. Thus large numbers of adult fish are found in the upper inlet in July to September. This period of intensive adult use corresponds to Somass River low flows: the Somass reaches its peak in November-January with minimums in July-September, and effluent concentrations and D O levels in the upper inlet follow a corresponding pattern [Birch 1989].  Salmon are susceptible to any disturbance in natural D O conditions, especially since they tend to be more demanding in terms of DO requirements than many other fish. Table 4.2 summarizes the D O guidelines for salmonids suggested by Davis [1975]. These guidelines are widely used for both freshwater and marine fish  72  Table 4.2. Dissolved Oxygen Criteria for Salmonids (mg/L and % saturation)  1  Freshwater Environment  Marine Environment  mg/L  % Sat.  mg/L  9.75 8.00 6.50  98-100 76-95 54-78  -  7.75 6.00 4.25  76-93 57-72 38-51  9.00 6.50 4.00  2  % Sat.  2  Early Life Stages  3  Level A Level B Level C  Other Life Stages Level A Level B Level C  1  2  3  100 79 57  Level A - little or no harm; Level B - possibility of moderate harm if exposed for more than a few hours; Level C - possibility of severe harm, especially if exposed for more than a few hours. Percent saturation over a range of 0-20°C and assuming 2.8% salinity for marine environments. Early life stages defined as eggs and larvae (considered to be up to 30 days after hatching).  Source: Davis [1975].  in the Canadian environment and are differentiated by 3 levels of protection afforded (Level A - little or no harm, Level B - moderate harm, Level C - severe harm) and temperature, and expressed in terms of both mg/L concentration and percent saturation. Davis stressed that meaningful D O criteria should consider the availability of oxygen (measured by concentration in water), the capability for diffusion into aquatic organisms through the external/internal oxygen pressure gradient (which depends on percent saturation), and the metabolic demands for oxygen (which increase non-linearly as temperature rises). Official Canadian water quality guidelines correspond to a slight level of impairment for salmonids in freshwater, according to the levels of Davis [CCREM 1987]. No guidelines take into account the impact of contaminants on oxygen  73 requirements, since little is known about the relationship.  Although pulp production at Port Alberni has a long history, problems with D O did not appear until the mid1960s. Waldichuk [1956] reported no D O problems in the mid-1950s in a study undertaken to delineate existing impacts, given proposals to more than double the output of the mill. Efforts were made to ensure that the expansion of the mill did not lead to severe disruption of DO levels, including construction of a dam to help maintain adequate Somass River flow in the critical summer period. Despite efforts to regulate discharge to ensure appropriate conditions, potentially harmful DO levels were observed during the warm dry season of 1965 [Waldichuk 1987]. The mill constructed primary clarifiers in 1966 and began secondary treatment of 50-60% of its effluent in 1971 [Parker et al. 1972, Birch 1989, Koning et al. 1993], but DO continued to drop below levels which cause severe harm or mortality to salmonids at certain periods of the year and for several kilometres down-inlet [Sullivan 1978, 1980, 1981].  Fortunately, the extent of the DO decline in Alberni Inlet can be assessed easily since excellent detailed data are available for D O conditions both prior to construction of the mill and for recent years. The mill is not the only source of pollution in the inlet ~  for example, Port Alberni discharges a small amount of treated  municipal effluent ~ but there is no doubt that the mill is by far the principal contributor of deoxygenating substances.  Tully [1949] provided a comprehensive review of water quality conditions for the early 1940s.  In the upper inlet, DO levels were generally at full saturation or higher in the surface layer and the halocline, and corresponded to Davis Level A conditions. In 1941, DO levels above and below the pycnocline (to a depth of 10 m) averaged 9.5 mg/L and 7 mg/L, respectively [Birch 1989].  Mill effluent has caused a  progressive deterioration over time ~ by 1990, D O had declined by roughly 60% (4 mg/L) below the halocline and by roughly 10% (1 mg/L) in the surface layer [Stucchi et al. 1990]. For example, in 1987, DO levels above and below the pycnocline averaged 8.2 mg/L and 1.9 mg/L, respectively [Birch 1989]. DO conditions over the last two decades have typically corresponded to Davis Level B or C conditions of impairment. As Figure 4.2 shows, between 1977 and 1988, 50% of DO observations in the surface 2 metres  74  oo GO i  ON o o  c  s Q >-»  X!  S  c  o E  CO Ul  3  C3  O  C CU o  O T3  c CO  *©  c« t/5 as  o  cN  co  CD  CM  cu  CO  ~  u 3  05  at  O  WD  C D CO  £  o  CNJ  (ssiieui)  L| id Ba  75 at Hohm Island 2 km down-inlet from the mill were below Davis' Level B (79% saturation). Below 2 metres, the median concentrations were below Level C (57% saturation), implying that fish would suffer severe effects or death if exposed for more than a few hours.  Studies done in the 1970s suggested that the naturally lower oxygen conditions below the halocline were exacerbated by effluent in the surface layer [Parker et al. 1972, Parker and Sibert 1973]. The effluent was thought to block light from reaching the photosynthetic algae below the halocline, causing oxygen depletion at the halocline and in the lower layer, and in the surface layer through entrainment. This has been discounted as a major problem [I. Birtwell, DFO Vancouver, pers. comm., 24 Aug. 1994] and attention has turned to the sediment oxygen demand (SOD) created by TSS deposits [Stucchi et al. 1990]. Mill investigation of SOD in the late 1980s concluded that it was not of great significance but DFO and Environment Canada scientists now believe that hypoxic waters below the halocline in the upper inlet reflect primarily the SOD resulting from decades of TSS deposition by the pulp mill and from log booming, exacerbated by a naturally slow flushing rate. Deposits from the mill cover about 40% of the harbour bottom, while about 25% of the harbour is covered by bark beds associated with log booming [Stucchi et al. 1990].  In any case, it is clear that significant problems with D O reductions exist and these result primarily from mill effluent. The difference between anthropogenically created and natural D O conditions suggests strongly that without a sustained and significant enhancement program, the important chinook and sockeye salmon stocks in Alberni Inlet would have been reduced. Recent assessments note that "dissolved oxygen conditions in the harbour are frequently below acceptable limits and pose an increasing risk to the maintenance of sustainable fisheries" [Stucchi et al. 1990, p.i] and that "warmer climatic conditions are considered to exacerbate the existing habitat degradation of upper inlet waters, thus posing a perennial risk to the health and survival of salmonids" [Piercey et al. 1993, p.l]. DO conditions below the halocline can be key during the late summer period when reduced rainfall and low river flows mean that the upper layer becomes thinner. Low levels of oxygen in the deeper water restrict salmon to the surface layer but it is that layer which has the highest  76 effluent concentration and seasonably stressful temperatures.  The most dramatic consequence of this habitat degradation occurred in August-September 1990 when an estimated 100,000 sockeye were lost during the spawning migration [K. Hyatt, D F O Nanaimo, pers. comm., 7 Dec. 1994]. D O levels were 3 mg/L below a depth of 5 m and the usually adequately oxygenated surface layer averaged only 2 m in depth and was sometimes absent [Stucchi et al. 1990]. Approximately 450,000 sockeye had arrived as expected in the outer inlet by early August but severe drought conditions (high temperatures, no rain for an extended period) resulted in unusually low Somass River flows, effectively creating a barrier to migration. As a result, the sockeye were forced to pool in the inlet for a much longer period than usual in stressful conditions of low DO and high surface temperatures. Average residence time in the inlet increased from 1-2 weeks to up to 2 months. Pooling occurs naturally with or without effluent but over an extended pooling period water quality conditions become critical.  In addition, in 1990 a  prolonged period in the inlet exposed the fish to a rapidly expanding parasite population. These stresses would cause direct mortality of some fish and weaken others, reducing survival rates during the final stage of the migration. It would also reduce the viability and quantity of gametes, affecting reproductive success rates. Because of the severity of the conditions, MacMillan Bloedel closed part of its mill for 10 days in early September to protect the salmon at the head of the inlet. However, virtually no D O improvements resulted despite a 40% reduction in BOD discharge, a fact which supports the theory that SOD creates a strong continuing oxygen demand.  The effect of the 1990 sockeye loss extended beyond the loss of the adult spawners in that year. Reduced returns in later years will result — an estimated additional loss of 100,000 adult equivalents, after adjusting for expected marine and freshwater survival rates — due to the reduced size and viability of the 1990 brood stock [K. Hyatt, DFO Nanaimo, pers. comm., 7 Dec. 1994]. Most of these fish would have returned in 1994 and 1995.  77 The climatic conditions in the summer of 1990 were particularly severe but similar types of conditions have occurred in other years, leading to extended pooling periods in Upper Alberni Inlet. For example, hot dry conditions resulted in abnormal holding periods in 1992 and 1993, and to a lesser extent in 1991, although in none of these years were conditions as severe or prolonged as in 1990. There have been previous mortality incidents prior to 1990 involving observations of several hundred to several thousand dead fish associated with such climatic events [K. Hyatt, DFO Nanaimo, pers. comm., 26 Jan. 1995]. The difference in 1990 was that procedures to fairly accurately assess sockeye stock size were in place. Only about 7,000-12,000pre-spawned dead sockeye were observed in Alberni Inlet and the lower Somass River in 1990, but the total loss could be inferred with some confidence.  Thus it is quite possible that there have been large losses related to a  combination of climatic conditions and reduced water quality in other years. Observed changes in total stock size would have been ascribed to other factors such as natural variability in marine and freshwater survival rates.  In general, the likelihood of serious disruptions of chinook is less than that for sockeye since peak chinook migration occurs at a slightly later time in the year when temperatures are lower and early autumn rains usually have raised the Somass River and increased the depth of the upper layer in Alberni Inlet. Aside from sockeye and chinook, other finfish and shellfish which use the inlet very likely have been affected by water quality degradation.  The 1990 incident, and concerns of DFO and D O E scientists regarding the continuing health of salmon stocks which use Alberni Inlet, led to specific Federal regulations to govern effluent from the Port Alberni mill, effective December 1992 [Canada Gazette, Part II, 126(11), 12 Nov. 1992, pp. 4603-4621]. The regulations are more stringent than the 1992 Federal Pulp and Paper Regulations which apply to all other mills. At the beginning of 1993 the mill began operating full and updated treatment facilities [Koning et al. 1993].  78 4.2. Economic Valuation  A variety of economic values are adversely affected by D O reductions in Alberni Inlet: here the focus is on the chinook and sockeye sport fisheries. Table 4.3 shows chinook and sockeye sport fishing data for 1956-94. Sockeye fishing is concentrated in the upper inlet and occurs primarily from mid-June to mid-August. Chinook fishing occurs over a larger part of the inlet with greatest effort during August and early September. Sport activity is regulated by DFO through licenses, catch limits and short-term closures of specific areas. Historically, chinook has been the most important sport fish while sockeye catches have been low or nonexistent in the inlet. The lack of sockeye catch reflected in large part a lack of angler knowledge about how to catch the fish, although they have always been sought after for canning purposes [T. Gjernes, D F O Nanaimo, pers. comm., 13 Feb. 1995; L . Gordon, DFO Port Alberni, pers. comm., 16 Feb. 1995]. With new fishing techniques and gear, interest in the Alberni Inlet sockeye stocks has grown tremendously in the last 5 years, to a point where the sport catch rivals total Area 23 sport chinook catches and may represent the most important sockeye sport fishery in the province [L. Gordon, DFO Port Alberni, pers. comm., 16 Feb. 1995].  Enhancement efforts have increased the total sockeye run despite reduced habitat quality and quantity, and pooling in some years has improved fishing opportunities, such as in 1992 and 1993. Conversely, relatively cool and wet weather, which facilitates quick passage of sockeye through the inlet, reduced fishing opportunities in 1994.  Overall, sport fishing has not been reduced by mill effluent, and in fact has reached  record levels in recent years. However, in terms of evaluating the potential economic cost of the effluent, the relevant comparison is between actual sport fishing activity and that which could have occurred in the absence of habitat degradation, but including enhancement programs.  DO impairment should be evaluated economically in terms of major destructive events and in terms of chronic, but less obvious, mortality and reproductive effects. Ideally we would use an historical model of salmon  79  Table 4.3. Sport Catch and Effort, and Components of Sockeye Stock, Alberni Inlet, 1956-94 Sockeye  Total Angler Days  1  Chinook Sport 2  1956-60 1961-65 1966-70 1971-75 1976-80 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994  1  2  3  4  5  Escapement Commercial  Native  Sport  3  2  4  1,500 2,000 3,400 6,700 9,100  42,600 35,000 84,000 149,800 234,000  9,000 15,600 38,300 163,500 698,500  2,200 2,300 3,400 5,900 10,000  0 0 0 0 na  na na na na na  11,200 17,100 28,000 28,000 11,400 9,100 11,800 11,500 17,300 19,500 43,400 8,900 10,200 14,200  324,400 384,800 655,300 214,000 258,500 318,600 438,500 443,500 407,900 285,200 645,300 402,400 420,800  1,009,400 475,900 919,300 919,000 440,000 44,100 209,600 293,000 4,100 11,000 1,042,700 412,400 568,700  17,000 23,500 30,000 21,000 16,000 12,800 23,400 21,300 23,400 10,500 36,500 53,600 58,000  na na na na 1,700 .2 18,200 300 100 24,800 78,600 101,400 101,500 30,300  na na na 155,700 90,500 62,800 72,800 50,400 67,600 90,800 107,800 132,500 99,000 84,800  s  s  Not shown are DFO catches for purposes of monitoring the stocks, and escapements to Henderson Lake (which averaged 5-10% of total sockeye escapements in 1971-94). Sport fishing data are based on creel surveys which have not been conducted during the same period every year. Thus there is some inconsistency in the data. Most chinook are caught in Alberni Inlet in August and September while June-August are the most important months for sockeye. 1987 and 1991-1994 provide the most complete coverage as the surveys ran from mid-June to the end of September. Other surveys always included August and September and thus are relatively good estimates of total chinook catch. After 1990, native catches are for all of Alberni Inlet; in 1990 and earlier years data refer to Somass River catches only. 1976-80 is an estimate. Data refer to days spent angling for all species. Creel survey data on boating trips are converted to angling day estimates using a standard conversion factor of 2.5. 1990 sport sockeye catch includes an estimated 15,000 fish caught prior to the creel survey in that year. Angling days in 1990 are adjusted upward to account for mid-June to midJuly activity by assuming a CPUE of 1.5, a rate fairly typical for this part of the summer.  Source: based on data from Knapp and Cairns [1978] for chinook, 1956-75 and sockeye, 195670; Starr et al. [1984] for sockeye 1971-80; and W. Luedke and T. Gjernes, DFO Nanaimo, for other years.  80 population dynamics which accounts for environmental conditions, water quality and climate. With such a model, the impact on stocks of water quality changes could be estimated over time. However, no such model currently exists for Alberni Inlet salmOn or similar stocks . An alternative is to retrospectively analyze stock 5  data. Current stocks represent the combined influences of natural production, stock enhancement programs and effluent impairment of the habitat. If the contribution of enhancement can be determined then it can be subtracted from observed production and the residual examined for trends.  For chinook, estimates of the  enhanced component of the stocks are available [Sue Lehman, DFO Vancouver, pers. comm., 7 Dec. 1994], but early chinook natural production data (pre-enhancement) are thought to be considerably underestimated, although by an unknown amount [Sue DiNovo, DFO Nanaimo, pers. comm., 25 Jan. 1995]. This means that no firm conclusions about chronic effects on chinook can be made. With respect to sockeye, very substantial natural stock variations make a retrospective analysis of stock data meaningless [K. Hyatt, DFO Nanaimo, pers. comm., 26 Jan. 1995]. In any case, the lack of technical knowledge among sport anglers about how to catch sockeye implies that any reductions in the sockeye stock in earlier years would have had little or no economic significance for anglers, although reductions would have had significance for other user groups.  For the purposes of damage valuation, then, only the 1990 sockeye mortality event can be considered. While chronic effects and previous severe mortality events are likely, the magnitude of these effects can not be determined. Thus we wish to value the loss of200,000 sockeye salmon (100,000 in 1990 and 100,000 in later years) in the context of sport fishing. Table 4.4 summarizes the various factors which must be considered in the valuation, along with a range of assumptions corresponding to low, intermediate and high value estimates. The assumptions are described in more detail below.  First, any economic evaluation of the effect of the lost stocks must start with a consideration of what could have been expected in 1990 without severe reductions in lower layer D O in upper Alberni Inlet. Pooling of  D F O is currently part-sponsor of a proposal to MacMillan Bloedel to create a sockeye production dynamics model incorporating water quality, climatic conditions and bioenergetics. It would be used to assess sockeye mortality risks on an on-going basis [K. Hyatt, DFO Nanaimo, pers. comm., 9 Dec. 1994]. 5  81  Table 4.4. Factors, Alternative Assumptions and Estimated Economic Cost to Sport Fisheries of 1990 Sockeye Mortality in Alberni Inlet Assumption Low  Intermediate  High  1. Additional fish if no effluent 1990 1994-95  100,000 100,000  100,000 100,000  100,000 100,000  2. Fish available to user groups 1990 1994-95  35,000 65,000  35,000 65,000  35,000 65,000  3. Fish caught by anglers 1990 1994-95  24,100 7,500  24,100 7,500  24,100 7,500  0.46  0.60  0.79  5. CPUE  1.9  1.5  1.1  6. WTP value for a sockeye fishing day (1994$)  40  60  80  7. Marginal WTP for additional sockeye (1994$)  4  6  8  306  758  1,816  58  114  200  Factors  4. Proportion of B.C. residents in fishery  Estimated Econmomic Cost (1994$ thousands) 1. Using WTP for a fishing day 2. Using marginal WTP for an additional fish  the sockeye during July-September in upper Alberni Inlet is a natural occurrence but in the absence of effluent impacts the amount of high quality habitat in the pooling area would be much greater, although it is possible that density effects (due to enhancement which creates large returns) would result in some stress, such as increased incidence of disease. I assume that such effects would be small so that the full 100,000 fish would have been available to user groups or escapement in 1990. attributed to effluent.  In other words, I assume that all losses can be  With respect to the lost 100,000 adult equivalents, I assume for simplicity that all of  82 these fish would return in 1994 and 1995. The number of fish which would actually be available depends on what proportion of the lost 1990 fish would have escaped to spawn.  Second, the actions of DFO managers must be considered. Obviously, DFO managers cannot mechanistically allocate stocks ex ante: management decisions occur on a week-by-week basis during the migration period as the stock size is assessed. However, management priorities give some guidance in determining how extra fish would be distributed. DFO's first priority is to achieve escapement objectives to ensure adequate returns in later years. Escapement objectives for Somass River sockeye in recent years, including 1990, have been 350,000 fish [DFO various years].  Ensuring fish for the First Nations food fishery is the next priority,  followed by meeting sport fishery needs. Finally, DFO managers allocate any remaining fish to commercial fishers in Barkley Sound and lower Alberni Inlet. In 1990, DFO managers forecast low total returns based on low escapements in the brood years of origin (1985, 1986) and low marine survival rates. Thus a complete closure of the Barkley Sound commercial net fishery was put into place — the catch numbered only 11,000 fish, well below typical catches (see Table 4.3).  Table 4.3 shows that 1990 Somass River sockeye escapement was 285,200 fish, approximately 65,000 less than the escapement objective. This implies that an additional 35,000 fish would have been available to user groups in 1990, once the escapement objectives had been met. For 1994-95, the key question concerns the size of the total run in these years. The brood years of 1989 and 1991 (moderate to high escapements) will affect the returns as will marine survival conditions. These have been quite poor so that 1995 is expected to have very low returns [K. Hyatt, D F O Nanaimo, pers. comm., 7 Dec. 1994].  However, I assume that  escapement objectives will be met and that an additional 65,000 fish will be allocated among the three user groups (ie. I assume that 65,000 of the lost 1990 fish would have escaped to spawn, yielding 65,000 fish in later years).  Third, the proportion of the extra fish available to anglers must be estimated. First Nations catches from the  83 Somass River averaged 20,900 fish in 1981-1989 while actual catch was 10,000 in 1990. This suggests that about 10,900 extra fish might have been caught by these fishers, leaving 24,100 fish. It is unlikely that DFO would have allowed any opening of the commercial sockeye net fisheries — in 1989, a closure occurred when the total run amounted to 436,000 fish, similar to the total run (before mortality) in 1990. that an additional 24,100 fish would have been available to anglers in 1990.  Thus, I assume  In 1994-95,1 assume that an  estimated 7,500 of the additional 65,000 available fish would have been available to anglers, based on the 1991-93 experience (11.5% of fish caught by the three user groups were caught by anglers).  Extra production can be assumed to be targeted by anglers only if interest in any additional fish exists. Chestnut [1993] noted a substantial growth in the number of tidewater angling licenses sold from the Alberni Inlet/Barkley Sound area over the past decade. He suggested several trends which explain such growth and imply that tidewater angling activity will increase in B . C . in general, and the Port Alberni area in particular. These include a rapidly expanding population in the Lower Mainland and Vancouver Island, the loss of fishing opportunities in the United States Pacific Northwest, and increases in leisure time. Specific to Port Alberni is its high and international profile as one of the major sport salmon fishing areas in southern B . C . Evidence during 1990 also suggests that a greater catch could have been made in that year. During much of the season the bag limit was lowered from the standard 4/day to 2/day, and the lower Somass River and 2 km out into Alberni Inlet were closed from mid-September to mid-November to protect the stocks [DFO various years]. Thus I assume that angler interest is sufficient to ensure that the estimated increases in available sockeye would be taken. The late summer of 1994 was relatively cool and wet at Port Alberni so that no pooling occurred ~ the sockeye moved quickly through the system and the sport catch was much lower than in the previous three years. However, I assume that the few thousand extra available fish in that year would have been caught.  Fourth, the extent of participation in the fishery by non-B.C. residents must be examined.  The value to  British Columbia of sport angling by non-residents will only be lost if those anglers do not come to the  84 province because of a reduction in fishing opportunity. This is not likely to be the case at Port Alberni since the quality of fishing has remained high, despite habitat degradation. I assume that non-resident anglers would have come in any case and that the province did not loose the benefit derived from them as a result of the 1990 mortality. A survey of chinook and sockeye anglers in Alberni Inlet and Barkley Sound in July and August 1993 found that B . C . residents and U.S. residents each accounted for 46% of the total (there were very few non-Canadians from countries other than the U.S.), with other Canadians accounting for the remaining 8% [Chestnut 1993]. In contrast, a 1990 survey of recreational tidewater angling in Canada found that B . C . residents comprised 79% of the province's tidewater anglers.  Similarly, a 1984 survey of over  4,000 tidewater anglers at major southern B . C . fishing locations, including Port Alberni, found that 73 % were from B . C . [Cameron and James 1987]. The proportion of B . C . residents in the Alberni Inlet sockeye fishery may be closer to these higher values since residents are relatively less likely to fish in Barkley Sound than in Alberni Inlet and, conversely, non-residents are more likely to fish in Barkley Sound than in Alberni Inlet. I assume that an intermediate value of 60% represents the proportion of B . C . anglers in the Alberni Inlet sockeye fishery. I also assume that catch rates of residents and non-residents are equal. This means that 40% of any additional catch (the share of non-residents) does not represent an increase in the province's welfare in terms of sport fishing. I use 46% and 79% as low and high estimates of the proportion of B . C . residents in the Alberni Inlet sockeye fishery.  Fifth, increased catch could occur in two ways. Catch is the product of fishing effort (angler days) times catch per unit effort (CPUE). Either effort or CPUE could increase (some combination of the two is most likely). Bag limits represent an upper bound to increases in CPUE.  Table 4.3 shows total fishing effort by anglers  fishing for all species of fish in Alberni Inlet — these data are of little use in determining typical C P U E rates for sockeye in the Inlet . Moreover, there are systematic differences in catch rates between locations so that 6  The creel survey data for Alberni Inlet report boating trips. I follow the standard approach in assuming that each boating trip corresponds to 2.5 angling-days (ie. one boating trip occupies one day, and an average of 2.5 anglers go on each trip) [T. Gjernes, DFO Nanaimo, pers. comm., 27 Jan. 1995]. Cameron and James [1987] found an average of 2.5 anglers per boat in their 1984 survey of angling at southern B . C . locations, including Port Alberni. 6  85 results from other locations are not likely to be applicable. However, during the mid-June to early August period almost all sport fishing in Alberni Inlet is for sockeye and I assume that the corresponding catch rates represent rates that apply to the sockeye sport fishery. For 1991-93, C P U E ranged from 1.6 to 1.9 sockeye per angler per day during the period. Pooling was evident in 1992 and 1993 and clearly raised the CPUE. In addition, at the height of the fishing season, anglers sometimes reach their bag limit of 4/d [L. Gordon, DFO Port Alberni, pers. comm., 16 Feb. 1995]. However, when weather conditions are such that little or no pooling occurs then the C P U E can be much lower ~ in 1994 it was 1.1.  Thus a C P U E range of 1.1 to  1.9 seems likely, with 1.5 representing an intermediate value.  Sixth, the appropriate type of economic surplus measure must be chosen.  As discussed in Chapter 3, the  choice of WTP and W T A theoretically should rest on an assessment of entitlements. I assume B . C . residents have an entitlement to sustainable fishing in natural conditions (no enhancement or pollution). Thus WTP for increased sport angling is the appropriate measure since enhancement programs have permitted fishing levels to greatly exceed levels supportable by natural conditions. As discussed in Section 3.4, the motivation behind sport fishing (and which justifies the use of angling-days as a unit of value) lies not just in the catch but also in the variety of other experiential benefits of the activity. In situations of large readily accessible pools of fish, motivation may be more closely related to food gathering for some people. In his study of the 1993 Port Alberni/Barkley Sound sport fishery, Chestnut [1993] found evidence that at least some local anglers displayed behaviour which suggested this motivation (extended fishing visits of 30 days or more, very large takes). However, as there is no means of quantifying the implications of this, I assume that WTP for angling-days represents the appropriate measure of sport fishing value in Alberni Inlet.  This analysis has not undertaken original estimates of the value of angling in Port Alberni so that WTP values identified in other studies must be used. Angling-day estimates for several studies of relevance are shown in Table 4.5.  These values have been converted to Canadian dollars where appropriate and inflated to 1994  dollars using the recreational spending component of the Canadian Consumer Price Index. This assumes that  86  Table 4.5. Estimates of WTP Values Relevant to Alberni Inlet Sport Salmon Fishing Original Study Source of Estimate and Type of Fisheries  Value  1994 Cdn$ Estimate  1976 $Cdn WTP  $15 25  $38 64  1977 US$ WTP  $26  $67  1984 $Cdn WTP  $44 25 43 50 66  $66 38 64 75 100  1984 $Cdn WTP  $6  $9  1987 US$ WTP  $32^6  $56-133  Description  Masse and Petersen [1977] Angling day value B.C. general marine salmon B.C. specialized marine salmon  Sorhus [1981] Angling day value Oregon marine salmon  Cameron and James [1987] Angling day value Southern B.C. marine salmon Victoria Campbell River Sechelt Port Alberni An additional fish Southern B.C. marine chinook  Walsh et al. [1992] Angling day value Anadramous fishing in U.S.  Masse and Petersen [1977]  based their estimates on a review of previous estimates of average WTP for salmon fishing days in B.C., Washington and Oregon.  Sorhus [1981]  surveyed Oregon anglers as to their travel and other costs incurred in participating in salmon  fishing.  Cameron and James [1987] surveyed 4,161  anglers at Victoria, Sechelt, Campbell River and Port Alberni during the second half of 1984. Chinook and coho were the preferred species. Results refer to the value of the current day of fishing (the marginal day) and the value of an additional fish caught during the current day. The figures for individual locations are estimates based on information provided in the study.  Walsh et al. [1992] undertook a meta-study of 287 studies in 1968-88 of the value of outdoor recreational activity in the U.S.. Nine studies involved anadramous sport fishing, with a mean WTP value of $54 (1987 US dollars), a range of $17-127 and a 95% confidence interval as shown here. No adjustments were made for differences in the studies.  87 values have grown at the same rate as the general cost of recreational goods and services.  The estimates  shown in Table 4.5 suggest that a salmon angling-day has a value in the range of $40-$80. For actually valuing a change in stock size, the common approach is to use average values ~ angler day value divided by C P U E . For example, for his study of the Salmonid Enhancement Program, Pearse [1994] used a value per sockeye equivalent to $26.5 in 1994 dollars to assess the impacts of the program on recreational sockeye fishing. The value he used for coho was much lower, while that for chinook was much higher. These values were derived from DFO analysis of a variety of studies and were meant to be used for province-wide broad analyses, with the caveat that the values were highly uncertain and would vary from location to location. Using a C P U E of 1.5 implies a corresponding fishing day value of $40, consistent with the range shown in Table 4.5.  It is worthwhile exploring whether Port Alberni sockeye sport fishing values may differ. One estimate of WTP values for angling-days at Port Alberni has been derived. A C V M study by Cameron and James [1987] involved personal surveys of 4,161 anglers in 1984 at Port Alberni, Sechelt, Victoria and Campbell River. They found that WTP (after expenditures) for the current day of fishing across all locations was $66 (in 1994 dollars); Port Alberni anglers were found to have a WTP of roughly $100 , or 50% higher. 7  The chief  difficulty with these results for the analysis here is that they refer to a period when sockeye catch was small or non-existent — chinook and coho represented almost all of the salmon catch and fishing for non-salmon species occurred as well.  Nevertheless, the study suggests that Port Alberni fishing may be more highly  valued than fishing at other south coast locations, and I assume that this applies to sockeye today. Thus I assume that Alberni Inlet sockeye has a 50% higher value than the $40 figure consistent with Pearse's study, yielding an angling-day value of $60. As low and high estimates I use $40 and $80, consistent with the range suggested by Table 4.5.  Cameron and James included location dummies in the demand function estimated using their survey results, with Victoria (no dummy) as the base. The estimates here for WTP values at different locations are derived by using the overall mean WTP value of $44 (1984 dollars) , the distribution of anglers across locations and the additional amounts which anglers at Port Alberni ($41), Sechelt ($25) and Campbell River ($19) were willing to pay above the base amount of Victoria anglers. 7  88 Seventh, these angling-day values are an appropriate measure only if all additional fish are caught as a result of additional angling effort (i.e. new angling-days). However, it is likely that increased C P U E would account for at least some, if not most, of the additional fish. There are no directly relevant estimates of the marginal value that increased sockeye catch contributes to the economic value of a fishing day. In their study, Cameron and James found that anglers at B . C . south coast locations had a WTP value for an additional chinook of about $9. As noted in Section 3.4, such estimates generally suggest that a marginal fish has an economic value to anglers that is an order of magnitude lower than fishing day value, although this obviously will vary according to catch success. I assume this relationship holds true for Port Alberni sockeye fishing with catch rates in the 1.1-1.9 range. Thus I assume the marginal WTP for an additional sockeye is $6, with low and high values of $4 and $8.  In summary, I assume the 1990 sockeye sport catch was reduced by 24,100 fish while the catch in 1994-95 will be reduced by 7,500 fish, for a total of 31,600 fish. The foregone catch is valued in 1994 dollar terms. The range of years covered (1990-1995) is so small that current values in 1994 dollars will not differ greatly from present values in 1994 dollars. The intermediate assumptions in Table 4.4 show the economic cost of the reduced sockeye sport fishing opportunities to be $114 thousand under the assumption that increased catch rates would account for all the fish, or $758 thousand if only effort increases. This represents the quantifiable cost to British Columbia of the impact of mill effluent in 1990 on the Port Alberni sockeye sport fishery. Varying the assumptions suggests the cost lies in the $58-$l,186 thousand range. Since the 1990 event was an extreme case, this range represents the upper limit to the annual cost of any other significant mortality events, chronic mortality or reduced reproductive success as they affect sockeye. Chinook move through the system later in the year during (usually) cooler and wetter conditions, so that they are less likely to encounter habitat degradation inimical to their survival or reproduction. Thus economic damages to sport chinook fishing would likely be much lower than those to sockeye sport fishing.  89  4.3. Summary  A history of BOD, TSS and colour discharges by the Port Alberni pulp mill have resulted in very significant DO reductions in upper Alberni Inlet. Important chinook and sockeye stocks pass Port Alberni during their annual spawning migration to the Somass River system. Drought conditions cause access to the Somass River to be obstructed so that the salmon pool in the upper inlet in conditions of impaired water quality due to effluent. The chronic effects on salmon stocks cannot be determined but one major destructive incident caused mortality of 100,000 sockeye in 1990, with a loss of an additional 100,000 adult equivalents in later years as a result of the 1990 loss. Estimation of the economic cost of these losses to British Columbia sport fishing must take into account the management regime of DFO and the proportion of non-B.C. residents in the sport fishery. The use of WTP estimates from other studies is a problematic exercise but some relevant estimates suggest that the lost sockeye sport fishing opportunities cost $58-$l ,816 thousand 1994 dollars. I discuss this case study further in Chapter 8.  90  5. DIOXIN CONTAMINATION AND SHELLFISHERY CLOSURES - HOWE SOUND  5.1. Background  Howe Sound covers an area of approximately 285 km on the southern mainland coast of British Columbia 2  north of Vancouver (see Figure 5.1).  A shallow sill just north of Anvil Island partially isolates inner Howe  Sound, creating an estuarine system strongly influenced by the Squamish River which flows into the head of the Sound. The southern portion is a typical sound, with a broad body of water and several islands. The Sound watershed is home to a variety of communities (including Horseshoe Bay, Squamish, Whistler and Gibsons) as well as port facilities, two chemical plants, mining and forest-related activities, all of which affect the environment [Environment and Land Use Secretariat 1980, Harding 1992].  The pulp and paper mills at Port Mellon and Woodfibre are by far the largest sources of effluent to Howe Sound. The Howe Sound Pulp and Paper Co. mill at Port Mellon is equally owned by Oji Paper Co. of Japan and Canfor Corporation of Vancouver, with current production capacity of 1,000 ADt/d of market bleached kraft softwood pulp and 585 ADt/d C T M P newsprint [Pulp and Paper Canada 1994]. The mill at Woodfibre is owned by Western Pulp Limited Partnership, which in turn is owned by Doman Industries.  Current  production capacity stands at 700 ADt/d of market bleached kraft softwood pulp [Pulp and Paper Canada 1994].  The Port Mellon mill completed a considerable expansion and modernization program in 1992,  including the installation of a secondary treatment system in 1990, the first full system on the coast. At Woodfibre, a secondary treatment system was put into operation in 1992.  Howe Sound is home to various finfish and shellfish stocks which have supported recreational and commercial fisheries [Environment and Land Use Secretariat 1980, Harding 1992]. These fisheries are managed by DFO as part of Management Area 28 which includes Howe Sound, Burrard Inlet and Indian Arm.  Table 5.1  summarizes Area 28 commercial catches of crab, shrimp and prawn in 1981-94. Shrimp fishing usually  Source: Nelson [1979].  92 occurs all year in Area 28 while prawn and crab fishing occur in-season only. Total British Columbia catches are shown in Table 5.1 for comparison. Mercury contamination related to a chemical plant at Squamish led to a closure between 1970 and 1978 of inner Howe Sound to commercial and sport fishing for crabs, shrimps, prawns and all fish except salmon, trout and herring. A large commercial shrimp fishery existed before the 1970 closures. In the late 1980s, fairly small commercial crab, shrimp and prawn fisheries existed, primarily in outer Howe Sound. Recreational shellfishing also occurred but there was virtually no native shellfishing [B.Ionson, D F O Squamish, pers. comm., 16 Jan. 1995]. Recreational crab fishing was fairly limited with some occurring throughout the Sound but mainly in sub-areas 28-3, 28-4 and 28-5 (see Figure 5.1). Recreational prawn and shrimp fishing was somewhat more active, most commonly in sub-areas 28-4 and 285.  Only a few people fished regularly while others fished opportunistically, and angling for shellfish may  often have occurred in conjunction with other activities (boating, finfishing).  Both pulp mills use chlorine-based bleaching and consequently have discharged a variety of chlorinated organics, including chlorinated dioxins and furans, leading to contamination of Howe Sound shellfish. The characteristics of chlorinated dioxins and furans which have made them of particular concern were summarized in Section 2.3. Concerns with respect to British Columbia coastal pulp mills first arose in January 1988 when Greenpeace Canada issued a press release advising Canadians not to eat fish caught near these mills. Greenpeace based its warning on analysis of a sediment sample taken near the Harmac mill which showed a high chlorinated dioxin contamination. As a result, DFO accelerated its fish and sediment sampling program begun in 1987: as part of this program, fish and shellfish tissues were taken from the vicinity of 8 British Columbia pulp mills and sediment samples were extracted at 17 sites near bleached kraft pulp mills or sawmills which had recently used pentachlorophenol for sapstain protection of lumber. Preliminary results were released in May 1988 [DFO 1988a] and led to a priority focus on prawns and crabs near the Port Mellon and Woodfibre mills and on crabs near the Prince Rupert mill. It also led to an expansion of the sampling program to all 47 Canadian pulp mills using chlorine bleaching.  93  p Cu  + O » H « t-; tN VO tN 0 0  t y »  o s oa  00 rtbSinioNnpiMvStc VO m o g ^ - H - H n v o v n v n - H O tN  rl  r~ 00 O 00 -H  8  vn ro vo vo ro vn ro ro vo i n  T f ^ H r o o o e j s T t r o v n v n  H N ^ v i o e J o j i n v o OH  </»  3§ 5  r~ r~ oo' oo'  ON  t-~ vo'  ON  o\  r - O v O N M < t e \ + v O N t N H ; f )  U  IS  vn T f o o o o , ,, n N n "m iOn -m vI OO f hj o oMr -vOg «N vo l ' o lr  "  U  o  vn<NvoTt-ro>nr~(NtN pvo-Hr^ovmov-H^t ri-HcN-H^r^rtt>ics'  | CO  i n T t oo o —< vn m m M oo vo r-  u 3 g  cN • * m —• r- r-. in in m  ovtHaomn\to TJ- m ~ H oo oo( — - H « - H —I © t - VS O4 v O O \ * v D c o N v r i • * ^ v 9 ft: r - w vn o l * N vo, c | eo tN fN tN - H ro o f ro tN  vo  -HoiroTj-mvor^oooNO—H OOOOOOOOOOOOOOOOOOONON ONONONOvONONONONONONON  OV ON  W  •13 OH  i Q  a  94 At the end of November 1988, DFO announced the results of an analysis of fish and shellfish samples from the three priority mills, and the closure of certain prawn, shrimp and crab fishing grounds to all fishing (i.e. sport, commercial and First Nations food fisheries) [DFO 1988b]. These were the first in a serious of closure and consumption advisory announcements that have continued until the present (see Appendix A). As a result of the 1988 announcements, an area near the Prince Rupert mill was closed to all crab fishing and the northern and western portions of Howe Sound (DFO management sub-areas 28-3,28-4 and 28-5 — see Figure 5.1) were closed to all crab, shrimp and prawn fishing. The greatest effect was expected to be on prawn fishers — 14 boats were licensed to fish for prawn in the Sound. Half were capable of fishing outside the Sound while smaller day-vessels found themselves restricted to the areas still open [DFO 1988b]. The closed areas were not important fishing grounds for shrimp and none of Howe Sound was a particularly important area for crab. Thus relatively small effects were predicted for shrimp trawlers (12 full-time vessels operated in all of Area 28) and crab vessels (3-6 vessels operated in Area 28).  The closures were based on a health assessment which considered the observed levels of contamination (in the ppt to ppb range), the total toxicity of the various congeners observed, and the possibility of adverse health effects if the shellfish were eaten regularly. The greatest concentrations were found in crab hepatopancreas, a digestive organ that is not usually eaten, but is consumed by some people [Nassichuk 1992]. Crabs have relatively poor ability to metabolize chlorinated dioxins and furans [Nassichuk 1992], and tend to remain for long periods in relatively small areas on the sea bed [DFO 1988b]. Similarly, prawn and shrimp tend to remain within small geographic areas for long periods. Thus these shellfish species are more likely than most finfish to display high contaminant levels. It should be noted that commercial fisheries are treated somewhat differently than other fisheries  [Nassichuk 1992].  Where a commercial fishery exists,  site-specific  consumption advisories are of little use since consumers have no way of knowing where purchased fish come from. Thus DFO enacts complete closures to reassure consumers of the safety of fish. Where sport or native food fisheries exist, then a consumption advisory can be effective and sufficient since individuals can monitor their fishing locations and their consumption so as to ensure that consumption does not exceed advised  95 maximum intakes.  In mid-June 1989, D F O closed the western portion of sub-area 28-1 (north and west of Keats Island ~ see Figure 5.1) in Howe Sound to all crab, shrimp and prawn fishing as a result of analyses of new samples [DFO 1989]. In addition, prohibition of commercial crab fishing was expanded to all of Howe Sound (i.e. sub-area 28-2 and the remainder of sub-area 28-1 were closed) and an advisory to sport and native food fishers to avoid consumption of crab hepatopancreas was issued for all of the Sound. Five crab vessels, four prawn vessels and up to 14 shrimp vessels were affected by the closures [DFO 1989].  In November 1989, fishing closures and consumption restrictions for various shellfish species were issued for seven more coastal British Columbia sites and the closures at Prince Rupert were also expanded [Government of Canada 1989]. Not all of the sites were in the vicinity of chlorine-using pulp mills.  A year later, in  November 1990, and again in August 1991, these sites were affected by additional consumption advisories and by expansion of area closures.  In May 1992, the Federal government released a new pulp and paper  regulatory package [Canada Gazette, Part II, 126(11), 1940-2006, 7 May 1992].  This package included  regulations under CEPA which required mills using chlorine compounds to eliminate measured concentrations of 2,3,7,8-TCDD and 2,3,7,8-TCDF by 1 January 1994 at the latest and to eliminate the use of defoamers containing recycled oil and wood chips treated with pentachlorophenol.  In August 1993, closure expansions and new consumptions advisories affected 7 of the areas already subject to government announcements [Government of Canada 1993]. In Howe Sound, the existing commercial crab closure was expanded slightly south into a portion of DFO Management Area 29 (Strait of Georgia off Vancouver and Howe Sound). The consumption advisory for crab hepatopancreas was expanded into these areas as well.  In February 1995, the effects of mill reductions in chlorinated dioxin discharges became evident in the  96 removal of closures and advisories in some areas, actions taken on the basis of 1993 tissue samples. However, expansions in other areas occurred. In Howe Sound, the Federal government rescinded the shrimp and prawn commercial fishing closures and the closure of non-commercial crab fishing in sub-areas 28-4 and 28-5 was replaced by a hepatopancreas consumption advisory. Commercial crab fishing remained closed and the portion of Area 29 closed was expanded about 20 km along the coast north of Howe Sound.  In summary, in late 1988 the northern and western portions of Howe Sound were closed to all crab, shrimp and prawn fishing, the first in a series of actions taken at 10 coastal sites. Since mid-1989 all of Howe Sound has been closed to commercial crabbing. Sub-areas 28-3, 28-4, 28-5 and part of 28-1 were closed to all shrimp and prawn fishing and to sport and native crab fishing from late 1988 to early 1995.  In 1993 and  again in 1995, the commercial crab closure was expanded slightly into areas outside Howe Sound. The series of closure and advisory expansions probably does not represent an increase in contaminated area, but rather a further delineation and discovery of contaminated areas. The Howe Sound pulp mills now use a high degree of chlorine dioxide substitution and have adjusted the use of defoamers and wood chips so as to comply with the May 1992 Federal regulations governing discharges of 2,3,7,8-TCDD and 2,3,7,8-TCDF. However, while chlorinated dioxins are now virtually non-detectable in effluent from the mills, past deposition will survive for years or decades in the benthic environment and, if it remains biologically available, will continue to contaminate aquatic organisms for a considerable time [Nassichuk 1992]. Thus the closures and advisories in Howe Sound may not be fully rescinded for several years.  5.2. Economic Valuation  Water quality reductions reduce the use and non-use values associated with the Howe Sound. These impacts are seen most concretely in chlorinated dioxin-related closures and advisories to protect against increased cancer risk in humans due to consumption of contaminated shellfish. The economic cost assessed here relates  97 to the commercial extinction of shellfish in part of Howe Sound (shrimp, prawns) or all of the Sound (crab). It has proven impossible to assess the economic cost associated with reduced recreational shellfishing opportunities due to the closures.  Recreational shellfishing is not regulated by licenses (although there are  trap limits) so that interest in the fishery is difficult to monitor. Surveys of recreational shellfishing activity have never been undertaken in Howe Sound or at other similar locations.  Thus the only data available are  anecdotal. For example, according to a local fisheries officer, 6 Squamish area people who angled regularly (4+ times a month) for shrimps and prawns stopped after the 1988 closure [B.Ionson, DFO Nanaimo, pers. comm., 16 Jan. 1995]. Typical recreational catches are unknown. In general, a comparison of popular shellfishing areas with the closure areas suggests that the primary recreational locations have not been available since late 1988.  However, some angling still goes on in other locations.  Overall, then,  quantification of activity in terms of catch or effort is impossible. Thus while anecdotal evidence suggests loss in social welfare due to a reduction in recreational shellfishing opportunities, only the losses in the context of the commercial fishery can be evaluated here.  Table 5.2 summarizes the various factors that must be considered in analyzing the economic cost to British Columbia of the Howe Sound commercial shellfishing closures. A range of assumptions corresponding to low, intermediate and high estimates is shown. The various factors and assumptions are discussed below.  First, potential Area 28 catches in the absence of closures must be assessed. A comparison of commercial catches in 1981-88 and 1989-94 gives no clear indication that the Howe Sound closures have affected Area 28 catches for crab or shrimp, although the prawn catch appears to have fallen, a trend not observed in total B . C . prawn catches.  I assume that average 1981-88 catches are representative of the catches which would  have occurred from Area 28 in the absence of contamination. Thus I assume that annual potential Area 28 catches are 58.9 t shrimp, 25.1 t prawn and 19.0 t crab. I assume that these catches could have been made in each year after 1988.  Natural conditions mean that the catch could vary considerably from year to year  but I assume that these amounts are reasonable averages.  98  Table 5.2. Factors, Alternative Assumptions and Estimated Economic Cost to Commercial Shellfishers of Chlorinated Dioxin-Related Closures, Howe Sound Assumption Low  Intermediate  High  58.9 25.1 19.0  58.9 25.1 19.0  58.9 25.1 19.0  25 65 5  25 65 12  25 65 20  Economic costs? Fishers Processors Wholesalers Retailers Consumers  Yes No No No No  Yes No No No No  Yes No No No No  Lost producer surplus due to: Increased costs Foregone potential catch  Yes No  Yes Partial  Yes Full  0.10  0.25  0.50  Shrimp (1989-94) Prawn (1989-94) Crab (1989-96)  21 94 3  54 236 18  107 472 59  Total  118  308  638  Factors 1. Annual potential Area 28 catch (t) Shrimp Prawn Crab Proportion of Area 28 catch forgone (%) Shrimp Prawn Crab  Lost producer surplus as proxied by proportion of gross revenue associated with forgone potential catch  Estimated Economic Cost (1994$ thousands)  Second, the extent of the catch reductions in Howe Sound must be estimated. The substantial year-to-year variability in catch statistics mean that they provide little help in assessing the effects of closures: of greater  99  Table 5.3. Estimated Geographical Distribution of Area 28 Shellfish Catch Prior to Closures' (percent of total Area 28 catch) Area  Shrimp  Prawn  28-1 28-2 28-3 to 28-5 Total Howe Sound  60 5 10 75  20 5 60 85  na na na 5-20  Other  25  15  80-95  Total Area 28  100  100  100  1  Crab  Based on DFO estimates made at the time of the December 1988 and June 1989 closures, supplemented by data from the prawn and shrimp log-book programs. Note that these two sources display considerable inconsistency — thus estimates must be taken only as a very approximate.  Source: Estimates based on data from J. Boutillier and C. Wallace, Pacific Biological Station, DFO Nanaimo, and DFO [1988b, 1989].  relevance is the distribution of the catch across the area. The data shown in Table 5.1 are official catch statistics for Area 28 based on sales slips recorded at the time of the dockside sale. Data for individual subareas of Howe Sound are not recorded in these official statistics. However, the division of the commercial catch between sub-areas can be approximated using data from a prawn and shrimp fishing log-book program maintained by D F O , combined with estimates provided by DFO at the time of the closure announcements. The distribution of the pre-closure harvests shown in Table 5.3 is estimated using these sources. The salesslip and log-book data are often inconsistent, reflecting the fact that 1) the log-book data record catch location and a rough estimate of the weight at the time of catch; and 2) sales are reported for the management area in which the catch is landed, not where it is caught [C. Wallace, DFO Pacific Biological Station, Nanaimo, pers. comm., 17 Feb. 1995]. The log-book data are generally expected to be higher than the sales-slip data for Area 28 suggesting that the catch data shown in Table 5.1 underestimate the true catch, but by an unknown  100 amount. Moreover, there are considerable inconsistencies between the log-book data and estimates of the catch distribution made at the time of the closures.  These difficulties mean that the distributions shown in  Table 5.3 for shrimp and prawn must be considered quite rough. No log-book data exist for commercial crab fishing. However, I assume that 5-20% of Area 28 catch was taken from the Sound, a range regarded as reasonable by B. Ionson, a DFO official in Squamish [pers. comm., 16 Jan. 1995].  In 1988, the fishing grounds were closed only for the last month of the year and I assume that the foregone catch was relatively minimal, although fishing activity varies considerably by month for each species.  I  assume that the relative productivity of the various sub-areas in Area 28, and the effort by fishers in each, would have remained the same in the post-1988 period as in the earlier period. Thus I assume that the catch distribution prior to 1989 is representative of what could have occurred in later years (see Table 5.3). For shrimp, I assume that sub-areas 28-3 to 28-5 accounted for about 10% of the Area 28 harvest. In mid-June 1989, approximately 25% of sub area 28-1 was closed and I assume that the same proportion of the shrimp catch was lost in that year and in following years from the sub-area. For prawn, sub-areas 28-3 to 28-5 accounted for about 60% of the pre-closure Area 28 harvest. I assume that 25% of the 28-1 harvest was lost in 1989 and later years. Finally, I assume that all of Howe Sound was lost to commercial crabbing from the beginning of 1989.  I ignore the closures of the small areas outside Area 28. Together these assumptions  imply that 25% of shrimp and 65% of prawn pre-1989 catches from Area 28 can be assumed to have been lost in each year as a result of the Howe Sound closures between 1989 and 1994. For crab, I assume a 5-20% loss (with an intermediate value of 12%) in each year between 1989 and 1996 — I assume the commercial crab closure in Howe Sound will be rescinded in 1997.  Third, the economic cost of foregone harvests should be considered in terms of the lost producer and consumer surplus of the fishers and others in the shellfish market. Howe Sound and Area 28, even when combined with other closure areas, represented only a relatively small portion of the total supply of crab, shrimp and prawn to British Columbia markets (see Table 5.1).  No wholesalers or processors relied  101 predominantly on the Howe Sound catch [B. Ionson, DFO Squamish, pers. comm., 16 Jan. 1995]. Thus I assume that the closure of Howe Sound has had no effect on market conditions or on market prices, and that wholesalers, processors, retailers and consumers have not been affected.  I assume that only commercial  harvesters have been adversely affected. There might be exceptions within the communities of Howe Sound itself.  For example, a seafood store opened in Squamish in late 1988 with the intention of selling local  harvests (at least in part) ~ it closed shortly after the first closure announcement, perhaps a casualty of the contamination [B. Ionson, DFO Squamish, pers. comm., 27 June 1994]. Local consumers might have lost the benefit of local shellfish sold at a cheaper price.  Fourth, the extent to which the potential lost harvests actually create losses in the welfare of commercial fishers will reflect the extent to which substitute locations or fisheries have been utilized by the effected fishers. The structure of the commercial shellfishery in Howe Sound prior to the closures was relatively simple. Only wild stocks were harvested (there was no aquaculture) and area licenses were not used in the Sound or in adjacent areas.  Switching fisheries was likely not an option for at least some boats since  specialized equipment and licenses could make the transition prohibitively expensive. For example, one crab fisherman noted that he could not afford to switch to salmonfishingas the license was $80,000 and equipment would cost an additional $20,000 [Vancouver Sun, 16 June 1989, pp.Al,A14].  Catch statistics prior to 1989 for Area 29 show crab catches 10-20 times those of Area 28, shrimp catches roughly comparable to those of Area 28 and prawn catches somewhat lower. Thus, in terms of stock sizes, relocation of crab vessels outside Area 28 should have posed little problem. Relocation for shrimp vessels and especially for prawn vessels would have been more difficult if catch statistics were indicative of stock size. However, relocation within Area 28 was also possible. I assume that all boats were able to relocate. I also assume that Howe Sound was the optimal economic choice for those who choose to fish there, so that relocation to another area involved greater costs (in fuel, travel time) and/or reduced catch (due to unfamiliarity with a new area, competition with existing fishers for stocks in the new areas) and hence reduced  102 producer surplus. Some of these effects, those related to becoming familiar with new areas, would decline over time.  In addition, to the extent that adjacent areas were already heavily fished or supported small stocks, the addition of fishers who formerly operated in Howe Sound could adversely affect the welfare of all participants. This would certainly occur if fishing in adjacent areas operated at or near the sustainable harvest level, and additional fishing pressure caused stocks to decline, or to decline faster than previously.  With these considerations in mind there are two possible extreme outcomes as a result of the closures. First, the displaced boats moved to other areas and harvested amounts that replaced the catch they would otherwise have caught in the closed areas, and this catch did not reduce the catch that other boats would normally make. In this case, the relocating fishers would loose some amount of producer surplus due to increased costs, under the assumption that Howe Sound was an economically optimal location for them. This possibility places a lower bound on lost producer surplus. Second, at the other extreme is the possibility that all of the producer surplus associated with the foregone catch is lost. This would occur if displaced boats harvested an equivalent amount in adjacent areas but at the expense of other boats. The displaced fishers would continue to suffer a loss of producer surplus from increased costs due to relocation. This possibility places a upper bound on lost producer surplus.  Fifth, it is clear that the evaluation of lost producer surplus associated with the closures requires considerable information and analysis. It means an analysis of fishing costs and relocation possibilities, and perhaps even consideration of the state of the shellfish stocks in the area. Unfortunately, shellfishing cost data are not available so that changes in producer surplus cannot be evaluated. Only the gross revenue associated with lost catches (using dockside prices) can be assessed although, as explained in Chapter 3, revenue and producer surplus changes can move in opposite directions. However, since I assume that shellfish prices are unaffected by the closures, gross revenue clearly falls here and this change places an upper bound on lost producer  103 surplus. Other than this, there is no way of knowing how gross revenue change relates to the lost producer surplus associated with increased fishing costs and/or decreased catch. At best, we know that lost producer surplus is greater than zero but below the landed value of the foregone catch. To gain some insight into the possible range of costs, I assume that producer surplus lost lies in the 10-50% range of forgone gross revenue, with 25% as an intermediate value.  In summary, I assume that chlorinated dioxin contamination resulted in a annual loss of 14.7 t shrimp (25% of 58.9 t) and 16.3 t prawns (65% of 25.1 t) that could have been harvested in the closed areas of Howe Sound between 1989 and 1994. For crabs, I assume the lost harvest is 1.0-3.8 t per year (5-20% of 19.0 t) between 1989 and 1996. Valuing the lost potential catches at the 1994 landed prices shown in Table 5.1, and assuming that the producer surplus loss equals 25% of the gross revenue loss, gives a cost of $308 thousand 1994 dollars over the whole 1989-96 period. The low and high assumptions give a range of $118-$638 thousand. Because the range of years covered is small, current values in 1994 dollars are used. Present value estimates will not be markedly different.  5.3. Summary  The Port Mellon and Woodfibre mills discharge their effluent to Howe Sound. Prior to recent efforts to eliminate chlorinated dioxins and furans, these toxic compounds were discharged and contaminated shellfish. In 1988, concentrations of these substances were found by Health and Welfare Canada to be at levels considered to pose a potential threat to human health. DFO instituted a series of closures to protect human health and consumer confidence in commercially harvested shellfish, starting with Howe Sound. All of Howe Sound has been closed to commercial crab fishing since mid-1989. Shrimp and prawn commercial fishing was restricted from late 1988 to early 1995. determined.  The effect of the closures on recreational fisheries cannot be  For commercial fisheries, estimation of the cost should consider relocation possibilities and  104 increased costs to fish elsewhere but only gross revenue losses associated with the lost potential catch are quantifiable. Using various assumptions suggests the economic cost of the closures in terms of commercial fishing in the  1989-1996 period lies in the range of $118-$638 thousand 1994 dollars. I discuss this case study  further in Chapter 8.  105 6. DIOXIN C O N T A M I N A T I O N A N D C O N S U M P T I O N ADVISORIES - C O L U M B I A R I V E R  6.1. Background  A bleached kraft mill was constructed in 1961 on the Columbia River 6 km upstream of the city of Castlegar in the West Kootenay region of British Columbia (see Figure 6.1). The current owner, Celgar Pulp Company, is a joint venture of CITIC B . C . Inc., Stone Consolidated Inc. and Venepal Canadian Investments Ltd. Production capacity stands at about 1,200 ADt/d of bleached and semi-bleached kraft softwood pulp [Pulp and Paper Canada 1994].  The Columbia River system is one of the most highly regulated watersheds in North America.  A large  number of dams regulate flow on the U.S. side of the river while on the Canadian side the Hugh Keenleyside Dam was built in 1968 about 3 km upstream of the mill to provide flow control for the American dams downstream. Two additional dams were built further upstream in later years, leaving the 55 km stretch of the Columbia between the Keenleyside Dam and the Canada-U.S. border (henceforth referred to as the Lower Columbia) the only readily accessible free-flowing portion of the river in Canada. There are also a series of dams on the Kootenay River, which joins the Columbia at Castlegar, and on the Pend d'Oreille River which joins the Columbia just before the border. On the United States side, the artificially created Lake Roosevelt starts 25 km from the border and stretches 250 km to the Grand Coulee Dam.  The cities of Castlegar and Trail discharge relatively minor amounts of treated municipal effluent to the Lower Columbia and there are a number of other small effluent sources, but the mill at Castlegar and Cominco's lead-zinc smelter and fertilizer plant at Trail are the only major sources of water pollution in the river [Butcher 1992].  The Comincb discharges, about 15 km from the international boundary, result in elevated  concentrations of zinc, lead, copper, arsenic, cadmium and mercury [Serdar et al. 1994]. contaminated sediments and fish in the river.  These have  Source: ARA Consulting [1992b].  107 The Celgar mill has operated for most of its life without significant effluent treatment and since the early 1970s it has had a history of deferment of effluent permit requirements [Derksen and Lashmar 1981a, Butcher 1992].  The current owners purchased the mill in 1986 and in 1989 proposed a major expansion and  modernization program. The Celgar Expansion Review Panel [1991], created by D F O , Environment Canada and B . C . Environment, reviewed the environmental and social effects of the proposal. It set forth over 40 recommendations in its report and advised that the project be approved. It also noted that the mill's historical non-compliance with its effluent permit conditions, and deferment by the British Columbia government of those requirements, were unacceptable and that future enforcement should be rigorous. The old mill was completely replaced with a state-of-the-art mill by mid-1993 — full primary and secondary treatment facilities were included [Mackay and Galloway 1994].  The Lower Columbia is home to 24 species of fish, of which 11 have been classified as sport fish [Birch et al. 1994]. Sampling studies have shown considerable changes in the number and relative abundance of fish in the past 15 years. Among sport fish, mountain whitefish declined from 80% in 1980-84 samples to about 57% in 1990-93, still the most abundant of the sport fish but well down from the earlier period. Burbot have almost disappeared from recent samples while rainbow trout have increased from 2% to 20% and walleye have increased from 0.1 % to 24%. Walleye are not indigenous to the area but were introduced to Lake Roosevelt in the early 1970s and migrate upstream [B.C. Environment 1992]. The growth in the walleye population may account, at least in part, for the decline in relative abundance of a number of other species.  Along with shifts in species availability there have also been significant shifts in angler interest.  Using a  species preference index in which values sum to 10, a 1990/91 survey found that interest concentrated primarily on rainbow trout (6.0) and walleye (2.9) [ARA Consulting 1992b]. Other species of interest were white sturgeon (0.5), whitefish (0.1) and kokanee (0.1). Interest was reflected in catch statistics for 1990/91 -rainbow trout, 47%; walleye, 48%; white sturgeon, 3%; and mountain whitefish, 1%. Creel surveys in the late 1970s and early 1980s implied that interest in rainbow trout was dominant then as well (index value of  108 7.6) but that kokanee (1.0, with highly variable catches) and Dolly Varden (1.2) were important at that time. In the past few years, the sturgeon fishery has become very active [B. Lindsay, B . C . Ministry of Agriculture, Fisheries and Food, Nelson, pers. comm., 15 Jan. 1995]. The Columbia River white sturgeon was listed as Vulnerable in Canada in 1990, and in 1993 the Lower Columbia fishery was designated catch and release only [Hildebrand etal. 1994].  Most of the Lower Columbia is fished actively — one particularly popular area is the stretch between Castlegar and the Keenleyside Dam where the water is relatively slow moving. Both boat and shore angling occurs and both are dependent upon water levels. Levels which are too low make boating dangerous and reduce access points, while levels which are too high can be dangerous due to submerged rocks, and also prevent shore fishing [Mallette and Baker 1994]. In large part, water levels are controlled by the Keenleyside Dam and the dams on the Kootenay River. Fishing effort peaks in June to September but some fishing occurs all year round [Andrusak and Martin 1983, A R A Consulting 1992b], subject to provincial government temporary closures to protect spawning fish.  Aside from the Lower Columbia there are a number of other popular fishing locations in the area including the Arrow Lakes behind the Keenleyside Dam, a reservoir created behind the Brilliant Dam on the Kootenay River near Castlegar, and Kootenay Lake. However, the Lower Columbia fishery is considered unique in a number of respects [Butcher 1992,  A R A Consulting 1992b, Mallette and Baker 1994]. It supports good  walleye and sturgeon fishing. It also is the only remaining large free-flowing river in western North America with a sizeable rainbow trout population [Mallette 1991], and has a reputation as a first-class rainbow trout and walleye river.  The value of the sport fishery in the Lower Columbia has been compromised by consumption advisories related to chlorinated dioxin and mercury contamination. Appendix A shows the chronology of chlorinated dioxin-related advisories at interior B.C. locations. Advisories with respect to contamination of walleye by  109 mercury have existed for some time [Province of British Columbia 1994]. The Cominco operations are the only known major source of mercury in the river while the pulp mill is the only source of chlorinated dioxins above the Grand Coulee Dam in the United States.  In 1988, an Environment Canada study of fish and  sediment taken near ten interior pulp and paper mills first revealed the presence of elevated chlorinated dioxin concentrations [Mah et al. 1989]. As a result, in May 1989 consumption advisories were issued with respect to certain fish species caught in the vicinity of mills at Prince George, Quesnel and Kamloops. At Castlegar, Health and Welfare Canada advised anglers who caught lake whitefish in the Celgar area and 7 km downstream to limit consumption to 40 g/week [Butcher 1992].  In June 1990, the results of further studies led the British Columbia government to advise restriction of consumption of mountain whitefish taken from the Lower Columbia to 205 g/week (one serving).  In July  1992, analysis of 1990/91 monitoring results led the local provincial Medical Health Officer to warn anglers to limit consumption of both mountain and lake whitefish caught anywhere on the Lower Columbia to one meal per week [B.C. Environment 1992]. In December 1994, after reviewing the most recent (1992/93) monitoring results, the Medical Health Officer advised the public that consumption restrictions were still required [Province of British Columbia 1994]. Canadian analyses of fish tissue have found that chlorinated dioxin levels in fish other than whitefish were either zero (non-detectable) or not sufficient to warrant advisories. Both whitefish species are bottom feeders and relatively sedentary compared to species such as rainbow trout. Both factors may explain why levels have been more elevated in whitefish than in other fish.  Meanwhile, the concerns about health risks related to chlorinated dioxins had led to monitoring by the State of Washington Department of Health [1991]. Analysis of fish tissue resulted in a 1990 advisory that children not eat whitefish from Lake Roosevelt.  On further analysis, the advisory was changed in 1991 to a  recommendation that, while anyone could eat fish from the lake, certain precautions should be taken in terms of preparation and cooking of any fish and that consumption should be limited.  110 The installation of new production and bleaching facilities, including 100% C10 substitution, as well as 2  efficient primary and secondary treatment facilities, has reduced chlorinated organic discharges from the Castlegar mill. 2,3,7,8-TCDD and 2,3,7,8-TCDFhave been non-detectable in the mill's effluent since at least September 1993 [Mackay and Galloway 1994]. Contamination of fish appears to have declined since 1990 in both the Lower Columbia and Lake Roosevelt [Province of British Columbia 1994, Serdar et al. 1994]. Given the declining trend, it is likely that the consumption advisories will be removed in the next year or two [B. Lindsay, B . C . Ministry of Agriculture, Fisheries and Food, Nelson, pers. comm., 25 January 1995].  6.2. Economic Valuation  The first consumption advisory occurred in 1989, affecting an insignificant sport species, albeit in a very popular fishing reach of the river. The 1990 advisory affected a somewhat more popular species but it was not until 1992 that the whole of the Lower Columbia was affected by chlorinated dioxin-related advisories. However, while the consumption advisories have been quite specific, detailing the maximum consumption level considered safe, referring only to two species and initially encompassing only a portion of the river, the economic effects likely are much wider.  A consumption advisory, no matter how limited, can create a  generalized perception of poor quality fish and degraded environmental conditions.  From an economic  standpoint a number of effects result. Some anglers may stop recreational fishing activity altogether so that the total level of activity in the region declines. Other anglers may stop fishing on the Lower Columbia and start to fish, or fish more, at other locations in the region. Still others may continue to fish on the Columbia but reduce their activity. Even if they do not reduce their activity the benefit they derive from fishing on the Lower Columbia could be reduced by the knowledge that fish are contaminated.  Valuation of the economic cost of the consumption advisories ideally should consider all of the potential economic impacts. Here I focus only on a measure of reduced activity. Table 6.1 summarizes the various  Ill  Table 6.1. Factors, Alternative Assumptions and Estimated Economic Cost of Reduced Sport Fishing Due to Chlorinated Dioxin-Related Consumption Advisories, Lower Columbia River Assumption Low  Intermediate  High  Factors 1. Baseline 1990/91 Lower Columbia angler days (assumed to reflect a reduction due to pollution concerns)  10,180  10,180  10,180  0.80  0.88  0.95  3. Annual growth rate of angler days (%)  Zero  0.3  0.7  4. Reductions in fishing activity due to chlorinated dioxin-related advisories (%)  10.0  20.0  30.0  5. Average WTP value of a Lower Columbia fishing day (1994 $)  .28  34  40  2. Share of B.C. residents in Lower Columbia angler days  Estimated Economic Cost 1. Reduced fishing days in 1990/91 to 1995/96 due to advisories  5,429  13,540  152  460  2. Cost of reduced fishing days in 1990/91 to 1995/96 (1994 $000)  25,309  1,012  factors considered along with a range of assumptions corresponding to low, intermediate and high cost estimates. The assumptions are described in more detail below.  First, it is necessary to determine the level of fishing activity in the Lower Columbia. The only available creel survey data refer to a few months in 1976, the 1980/81 season (May to April), a few months in 1982 and the 1990/91 season [Andrusak and Martin 1983, A R A Consulting 1992b].  The 1990/91 survey,  conducted for B . C . Hydro, appears methodologically sound and covers the Lower Columbia, the lower  112 Kootenay River and Brilliant Reservoir. The earlier surveys, conducted for the B . C . government, suffered from incomplete angler coverage and assessed activity only in the Lower Columbia between the Keenleyside Dam and Trail. Thus the only reliable data refer to 1990/91. In 1990/91 an estimated 10,180 fishing days (shore and boat fishing) were spent on the Lower Columbia by local residents and anglers from elsewhere in the province or from outside the province. Just over half of the activity was spent in the Castlegar area. I assume that the effects of the chlorinated dioxin-related advisories first had an effect in 1990 (potentially against a background of less specific pollution concerns which affected angling activity). Thus I assume that the 1990/91 data represent a level of activity which was lower than it would otherwise have been. A study by A R A Consulting [1992a] supports this assumption, and is discussed further below.  I use the 1990/91  fishing effort as a base from which to estimate the extent of reduced fishing activity in that season and in later years. I further assume that the chlorinated dioxin-related consumption advisories will continue to the end of the 1995/96 angling season.  Second, the proportion of the total fishing effort accounted for by British Columbia residents must be determined, since, as described in Chapter 3, calculations of economic value must distinguish between residents and non-residents. In its 1990/91 creel survey, A R A Consulting [1992b] found that 80% of anglers in the area were B . C . residents (70% local plus 10% from other parts of the province), 3% were from other provinces and 7% were from the United States. Over time is it likely that the proportion of non-resident anglers has grown as the fishing opportunities have become better known outside the area [Mallette and Baker 1994]. The earlier creel surveys suggested that over 95% of anglers were from the local area [Andrusak and Martin 1983], but this figure may be biased upward by the fact that a relatively higher proportion of U.S. anglers fish in the stretch of the Lower Columbia below Trail, an area not included in these earlier surveys. On the other hand, local anglers do more fishing in a year than do non-residents, implying that their share of fishing days would be higher than 70%. I use the 80% and 95% figures as the low and high estimates of the proportion of fishing days accounted for by B . C . residents, and use 88% as an intermediate value. I assume that most non-residents generally are unaware of the fish consumption advisories in the area so that  113 their activity has not been affected appreciably by the advisories.  Third, using the 1990/91 baseline estimate of fishing activity of 10,180 angling-days, the next step is to estimate what level of fishing activity would occur in later years. Note that because I have assumed that the 1990/91 estimate refers to a level of activity reduced by pollution concerns, these estimates will represent the activity in later years under the same assumption. The 1980/81 creel survey did not estimate fishing days, but fishing effort as measured by angler-hours was marginally higher than the 1990/91 estimate [Andrusak and Martin 1983, A R A Consulting 1992b]. However, the earlier estimate was based on an orf hoc assumption that a 40% under-count had occurred ~ it is difficult to know whether the estimate can be meaningfully compared with the 1990/91 estimate. Moreover, both studies suggest that fishing effort varies greatly from year to year, due at least in part to water levels. Thus the available data give no indication as to what trend might characterize fishing effort.  Since most fishing effort on the Lower Columbia by B . C . residents is  undertaken by local residents, I assume that fishing effort by B . C . residents will grow at the rate of local population growth. In the 1990-93 period, population growth in the 6 major municipalities of the immediate area grew at an average annual rate of about 0.3% [British Columbia Ministry of Finance and Corporate 8  Relations 1994]. These municipalities account for over 75% of the local population, and roughly 75% of local anglers [ARA Consulting 1992a, 1992b].  The use of the 0.3 % figure to determine annual angling-day growth assumes that the average number of days spent fishing by each angler does not change significantly from year to year and that the proportion of the population which engages in recreational angling remains unchanged.  A 1981 survey of B . C . freshwater  fishing found that about 22% of Kootenay Region residents fished in the region [Reid 1986]. A similar 1985 survey found that proportion had dropped to 17%, although fishing licenses had doubled in price in the interim [Stone 1988a]. Both studies found that the average number of days spent fishing was about 22. While not  The municipalities are Castlegar, Trail, Fruitvale, Montrose, Rossland and Warfield. Castlegar and Trail are by far the largest cities in this group and account for the majority of anglers. All of these communities are within 30 km of the Lower Columbia. 8  114 directly comparable to these earlier results, the 1990/91 survey found that 12% of Lower Columbia area residents fished locally and that the average days spent fishing was 20. It does appear that the average time spent fishing has remained fairly steady.  With respect to participation in fishing, it is hard to draw any  conclusion from these results. However, to provide a lower bound to estimated post-1990/91 fishing effort, I assume that a decline in local angler participation offsets population growth so that no increase in growth in fishing effort occurs. On the other hand, there are a number of reasons to think that 0.3% growth could be too low. One is that when the local area is defined somewhat more broadly the population growth is much higher — when the city of Nelson is included the rate is 0.7% per year, and while Nelson anglers are more likely to fish in Kootenay Lake, the growth in the Nelson area will also increase effort on the Lower Columbia. Other reasons to expect a higher growth rate are that the provincial population as a whole grew at over 2% per year in the 1990-93 period and that interest in the Lower Columbia may increase, despite the consumption advisories, due to the river's growing reputation as a prime rainbow trout fishing area and as the only local walleye and sturgeon fishing area. I use 0.7% as a high value to account for these possibilities.  Fourth, the use of the base 1990/91 angling-day estimate and the above growth rates give estimates of annual angling activity for the 1990/91 to 1995/96 period assuming that activity has been reduced from some unknown level ~ now the key question is to what extent were 1990/91 and later years actually affected by pollution concerns?  Clearly, perceived water quality plays an important role in an angler's enjoyment of  fishing. For example, in the 1985 freshwater fishing survey, Stone [1988a] found that 95% percent of anglers residing in the Kootenay Region cited water quality as being important or very important for their enjoyment of angling. The only other factor to rank so highly was the natural beauty of the surroundings.  To address the question of the impact of contamination concerns on angling activity A R A Consulting [1992a] conducted a telephone survey in early 1991 of anglers who had fished on the Lower Columbia, Kootenay River or Brilliant Reservoir in 1990. A total of 307 anglers were surveyed including 134 who had fished the Lower Columbia (118 local, 12 others from B . C . and 6 from outside the province). A summary of some of  115 the results is shown in Table 6.2.  All anglers were asked to list factors which had affected their frequency  of fishing the Lower Columbia (an indirect question) and whether pollution concerns had affected the frequency (a direct question). Using the intermediate values between the responses to the indirect and direct questions suggests pollution concerns in 1990 meant that: 1) 23% of Columbia River anglers reduced their fishing effort on the river; 2) 50% of anglers who fished locally (but not on the Lower Columbia) decided not to fish the Lower Columbia; and 3) 20% of local anglers who fished outside the region decided not to fish the Lower Columbia. Overall, about one-third of anglers reduced their frequency of fishing on the Lower Columbia due to pollution concerns. Responses to other questions in the survey supported these results. Both chlorinated dioxin and mercury contamination could have been of concern to anglers.  However, among  Lower Columbia anglers, 67% identified the pulp mill as the source of pollution most affecting their activity, while only 13 % said Cominco. Recall that only whitefish were affected by the chlorinated dioxin consumption advisory — these fish were not of great interest to anglers but concern among local anglers had clearly spread to fishing in general.  These results reveal that a great many anglers reduced their activities and that pulp mill pollution was foremost in their minds. However, the results say nothing about the extent of actual reductions in fishing related to the chlorinated dioxin-related consumption advisories. In its study of the potential benefit to be derived from lifting chlorinated dioxin-related consumption advisories, the EPA assumed that such advisories reduce angler participation in a fishery by 20% [U.S. EPA 1993]. This figure was based on a review of 6 studies of angler behaviour in response to consumption advisories in Wisconsin, Michigan and New York State. I assume that this figure can be applied to the Lower Columbia fishery. I use the EPA's 20% figure as an intermediate value, and use 10% and 30% as low and high values to suggest a possible range for the economic cost of the advisories.  Thus under the intermediate assumption the estimated angler days are 80% of the level that would  have occurred without the advisories.  Obviously, some degree of movement to alternative fisheries occurs  although these fisheries are qualitatively different and the economic value derived may be less. possibility can not be directly accounted for. In addition, the lifting of the chlorinated dioxin-related  This  116  Table 6.2. The Effects of Pollution on Lower Columbia Sport Fishing, 1990* 1. Decreased fishing activity a) Percentage of anglers who decreased fishing on the Lower Columbia due to pollution concerns: Area of Angling Activity Other Columbia Local River Areas  Outside Local Area  All Anglers  When asked indirectly When asked directly  6 40  45 54  8 33  23 44  Average  23  50  20  33  2  2  b) Percentage of Lower Columbia anglers who said that pollution had reduced aspects of their fishing-related activity: Frequency of fishing in the river Frequency of eating fish caught Other aspects of fishing-related activity At least one of the above  39 46 1 54  c) Percentage of Lower Columbia anglers who, "if pollutants were substantially reduced," would: Fish more often on the Lower Columbia Fish less frequently elsewhere in order to fish the Lower Columbia more often  61 27  2. Primary pollution concerns Percentage of Lower Columbia anglers who indicated the following pollution sources to have most affect on their fishing activity:  1  2  Celgar pulp mill at Castlegar None Cominco smelter and fertilizer plant Hugh Keenleyside dam Other  62 21 13 1 3  Total  100  Based on a telephone survey of 289 local anglers, plus an additional 18 from outside the local area. 134 of the anglers fished on the Lower Columbia. For the indirect question, anglers were asked what factors had affected their fishing frequency on the Lower Columbia. They were free to list anything they wished. For the direct question, anglers were asked if pollution concerns had affected their frequency of fishing the Lower Columbia.  Source: ARA Consulting [1992a].  117 advisories would not necessarily render the fishery contaminant-free in the eyes of anglers, if only because the mercury-related consumption advisory would still be in place. In 1990, as already noted, it appears that anglers were chiefly concerned with pulp mill effluent, a concern that may have been reinforced by the 1990 public environmental impact assessment of the proposed mill expansion [Celgar Expansion Review Panel 1991]. These considerations suggest that a 20% reduction could be too high in subsequent years. On the other hand, concern likely has been reinforced by the series of announcements over the years regarding chlorinated dioxin contamination.  Fifth, the assumed reduction in fishing activity must be valued. The question of WTP versus W T A arises. In this case, W T A seems most appropriate since I assume that British Columbia residents are entitled to fish in a clean environment. However, only two relatively recent estimates exist for the value of a freshwater fishing day in the Kootenay Region, and both refer to WTP. Reid [1986] found that in 1981 B . C . residents derived a net economic value from a fishing day in the region of about $40 (in 1994 dollars). Stone [1988b] found that in 1985 the corresponding value was substantially lower, at $28 in 1994 dollars. These values were inflated to 1994 dollars using the recreational spending component of the Canadian Consumer Price Index. This assumes that values have grown at the same rate as the general cost of recreational goods and services. Both estimates were based on large-scale C V M surveys with large sample sizes, almost identical questions and similar administration procedures so that design differences can not account for the divergence of estimates. Stone suggests that the sharp province-wide drop in economic activity in 1982-85, and the consequent drop in real incomes, accounts for part of the difference. The doubling of freshwater angling license fees in 1982 also likely played an important part. Thus the 1985 figures may be underestimates.  Another problem with  the figures is that they refer to the whole of the Kootenay Region ~ angling-day net economic values for activity on the Lower Columbia could be somewhat different. However, these two estimates suggest that a B . C . resident's average WTP for a fishing day in the Lower Columbia lies in the $28 to $40 range in 1994 dollars, with $34 as an intermediate value.  118 The results of the analysis are summarized in Table 6.1. Under the intermediate assumptions, reductions in fishing activity amount to about 13,540 angler days over the 1990/91 to 1995/96 period, valued at $460 thousand in 1994 dollars. This represents the economic cost of the chlorinated dioxin discharges from the Castlegar mill in terms of a reduction in B . C . residents' angling activity on the Lower Columbia due to consumption advisories. Varying the range of assumptions yields a cost range of about $152-$1,012 thousand 1994 dollars.  6.3. Summary  Fish consumption advisories have been issued for the receiving environment at most of the interior mills in British Columbia due to contamination by chlorinated dioxins. The Castlegar mill discharges to the Lower Columbia and whitefish have been subject to advisories since 1989.  Although whitefish are not of great  interest to recreational anglers on the river, it is clear that concerns about contamination have caused angling activity to decline. The contamination of whitefish in the Lower Columbia has been observed to fall over the period since 1990 and the advisories should be lifted in the next year or two. The economic cost to British Columbia of the advisories in the 1990/91 to 1995/96 period is estimated to lie in the range of $152-$1,012 thousand 1994 dollars. Reductions in the activity of non-residents, the effect of switching to other potentially less valued fisheries, or possible reductions in the value of fishing that continued on the Lower Columbia, were not examined. I discussed this case study further in Chapter 8.  119  7. FISH TAINTING AND FOOD FISHING - KITIMAT  7.1. Background  Eurocan Pulp and Paper Co., a joint venture of Enso Forest Products Ltd. of Finland and West Fraser Mills Ltd., produces unbleached kraft and chemomechanical softwood pulp at its mill near Kitimat in north-western British Columbia. All of the pulp is made into paper, with current production capacity of about 1100 ADt/d [Pulp and Paper Canada 1994]. Mill effluent has received primary and secondary treatment since start-up of the mill in late 1970. It is discharged to the Kitimat River about 3.2 km from the Kitimat Arm of Douglas Channel (see Figure 7.1).  The period of least effluent dilution in the river occurs in January-March,  corresponding to the low river flow period [Derksen 1981a].  The town of Kitimat was created in 1951 as part of the construction of an Aluminum Company of Canada (Alcan) smelter just south of the town site [Bell and Kallman 1976]. Today the Alcan smelter, the pulp mill and a Methanex methanol and ammonia plant (which started production in 1982) are the only major employers in the town. Each of these major operations, as well as the District of Kitimat, holds a permit to discharge effluent to the Kitimat River, but by far the largest effluent sources are the pulp mill and the smelter [Warrington 1987].  The Kitimat River estuary and the lower few kilometres of the river provide spawning grounds for eulachon, a small smelt-like anadramous fish that ranges from Northern California to Alaska. In British Columbia, the Kitimat, Skeena, Nass and Fraser river estuaries are the four major eulachon spawning grounds [Rogers et al. 1990], although another 10 rivers are known to support the fish [Langer et al. 1977].  Millions of eulachon migrate into the Kitimat estuary and lower river to spawn for a brief period between midMarch and mid-April, after 3-4 years of life in coastal waters [Bell and Kallman 1976, Langer et al. 1977].  120  Source: Derksen [1981].  121 Several weeks later the newly hatched fry migrate downstream to the sea. Historically, eulachon taken from the Kitimat River at the time of migration have been the single most important food fish species for the Haisla people of Kitimaat Village just south of Kitimat, although collectively the six salmon species (Pacific salmon plus steelhead) represented a greater food source [Knapp and Lashmar 1978]. Eulachon harvesting occurred not only in the Kitimat River but also to a lesser extent in the estuary of the Kemano River, which empties into Gardner Channel (see Figure 7.1), and occasionally in several other river estuaries in the area. However, the Kitimat River run was the largest and was of greatest importance.  Eulachon contain a very high oil content, about 15% wet weight [Rogers et al. 1990], and this has made them highly valuable to the Haisla and other coastal First Nations groups, both in the past and today [Collinson 1941, Lawrence 1977, Pritchard 1977, Horwood 1994]. Some fish are eaten fresh, or smoked, salted or dried, but most are rendered into grease for use in cooking and for ceremonial purposes. Because of their high oil content the fish have sometimes been referred to as "candle-fish," since, when dried, they could be used as torches. They have also been known as the "salvation fish" because of their great abundance at a critical time in the late winter when other food was scarce and food reserves were low. As well, historically the fish and grease were an important trading item for the Haisla and other groups with hereditary control of the spawning areas. People from the interior travelled on "grease trails" to the fishing areas and exchanged furs for fishing rights, equipment and oil. People from other coastal areas traded seaweed and herring roe, and the Haida of the Queen Charlotte; Islands traded their renowned cedar boats.  Commercial fishing for eulachon also occurs in British Columbia, having first started roughly 120 years ago [Langer et al. 1977]. By the early 1900s, eulachon were the fifth most important commercial species in the province in terms of value.  However, since the mid-1950s commercial fishing has been restricted to the  lower Fraser River [Langer et al. 1977, DFO 1993], averaging about 22 tonnes per year in the past ten years. There appears now to be no great commercial demand for the fish, and they do not enter the market economy except to a limited extent in the Vancouver area.  122 The high oil content of the eulachon makes them particularly susceptible to bioaccumulation of lipophilic compounds [Rogers et al. 1990]. This predisposition, coupled with the coincidence of their spawning period with high effluent concentrations in the Kitimat River in mid to late March, has caused tainting since shortly after the start-up of the mill in late 1970. As a result, the Haisla stopped harvesting eulachon from the Kitimat River after 1972. The Haisla Band filed a Writ of Summons against the mill and its owners in 1992, claiming damages for pollution of their water supply, tainting of their food resources and harm to their traditional lifestyle [Supreme Court of British Columbia, Vancouver Registry, No. C921885]. The Band did not present specific damage estimates in its writ, although a figure of $16 million has been cited by various sources as one estimate of past damages resulting from tainting. This figure may have been based at least partially on assumptions as to the value of eulachon grease, but its origin and derivation remain unclear. The suit is currently dormant (it can be reactivated at any time) while the Haisla pursue negotiations to resolve the tainting problem and obtain compensation for past damages. Mediated negotiations between Eurocan and the Haisla on the subject of compensation began toward the end of 1994.'  It is important to note that while eulachon tainting represents the most obvious problem, the Haisla have emphasized that they have broader concerns about the contamination of the environment upon which they rely and about the safety of eating eulachon taken from the Kitimat River, even if the flavour were not impaired [M. Gordon, Technical Consultant to the Haisla, pers. comm., 3 April 1995]. This broader concern is evident in the law suit. It suggests that, even when tainting stops, the people will not begin immediately to once again harvest the Kitimat River eulachon.  The mill has undertaken a variety of technical studies since 1991 to try to address the tainting problem [Eurocan 1994]. These include analyses of various in-mill effluent streams to try to identify the specific compound(s) causing tainting and its (their) source, analysis of secondary treatment lagoon operating efficiency, analysis of lagoon solids and consideration of moving the effluent discharge to a marine outfall in Douglas Channel. Terpenes have been considered a likely cause of tainting, based on studies that attempt to  123 correlate chemical analyses of eulachon tissue with the results of taste tests [Eurocan 1994, Rohr 1994]. As a result, attention has focused on reducing these substances, although efforts have yet to prove completely successful.  The tainting propensity of effluent remains highly unpredictable and variable.  Since 1991, Beak Consultants [1993] has undertaken a series of studies in an attempt to assess the river effluent dilution required to prevent tainting and to assess the effect of in-plant modifications on tainting. In each successive report they have concluded that a still higher dilution of effluent is required to ensure that tainting does not occur. One possible-explanation for the observed variability in tainting propensity, despite reduction in terpenes in effluent, is that these substances accumulate in floating oils and foam in the treatment system so that wind or rain, which mixes the oil and foam with the rest of the effluent, can result in substantial and rapid variation in tainting potential [Eurocan 1994].  Currently, the mill managers are  considering various forms of tertiary treatment for the terpene-containing effluent stream. Alternatively, other as yet unidentified substances may contribute to tainting.  Eurocan's effluent permit was modified by the provincial Ministry of Environment, Lands and Parks (MELP) in 1992 so as to prohibit tainting as of 1 March 1994 [Northern Sentinel, 15 Feb. 1995, A l ] . The company requested an extension of the deadline in early 1994 but this was not forthcoming by the time of the eulachon run in mid-March. The mill had scheduled a three-week shut-down for maintenance in March-April but was forced to close four days early to ensure compliance with its permit once eulachon were seen in the Kitimat River [Northern Sentinel, 16 March 1994, A l ] .  At the end of March 1994, after the eulachon run, the deadline for taint-free discharges was extended to 31 July 1994 and then later to 31 October 1994, pending a hearing of Eurocan's appeal of the non-tainting clause of the permit [S. Lakhani, Environment Canada, North Vancouver, pers. comm., 26 August 1994]. M E L P eventually cancelled the appeal and, at the beginning of February 1995, created a Technical Committee to examine the tainting issue [S. Lakhani, Environment Canada, North Vancouver, pers. comm., 23 March  124 1995].  In the meantime, the non-tainting clause has been held in abeyance.  The Technical Committee,  chaired by an independent Chairman, comprises representatives of the Haisla, Eurocan, M E L P , Environment Canada, D F O and the District of Kitimat. It is to meet monthly, starting in April 1995, to discuss actions taken to-date and to try to build agreement about possible future actions regarding studies to identify tainting compounds, remedial actions to eliminate tainting, and the most appropriate test to use as a standard on-going tainting indicator [Northern Sentinel 15 Feb. 1995, A l ] .  7.2. Economic Valuation  Table 7.1 summarizes various factors considered in the valuation exercise, along with a range of assumptions corresponding to low, intermediate and high value estimates. The assumptions are described in more detail below.  First, the potential use of the Kitimat River by the Haisla must be assessed. Historical data on the Kitimat River harvest are sketchy. The earliest estimates show a harvest of 60 tonnes from the combined Kitimat and Kemano Rivers in 1965.  Data for Kitimat River harvests for later years are 55 t in 1969, 30 t in 1970, 89  t in 1971 and 25 t in 1972 [W. Knapp, D F O Vancouver, pers. comm., 6 Sept. 1994]. During this same period, the Kemano River harvest ranged between 30 t and 50 t per year.  The Haisla believe these data  grossly underestimate the actual catch. The Kitimat River run in 1972 was unusually large but the harvest was reduced because of complaints of tainting by mill effluent.  Eulachon have not been harvested on the  Kitimat River since 1972.  There has been some question as to the true extent of the use of the Kitimat River by the Haisla in the two decades prior to the start-up of the mill, since the District of Kitimat sewage plant discharged raw sewage into the river [B. Stevens, Technical Superintendent, Eurocan mill, pers. comm., 6 June 1994]. However, sewage  125  Table 7.1. Factors, Alternative Assumptions and Estimated Economic Cost to the Haisla of Eulachon Tainting at Kitimat, 1973-1995 Assumption Low  Intermediate  60,000  80,000  High  Factors 1. Potential annual Kitimat River eulachon harvest (kg) 2. Price of eulachon oil (1994$/gallon) 3. Weight of fish required to produce one gallon of eulachon oil (kg)  100,000  90  130  170  50.4  50.4  50.4  2,464  4,746  Estimated Economic Cost, 1973-1995 (1994 $ thousand) Cost of tainting in terms of local exchange value of eulachon oil  7,758  has been treated since 1970 [Bell and Kallman 1976], and the reduction in the quality of the eulachon is now due solely to mill effluent. Of greater relevance than conditions prior to the mill are conditions prior to large scale effluent discharges to the Kitimat River: it is comparison with use in these conditions that reflects the extent of the potential harvest lost, and as the mill is currently the sole cause of any such loss, the damage can be attributed to mill effluent.  Historically, almost the whole Haisla community would move to the fishing grounds, spending about a month catching the fish and preparing the oil [Pritchard 1977]. While involvement in the fishery has fallen, it is clear that interest in it remains profound among the Haisla. In speaking of the harvests on the Kitimat, Nass and Skeena Rivers, Horwood [1994] notes that "the [eulachon] runs today remain strong and continue to  126 generate considerable excitement" [p.Bl, 30 March] and that "the capture of eulachons and the resulting oilbased economy remains an invaluable component of coastal native culture" [p.Bl, 6 April].  The available  Kemano estuary harvest data confirm this, and provide evidence that the catch remains tremendously important. In the 1976-80 period the Haisla harvest on the Kemano ranged from 70 to 200 tonnes per year [W. Knapp, DFO Vancouver, pers. comm., 6 Sept. 1994]. In the 1988-92 period, the annual harvest ranged from 70 to 120 t [Triton Environmental Consultants 1993]. No data are available for other years.  I assume that at least some of the Kemano harvest would have occurred even if the Kitimat eulachon were taint-free.  The data for the Kemano harvest, and the few available data for the Kitimat harvest prior to  tainting, then suggest that the Haisla would have harvested 60-100 t of eulachon annually from the Kitimat River if tainting had not occurred. I assume this range encompasses a reasonable lower and upper bound, and I use 80 t as an intermediate estimate.  Second, an especially difficult question concerns how to undertake an economic valuation of something so inextricably linked to social and cultural life. The products of the Kitimat River eulachon harvest were used for trading purposes and for food, but the harvest was also an important occasion of experiential, social and ceremonial value. These benefits and functions of the fishery can not be quantified easily in economic terms. Here I focus on the quantifiable economic value of the eulachon fishery as a food source. In Chapter 3 I reviewed several methods used in the valuation of the food component of subsistence fisheries. The use of commercial landed or wholesale prices clearly has no relevance here since the Haisla have no access to a commercial eulachon fishery, even if they wanted to sell the fish commercially.  Another methodology  involves valuation using retail prices of eulachon, eulachon-based products or similar commodities in the Kitimat area.  However, eulachon and eulachon oil are not sold at retail outlets (seafood stores, grocers,  delicatessens) in Kitimat or other communities in the region (Terrace, Prince Rupert). Nor are they sold at the general store in Kitimaat Village. The only way to obtain either the fish or the oil is via personal contacts  127 with individualfishermenor their families . Retail prices of similar types of commercially available products 9  used for similar cooking purposes, such as lard, represent a possible proxy for valuing the grease, although eulachon oil is quite unique.  The most appropriate valuation methodology in this case involves the use of local exchange values. Monetary values for eulachon oil within Nass-Skeena-Kitimat First Nations communities have been quoted by various authors, based on actual cash sales or on estimates of the value of the items traded for the oil. In the early 1960s, prices around $5/gallon were common, while in the early to mid 1970s prices in the range of $25-45 per gallon were recorded [Kopas 1974, Langer et al. 1977, Lawrence 1977, Pritchard 1977]. Although exchanges likely have become more prevalent among the Haisla in the past two decades, I have been unable to obtain a later estimate of exchange values within Haisla or other communities. I assume that the inflation rate implicit in the food component of the Canadian Consumer Price Index can be used to approximate the growth in eulachon local exchange value. Using this rate for 1973-1995, and the observed price range of $2545 in the first half of the 1970s, yields a price range of $90-170 per gallon in 1994 dollars. I use this price range to value the lost Kitimat harvest over the 1973-95 period in 1994 dollar terms, with $130 as an intermediate value.  Third, valuing the harvest in terms of the oil produced requires that the harvest estimates measured in tonnes be converted to estimates of the amount of oil produced. A portion of the harvest would not be used for production of oil, but I assume that this portion is small. Horwood [1994] notes that some grease makers can render 45 gallons of oil from 2.5 tons of fish (or 1 gallon from 50.4 kg of fish). I use this conversion factor.  Fourth, an especially difficult and significant question in the valuation exercise is how to take account of the  In fact, in all of British Columbia it is likely that eulachon can be purchased fresh at only a very few retail outlets in Vancouver, at between $3.3-6.6/kg [Seafood City, Granville Island Market; Salmon Shop, Granville Island Market; Vancouver, pers. comm., March 1995]. Eulachon oil, or salted, smoked or dried eulachon likely do not enter the market economy anywhere in the province. 9  128 fact that the people of Kitimaat Village have the option of fishing in other, albeit more distant, locations. Fishing at these other locations, like the Kemano estuary, reduces individual and community involvement [Pritchard 1977; M . Gordon, Technical Consultant to the Haisla, pers. comm., 3 April 1995]. The Kemano estuary lies 80 km from Kitimaat Village so that the harvest has been limited to those with access to a reasonably large boat. Participation is also limited for those who are employed in the Kitimat area and cannot take the time to travel the distance, whereas prior to 1973 the harvest occurred in the immediate vicinity of Kitimat and Kitimaat Village. Many families can no longer afford to fish for eulachon. In addition, rights to harvests in specific areas are held by certain families, and this too limits participation in the Kemano harvest. Some families now must rely on trading or buying grease and fish from fishermen, or on receiving gifts.  These considerations imply that the Kemano harvest cannot be considered a replacement for the Kitimat harvest. The reliance on the Kemano has altered traditional harvesting patterns, reduced participation and affected inter-family relationships. I assume that the calculable economic value of the forgone Kitimat harvest in terms of the value of eulachon oil represents an appropriate measure of the cost of tainting.  This  assumption may go some way toward accounting for the cultural and social effects of the tainting.  Table 7.1 shows the estimated value of the lost Kitimat harvest to the Haisla in the 23 year 1973-1995 period. The tainting problem has yet to be solved and there is no basis for assuming that it will be solved soon. Even when eulachon flavour is improved, the Haisla will continue to have some concerns with respect to contamination of the eulachon and likely would not immediately begin to harvest on the Kitimat.  The  assumptions used in the analysis imply that the economic cost of the tainting in terms of the estimated value of eulachon oil from the foregone 1973-95 harvest amounts to $4.7 million 1994 dollars under the intermediate assumptions, with the low and high sets of assumptions yielding a range of $2.5-7.8 million dollars.  129  7.3 Summary  The pulp and paper mill at Kitimat l»egan operation in 1970 and has caused tainting of eulachon since that time. These fish migrate into the Kitimat River estuary to spawn in mid-March to early April, a period of low river flow and relatively low effluent dilution. Eulachon have a very high oil content, and this fact, as well as the extraordinary abundance of the fish, has made them of great importance to many coastal First Nations groups, including the Haisla of Kitimaat Village near Kitimat. However, the high oil content of the fish also makes them particularly susceptible to tainting.  The mill has undertaken a variety of steps since 1991 to solve the tainting problem, with only partial success, and tainting potential remains highly variable and uncertain. Currently (April 1995), a number of discussions and negotiations are underway in attempts to reach agreement on technical issues, remediation and compensation for past damages.  Under the assumptions used here, the cost of the tainting in terms of  foregone Kitimat eulachon oil production lies in the $2.5-7.8 million 1994 dollar range, over the 1973-1995 period.  The analysis is limited by the difficulty of adequately assessing in economic terms the effect of  tainting on the important non-food elements of the harvest. I discuss this case study further in Chapter 8.  130  8. DISCUSSION  In the preceding 4 chapters I presented estimates of the economic cost of specific pulp mill environmental impacts in studies of mills at Port Alberni, Howe Sound, Castlegar on the Lower Columbia, and Kitimat. In this chapter I assess these studies and discuss what they suggest about the problems inherent in applied environmental valuation.  8.1. Case Study Results  Table 8.1 summarizes the case study results. Four important questions with respect to these results arise: 1) How credible are they? 2) How could they be improved? 3) What other economic costs may be important at each site, aside from those estimafesd? and 4) To what extent can the estimates be generalized?  8.1.1. Credibility  I define credibility in terms of validity and reliability. Validity addresses the question of the appropriateness of the analytical methodology in terms of conformance to theory, the choice of the influences assumed to be most relevant to the problem at hand, and how these influences are measured. Each estimate was based on varying degrees of incomplete data but attempted to conform to economic theory as much as possible. Each required judgement to resolve important difficulties. In the Alberni Inlet study, relating the 1990 sockeye mortality incident to sport fishing economic values required important assumptions regarding how the lost fish would have been partitioned among escapement and user groups, the likelihood that any additionalfishwould have been caught, and the magnitude of the relevant WTP value. In the Howe Sound study, the use of gross revenue and the actual extent of relocation both present potentially serious problems. The unknown  131  Table 8.1. Summary of Case Study Cost Estimates ($000 1994 Dollars) Mill Site and Impact Analyzed  Port Alberni - cost to sport fishery  Low  Intermediate  High  58  436  1,816  118  308  638  152  460  1,102  2,464  4,746  7,758  of 1990  sockeye mortality  1  Howe Sound - cost to commercial shellfishing of 1989-96 dioxin-mediated closures  Lower Columbia  - cost of reduced sport fishing due to 1990-96 dioxin-mediated advisory  Kitimat  - cost to Haisla of loss of oil due to tainting of Kitimat River eulachon, 1973-95  1  Average of estimates based on WTP for an angling-day and marginal WTP for an additional fish.  relationship between gross revenue changes and producer surplus changes necessitated an ad hoc assumption for illustrative purposes. For the Lower Columbia study, a serious drawback was the lack of creel survey data except for 1990/91. In the absence of such data it was necessary to use the available 1990/91 fishing data despite the fact that fishing activity varies considerably from year to year in the area. Another problem was that only WTP value estimates were available when W T A estimates were more appropriate. In the Kitimat study, the implication in economic terms of the ability of Haisla fishers to catch eulachon at other locations, albeit much more distant ones, was a difficult issue.  Reliability refers to the robustness of an estimate. Ideally we would consider an estimate to be reliable when repeated measurements using a variety of methodologies yield similar results. This is not possible here. What is most striking about Table 8.1 is that the estimate ranges are quite substantial, with the ratio of the high to low estimates ranging from about 5 for the Howe Sound study to over 30 for the Alberni Inlet study. Such  132 ranges are typical of valuation analyses and are simply a measure of the high degree of uncertainty involved. For the Port Alberni study, the especially large range reflects ignorance about the extent to which the assumed increased catch would lead to more angling days or greater catch rates. The use of a range of assumptions for each factor considered important in the analyses is an attempt to provide a degree of robustness by explicitly accounting for uncertainty.  While it is impossible to measure the statistical confidence interval  associated with an estimate without assigning subjective probabilities to each element in the analysis, I expect that the range for each estimate should encompass the "true" cost, subject of course to methodological validity and the appropriateness of the assumptions themselves.  Overall, the reliability and internal validity of the estimates remains primarily a subjective judgement. Comparison with external criteria provides another means of assessing credibility. To my knowledge, no studies in Canada have undertaken economic analyses of environmental impacts similar enough to those in this thesis as to be useful for comparison purposes. In the United States, case studies by the U.S. E P A [1993] of pulp mill pollution show estimates which are qualitatively similar.  8.1.2. Possible Improvements  With more time and financial resources, improvements could have been made in each of the case studies by addressing in more detail some of the major difficulties noted in the previous sub-section, as well as others. In every case, more primary research would have helped to add greater precision to the estimate ranges as well as adding to their credibility. With respect to recreational valuation issues, it is likely that surveys of anglers would give a better sense of how they value their activity and the factors which influence their value. With respect to Howe Sound commercial shellfishing, more in-depth study of cost conditions and relocation might have been fruitful. Much of the information used in this study relied on discussions with provincial and federal government officials charged with the responsibility of overseeing pulp mills, pollution and the fish  133 stocks in each case study area. I assumed that these people would be broadly knowledgeable about most facets of the issues I addressed. Clearly, given more time, discussion with other people including, most importantly, mill managers and user group representatives, would have added a fuller perspective and helped to check and fine-tune some of the assumptions used.  8.1.3. Other Important Costs  The economic cost estimates shown in Table 8.1 for each site represent only a portion of the potential total cost of effluent at each site.  Table 8.2 summarizes various other types of economic cost which can be  presumed to exist at each location.  In each study I considered only very specific types of costs and found that some of these could not be estimated.  The limits imposed on the analysis by data deficiencies and uncertainty imply that the Alberni  Inlet, Lower Columbia and Kitimat results underestimate the cost of effluent to the user groups studied. With respect to the Alberni Inlet study, I was able to assess only the cost of a single sockeye mortality incident although it seems likely that other less severe mortality events have occurred and that stocks have been affected by chronic mortality and reduced reproductive success. Nothing could be said about chinook other than the likelihood that economic costs would be lower than for sockeye given that chinook move through the system at a slightly different time. In the Howe Sound study, the cost of chlorinated dioxin-related closures could only be estimated in terms of commercial fishing ~ the cost in terms of recreational fishing could not be estimated due to insufficient information on the extent of activity prior to the closures.  In the Lower  Columbia study, I estimated the cost of the consumption advisories only in terms of reduced fishing activity. Other effects on the value of sport fishing ~ reduced value for those who continued to fish on the Lower Columbia, for example — could not be estimated.  Finally, in the Kitimat study I focused only the food  component of the native harvest. The relationship between the resulting economic cost estimates and impacts  134  Table 8.2 Potential Economic Costs of Pulp Mill Effluent at Case Study Sites ($000 1994 Dollars) Lower Columbia  Alberni Inlet  Howe Sound  Commercial fishing  X  $118-658  Recreational fishing Aesthetic value Subsistence fishing  $58-1,816 X X  X X 0  Non-use services  X  X  X  X  Human health risks  X  X  X  ?  Economic Cost  1  X - a likely cost 1  0 $152-1,012 X ?  Kitimat  X X X $2,464-7,758  ? - uncertain  The costs shown refer to specific time periods and specific damages. In particular, the cost to recreational anglers in Alberni Inlet and on the Lower Columbia are underestimates of the full costs to this user group.  on cultural and social functions, assuming it could be estimated, is not known.  The costs examined in the case studies occurred only in the context of economic value as it related to specific user groups. Thus some elements of a particular environmental impact had no economic consequence. For example, in the Howe Sound study the contamination by chlorinated dioxin mattered, in the context of  commercial shellfishing, only to the extent that thefisherslost producer surplus and could notfindcomparable catches elsewhere.  For some vessels, it might have been the case that there was no economic cost from the  closures, although I assumed otherwise in assuming that Howe Sound was the optimal fishing location for all those who had fished there. Similarly, I assumed that a portion of the lost Alberni Inlet 1990 sockeye stock did not represent an economic cost in the context of sportfishingbecause I assumed that non-B.C. anglers would not have reduced their participation in the fishery. Moreover, under the assumptions used, it also did not represent a cost in terms of commercial and native food fishing. However, it would have represented an economic cost in terms of other values.  135 Some of the unestimated economic costs shown in Table 8.2 could be quite considerable. Most important may be non-use values, especially as affected by chlorinated dioxin contamination. One rule-of-thumb, that non-use services of a particular site have a value to recreational users that can be approximated as at least 50% of recreational use value (see Section 3.4.1), provides one possible means of estimating these costs with respect to the Alberni Inlet and Lower Columbia studies. The economic cost of increased health risks associated with several decades of chlorinated dioxin contamination likely also would be important. For example, the U.S. EPA [1993] study of the economic benefits of pulp mill pollution control found that reducing health risks by limiting chemical contaminants accounted for a large share of the benefits. As noted in Section 3.4.4, such estimates are likely to be highly controversial.  8.1.4. Generalizations  A key question about the case study results is the extent to which they can be generalized to environmental impacts at other mills in British Columbia.  For example, can the Lower Columbia results be used to  approximate the economic cost of consumption advisories at other interior locations? In a number of respects, the potential for generalization seems good. One reason is the range of analysis. I examined four different economic impacts affecting three user groups at five mills. An alternative approach, to attempt a thorough analysis of all economic impacts at a single mill, would not have provided the same richness of analysis. As Chapter 2 indicated, no mill in the province has given rise to more than two of the impacts ~ severe DO reductions, chlorinated dioxin contamination leading to closures or consumption advisories, tainting — described in the case studies, although there are many other possible types of economic cost, including potentially important non-use and health risk costs.  Broadly speaking, the case study mills are fairly representative of mills in British Columbia. The five mills vary in size but all can be considered medium-sized among those in the province. The Port Alberni and Howe  136 Sound mills discharge to marine environments. The Castlegar mill discharges to a freshwater environment as does the Kitimat mill, although the latter sits only a few kilometres from salt water. The Howe Sound and Castlegar mills discharge chlorinated organics as did the Port Alberni mill until relatively recently in its history.  Like other coastal mills, the three coastal mills in the case studies installed full primary and  secondary treatment systems only in the 1990s, although the Port Alberni mill has had partial treatment since 1966. The Kitimat mill has had treatment since its inception as have most interior mills. The Castlegar mill is an anomaly, having discharged untreated effluent to a freshwater environment for many years. Like most mills in the province, all of the study mills produce kraft pulp and all but the Woodfibre mill also produce C M P or C T M P . The Port Mellon and Castlegar mills began producing these chemomechanical pulps after considerable expansion and modernization programs in the early 1990s.  As described in Section 3.3, transferring per unit economic values such as angling-day value can be a valid procedure, especially if differences in the determinants of value are explicitly taken into account ~ however, the wholesale transference of a site damage estimate may yield quite misleading results.  While the mills  themselves may be roughly representative, many environmental and economic characteristics of the case studies are highly site-specific.  For instance, the Port Alberni receiving environment is qualitatively similar  to that of Port Alice and Gold River, and in each case reductions in dissolved oxygen pose a danger to aquatic life. Despite these similarities, the general result of the Alberni Inlet study cannot be applied to these other locations since the size of the fish stocks and the economic values derived from Alberni Inlet in terms of fishing activity are considerably greater than at the other locations. Similarly, it would be incorrect to simply assign the range of costs estimated for Howe Sound to other coastal closure areas.  At other locations,  different species have been subject to closures, some of the dioxin-mediated closures coincide with closures for other reasons, and the importance of the fishery varies considerably. For example, oyster harvesting in the Crofton and Elk Falls areas is commercially quite important but, on the other hand, parts of the affected areas were already the subject of sanitary closures at the time of the initial chlorinated-dioxin closure announcements.  137 When the objective of analysis is to obtain province-wide estimates of costs then two possible approaches can be used. One is to analyze every relevant site. This allows the important site-specific elements at each site to be considered. Clearly, the drawback is that such an analysis will be time-consuming and potentially very expensive.  Moreover, there are theoretical reasons to expect that aggregate estimates which result from  adding-up across sites and types of value could sometimes lead to misleading estimates [Mitchell and Carson 1989, Randall 1992].  Site-by-site valuation and aggregation ignores the potential for substitution and  complementarity relationships between environmental services at each site and across sites. As well, the total availability of environmental services must be considered — for example, while site-specific estimates may incorporate an assumption about relocation, as in the Howe Sound study, when a large number of sites become contaminated the appropriateness of this assumption becomes questionable.  As an alternative to site-by-site analysis, a number of U.S. studies have attempted nationally-based economic evaluation of pulp mill effluent in the context of the Federal Clean Water Act. These studies make broad sweeping assumptions, but broadly defining the economic service to be valued as water quality that is boatable, fishable and swimmable (categories consistent with the Clean Water Act itself), as some of these studies have done, provides one means of empirically evaluating the national cost of reductions in amenity and non-use services (for example, see Carson and Mitchell 1993). Similarly, nation-wide studies which rely on the transfer of values taken from the literature and broad characteristics of impacts, user groups and receiving environments have been done in the United States. Most notable are the studies by Luken [1990] and the U.S. E P A [1993] discussed in Chapter 3.  While the estimates derived in the case studies in this thesis can not be generalized, the methodologies themselves can be broadly applied. If more and better information were available, the methodologies could be used to generate province-wide estimates. For example, the approach of the Lower Columbia study of the cost of advisory-related reductions in fishing activity could easily be applied to the Fraser River or elsewhere. Similarly, if there were available good data on the sub-areas affected by each shellfishing ground closure, and  138 estimates of harvesting costs, then the approach used in the Howe Sound study could be generalized fairly readily.  8.2.  Difficulties in the Economic Analysis of Pulp Mill Pollution  Each case study provided a somewhat different context for translating environmental impacts into economic costs, and thus provided unique insights. The Port Alberni study was the only one where consequences for a fish population had to be translated into potential economic costs. As a consequence it was also the most difficult since a variety of influences obscure the relationship between mill effluent and the economic focus point, salmon stocks. These include the complex oceanographic characteristics of Alberni Inlet, a history of substantial stock enhancement programs, and the importance of variable weather conditions in mediating the impact of effluent. Only a single clear mortality event could be considered but these complicating factors had still to be kept in mind. In contrast, both the Howe Sound and Lower Columbia studies involved instances where economic use clearly has been affected.  These effects arose because of human health concerns,  irrespective of what effects on the fish have occurred, if any. Chlorinated dioxin contamination has been observed to cause sublethal effects in both shellfish and finfish, but these effects and their consequence for fish populations are uncertain and, in the Howe Sound study, were not of relevance for the types of value examined. A particularly important element of the Lower Columbia study was the importance of perceptions in shaping economic value. Finally, the Kitimat study involved an environmental impact where the economic cost related not to any known or suspected harmful impact on the biota or on human health, but to subjective assessments of fish quality. A second noteworthy aspect of the Kitimat study was that it concerned impairment of native use of the environment, creating particular difficulties for economic valuation.  These case studies, and the literature reviews in Chapters 2 and 3, demonstrate five significant problems in the economic analysis of pollution. These are complexity, uncertainty, site-specificity, limited information  139  and a group of problems that relate to the general difficulty of meaningfully describing environmental impacts in economic terms. Thefirstfour difficulties are by no means limited to the economic analysis of pollution - they infuse every problem in environmental studies ~ but economic analysis adds a new layer of difficulty to each.  8.2.1.  Complexity  The environment is a complex set of often poorly understood systems. As Chapter 2 showed, the impacts of effluent may themselves occur in complex ways. The natural variability of environmental systems, especially when combined with human interventions of various sorts, poses a particularly difficult but relevant issue related to environmental complexity.  At the level of economic analysis, there is a very substantial theoretical and methodological literature which addresses environmental valuation, some of which has been touched on in this thesis. The complex issues in this literature, the intricacies and caveats, are no less profound than the environmental complexities, and often form the subject of extended debates. Any applied economic analysis must attempt to not only make judgements about the environmental uncertainties but must also consider which of the economic issues is most relevant and important, and what might be the implications. Complexity means that any applied analysis, whether involving environmental science or economics, inevitably will be a simplification of the "real" world.  8.2.2.  Uncertainty  Complexity creates uncertainty. All applied economic analyses suffers from uncertainty to a greater or lesser extent but it likely becomes most serious in environmental economic studies because of the uncertainties in  140 environmental impacts and the methodological difficulties inherent in attempting to value services not traded in markets. Environmental responses to effluent occur at levels ranging upward from the sub-cellular to the level of populations and communities.  Given the diverse composition of effluent, it is often difficult to  determine what actually causes a given effect so that meaningful dose-response relationships can be difficult to identify.  Furthermore, relating these effects, or even effects at the organism level, to the higher level  effects usually relevant for economic use value adds another level of uncertainty. Finally, the uncertainties in the valuation process itself can be immense, as hinted in Chapter 3 and as demonstrated in the case studies.  Economists can seldom do repeated experimentation to test theories and evaluate the reliability and validity of results regarding actual environmental damages. Thus even measures of uncertainty in applied analysis, the "reasonable" estimate ranges, are themselves uncertain.  At best, repeated studies of similar sites,  environmental services and types of pollution justify a degree of confidence as to what likely constitutes a reasonable range of values.  8.2.3. Site-Specificity  The survey in Chapter 2 demonstrated the site-specific nature of effluent impacts.  A wide range in the  characteristics and volume of effluent discharges, varying types of other human influences, diverse hydrological features of the receiving environment, different species use and abundance, and even different weather patterns, create unique conditions and environmental impacts at each site. At the economic level, each site potentially provides all of the various environmental aquatic environmental services described in Chapter 3, but the relative importance of the services varies considerably from site to site and over time, according to specific environmental features and the characteristics of the local user groups. Thus site-specific results of economic analysis are not likely to be broadly generalizable, as discussed above with respect to the case studies in this thesis. What they can provide are general impressions and qualitative characterizations.  141 8.2.4. Limited Information  Valuation can only be as credible as the information it relies on. The case studies in this thesis are indicative of a general problem in environmental economic analysis, that of severe data deficiencies.  Data shortages  were apparent even in the case of important sport fisheries such as in Alberni Inlet and on the Lower Columbia. In the Kitimat case study, limited information reflected to some extent the fact that the issue of compensation currently forms the subject of negotiation and a law suit.  Limitations in data and other  information mean that the analyst must make sometimes considerable assumptions in what essentially involves the construction of an artificial data set for analysis. This was the case to a greater or lesser extent with each case study, and added considerable uncertainty.  One important implication of limited information at the  economic level is that theoretical and methodological rigour must often be compromised, and it is here that the judgement of the analyst becomes especially important.  8.2.5. Economic Difficulties  A variety of important difficulties and issues arise in attempting to translate reductions in water quality into changes in the quantity or quality of potential aquatic environmental services, and thence into economic costs. One of the most significant is that even when environmental impacts are well-characterized and understood, the technical or scientific measures of the impacts may not be easily related to use and non-use values [Luken 1990, Russell and Smith 1990]. Water quality or environmental health indicators usually describe items relevant to scientific understanding of aquatic biological and physical characteristics rather than those directly relevant for valuation. Moreover, how people themselves perceive or understand water quality may bear little or no relation to technical measures, an important point especially when considering damages to environmental amenity services. For example, a clear lake, one in which life has been effectively destroyed by acid rain or acid rock drainage, may be viewed by some people as having higher water quality, and consequently higher economic value for certain purposes, than a lake which may in fact be pristine. People seem to value criteria  142 such as water clarity and colour as indicators of water quality when in fact these may have little relationship to technical parameters [Russell and Smith 1990]. different than public perceptions.  Similarly, scientific assessments of risk may be quite  This effect was important in the Lower Columbia study where concerns  about the safety of fish consumption clearly extend beyond the scientific concern behind the advisories.  The concept of water quality itself as a basis for thinking about the economic value of the environment has been criticized. Lant and Mullens [1991] argue that water quality indicators are valid when applied to nonrecreational uses and human health concerns, but likely do not adequately encompass what it is that recreationalists actually value.  In part, this relates to a discrepancy between technical indicators of water  quality and perceptions of water quality. More importantly though, Lant and Mullens argue that what really matters to recreationalists is not only perceived water quality but also a large variety of other factors involving the scenic aesthetics of the surrounding land, the degree of congestion and the existence of recreational facilities. Similarly, Green and Tunstall [1991] believe that people can only judge water quality on the basis of the absence or presence of physical indicators such as smell or debris, users presumed to be knowledgeable (such as anglers), or biota such as plants and waterfowl.  Aside from the disjunction between scientifically meaningful measures of risk or environmental quality and the perceptual determinants of economic value, a further difficulty lies in the fact that it simply may be impossible to value many environmental impacts. In several of the case studies the combination of uncertainty and data limitations precluded valuation. A more fundamental problem may be that some environmental impacts are not obviously relevant in economic terms, or cannot be readily described in economic terms. This is the case with the cultural significance of native food fisheries. As another example, the destruction or alteration of benthic habitat near mill outfalls, or the sublethal effects of contaminants on fish physiology, do not lend themselves easily to economic valuation. The impact on fisheries may be virtually zero or at best the impacts on fish abundance are likely to be unquantifiable. Theoretically, C V M studies of non-use value could be used, but valuation of such specific (and sometimes uncertain) effects at non-unique sites might be  143 quite difficult for survey respondents. This certainly does not mean that a concern to limit such damage is unwarranted. meaningful.  Valuation of a more general concept of biotic health over a larger area might be more This would be useful for broad province-wide studies but not for evaluation of costs at a  particular site, as was done in this thesis.  Another important issue that arises in the translation of environmental impacts into economic values is the question of entitlements.  In Section 3.2 I stressed the importance of this question in the choice of WTP or  W T A . Entitlements may be controversial and there are problems in estimating W T A , although as I suggested in Chapter 3 these may not be always as serious as sometimes implied in the empirical literature. Practically speaking, however, an applied analysis based on value transfers must almost invariably use WTP estimates even where the analyst deems W T A to be appropriate. There are simply too few W T A estimates to do otherwise.  The time and cost of a valuation analysis can be crucial in determining how it is performed. Often, transfer analysis represents the only option and I have noted several times the potential for distortion created by value transfers. In this thesis, the Alberni Inlet results relied on angling-day value estimates for marine sport salmon fishing in the Pacific Northwest taken from other studies.  How well these estimates actually apply to Port  Alberni in the 1990s is unknown, especially as all of the estimates were derived in the context of fisheries with quite different characteristics (different species, no large pools of fish). Similarly, Kootenay-wide estimates for 1981 and 1985 bear an unknown relationship to current Lower Columbia values since the Kootenay area includes a large number of fishing sites other than those on the Lower Columbia. In each of these cases we can probably safely assume that the transfers give a valid order-of-magnitude estimate ~ for example, a Port Alberni sockeye fishing day certainly has net value to anglers of more than a few dollars, but less than one hundred dollars. Only sound original research could really give us a better idea. Even more detailed analysis of the transfer studies could help although often, as in the case of the Kootenay angling-day value estimates, information which would allow this is not reported in the published studies.  144  9. CONCLUSION  In British Columbia, as in much of the rest of Canada, pulp and paper mill effluent represents one of the most significant point-sources of water pollution. Extremely large quantities of the "conventional" pollutants, BOD and TSS, are discharged every day in British Columbia, although substantially less now than even 5 years ago. Chemical contaminants from the mills, especially chlorinated organic compounds, have become of great concern in the last 10 years. The impacts of these discharges were summarized in Chapter 2 ~ they were seen to be diverse, often uncertain and often quite site-specific.  I have spent considerable time in discussing the environmental impacts of pulp mill effluent and the environmental and economic context of each case study in this thesis.  It is important that any economic  analysis be based on a sound understanding of the non-economic dimensions of the problem at hand, and the way that other disciplines look at the problem. A public health officer views a water pollution problem as a health issue, an engineer sees it as a challenge of environmental engineering, a biologist sees it as an ecological concern, and an economist regards it as an economic problem. In fact it is all of these and its resolution must consider all of these facets, the unique perspective that each discipline brings, and the limitations of each approach.  With respect to economics, an extensive body of literature has developed over the last 30 years regarding the economic theory relevant to measurement of the cost of environmental damage or the benefit of reductions in pollution. The difficulties and issues discussed in the literature were touched upon only briefly in Chapter 3, and some of the difficulties in applied analysis should be clear from the case studies.  Of course, it is the nature of applied economic analysis, perhaps more so in the realm of environmental analysis than any other, that severe data shortages and uncertainty require approximations, sometimes elaborate assumptions, rules-of-thumb and departures from the dictates of economic theory. Disagreement  145 often exists with respect to the scientific "facts" or their interpretation. The same basic economic theory and methodology applied to a given problem may yield different results depending on the assumptions used. Clearly then, environmental valuation involves great complexity and uncertainty, and what is chosen as the basis of analysis reflects as much the analyst's judgement as it does any "objective" criteria.  When one considers the extent to which applied economic analysis of environmental impacts must depart from theoretical soundness, the often (or usually) poor data quality and availability, the methodological simplifications needed, the high degree of uncertainty, and the elemental problem of casting environmental impacts in economic terms, the conclusion regarding the use of applied environmental valuation in the analysis of pulp mill or any other form of pollution may seem pessimistic.  There is no doubt that there are many  instances where credible economic value estimates cannot be derived. On the other hand, credible estimates are possible in many other cases, especially involving the natural resource and amenity services of the environment.  It is important to recall why analyses of environmental impacts can be useful. These analyses have relevance for economic efficiency in terms of comparing the economic values associated with alternative combinations of environmental services. Efficiency concerns the allocation of all things which provide economic value in such a way as to maximize the benefit to society. While individual estimates may be deficient in various respects, hundreds of studies have collectively confirmed that environmental services, traditionally not included in market transactions, and conventionally valued at zero or next to zero, do in fact have economic value that can be quite significant. It is vital that public and private decision makers take these values into account. If they do not, the well-being of society cannot be optimized. Even partial and incomplete estimates can be helpful in demonstrating that environmental services have economic value.  Moreover, while the  shortcomings of applied analysis can sometimes be serious, economic analysis provides another structure for integrating, organizing and presenting information, as well as for making uncertainty explicit.  146 It is also important to keep in mind several factors which should temper our use of economic analysis of environmental impacts. We should not forget that estimates which are both scientifically and economically credible are difficult to derive, and are usually highly site-specific in nature. As well, economic valuation is by definition reductionist, reducing all things to monetary values because of an analytical need for a single commensurable metric of measurement. Economic analysis cannot fully account for the multifunctionality and complexity of the environment, nor for our incomplete comprehension of how we affect the environment.  Finally, we should keep in mind that society has many goals including social, environmental and economic sustainability, societal cohesion, income redistribution and other equity concerns. Economic efficiency matters only in helping to choose among alternative ways of reaching these goals ~ efficiency is not a primary goal in and of itself (although considerations of costs and benefits, when they can be measured, may be very useful in establishing societal goals). Moreover, estimates of economic value, and thus of measures of efficiency, are themselves not independent of distributional equity considerations, since individuals' net economic values depend on their income level.  147  REFERENCES CITED Adamowicz, W . L . , V . Bhardwaj and B. Macnab [1993]. "Experiments on the difference between willingness to pay and willingness to accept." Land Economics 69(4), 416-27. Alderdice, D . F . and J.R. Brett [1957]. "Some effects of kraft mill effluent on young Pacific salmon." Journal of Fisheries Research Board of Canada 14(5), 783-795. Anderson, E . [1983]. "Effects of kraft mill effluent on phytoplankton and zooplankton." In W . M . Pomeroy (ed.), Proceedings of Pulp Mill Effluent Monitoring, Regional Program Report 83-15, Environmental Protection Service, Environment Canada, Vancouver. Andrusak, H . and A . D . Martin [1983]. "Angler use and harvest of fish on the Columbia River from the Hugh Keenleyside Dam downstream to Trail: Summary report on creel census data collected from 1976 to 1982." B . C . Fish and Wildlife Branch, Nelson, B . C . . A R A Consulting [1992a]. "Factors affecting angler use of the Columbia River." Prepared for B . C . Hydro, Vancouver. A R A Consulting [1992b]. "Lower Columbia River sport fishing and boating survey." Hydro, Vancouver.  Prepared for B . C .  Arrow, K . , R. Solow, E . Learner, P. Portney, R. Radner and H . Schuman [1993]. "Report of the N O A A Panel on Contingent Valuation." Federal Register 58(10), 4602-4614. Axegard, P, O. Dahlman, I. Haglind, B. Jacobson, R. Morck and L . Stromberg [1993]. "Pulp bleaching and the environment - the situation 1993." Nordic Pulp and Paper Research Journal 4(8), 365-378. B . C . Environment [1992]. "Lower Columbia River dioxin study results released: Consumption advisory on whitefish." Information Bulletin 28 July. Nelson, B . C . . Beak Consultants [1993]. "Tainting evaluation of Eurocan final effluent." Prepared for Eurocan Pulp and Paper Co., Kitimat, British Columbia. Bell, L . M . and R.J. Kallman [1976]. The Kitimat River Estuary: Status of Environmental Knowledge to 1976. Special Estuary Series No. 6. Estuary Working Group, Environment Canada, Ottawa. Berger, T.R. [1977]. Northern Frontier, Northern Homeland: Report of the Mackenzie Valley Pipeline Inquiry, Volume One. Government of Canada, Ottawa. Bergstrom, J.C. [1990]. "Concepts and measures of the economic value of environmental quality: A review." Journal of Environmental Management 31, 215-228. Birch, G.J., L . R . Hildebrand and G.R. Ash [1994]. "Fish population composition, distribution and habitat in the Columbia River in British Columbia downstream of the Hugh Keenleyside Dam." In Canada/United States Technical Workshop on the Upper Columbia River Basin: Proceedings from the Technical Section. Spokane, 15-16 Nov. 1994. State of Washington Water Research Center, 91-96.  148 Birch, J.R. [1989]. "Trends in dissolved oxygen levels in Port Alberni Harbour and correlation with Somass Rover flow." Prepared by Arctic Sciences Ltd. for Institute of Ocean Sciences, Department of Fisheries and Oceans, Sidney, British Columbia. Birtwell, I.K. [1989]. "Comments on the sensitivity of salmonids to reduced levels of dissolved oxygen and to pulp mill pollution in Neroutsos Inlet, British Columbia." Canadian Technical Report of Fisheries and Aquatic Sciences, 1695.  Bishop, R.C. and M.P. Welsh [1992]. "Extinction values in benefit-cost analysis and damage assessment." Land Economics 68(4), 405-417. Bishop, R . C , K.J. Boyle and M.P. Welsh [1987]. "Toward total economic value of Great Lakes fishery resource." Transactions of the American Fisheries Society 116, 339-345.  Bockstael, N.E., W.M Hanemann and C L Kling [1987]. "Estimating the value of water quality improvements in a recreational demand framework." Water Resources Research 23(5), 951-960. Bonsor, N., N. McCubbin and J.B. Sprague [1988]. Kraft Mill Effluents in Ontario. Expert Committee on Kraft Mill Toxicity for the Pulp and Paper Sector of Municipal-Industrial Strategy for Abatement (MISA), Ontario Ministry of the Environment, Toronto. Bothwell, M.L. [1992]. "Eutrophication of rivers by nutrients in treated kraft pulp mill effluent." Water Pollution Research Journal of Canada 27(3), 447-472.  Boyle, K.J. and J . C Bergstrom [1992]. "Benefit transfer studies: Myths, pragmatism and idealism." Water Resources Research 28(3), 657-663. Brand, D.G., R.G.B. Reid and M.L. Harris [1992]. "Prevalence of idiopathic gill and liver lesions in English and Flathead sole collected near pulp mills of Howe Sound." In C D . Levings, R.B. Turner and B. Ricketts (eds.), "Proceedings of Howe Sound Workshop." Canadian Technical Report of Fisheries and Aquatic Sciences 1879, 239-250.  British Columbia Ministry of Environment, Lands and Parks [1994]. B.C. Pulp Mills: Effluent Status Report. Victoria. British Columbia Ministry of Finance and Corporate Relations [1994]. British Columbia Financial and Economic Review. 54th Edition, Victoria. Brookshire, D.S., L.S. Eubanks and C.F. Sorg [1987]. "Existence values and normative economics." In G.L. Peterson and C.F. Sorg (eds.), Toward the Measurement of Total Economic Value. General Technical Report RM-148. Rocky Mountain Forest and Range Experiment Station, Forest Service, United States Department of Agriculture. Brookshire, D.S. and H.R. Neill [1992]. "Benefit transfers: Conceptual and empirical issues." Water Resources Research 28(3), 651-655. Brown, T.C. and E.S. Birch [1992]. "Estimating the economic value of subsistence harvests of wildlife in Alaska." In G.L. Peterson, C.S. Swanson, D.W. McCollum and M.H. Thomas (eds.), Valuing Wildlife Resources in Alaska. Westview Press, Boulder, Colorado.  149 Brown, R.F., M . M . Musgrave, D.G. Demontier, D.E. Marshall and M J . Comfort [1979]. "Catalogue of salmon streams and spawning escapements of Statistical Areas 22 and 23 (Nitinat and Barkley Sound)." Canadian Data Report of Fisheries and Aquatic Sciences 167. Brox, G.H. [1976]. "Water quality in the lower Fraser River Basin: A method to estimate the effect of pollution on the size of a salmon run." M.A.S. thesis, Department of Civil Engineering, University of British Columbia, Vancouver. Butcher, G.A. [1992]. Lower Columbia River: Hugh Keenleyside Dam to Birchbank Water Quality Assessment and Objectives. Technical Appendix. Water Quality Branch, B.C. Ministry or Environment, Lands and Parks, Victoria. Cameron, T . A . and M.D. James [1987]. "The determinants of value for a recreational fishing day: Estimates from a contingent valuation survey." Canadian Technical Report of Fisheries and Aquatic Sciences 1503. Carey, J.H., P.V. Hodson, K.R. Munkittrick and M.R. Servos [1993]. "Recent Canadian studies on the physiological effects of pulp mill effluent on fish." Environment Canada and Fisheries and Oceans Canada CCREM [1987]. Canadian Water Quality Guidelines. Prepared by the Task Force on Water Quality Guidelines, Canadian council of Resource and Environment Ministers, Ottawa. Carson, R.T. and R.C. Mitchell [1993]. "The value of clean water: The public's willingness to pay for boatable, fishable and swimmable quality water." Water Resources Research 29(7), 2445-2454. Celgar Expansion review Panel [1991]. Celgar Expansion Review Panel Final Report. Submitted under the Federal Environmental Assessment and Review Process and B.C. Major Project Review Process. CEPA [1990]. Priority Substances List Assessment Report No. 1: Poly chlorinated Dibenzodioxins and Polychlorinated Dibenzofurans. Environment Canada and Health and Welfare Canada, Ottawa. CEPA [1991]. Priority Substances Assessment Report List No. 2: Effluents from Pulp Mills Using Bleaching. Environment Canada and Health and Welfare Canada, Ottawa. Chestnut, C.W. [1993]. "Tidal water recreational fishery study: Alberni Inlet-Barkley Sound." Unpublished report submitted to Recreational Fisheries Division, Department of Fisheries and Oceans, Vancouver. Collinson, H.A. [1941]. "Theoolachan fishery." The British Columbia Historical Quarterly V, 25-31. Colodey, A . G . and M . Pomeroy [1986]. Low dissolved oxygen in Neroutsos Inlet, British Columbia: Observed impacts of sulphite mill effluent, July-October 1985." Draft Interim Report, Environmental Protection Service, Environment Canada, West Vancouver. Colodey, A . G . and J.G. Tyers [1987]. "Marine monitoring of the Crofton B.C. pulp mill: November 1983 and April 1986." Regional Program Report 87-10, Pacific Region, Environmental Protection, Environment Canada, Vancouver.  150  Colodey, A . G . , L . E . Harding, P.G. Wells and W.R. Parker [1990]. "Effect of pulp and paper mill Effluents and their constituents on estuarine and marine environments in Canada: A brief review." Regional Program Report 90-08, Pacific Region, Environment Canada, Vancouver. Colodey, A . G . and P.G. Wells [1992]. "Effects of pulp and paper mill effluents on estuarine and marine ecosystems in Canada: A review." Journal of Aquatic Ecosystem Health 1, 201-226. Cropper, M . L . and W.E. Oates [1992]. "Environmental economics: A survey." Journal of Economic Literature XXX (June), 675-740. Cross, C . L . , L . Lapi, and E.A. Perry [1991]. "Production of chinook and coho salmon from British Columbia hatcheries, 1971-1989. Canadian Technical Report of Fisheries and Aquatic Sciences  1816. Dahlman, O., I. Haglind, R. Morck, F. de Sousa and L . Stromberg [1993]. "Chemical composition of effluents from chlorine dioxide bleaching of kraft pulps before and after secondary treatment." EUCEPA International Environmental Symposium, 27-29 April, Paris. Davis, J. C. [1975]. Waterborne Dissolved Oxygen Requirements and Criteria with Particular Emphasis  on the Canadian Environment. Report no. 13, National Research Council of Canada, Associate Committee on Scientific Criteria for Environmental Quality, Ottawa. Department of Fisheries and Forestry [1970]. "Proposed chinook and coho hatching development program for British Columbia." Pacific Region, Resource Development Branch, Vancouver. Department of Fisheries and Oceans [various years]. "Record of Management Strategies: Salmon, Area 23." Pacific Region, Vancouver. Department of Fisheries and Oceans [1988a]. "Preliminary results of fish and sediment sampling program for dioxins and furans announced." News Release NR-HQ-88-032 and Backgrounder, 16 May. Department of Fisheries and Oceans [1988b]. "Closure of fisheries areas." News Release NR-PR-88-36 and Backgrounder, 30 November. Department of Fisheries and Oceans [1989]. "Shellfish closures announced for Howe Sound and Prince Rupert." News Release NR-PR-89-017 and Backgrounder, 14 June. Department of Fisheries and Oceans [1990]. "Stream summary catalogue: Alberni." South Coast Division, Fisheries Branch, Vancouver.  Subdistrict #23, Port  Department of Fisheries and Oceans [1993]. "1994 Management Plan: Eulachon." Pacific Region, Vancouver. Derksen, G. [1980]. "Environmental review of the Finlay and Mackenzie pulp mills at Mackenzie, B.C." Regional Program Report 80-7, Pacific Region, Environment Canada, Vancouver. Derksen, G. [1981a]. "Environmental review of the Eurocan pulp mill at Kitimat, B.C." Program Report 81-27, Pacific Region, Environment Canada, Vancouver.  Regional  151 Derksen, G. [1981b]. "Environmental review of the Northwood, Intercontinental and Prince George pulpmills at Prince George, B.C." Regional Program Report 82-4, Pacific Region, Environment Canada, Vancouver. Derksen, G. and M . Lashmar [1981a]. "Environmental review of the B.C. Timber (Celgar pulp division) pulpmill at Castlegar, B.C." Regional Program Report 81-20, Pacific Region, Environment Canada, Vancouver. Derksen, G. and M . Lashmar [1981b]. "Environmental review of the Crestbrook pulpmill at Skookumchuk, B.C." Regional Program Report 81-24, Pacific Region, Environment Canada, Vancouver. Desvousges, W.H., M.C. Naughton and G.R. Parsons [1992]. "Benefit transfer: Conceptual problems in estimating water quality benefits using existing studies." Water Resources Research 28(3), 675-683. DREE, [1979]. Single Sector Communities. Department of Regional Industrial Expansion, Ottawa. Dwernychuk, L.W. [1990]. "The receiving environment of the Upper Fraser River: A pilot environmental effects monitoring program examining the physical/chemical/biological elements of the system related to pulp mill effluents, 1989." Two volumes. Hatfield Consultants Ltd., West Vancouver. Edwards, S.F. and C. Carlson [1989]. "On estimating compensation for injury to publicly owned marine resources." Marine Resource Economics 6, 27-42.  El Serafy [1991]. "The environment as capital." In R. Costanza (ed.,)•Ecological Economics. Columbia University Press, New York. Ellis, G.M. and A . C . Fisher [1987]. "Valuing the environment as input." Journal of Environmental Management 25, 149-156. Environment and Land Use Secretariat [1980]. Howe Sound Overview Study. Columbia, Victoria.  Province of British  Eurocan [1994]. "Eurocan effluent quality program." Quarterly reports prepared by D.A. Kelly, Director, Technical and Environmental Services, 31 May, 26 July, 15 September and 19 December. Fisher, A . and R. Raucher [1984]. "Intrinsic benefits of improved water quality: Conceptual and empirical perspectives." In V . K . Smith and A . D . Witte (eds.), Advances in Applied Microeconomics, Vol. 3. JAI Press Inc., Greenwich, Connecticut. Frankel, M . and C M . Linke [1992]. "The value of life and hedonic damages: Some unresolved issues." Journal of Forensic Economics 5(3), 233-247. Freeman, A . M . [1993].  The Measurement of Environmental and Resource Values:  Methods. Resources for the Future, Washington, D . C .  Theory and  152 Government of Canada [1993]. "Dioxin levels in shellfish dropping but contamination closures, consumption advisories continue." News Release NR-PR-93-48E, 27 August. Government of Canada [1989]. "Fishing closures and consumption restrictions issued for eight British Columbia coastal sites." News Release, 23 November. Government of Canada [1995]. "Reduced dioxin levels lead tofisheriesreopenings." News Release NR-PR-95-02E, 10 February 1995. Green, C.H. and S.M. Tunstall [1991]. "The evaluation of river water quality improvements by the contingent valuation method". Applied Economics 23, 1135-1146. H.A. Simons Ltd. [1992]. "Assessment of industry costs to meet British Columbia's new AOX regulations." Prepared for the British Columbia Ministry of Environment, Lands and Parks and the Ministry of Economic Development, Small Business and Trade., Victoria. Harding, L . E . [1992]. "Overview Of the marine ecosystem of Howe Sound." Environmental Science Workshop, Bowen Island, 30 Sept.-3 Oct. 1991.  Howe Sound  Harrison, K. [1993]. "Passing the buck: Federalism and Canadian Environmental Policy." PhD thesis, Department of Political Science, University of British Columbia, Vancouver. Hausman, J.A.(ed.) [1993]. Contingent Valuation: A Critical Assessment. North-Holland, Amsterdam. Hicks, D.B. and J.W. DeWitt [1971]. "Effects of dissolved oxygen on kraft pulp mill effluent toxicity." Water Research 5, 693-701.  Hildebrand, L.R., G.R. Ash and G.J. Birch [1994]. "Status of white sturgeon in the Columbia River downstream of the Hugh Keenleyside Dam." In Canada/United States Technical Workshop on the Upper Columbia River Basin: Proceedings from the Technical Section. Spokane, 15-16  Nov. 1994. State of Washington Water Research Center, 97-101. Horwood, D. [1994]. "Eulachon — Ancient grease of the coast." The Northern Sentinal. Two part series, 30 March, BI and 6 April, BI Hyatt, K.D. and G.J. Steer [1987]. "Barkley Sound sockeye salmon (Oncorhynchus nerka): Evidence for over a century of successful stock development, fisheries management, research and enhancement effort." Canadian Special Publication of Fisheries and Aquatic Sciences 96,435-  457. Inch, H . [1993]. "Ex-ante assessment of secondary impacts of environmental regulation: Case study — organochlorines in the Canadian pulp and paper industry." M.Sc. thesis, Resource Management Science/Interdisciplinary Studies, University of British Columbia, Vancouver. Johnson, N.S. and R.M. Adams [1989]. "On the marginal value of a fish: Some evidence from a steelhead fishery." Marine Resource Economics 6, 43-55. Just, R . E . , D . L . Hueth and A. Schmitz [1982].  Applied Welfare Economics and Public Policy.  Prentice-Hall Inc., Englewood Cliffs, New Jersey.  153  Kahneman, D. and A . Tversky [1979]. Econometrica 47(2), 263-291.  "Prospect theory:  An analysis of decision under risk."  Kahneman, D. and A. Tversky [1984]. "Choices, values and frames." American Psychologist 39(4), 341-350. Kahneman, D. and J.L. Knetsch [1992]. "Valuing public goods: The purchase of moral satisfaction." Journal of Environmental Economics and Management 22, 57-70. Kay, B.H. [1989]. Pollution in British Columbia's Marine Environment:  A Status Report. State of  the Environment Report No. 89-1. Environment Canada, Ottawa. Knapp, W. and I. Cairns [1978]. "MacMillan-Bloedel pulp mill - Port Alberni: Fisheries resources of Fisheries and Marine Service Statistical Area 23". Internal Report, Fisheries and Marine Service, Pacific Region, Department of Fisheries and Oceans, Vancouver. Knapp, W. and M . Lashmar [1978]. "Fisheries resources of Fisheries and Marine Service Statistical Area 6." Internal Report, Fisheries and Marine Service, Pacific Region, Department of Fisheries and Oceans, Vancouver. Kneese, A.V. [1984]. Measuring the Benefits of Clean Air and Water. Washington, D . C , Resources for the Future, Inc. Knetsch, J.L. [1990]. "Environmental policy implications of disparities between willingness to pay and compensation demanded measures of values." Journal of Environmental Economics and Management 18, 227-237. Koning, A.R, L . Cross, J.E.G. Sikes and O. Turk [1993]. "From conception to resolution: Effluent treatment upgrade at Alberni Pulp and Paper to meet special quality regulations." Paper presented at the 1993 CPPA Environmental Conference, October 26-28, Thunder Bay. Kopas, L . [1974]. "Eulachon the saviour fish." Wildlife Review Vll(2), 11-12. Kovacs, T. [1986]. "Effects of bleached kraft mill effluent onfreshwaterfish: A Canadian perspective," Water Pollution Research Journal of Canada 21(1), 91-118.  Kringstad, K.P. and K. Lindstrom [1984]. "Spent liquorsfrompulp bleaching." Environmental Science and Technology 18(8), 236A-247. Langer, O., B.G. Shepherd and P.R. Vroom [1977]. "Biology of the Nass River eulachon (Thaleichthys pacificus)." Technical Report T-77-10, Pacific Region, Department of Fisheries and Oceans, Vancouver. Lant, C L . and J.B. Mullens [1991]. "Lake and river quality for recreation management and contingent valuation." Water Resources Bulletin 27(3), 453-460. Lebrasseur, R.J., C D . McAllister, W.E. Barraclough, O.D. Kennedy, J.Manzer, D. Robinson and K. Stephens [1978]. "Enhancement of sockeye salmon (Oncorhynchus nerka) by lake fertilization in Great Central Lake: Summary Report." Journal of the Fisheries Research Board of Canada 35,1580-1596.  154 Lim, P. [1992a]. "Marine monitoring of Northumberland Channel near Harmac pulp mill, Harmac, B.C., 1986, 1989." Regional Data Report 92-04, Pacific Region, Environment Canada, North Vancouver. Lim, P. and A. Colodey [1992]. "Marine monitoring near the Gold River pulp mill on Muchalat Inlet, B.C., 1986, 1988, 1989." Regional Data Report 91-08, Pacific Region, Environment Canada, North Vancouver. Loomis, J.R. [1992]. "The evolution of a more rigorous approach to benefit transfer: benefit function approach." Water Resources Research 28(3), 701-705. Luken, R.A. [1990]. Efficiency in Environmental Regulation: A Benefit-Cost Analysis of Alternative  Approaches. Boston, Kluwer Academic Publishers. Luken, R.A., F.R. Johnson and V. Kibler [1992]. "Benefits and costs of pulp and paper effluent controls under the Clean Water Act." Water Resources Research 28(3), 665-674. Mackay F . M . and L.R. Galloway [1994]. "Start-up and initial operation of Celgar Pulp's low-rate air activated sludge treatment system." In Canada/United States Technical Workshop on the Upper Columbia River Basin: Proceedings from the Technical Section. Spokane, 15-16 Nov. 1994.  State of Washington Water Research Center, 81-87. Mah, F.T.S, D.D. MacDonald, S.W. Sheehan, T . M . Tuominen and D. Valiela [1989]. "Dioxins and furan in sediment and fish from the vicinity of ten inland pulp mills in British Columbia." Conservation and Protection, Environment Canada, Ottawa. Mallette, G.F. [1991]. "An investigation of some aspects of the dollar value of the environmental resources of the Columbia River in South-Central British Columbia." M.A. thesis, Department of Geography, University of Waterloo, Waterloo, Ontario. Mallette, G.F. and D.C. Baker [1994]. "A model for estimating the impacts of the operation of the Hugh Keenleyside and Brilliant Dams on downstream recreation use on the upper Columbia reach." In Canada/United States Technical Workshop on the Upper Columbia River Basin: Proceedings  from the Technical Section. Spokane, 15-16 Nov. 1994. State of Washington Water Research Center, 219-223. Masse, W.D. and K. Petersen [1977]. "Evaluation of incremental recreational benefits from Salmonid enhancement." Department of Fisheries, Vancouver. McCollum, D.W., G.L. Peterson, C.S. Swanson [1992]. "A managers guide to the valuation of nonmarket resources: What do you really want to know?" In G.L. Peterson, C.S. Swanson, D.W. McCollum and M.H. Thomas (eds.), Valuing Wildlife Resources in Alaska. Westview Press, Boulder, Colorado. McConnell, K.E. and I.E. Strand [1989]. "Benefits from commercial fisheries when demand and supply depend on water quality." Journal of Environmental Economics and Management 17,284-292. McConnell, K.E. [1992]. "Model building and judgement: Implications for benefit transfers with travel cost models." Water Resources Research 28(3), 695-700.  155 McCubbin, N . , N.C. Bonsor and D. Owen [1990]. "Economic impact of proposed regulation of pulp and paper industry BOD, TSM, toxicity, organochlorines (AOX), dioxins and furans." Prepared by McCubbin Consultants Inc. for Regulatory and Economic Affairs Division, Environment Canada, Ottawa. McKinnon, L [1992]. "AOX as a regulatory parameter: A scientific review of AOX toxicity and environmental fate." Draft report, Environmental Protection Division, B.C. Ministry of Environment, Lands and Parks, Victoria. McLeay, D. [1987]. "Aquatic toxicity of pulp and paper mill effluent: A review". Report EPS 4/PF/l, Environmental Protection, Environment Canada, Ottawa. Mitchell, R.C. and R.T. Carson [1989].  Using Surveys to Value Public Goods: The Contingent  Valuation Method. Washington, D.C., Resources for the Future, Inc. Munkittrick, K.R., C.B. Portt, G.J Van Der Kraak, I.R. Smith and D.A. Rokosh [1991]. "Impact of bleached kraft mill effluent on population characteristics, liver MFO activity, and serum steroid levels of a Lake Superior white sucker (Catostomus commersoni) population," Canadian Journal of Fisheries and Aquatic Sciences 48, 1371-1380..  Munkittrick, K.R., M . E . McMaster, C.B. Portt, G.J Van Der Kraak and I.R. Smith [1992]. "Comparison of wild fish populations before and after secondary treatment of bleached kraft mill effluent," Canadian Technical Report of Fisheries and Aquatic Sciences 1863, 190-197.  Muth, R . M . , D.E. Ruppert and R.J. Glass [1987]. "Subsistence use of fisheries resources in Alaska: Implications for Great Lake Fisheries Management." Transactions of the American Fisheries Society 116, 510-518. Nassichuk, M.D. [1992]. Dioxin-mediated shellfish closures in Howe Sound." Environmental Workshop, Bowen Island, 30 Sept.-3 Oct. 1991.  Howe Sound  Naughton, M.C. and W.H. Desvousges [1986]. "Water quality benefits of additional pollution control in the pulp and paper industry." Prepared at Research Triangle Institute for United States Environmental Protection Agency, EPA contract 68-01-7033, Washington D.C. Neilson, A . H . , A.-S. Allard, P.A. Hynning and M . Remberger [1991]. "Distribution, fate and persistence of organochlorine compounds formed during production of bleached pulp." Toxicological and Environmental Chemistry 30, 3-41.  Nelson, H. [1979a]. "Pulpmill environmental impact assessment: Canadian Forest Products Ltd., Howe Sound Pulp Division." Regional Program Report 79-2, Pacific Region, Environment Canada, Vancouver. Nelson, H. [1979b]. "Pulpmill environmental impact assessment: MacMillan Bloedel Ltd, Powell River Division." Regional Program Report 79-14, Pacific Region, Environment Canada, Vancouver. Nener, J., D. Kieser, J.A.J. Thompson, W.L. Lockhart, D.A. Metner and R. Roome [1994]. "Columbia River mountain whitefish health: Initial findings of the July 1992 study." In Canada/United States Technical  Workshop on the Upper Columbia River Basin: Proceedings from the  Technical Section. Spokane, 15-16 Nov. 1994. State of Washington Water Research Center.  156 Norton, V . , T. Smith and I. Strand [1983]. "Stripers: The economic value of the Atlantic coast commercial and recreational striped bass fisheries." Sea Grant Program, University of Maryland, College Park, Maryland. O'Connor, B.I., T . G . Kovacs, R.H. Voss, P.H. Martel and B. van Lierop [1994]. "A laboratory assessment of the environmental quality of alternative pulp bleaching effluents." Pulp and Paper Canada 95(3), 47-55. OECD [1992]. Benefits Estimates and Environmental Decision-Making. Prepared by D.W. Pearce and K. Turner for Organization for Economic Co-Operation and Development, Paris. OECD [1989]. Environmental Policy Benefits: Monetary Valuation.  Prepared by D.W. Pearce and  A. Markandya for Organization for Economic Co-Operation and Development, Paris. Owens, J.W. [1991]. "The hazard assessment of pulp and paper effluents in the aquatic environment: A review." Environmental Toxicology and Chemistry 10, 1511-1540.  Packman, G.A. [1979a]. "Pulp mill environmental impact assessment, Ocean Falls Corp., Ocean Falls, B. C." Regional Program Report 79-6, Pacific Region, Environment Canada, Vancouver. Packman, G.A. [1979b]. "Pulp mill environmental assessment: Canadian Cellulose Ltd. Northern Pulp Operation, Port Edward, British Columbia." Regional Program Report 79-7, Pacific Region, Environment Canada, North Vancouver. Parker, R.R., D.H. Heller, C R . Harwood and J.G. Sanderson [1972]. "Some facets of the impacts of pulp mill effluent on the Alberni Inlet." Pulp and Paper Magazine of Canada 73(10), 89-99. Parker, R.R. and J. Sibert [1973]. "Effect of pulpmill effluent on dissolved oxygen in a stratified estuary. I. - Empirical observations." Water Research 1, 503-514. Pearce, D.W. and R.K. Turner [1990]. Economics of Natural Resources and the Environment. Johns Hopkins University Press, Baltimore. Pearse Bowden Economic Consultants [1975]. "Recreational values from fish and wildlife resources in British Columbia." Draft paper. Pearse, P. [1994]. "An assessment of the salmon stock development program on Canada's coast." Final Report of the Program Review of the Salmonid Enhancement Program, Internal Audit and Evaluation Branch, Department of Fisheries and Oceans, Vancouver. Pharand, N.L. [1988]. "Forest-sector dependent communities in Canada: A demographic profile." Information Report DAC-X-23, Canadian Forest Service, Ottawa. Piercey, G . E . , I.K. Birtwell, J.S. Korstrom, G . M . Kruzynski, C M . Langton, S. Spohn [1993]. "Behavioural responses of adult sockeye salmon (Oncorhynchus nerkd) under simulated estuarine conditions and salt water hypoxia."  906.  Canadian Data Report of Fisheries and Aquatic Sciences  157 Pomeroy, W.M. [1983]. "Environmental impacts of bottom deposits originating from B.C. coastal pulp mills: A general summary." In W.M. Pomeroy (ed.), Proceedings of Pulp Mill Effluent Monitoring, Regional Program Report 83-15, Environmental Protection Service, Environment Canada, Vancouver. Poole, N.J., D.J. Wildish, D.D. Kristmanson [1978]. "The effects of the pulp and paper industry on the aquatic environment." Critical Reviews in Environmental Control 8, 153-195.  Pritchard, J.C. [1977]. "Economic development and the disintegration of traditional culture among the Haisla." PhD thesis, Department of Anthropology and Sociology, University of British Columbia, Vancouver. Province of British Columbia [1994]. "Decrease in dioxins and furans in Columbia River whitefish reported." Information Bulletin, 14 December. Nelson, B.C. Pulp and Paper Canada [1994]. Annual and Directory, 1994. Randall, A. [1992]. "A total value framework for benefit estimation." In G.L. Peterson, C.S. Swanson, D.W. McCollum and M.H. Thomas (eds.), Valuing Wildlife Resources in Alaska. Westview ress, Boulder, Colorado. Reid, R. [1986]. "The value and characteristics of freshwater angling British Columbia." B.C. Ministry of Environment, Victoria. Rogers, I.H., J.A. Servizi and C D . Levings [1988]. "Bioconcentration of chlorophenols by juvenile chinook salmon (Oncorhynchus tshawytscha) overwintering in the upper Fraser River: field and laboratory tests," Water Pollution Research Journal of Canada 23(1), 100-113.  Rogers, I.H., C D Levings, W.L. Lockhart and R.J. Norstrom [1989]. "Observations on overwintering juvenile chinook salmon (Oncorhynchus tshawytscha) exposed to bleached kraft mill effluent in the upper Fraser River, British Columbia," Chemosphere 19(12), 1853-1868. Rogers, I.H., I.K. Birtwell and G.M. Kruzynski [1990]. "The Pacific eulachon (Thaleichthyspacflcus) as a pollution indicator organism in the Fraser River Estuary, Vancouver, British Columbia." The Science of the Total Environment 97/98, 713-727.  Rohr, A . C [1994]. "Application of semi-permeable membrane devices (SPMDs) to the monitoring of kraft mill effluent with an emphasis on potential fish-tainting compounds." Master of Applied Science Thesis, Department of Civil Engineering, University of British Columbia, Vancouver. Russell, C.S. and V.K. Smith [1990]. "Demands for data and analysis induced by environmental policy." In E.R. Berndt and J.E. Triplett (eds.), Fifty Years of Economic Measurement. Studies in Income and Wealth Volume 54, National Bureau of Economic Research. Schreier, H . , S.J. Brown and K.J. Hall [1991]. "The land-water interface in the Fraser River Basin." In A.H.J. Dorcey (ed.), Water in Sustainable Development: Exploring Our Common Future  in the Fraser River Basin. Vancouver.  Westwater Research Centre, University of British Columbia,  158  Scott, M.J., R.J. Moe, M.R.LeBIanc, and J.W. Currie [1987]. "Columbia river salmon: benefit cost analysis and mitigation." Northwest Environmental Journal 3(2), 121-152. Seiler, D. [1989]. "Differential survival of Grays Harbour Basin anadramous salmonids: Water quality implications." In C D . Levings, L.B. Holtby and M.A. Henderson (eds.), Proceedings of the National Workshop on Effects of Habitat Alteration on Salmonid Stocks. Canadian Special  Publication of Fisheries and Aquatic Sciences 105, 123-135. Serdar, D . , B. Yake and J. Cubbage [1994].  "Contaminant trends in Lake Roosevelt."  In  Canada/United States Technical Workshop on the Upper Columbia River Basin: Proceedings  from the Technical Section. Spokane, 15-16 Nov. 1994. State of Washington Water Research Center, 63-67. Servizi, J.A. [1989]. "Protecting Fraser River salmon (Oncoryhnchus spp.) from wastewaters: An assessment." In C D . Levings, L.B. Holtby and M.A. Henderson (eds.), Proceedings of the National Workshop on Effects of Habitat Alteration on Salmonid Stocks. Canadian Special  Publication of Fisheries and Aquatic Sciences 105, 136-153. Shuter, B.J. and H.A. Reiger [1989]. "The ecology of fish and populations: Dealing with interactions between levels." In C D . Levings, L.B. Holtby and M.A. Henderson (eds.), Proceedings of the National Workshop on Effects of Habitat Alteration on Salmonid Stocks. Canadian Special  Publication of the Fisheries and Aquatic Sciences 105, 34-49. Sinclair, W.F. [1990].  Controlling Pollution from Canadian Pulp and Paper Manufacturers:  A  Federal Perspective. Environment Canada, Ottawa. Smith, V . K . , and W.H. Desvousges [1986]. Measuring Water Quality Benefits. Boston.  Kluwer-Nijhoff,  Smith, V . K . , W.H. Desvousges and M.P. McGivney [1983]. "Estimating water quality benefits: An Econometric analysis." Southern Economic Journal 50(2), 422-437. Smith, V.K. [1989]. "Taking stock of progress with travel cost recreation demand models: Theory and implementation." Marine Resource Economics 6, 279-310.  Smith, V.K. [1992]. "On separating defensible benefit transfers from 'smoke and mirrors.'" Water Resources Research 28(3), 685-694. Solomon, K., H . Bergman, R. Hugget, D. Mackay and B. McKague [1993]. "A review and assessment of the ecological risks associated with the use of chlorine dioxide for the bleaching of pulp." Alliance for Environmental Technology. Sonnen, C A . and M . Laurence [1990]. "National and provincial economic effects of regulations to reduce emissions of the pulp and paper industry." Prepared by Informetrica for Environment Canada, Ottawa. Sorhus, C N . [1981]. "Estimated expenditures by sport anglers and net economic values of salmon and steelhead for specified fishers in the Pacific Northwest." PhD thesis, Oregon State University.  159  Stanbury, W.T. [1993]. "Regulating water pollution by the pulp and paper industry in Canada." FEPA Working Paper 192, University of British Columbia, Vancouver. Starr, P.J., A . T . Charles and M.A. Henderson [1984]. Reconstruction of British Columbia sockeye salmon stocks: 1970-1982." Canadian Manuscript Report of Fisheries and Aquatic Sciences  1780. State of Washington [1991]. "Health agency recommends caution on eating Lake Roosevelt fish." Department of Health, Olympia. Steinnes, D.N. [1992]. "Measuring the economic value of water quality: The case of lakeshore land." Annals of Regional Science 26, 171-176. Stone, M . [1988/a].  British Columbia Freshwater Results of the 1985 National Survey of Sport  Fishing. Fisheries Technical Circular 79. British Columbia Ministry of Environment and Parks, Victoria. Stone, M . [1988/b]. Economic Values and Impacts of Freshwater Sportfishing in British Columbia."  Fisheries Management Report 91. Victoria.  British Columbia Ministry of Environment and Parks,  Stucchi, D J . [1990]. "The pulp mill effluent problem in Neroutsos Inlet, B.C." Canadian Technical Report of Hydrography and Ocean Sciences 124.  Stucchi, D.J., A. Colodey, I. Birtwell, M. Waldichuk, K. Hyatt and W. Knapp [1990]. "Review of the water quality issue in Port Alberni Harbour." Unpublished report, Department of Fisheries and Oceans, Vancouver. Sullivan, D.L. [1979]. "Marine environmental surveillance monitoring of Muchalat Inlet, 1976-77." Regional Program Report 79-10, Pacific Region, Environment Canada, Vancouver. Sullivan, D.L. [1980]. "Marine environmental surveillance monitoring at B.C. pulp mills." Regional Program Report 81-16, Pacific Region, Environment Canada, Vancouver. Sullivan, D.L. [1981]. "Marine environmental surveillance monitoring at B.C. South Coast pulp mills, 1980." Regional Program Report 81-21, Pacific Region, Environment Canada, Vancouver. Sullivan, D.L. [1982]. "Marine environmental surveillance monitoring at B.C. South Coast pulp mills, 1981-82." Regional Program Report 83-17, Pacific Region, Environment Canada, Vancouver. Suntio, L.R., W.Y. Shiu and D. Mackay [1988]. "A review of the nature and properties of chemicals present in pulp mill effluents," Chemosphere 17(7), 1249-1290. Tully, J.P. [1949]. "Oceanography and prediction of pulp mill effluent pollution in Alberni Inlet." Fisheries Research Board of Canada, 83.  U.S. EPA [1993]. "Regulatory impact assessment of proposed effluent guidelines and NESHAP for the pulp, paper and paperboard industry." Final Report, EPA-821-R-93-020. Office of Water, Environmental Protection Agency, Washington D.C.  160  Usher, A.J. [1976]. "Evaluating country food in the northern native economy." Arctic 29, 103-120. Usher, A.J. [1987]. "Ontario Lake of the Woods fishery: Economic and social analysis." Transactions of the American Fisheries Society 116, 352-366. Vaughan, J. and C.S. Russell [1982].  Freshwater Recreational Fishing:  The National Benefits of  Water Pollution Control. Resources for the Future, Washington D.C. Viscusi, W.K. [1993]. "The value and risks to life and health." Journal of Economic Literature XXXI(December), 1912-1946. Waldichuk, M . [1956]. "Pulp mill pollution in Alberni Harbour, British Columbia." Industrial Wastes 28(2), 199-205.  Sewage and  Waldichuk, M . [1983]. "Water pollution from pulp mill effluent in British Columbia: A general overview." In W.M. Pomeroy (ed.) Proceedings of Pulp Mill Effluent Monitoring, Regional Program Report 83-15, Environmental Protection Service, Environment Canada, Vancouver. Waldichuk, M . [1988]. "Effects of solid wood wastes on marine benthic organisms and habitats." In D.D. Wolfe and T.P. O'Connor (eds.), Oceanic Processes in Marine Pollution. Volume 5. Robert E . Kreiger Pub. Co., Malabar, Florida. Walsh, R.G., D . M . Johnson and J.R. McKean [1992]. "Benefit transfer of outdoor recreation demand studies 1968-1988." Water Resources Research 28(3), 707-713. Warrington, P.D. [1987]. Skeena-Nass Area.  Lower Kitimat River and Kitimat Arm Water Quality  Assessment and Objectives. Technical Appendix. Environment, Lands and Parks, Victoria.  Water Quality Branch, B.C. Ministry of  161  APPENDIX A .  F I S H E R Y C L O S U R E S A N D FISH C O N S U M P T I O N ADVISORIES BRITISH C O L U M B I A , 1988 T O F E B R U A R Y 1995  162  Table A . l . Fishery Closures and Consumption Advisories Due to Chlorinated Dioxin Contamination, Coastal British Columbia Locations, 1988-1995 Date  30 Nov. 1988  14 Jun. 1989  23 Nov. 1989  29 Nov. 1990  General Location  1  Species  2  Users Affected  3  Action  Howe Sound (northern, western portions) Prince Rupert  C, S, P  All  Closure  C  All  Closure  Howe Sound Howe Sound (whole area) Howe Sound (whole) Prince Rupert Prince Rupert Prince Rupert  C, S, P C C(h)  C(h)  All C S, N C All S, N  s  All  Closure  c  C  Closure  C(h) Cl  C, P  S, N All All All All  Advisory Closure Closure Closure Closure  C  C  Closure expanded  C(h)  S, N All All  Advisory expanded Closure expanded Closure expanded  Prince Rupert Kitimat, Elk Falls, Harmac, Crofton, Cowichan Bay Kitimat, Elk Falls, Harmac, Crofton, Cowichan Bay Kitimat Powell River Powell River, Crofton Gold River Kitimat, Elk Falls, Powell River, Crofton, Harmac Kitimat, Elk Falls, Powell River, Crofton, Harmac Powell River Gold River All coastal mills  c s  c o  o c  Bottomfish (livers only)  All  Closure expanded Closure expanded Advisory Closure expanded Closure Advisory  Advisory  9 Aug. 1991  Elk Falls, Harmac, Crofton Elk Falls, Harmac, Crofton  C C(h)  C S, N  1 May 1992  Vancouver Harbour Victoria Harbour Victoria Harbour  C  c  C(h)  All C S, N  c  C  Closure expanded  C(h) S, P  S, N All  Advisory expanded Closure expanded  27 Aug. 1993  Howe Sound, Elk Falls, Gold River, Harmac, Powell River, Prince Rupert Howe Sound, Gold River, Harmac, Powell River, Prince Rupert Howe Sound  Closure expanded Advisory expanded Closure Closure Advisory  163  Table A . l . (continued) Date  10 Feb. 1995  1  2  3  General Location  1  Howe Sound, Elk Falls Howe Sound, Powell River, Gold River Howe Sound, Powell River . Howe Sound Crofton, Powell River Gold River Prince Rupert  Species  2  Users Affected  3  Action  C  C  Closure expanded  C C(h) S, P  S, N S, N All All All All  Closure rescinded Advisory expanded Closure rescinded Closure rescinded Closure rescinded Closure rescinded  o p  s  Area covered by closures, advisories and expansions vary according to the species and user group affected at any given location. Species codes: C - Crab, C(h) - Crab hepatopancreas, S - shrimp, P - prawn, CI - clams, O - oyster. User group affected codes: C - commercial fishers, S - recreational anglers, N - native anglers.  Source: Environment Canada.  164  Table A.2. Fish Consumption Advisories Due to Chlorinated Dioxin Contamination, Interior British Columbia Locations, 1989-1994 Date  19 May 1989  General Location  Castlegar (Columbia R.) Prince George, Quesnel (Fraser R.) Kamloops (Thompson River)  Species  Lake whitefish Largescale sucker, mountain whitefish Mountain whitefish, northern squawfish, peamouth chubb  27 Apr. 1990  Kamloops (Thompson R., Kamloops L.)  Dolly varden, rainbow trout  1 Jun. 1990  Castlegar (Columbia R.)  Mountain whitefish  29 Nov. 1990  Prince George (Fraser R.)  White sturgeon (livers only)  10 Apr. 1992  Fraser and Thompson R. Fraser R. (part), Quesnel R., North and South Thompson Rivers, Thompson R. Kamloops (Thompson R., Kamloops L.)  All fish livers  28 July 1992  Columbia R. (Keenleyside Dam to border) Kootenay River (pulp mill to border)  Lake and mountain whitefish All fish liver  25 Jan. 1994  Fraser R. (part), Quesnel R., North and South Thompson Rivers, Thompson R.  Mountain whitefish advisory changed to livers only  Source: Environment Canada.  Mountain whitefish Rainbow trout  THE  UNIVERSITY O F BRITISH C O L U M B I A 1956 Main Mall Vancouver, B.C., Canada V6T 1Y3  LIBRARY  Per missions  Dear ^  O f / i c e  r  ^  As a graduate student at the University of British Columbia, I am preparing my thesis, which will be microfilmed by the National Library of Canada, and copies of the film will be lent or sold. May I have permission to use in my thesis, and for the National Library to microfilm, the excerpts from your publication(s) described on the back of this letter. I would be very grateful for your favourable consideration of this request Would you please complete the form on the back of this letter, and return it to the address given on the form. Thank you very much.  Sincerely,  /VI.Sc.  CWtdU4c  £o9 ~^22 -92so  DE-11  (4/88)  (FAt)  THE  UNIVERSITY O F BRITISH C O L U M B I A 1956 Main Mall Vancouver, B.C., Canada V6T 1Y3  LIBRARY  2o  Dear S»s; rYlf. / I A J S I M .*.  my of the the  As a graduate student at the University of British Columbia, I am preparing thesis, which will be microfilmed by the National Library of Canada, and copies the film will be lent or sold. May I have permission to use in my thesis, and for National Library to microfilm, the excerpts from your publication(s) described on back of this letter.  I would be very grateful for your favourable consideration of this request Would you please complete the form on the back of this letter, and return it to the address given on the form. Thank you very much.  Sincerely,  CS-VUI'-OAH-V^./I tiA-1 Siu C'llil  D E - l l (4/88)  THE UNIVERSITY OF BRITISH COLUMBIA 1956 Main Mall Vancouver, B.C., Canada V6T 1Y3  Canada.  b i O i ^ r t m ^ Dear SiKf-  my of the the  fl[ir  .  De^Kv^n-.  As a graduate student at the University of British Columbia, I am preparing thesis, which will be microfilmed by the National Library of Canada, and copies the film will be lent or sold. May I have permission to use in my thesis, and for National Library to microfilm, the excerpts from your publication(s) described on back of this letter.  I would be very grateful for your favourable consideration of this request Would you please complete the form on the back of this letter, and return it to the address given on the form. Thank you very much.  Sincerely,  /Y[.Sc. (  W U l o i e  (Resource  yY[aiA.ayJLiT^.i\{  6 0 H - 7 3 ( c - 9 7 5 V  2 :  ton - roa - baay  D E - 1 1 (4/88)  aW  ( V A G ^ O Cojj^  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items