Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Perceptions of risk to water environments in the lower Fraser basin, British Columbia Cavanagh, Nigel Steven 1998

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

Item Metadata

Download

Media
831-ubc_1999-0015.pdf [ 3.6MB ]
Metadata
JSON: 831-1.0088884.json
JSON-LD: 831-1.0088884-ld.json
RDF/XML (Pretty): 831-1.0088884-rdf.xml
RDF/JSON: 831-1.0088884-rdf.json
Turtle: 831-1.0088884-turtle.txt
N-Triples: 831-1.0088884-rdf-ntriples.txt
Original Record: 831-1.0088884-source.json
Full Text
831-1.0088884-fulltext.txt
Citation
831-1.0088884.ris

Full Text

PERCEPTIONS OF RISK TO W A T E R ENVIRONMENTS IN T H E L O W E R FRASER BASIN, BRITISH C O L U M B I A by NIGEL STEVEN C A V A N A G H B.Sc. The University of Victoria, 1994  A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES (School of Community and Regional Planning) W e accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A November, 1998 © Nigel Steven Cavanagh, 1998  In  presenting  degree freely  at  this  the  thesis  in  partial  fulfilment  of  University  of  British  Columbia,  I agree  available for reference  copying  of  department publication  this or  thesis by  of this  for  his thesis  and  Date  DE-6 (2/88)  I further  scholarly purposes  or for  her  Columbia  requirements that  agree  may  be  It  is  representatives.  financial  permission.  The University of British Vancouver, Canada  study.  the  gain shall  not  be  that  the  for  an  advanced  Library shall  permission for  granted  by  understood  the that  allowed without  head  make  it  extensive of  my  copying  or  my written  Abstract The following thesis presents the results of a two year study that addressed lay perceptions of the risks to the water resources of the Lower Fraser Basin, British Columbia. Studies of this nature are important because by clarifying public perceptions, risk communication policies as well as land and water resource use plans that accommodate people's concerns can be developed appropriately. The study was based on a written survey that was administered to 183 lay subjects in four communities within the Lower Fraser Basin. Sixteen experts participated in a portion of the survey. Analysis involved the determination of how people perceive distinct human activities in terms of the risk each may pose to the health of aquatic ecosystems. Further analysis compared these perceptions to those of specialists in the aquatic sciences. The thesis is structured as a collection of three papers that examinedifferent aspects of the study. The First paper provides a review of the overall data set, while the remaining two papers address related subsets of the data. One paper reviews perceptions of activities that cause eutrophication problems and the other reviews perceptions of forest industry activities. The results demonstrated that people tend to view risks to water environments in a multi-faceted fashion. Four factors were found to influence people's general perception of risk and the need for regulation of the activities that were perceived to pose the risk. These factors were termed Ecological Impact, Human Benefit, Controllability and Knowledge. Another significant result was the fact that there were often striking differences between the views of the lay sample and those of the expert sample. Some activities were perceived by the lay sample as posing substantial risk while the experts did not view this to be the case. Alternatively, for some activities, the reverse scenario occurred. Accordingly, there were differences in judgements between the two groups as to the degree of regulation that should be imposed on the respective activities.  ii  PERCEPTIONS OF RISK TO WATER ENVIRONMENTS IN THE LOWER FRASER BASIN, BRITISH COLUMBIA TABLE O F CONTENTS  Page Abstract  —  ii  List of Tables  v  List of Figures  vi  Preface  vii  CHAPTER I. Introduction 1.1 Overview 1.2 Thesis Objectives- - 1.3 Thesis Study Approach 1.4 Thesis Structure  ---  1 1 2 3 3  —  CHAPTER II. Perception of Ecological Risk to Water Environments  4  2.1 Introduction---' 4 2.2Method 6 Participants --6 Questionnaire Development 7 Risk Perception Judgements: Item and Scale Development 9 Items — 9 Judgement scales9 Obtaining Expert Judgements of General Risk and Need to Regulate 12 2.3Results 12 MeanRatings 12 Differences in mean ratings among thefour lay sample groups Intercorrelations Among Scales— 14 Factor Analysis of Scale Intercorrelations 15 Risk Perception Map 18 Comparison of Ratings of General Risk and Need to Regulate among Lay and Expert Samples- - 19 Relationship Between Factors and Dependent Variables 22 2.4 Discussion and Conclusions 24 :  CHAPTER III. Perceptions of Ecological Risks Associated with Eutrophication Sources in the Lower Fraser Basin, British Columbia  27  3.1 Introduction 27 3.2Method 29 3.3 Results 30 Lay Perceptions of the Judgement Scales: Descriptive Analysis of Mean Scores 30 Agricultural waste disposal 30 Fertilizeruse 31 Septic systems 31 Sewage treatment 32 Comparison of Ratings of General Risk and Need to Regulate among Lay and Expert Samples- - 33 General risk 33 Need to regulate 34 3.4 Discussion and Conclusions— 35  iii  13  TABLE  OF CONTENTS  CONTINUED  Page C H A P T E R IV. Perceived Ecological Risks to Water Environments from Selected Forest Industry Activities- - - - -  --—  4.1 Introduction 4.2Method DataAnalysis 4.3 Results Comparison o f M e a n Scores Between L a y and Expert Samples for T w o Scales Risk in general Need to regulate Factor Analysis Results and Findings for Forestry Items- Factor Analysis Results Results of Six Forest Industry Activities Clearcut logging Selective logging Construction of logging roads Effluent from pulp mills Fertilizeruse — Pesticide use Risk Maps 4.4 Discussion and Conclusions -  37  — -  :  — — •- —  T  C H A P T E R V . Conclusion  --  REFERENCES-----  - -_'  iv  37 39 40 41 41 41 44 46 46 46 47 49 49 ^50 51 52 52 53 56  .  59  List of Tables  page  Table I. Description of the Judgement Scales and Response Categories in Order Presented in the Survey  11  Table II. Intercorrelations Among 17 Judgement Scales-  14  Table III. Rotated Factor Loadings for 15 Risk Characteristic Scales and Factor Performance Summary Data- - - - Table IV. Ten Highest and Ten Lowest Items for the Four Factors That Characterize Perceived Ecological Risk-  16 17  Table V. Lay and Expert Rating of 33 Items on General Risk to Water Environments of the Fraser Basin - -  20  Table VI. Lay and Expert Rating of 33 Items Regarding Society's Need to Regulate the Items  22  Table VII Multiple Regression Models Predicting Perception of General Riskiness and Need for Regulation Across 33 Hazards to Water Environments  23  Table VIII. Mean Scores on Each Judgement Scale for Items that are Sources of Eutrophication  30  Table IX. Mean Values of Lay and Expert Perceptions for the General Risk Associated with Each Event ' -'  34  Table X. Mean Values of Lay and Expert Perceptions for the Need to Regulate Each of the Events —  35  Table XI. Lay and Expert Ratings of the 33 Items regarding General Risk to Water Environments of the Fraser Basin  43  Table XII. Lay and Expert Ratings of the 33 Items regarding Society's Need to Regulate the Items  45  Table XIII. Overall Mean Judgement Scale Scores and Factor Analysis Scores for the Six Forest Industry Activities  47  v  List of Figures page Figure 1. The Lower Fraser Basin of British Columbia  —  Figure 2. A n Example of an Item Rated on One Scale (Controllability) Figure 3. Risk Perception M a p of Factor 1 vs. Factor 2  vi  7 -  —  8 18  Preface Versions of each of the papers that comprise this thesis appear in various peer reviewed scientific journals. A version of the second chapter of this thesis entitled "Perception of Ecological Risk to Water Environments" was published as T.L. McDaniels, L.J. Axelrod, N.S Cavanagh, and P. Slovic in Risk Analysis, Vol. 17(3): 341-352, 1997. That version was a joint collaboration by the author of this thesis and the three above named individuals. Dr. McDaniels was the primary researcher of a study from which data was used to produce this particular paper. The author of this thesis was a research associate for that study and contributed extensively to the development and administration of the survey instrument; data verification, interpretation and analysis; and, report writing and editing. A version of the third chapter of this thesis entitled "Perceptions of Ecological Risks Associated with Eutrophication Sources in the Lower Fraser River Basin, British Columbia" was published as N. Cavanagh and T.L. McDaniels in the Canadian Water Resources Journal, Vol. 22(4): 433-444, 1997. That version was a joint collaboration by the author of this thesis and Dr. T. McDaniels. The author of this thesis was the primary author of the published paper. Dr. McDaniels provided guidance and editorial input throughout several iterations of the production of the published version. A version of the fourth chapter of this thesis entitled "Perceived Ecological Risks to Water Environments from Selected Forest Industry Activities" by N. Cavanagh, T.L. McDaniels, L.J. Axelrod, and P. Slovic is due to be published in an upcoming issue of the journal Forest Science. The author of this thesis was the primary author of the published paper. Drs. T.L. McDaniels, L.J. Axelrod and P. Slovic provided guidance and editorial input throughout several iterations of the production of the published version.  Signed: Dr. Timothy McDaniels  vii  PERCEPTIONS OF RISK TO W A T E R ENVIRONMENTS IN T H E LOWER FRASER BASIN, BRITISH COLUMBIA  CHAPTER I. INTRODUCTION 1.1 OVERVIEW Clearly identifying public perceptions, values and preferences and then successfully integrating these views into the planning and policy making processes are important aspects of managing human activities that pose risk to water environments. This is not a novel concept, as Sewell (1971a) provided ample guidance in this regard nearly three decades ago. He stated that it was recognized by the 'technical elite' of the time that by including the public more fully then the process of policy development would act as a means of better informing themselves as well as the public (Sewell, 1971a). Additional work by Sewell underscored the relevance of perceptions and attitudes of technical experts (Sewell, 1971b) and the potential role of attitudinal research in early public involvement efforts (Sewell, 1974). Although this shift in attitude from an alienated to included public was evident in the early 1970s, as well as the fact that public participation has taken many forms in the interim, there continues to be problems associated with obtaining unbiased public perceptions. If perceptions and preferences are not soundly identified then there may continue to be substantial differences between the views of the public and technical experts as to the potential risks associated with particular courses of action. Under these circumstances, the impacts of some activities may be under-recognized, while others may be overemphasized. Since public sentiment plays a significant role in policy development, contrasting perceptions could contribute to disparities in risk management priorities. This pattern may be of great significance in a planning context. By clarifying where public perceptions and expert views differ, we should be able to identify potential conflicts and plan where risk  1  communication efforts should be focused. Successful risk communication may help reduce controversy over subsequent risk management approaches and policy development.  The following thesis presents the results of a social science research study that attempted to gain an understanding of the perceptions people have regarding risks to water environments in the Lower Fraser Basin of British Columbia. The study utilized techniques developed for earlier perception studies of risks to human health (Slovic, 1987). Those techniques were adapted here for the purpose of applying them in an altered context, that of perceptions of risks to environmental health. The thesis consists of three chapters that address various aspects of the study. Versions of each of these chapters appear as papers in recent or upcoming peer reviewed journals.  1.2 THESIS OBJECTIVES The three primary objectives of the thesis are: (1) To clarify how lay people "perceive" (think about and judge) various kinds of events or activities, in terms of the risk that each may pose to the health and productivity of water environments; (2) To compare lay views to the views of experts on the matters of general risk and regulation requirements; and, (3) to consider implications for policy development and water resource planning.  2  1.3 T H E S I S S T U D Y  APPROACH  The thesis reports the results of various aspects of a social science research study conducted at the University of British Columbia in 1996. The study was based on a written survey that was administered to 183 lay subjects in four communities within the Lower Fraser Basin. Sixteen experts participated in a portion of the survey. The survey involved respondents' ranking items (generally human activities) on a series of seven point scales. Each scale was a characteristic of the items that may influence the judgement of risk. Statistical tests (t-tests) were conducted to determine if there were significant differences in response patterns among respondents from the four communities and between lay and expert views. One chapter reports the results of the overall study while the two others report the results of subsets of the overall data.  1.4 T H E S I S  STRUCTURE  Each of chapters 2, 3, and 4 are research papers that have their own introduction, problem statement, methods, results and conclusion. Chapter 2, is entitled "Perception of Ecological Risk to Water Environments". It reports the overall results of the survey and outlines the research framework. Chapter 3 is entitled "Perceptions of Ecological Risks Associated with Eutrophication Sources in the Lower Fraser River Basin, British Columbia". Chapter 4 is entitled "Perceived Ecological Risks to Water Environments from Selected Forest Industry Activities". These chapters report the results of focused subsets of data taken from the overall study. Each of chapters 3 and 4 demonstrate how valuable information about perceptions of related activities can be drawn from a survey of this scope. Chapter 5 is the concluding chapter that discusses common findings, policy and planning implications, and possible further research efforts.  CHAPTER II. PERCEPTION OF ECOLOGICAL RISK TO W A T E R ENVIRONMENTS  2.1  INTRODUCTION Managing risks associated with water resources has been a fundamental concern  of humans since the birth of structured societies. Traditionally, the hazards receiving attention have been risk of unhealthy or inadequate supplies of water for municipal or agricultural uses, while the responses to these hazards have stressed engineering solutions based on technical judgements. Over the last three decades, the public debate associated with water resources has broadened dramatically. One important change has been increased attention to ecological risks in virtually all public policy contexts, including water management. Ecological risks are defined here as threats to the health and productivity of species and natural environmental systems. A second important change has been the recognition that risk management for water (and for all other issues) is not simply a scientific or technical enterprise, but rather is profoundly shaped by social judgements, attitudes, and values, as well as organizational and political processes (Slovic, 1992; O'Connor, et al, 1994). This chapter explores the intersection of these two important changes in the social interplay over water resource management. It analyzes lay and expert perceptions of the ecological risks associated with a range of human activities that could adversely affect water resource environments. To accomplish this task, the research draws from methods originally developed to characterize public perceptions of human health risks from technologies (Slovic, 1987). The psychometric paradigm characterizing perceptions of health risk from technologies (and applied here to ecological risks associated with water resources) involves three component activities: •  clarifying the items (i.e., risky activities) and scales (i.e., characteristics of activities) relevant for the study;  4  •  having subjects fill out a survey that involves rating the items on each of the scales;  •  applying multivariate statistical analysis to clarify the underlying factors  that  characterize variation in the responses.  The present chapter reports the results of the second of two closely related studies. Both studies employed the psychometric paradigm to characterize perceptions of ecological risks. The first study, published earlier, had participants (n = 68) rate each of 65 possible ecological hazards (items included technologies, human activities, natural hazards, and social, political and economic beliefs or systems) on 30 characteristics and one scale regarding general risk to natural environments (McDaniels, et al., 1995). A factor analysis of this data identified five general factors that contribute to the formation of risk judgements in lay subjects. The first factor, termed Impact  on Species,  reflected  concerns that people have regarding impacts of the hazards on nonhuman species and was positively correlated with overall perceived ecological risk. The second factor combined characteristics representing benefits to humans derived from hazardous activities. This factor, termed Human  Benefits,  was negatively associated with perceived  risk. Thus, the more a given risk item was perceived as a source of benefits to humans, the less risky the hazard was judged. The third factor, termed Impact  on  Humans,  included characteristics such as the number of people affected by the item, the scope of the impact, and the relevance to individuals' lives. The fourth factor reflected the Avoidability  that Knowledge  or controllability of the event. The fifth factor recognized the influence of the risk item has on risk judgements (McDaniels, et  al.,  1995).  In the second study, reported in this chapter, we sought to extend the initial results from the first survey. Three specific questions were of interest in designing the second study. First, would the factor structure outlined previously be replicated when applied in a more restricted topic area? Second, would the perceptions of a more diverse  5  sample, which included local community residents (as opposed to a sample that included only university students), provide a similar structure? Third, would the structure be of use in understanding the public's desire for the regulation of water-related ecological risks? In response to Question 1, we tested the structure by examining lay perceptions of ecological risk to water environments. To explore the second question, we collected information from four different groups, three of which were residents of different communities in the lower mainland of British Columbia. Finally, in response to the third question, we decided to obtain judgements on "general risk" to water environments, as well as public perceptions regarding the "need to regulate" risks. We also collected data regarding expert judgements for the "general risk" and "need to regulate" scales in order to compare lay and expert views.  2.2 METHOD Participants Participants in the study were 102 women and 81 men recruited from three residential communities in the Lower Fraser River Basin region and the student population of the University of British Columbia (n = 47) (Figure 1). The residential communities included Burnaby, a suburb of Greater Vancouver located at the western end of the Lower Fraser Valley (n = 49); Langley, a mixed rural/urban area located centrally in the valley (n = 49); and, Abbotsford, a primarily rural area located near the eastern end of the valley (n = 38). Ages of participants ranged from under 20 to over 60. The majority of the students in the sample (72%) were under the age of 25 whereas the majority of the participants from the residential communities (63%) were older than 25 years of age. The sample was recruited through advertisements seeking individuals interested in a few hours of paid work filling out a survey. Participants could be reasonably viewed as a (self selected) random sample in the three communities, in the same manner as all surveys rely on self-selected participants. Participants in the student  6  group were widely distributed in terms of age and ethnic background, and had no particular environmental orientation or training. They responded to advertisements in the student employment office. Participants required between 1 and 2 hours to complete the survey instrument and were paid $10.00 to $20.00, depending on the time taken.  Figure 1. The Lower Fraser Basin of British Columbia Questionnaire Development The survey instrument had four sections. The section of focus for this chapter addressed how people "perceive" (think about and judge) various kinds of hazards (items), in terms of the risk that each may pose to the health and productivity of water environments within the Lower Fraser Basin of British Columbia. The Lower Fraser Basin extends approximately 150 kilometers (90 miles) from highly urbanized  7  metropolitan Vancouver through small rural communities to the eastern end of the Lower Fraser Valley. The population of the region exceeds 1.8 million. Participants were instructed to rate survey items (i.e., possible ecological hazards) in terms of judgement scales (i.e., characteristics of the hazards that may influence risk judgements) that ranged from 1 to 7, with 4 being the midpoint. An example of one item rated on one scale is shown in Figure 2. A map showing the Lower Fraser Basin was provided; "water environments" was defined as the plant and animal life associated with a specific water body; a brief definition for each survey item was provided. The purpose was to ensure consistent understanding of terms among the survey subjects. A total of 33 items were rated on each of 17 scales. Subsequent sections asked participants about: (a) their perceptions of the water quality of nine water bodies located within the Fraser Basin of British Columbia, (b) their "worldviews" regarding the environment and environmental protection, and (c) demographic information (i.e., age, gender, type of residence, income).  Characteristic: Controllability  "Please rate how 'controllable' each item is, in terms of our ability to control its impacts on water environments in the Fraser Basin" NOT AT ALL CONTROLLABLE Item Acid rain  1  2  VERY CONTROLLABLE 4  3  5  6  .7  Figure 2. An example of an item rated on one scale (controllability)  8  Risk Perception Judgements: Item and Scale Development As a first step in developing the risk perception section of this survey instrument, a focus group was held to identify (a) human practices and activities (termed  items)  that  may be perceived as risks to the health and productivity of water environments, and (b) scales  reflecting characteristics of these items that may influence the judgement of risk.  Experts (individuals working in the aquatic sciences at the University of British Columbia) were selected to participate in the focus group to facilitate the selection of survey items based on their informed understanding of risks to water environments. In addition, the experts provided input regarding which of the scales used in the first ecological risk perception survey (McDaniels, et al., 1995) were appropriate for use in the current study given its narrower focus (water environments of the Fraser Basin).  Items Dozens of human activities and natural events were identified as posing some level of risk to the water environments in the Fraser Basin. Thirty-three of these items were selected for inclusion in the study. The vast majority were either direct human practices (e.g., commercial fishing, urban development, motor boating) or consequences that occur indirectly as a result of human endeavors (e.g., acid rain, leachates from landfills, loss of fish habitat). Two items were natural phenomena (drought and flooding). Items were included that were reasonably expected to be rated as not very risky (e.g., sports fishing; canoeing, kayaking and rafting) as well as items thought likely to pose substantial risk (e.g., acid rain; increased ultra violet radiation due to ozone depletion). The entire set of items is listed later in Table V, when discussing the results.  Judgement Scales The set of judgement scales used in this study was largely drawn from 30 characteristics examined in our initial survey, which attempted to broadly characterize  ecological risk perception (McDaniels, et al., 1995). Fifteen scales were selected to represent the diversity of those found previously to characterize ecological risk perceptions. The focus group was used to help specify which individual scales were most relevant to the formation of judgements ofriskto water environments, and to tailor them for clarity in this specifically defined domain. To test the reliability of the factor structure found in McDaniels, et al. (1995), we included a sample of scales from each of the five factors that were identified in the original study.  The set of scales in this study included characteristics such as: (a) impact on species (i.e., species  loss, rights  (b) human benefits (i.e., benefits  of non-human  people  affected  hazard  and availability  known  by experts,  impacts).  and risk  to human  to your  observability  Finally, the general  of the impacts,  of of number  impacts), benefits), of  controllability (i.e., controllability  and (e) knowledge of impacts (i.e.,  of impacts, risk posed  of the distribution  life, scope  health),{d)  of alternatives),  and reversibility  and equity  to society  (c) impact on humans (i.e., relevance  species,  predictability  by each  item  water environments in the Fraser Basin" and the need  of impacts  of impacts  and immediacy  of  to the "health and productivity of for  society  to regulate  each  item  was assessed. The complete set of rating scales is presented in Table I.  The reasoning for reducing the number of items and judgement scales, compared to the original study, is straightforward. We felt it imperative that the survey take less time to complete than the original study (which required up to 3 hours) in order to obtain reliable data from a larger, more diverse sample.  10  Table I. Description of the Judgement Scales and Response Categories in Order Presented in the Survey  Scale end points Scale  Description of scale  Please rate how much you feel is known by experts about the harmful impacts of each event on water environments within the Fraser Basin Controllability Please rate how controllable each event is, in terms of our ability to control its impact on water environments in the Fraser Basin Please rate the scope of the impacts of each event, in terms Scope of the area of water resources potentially affected within the Fraser Basin Observability Please rate how observable are the impacts of each event on water environments in the Fraser Basin Please rate how relevant each event is to your life in terms Relevance of its impact on water environments in the Fraser Basin Please rate how you think the event provides benefits to Benefit society People affected Please rate how many people are, or could be, affected by the impact each event may have on water environments in the Fraser Basin Please rate the impacts of each event in terms of its Species loss potential for causing loss of animal or plant species within the Fraser Basin ' Please rate the equity of each event, in terms of whether Equity those who receive the benefits from the event (or the practices that lead to the event) are the same respondents who incur the costs Please rate the immediacy of each event, in terms of how Immediacy soon possible harmful effects may occur to water environments in the Fraser Basin Please rate to what extent each event infringes on the Rights rights of non-human species within the Fraser Basin Please rate the extent to which impacts on water Reversibility environments are reversible (i.e., the ability of the water environment to return to pre-event conditions) Alternatives Please rate the extent to which there are alternatives to each event, or the practices that lead to the event Predictability Please rate how predictable are the impacts of each event on water environments in the Fraser Basin Risk to human Please rate the extent to which each event and its impacts on the water environments of the Fraser Basin pose a risk health to human health Risk (general) Please rate how risky in general you think each event is, in terms of its impacts on the health and productivity of water environments in the Fraser Basin Please rate the extent to which you feel the event (or Need to practices that lead to the event) requires regulation by regulate society Knowledge  11  Low(l)  High(7)  Very little known  Great amount known  Not at all controllable  Very controllable  Small area  Large area  Not at all observable No relevance  Very observable Direct relevance  No benefit  Great benefit  Very few people  Great number of people  Low potential for species loss  High potential for species loss  Not equitable  Very equitable  Occurs immediately  Occurs far in the future  Does not infringe Completely irreversible  Greatly infringes Completely reversible  Alternatives not available Not at all predictable Poses no risk  Alternatives are available Very predictable Poses great risk  Poses no risk  Poses great risk  Requires no regulation  Requires strict regulation  Obtaining Expert Judgements of General Risk and Need to Regulate Expert judgements were also elicited for two scales from the questionnaire for comparison with the judgements of the lay sample. The two scales, general need  risk  and  were selected because they are related directly to risk management  to regulate,  decisions. The expert sample consisted of sixteen individuals who work in fields associated with aquatic sciences. Eight are professors at the University of British Columbia. The remaining eight are resource management professionals at the Water Quality Branch of the B C Ministry of Environment, Lands and Parks. The objectives of the survey were explained to each expert and directions for completing the scales provided. Each expert completed the two scales in the same format presented to the lay sample. The data collected from experts were used only to make comparisons between experts and the lay sample for these two scales. Thus, results of analyses reported in this article are based on the data set collected from the lay sample, unless otherwise indicated.  2.3 RESULTS Mean Ratings The first step in analyzing the data was to create a data matrix of mean responses over all individuals, for each combination of scale and item. Several scales had means well dispersed from the scale midpoint of 4. The most extreme mean was for need to regulate  (M=5.36), followed by rights  people  affected  benefit  to society  of non-human  species  (M=5.15), and number  of  (M=4.94). At the opposite end of the spectrum was perceptions of (M=3.08). In general, participants perceived the items as having  detrimental effects on the environment and needing stringent regulation, while being of limited benefit to society.  12  Turning to the variability of responses across items, the highest standard deviation was observed for benefit  to society  (1.33), suggesting the existence of a large  range of viewpoints on this dimension. The next most variable was (1.03), followed closely by controllability need  to regulate  (0.98), number  risk  of people  to human  affected  health  (0.95) and  (0.95). Two scales had little variability between responses relative to  the other scales. These were predictability  (0.41) and knowledge  of impacts  (0.42).  Differences in mean ratings among the four lay sample groups Before conducting further analyses, we checked for possible differences among the four sample groups in the lay sample. Analysis of variance procedures using Tukey HSD post-hoc comparison tests were conducted to compare the overall mean responses (averaged over the 33 items) for each judgement scale. Significant differences among groups were found for only two scales. Residents from the three communities rated the items as more relevant  to their lives than did the students (F = 4.34,  also rated the items as posing slightly more risk  to human  health  .01). Residents  (F = 2.21, p < .05).  We also looked, in detail, at perceptions for two scales: general regulate.  p <  risk  and need  to  Marginal group differences were found for only three items. Not surprisingly,  residents of the developing community of Langley rated "urban development" as slightly less risky (M = 4.80) than did residents of the more rural community of Abbotsford (M = 5.57, F = 2.38, p < .08). Langley residents (M = 6.02) rated "disposal of liquid products in sewers" as more risky than students (M = 5.43, F = 2.83, p < .05). Finally, residents of the urban community of Burnaby rated "water demands for irrigated agriculture" as more risky (M = 4.96) than did the residents of Langley (M = 4.06), many of whom actively farm. No other differences were found for general differences were found for perceptions of need  13  to regulate.  risk  The overwhelming  and no  congruence among the four sample populations provided empirical support for using the complete sample as a single unit in examining the structure of risk judgements.  Intercorrelations Among Scales Table II presents the intercorrelations among mean ratings for all 17 scales. Some scales show high correlations while others exhibit little or no correlation. The most highly correlated scales were rights of non-human species and species loss (r = .95), people affected and relevance (r = .94) and, general risk and risk to human health (r = .94). Examples of scales that demonstrated no association include knowledge of impacts and availability of alternatives (r = .00) and predictability of impacts and benefits (r =.00). Table II. know cont scope observ rel ben people spelos equity mimed rights revers altem . prdict hhrisk risky reglat  know -.19 .16 .44 .23 .22 .21 .04 .32 -.40 .08 .30 .00 .69 .07 .08 .03  cont --  -.53 -.18 -.45 .57 -.63 -.46 37 31 -.22 .53 .76 .25 -.59 -.52 .23  scope  -  .48 .92 -.40 .93 .71 -.04 -.41 .65 -.72 -.15 .34 .76 .81 . .44  Intercorrelations Among 17 Judgement Scales  observ  —  .53 -.11 .50 .56 -.02 -.61 .55 -.41 -.07 .68 .30 .47 .24  ben  —  -.38 .94 .69 .00 -.54 .65 -.71 -.10 .47 .80 .83 .52  ~  -.46 -.69 .82 .52 -.62 .52 .00 .00 -.75 -.72 -.30  people spelos equity inurted rights  —  .76 . -.12 -.57 .67 -.75 -.27 .40 .85 .87 .43  — "  -.54 -.69 .95 -.79 .06 .47 .82 .92 .59  —  .48 -.48 .31 -.05 .12 -.47 -.48 -.19  —  -.67 .36 .04 -.55 -.63 -.70 -.39  —  -.77 .31 .57 .76 .89 .78  revers  — '  .06 -.17 -.72 -.80 -.57  altem  prdict  -.35 -.07 .01 .61  .28 .44 .56  —  hhrisk risky  reglat  —  .94 .51  .65  Note, know = Knowledge; cont = Controllability; scope = Scope; observ = Observability; rel = Relevance; ben = Benefits; people = People affected; spelos = Species loss; equity = Equity; immed = Immediacy; rights = Rights; revers = Reversibility; altern = Alternatives; prdict = predictability; hhrisk = Risk to human health; risky = Riskiness (general); reglat = Need to regulate  Several scales exhibited high positive correlations with general riskiness. These included scope (r = .81), relevance (r = .83), people affected (r = .87), species loss (r = .92), rights of non-human species (r = .89) and risk to human health (r = .94). High negative correlations with general riskiness were found for benefits (r = -.72), immediacy of impacts (r = -.70), and reversibility (r = -.80). Ratings of general riskiness  14  were not correlated with perceptions regarding the availability and knowledge  of impacts  (r =  of alternatives  (r =  .01)  .08).  Factor Analysis of Scale Intercorrelations As in previous risk perceptions studies, the correlation matrix in Table II shows a substantial degree of association among many pairs of scales. To test whether a pattern of underlying dimensions exists, and determine whether a structure consistent with previous studies would be found, we performed a factor analysis of the correlation matrix.  Table III presents the summary of a principle components factor analysis with varimax rotation performed on the interrelation among the mean responses for the 15 risk characteristics. Four orthogonal factors with eigenvalues greater than 1.0 emerged from the analysis. A comparison of this structure with the five-factor structure found in McDaniels, Axelrod and Slovic (1995) reveals substantial consistency. Three of the factors (Factor 2, Human Controllability)  Benefits;  Factor 3, Knowledge  of Impacts,  and Factor 4,  contained a set of individual scales as predicted. Factor 1 appears to be  an amalgam of the Impact  on Species  and Impact  on Humans  original study. Thus, we refer to this factor as Ecological  Impact  view that humans are part of the larger ecological system).  15  factors found in the (consistent with the  Table III. Rotated Factor Loadings for 15 Risk Characteristic Scales and Factor Performance Summary Data  Characteristic  Factor 1  Scope Relevance People affected Reversibility Risk to human health Species loss Rights  .92 .90 .86 -.88 .75 .72 .70  Factor 2  Factor 3  Factor 4  .97 .86  Equity Benefit Knowledge Predictability Observability Immediacy  .90 .83 .70 -.65  Alternatives Controllability  .97 .79  Factor performance  Factor 1  Factor 2  Factor 3  Factor 4  Explained variance Eigenvalue  .504 7.552  .184 2.759  .130 1.950  .083 1.242  Conceptually, one could argue for dividing Factor 1 into two sub-dimensions (Impact on Species and Impact on Humans), which would replicate the exact structure found in the original study. Supporting this argument is the observation that species loss and rights of non-human species, two prominent scales in terms of Impact on Species, were the scales with the lowest loadings on Factor 1. In addition, these two scales had the highest intercorrelation among all pairs of scales (r = .95). In sum, the 4-factor structure derived in this study closely parallels the original 5-factor model, providing evidence for its robust nature and general reliability.  16  Factor scores for each item were computed by weighting the ratings on each risk scale proportionally to the scale's importance in determining each factor and then summing across all scales, resulting in four factor scores for each item. Table IV shows the 20 extreme items (10 highest and 10 lowest) on each factor. For the first factor, Ecological  Impact,  "climate change" and "increased UV radiation" were rated as  having the highest impact. In contrast, "paddle boating" (i.e., canoeing, kayaking and rafting) and "sport fishing" were perceived to have the least ecological impact. For the second factor, Human  Benefit,  "Sewage treatment" and "chlorination of water",  followed closely by "paddle boating" were seen as most beneficial, whereas "acid rock drainage" and "loss offish habitats" were rated as offering the least benefits to humans. "Urban development" and "clearcut logging" were the hazards considered most controllable  (the third factor), while, as would be expected, the two natural hazards (i.e.,  "floods" and "drought") were seen as the least  controllable.  In terms of  knowledge  (the fourth factor), the impacts of "clearcut logging" and "flooding" were considered to be most known and predictable. Several hazards had impacts that were considered not well known. These included "introduced species", "increased UV radiation", "climate change", "leachates from landfills" and "paddle boating". Table IV. Ten Highest and Ten Lowest Items for the Four Factors That Characterize Perceived Ecological Risk Factor 1 Ecological impact Climate change Increased UV radiation Population growth Auto emissions Urban development Acid rain Urban water consumption Pesticide use Chlorinated water Liquid waste in sewers Drought Acid rock drainage Introduced species Water exports Const, of logging roads Flooding Selective logging Motor boating Sport fishing Paddle Boating  1.77 1.72 1.57 1.46 1.19 0.84 0.70 0.60 0.59 0.47  Factor 2 Human benefits Sewage treatment Chlorinated water Paddle Boating Urban water consumption Septic systems Selective logging Water use (irrigation) Population growth Urban development Hydro-power develop.  -0.48 -0.54 -0.56 -0.59 -0.96 -0.98 -1.15 -1.44 -1.95 -2.60  Urban development Clearcut logging Liquid waste in sewers Leachates from landfills Acid rain Pulp mill effluents Drought Flooding Loss of fish habitat Acid rock drainage  1.80 1.57 1.48 1.31 1.17 1.16 1.01 0.97 0.90 0.80  Factor 3 Controllability Urban development Clearcut logging Water exports Pesticide use Liquid waste in sewers Loss of fish habitat Auto emissions Commercial fishing Pulp mill effluents Motor boating  -6.46 -0.76 -0.84 -1.16 -1.21 -1.35 -1.36 -1.37 -1.47 -1.76  Sewage treatment Septic systems Population growth Paddle Boating Urban water consumption Water use (irrigation) Increased UV radiation Climate change Flooding Drought  1.50 1.44 0.90 0.87 0.81 0.81 0178 0.76 0.63 0.59  Factor 4 Knowledae Clearcut logging Flooding Auto emissions Drought Hydro-power develop. Urban water consumption Commercial fishing Agri. waste disposal Pulp mill effluents Sewage treatment  1.71 1.62 1.13 1.10 0.87 0.68 0.57 0.51 0.51 0.50  -0.27 -0.30 -0.37 >0.48 -0.73 -0.82 -1.11 -1.36 -2.86 -2.98  Septic systems Acid rock drainage Sport fishing Water exports . Urban runoff Paddle Boating Leachates from landfills Climate change Increased UV radiation Introduced species  -0.24 -0.49 -0.54 -0.67 -0.68 -1.24 -1.39 -1.83 -2.38 -2.55  Note. The table entries are factor scores calculated using regression procedures, as described in the text.  17  Risk Perception Map The relative position of each of the 33 items in terms of the first two factors can be seen in Figure 3. The vertical axis represents Factor 1 horizontal axis represents Factor 2 (Human  Benefits).  (Ecological  Impact);  the  Items at the extreme bottom of  Factor 1 are judged to have very little impact, whereas items near the top are perceived as having a high impact. Items at the far right along the horizontal dimension are seen as being very beneficial to humans, whereas items on the far left have been rated as offering little or no benefit.  Factor 1 Ecological impact •Climate change  UV-radiation*  • Population growth • Auto emissions  1.5  •Urban development • Acid rain  0.5  Urban consumption*  Pesticide use*  •Chlorination  Liquid waste*  Sewage _ °f * treatment* .Shoreline •Fertilizer use *Hvdro-development development •Commerical •Irrigation Mill effluents* Agricultural waste* fishing *Septic systems Acid, •Drought , •Water exports rock species drainage Flooding* Logging roads* Selective logging T anrifill lcarhatpi  U  r  b  a  n  r u n  f  Landtill hjacnate^ Loss of* Clearcuttin^ nsn namtat  .0.5  I n t r o d u c e d  1  Motor boating*  -1.5  • Sport fishing  -2  -2.5  Paddle boating*  -3 -2  -1.5  .  -1  -0.5  0  0.5  1  1.5  Figure 3. Risk Perception Map of Factor 1 vs. Factor 2  18  2  Factor 2 Human benefits  Items in the upper left quadrant represent those hazards that are perceived as most associated with ecological risk in that they have high ecological impact while offering little benefit to society. These hazards include "acid rain", "increased U V radiation from ozone depletion" and "loss of fish habitat". "Clearcut logging" and "pesticide use" also appear in this quadrant. In the upper right quadrant are hazards that have high ecological impact, but also present high benefits, and thus entail major environmental conflict. These include "sewage treatment", "chlorination of water" and "climate change". Surprisingly, "population growth" was also rated as somewhat high on the benefits factor. The lower left quadrant contains those items for which there is low ecological impact and low benefits. These include the two natural hazards, "flooding" and "drought", as well as "acid rock drainage". The last quadrant, lower right, contains the items that are perceived as least problematic in terms of ecological risk. These items include "paddle boating", "sport fishing" and "selective logging".  Comparison of Ratings of General Risk and Need to Regulate among Lay and Expert Samples Table V presents the 33 items ordered in terms of their mean rating of general risk to water environments of the Fraser Basin as judged by the lay sample. These mean values were very diverse, ranging from a low of 1.87 to a high of 6.10 (on a 7-point scale). In terms of specific items, canoeing, kayaking and rafting (1.87) and sport fishing (2.66) were perceived as posing little risk to water environments. Selective logging (3.42), motor boating (3.66) and introduced aquatic species (3.88) were perceived as posing moderate risk. Conversely, acid rain (6.10), air emissions from automobiles (5.98), increased ultra-violet radiation due to ozone depletion (5.74) disposal of liquid waste in sewers (5.64) and climate change (5.62) were perceived as posing the highest risk to water environments.  19  •  Table V . L a y and Expert Ratings of 33 Items on General Risk to Water Environments of the Fraser Basin  Lay Sample Expert Sample Item ' Mean Rank Mean Rank Acid rain 6.10 1*** 3.56 30 Air emissions from automobiles 5.98 2** 5.31 5(tied) 3*** Increased ultra-violet radiation due to ozone depletion 5.74 4.31 20 Disposal of liquid waste products in sewers 5.64 4 5.19 9(tied) Climate change (e.g., global warming) 5.62 5 5.06 12 Pesticide use 5.60 6 5.00 13(tied) Clearcut logging 5.56 7 5.19 9(tied) Effluents from pulp mills 5.51 8 5.13 11 9* Population growth 5.44 6.19 1 Acid rock drainage from mines 5.36 10 4.13 23 Loss of fish habitat 5.36 11 5.88 2(tied) Drought 5.28 12* 3.75 27 Leachates from landfills 5.19 13** 4.19 21 Urban development 5.17 5.88 2(tied) Fertilizer use 5.05 15 5.00 13 Flooding 5.03 16* 4.06 24 Agricultural waste disposal 5.02 17 5.25 8 Urban water consumption 4.99 18 4.44 18 Urban runoff 4.91 19 5.31 5(tied) Commercial fishing 4.77 20 5.00 13(tied) Alterations of shorelines for development 4.63 21 5.31 5(tied) Water demands for irrigated agriculture 4.54 22 4.38 19 Hydro-power development (dams) 4.52 23 5.00 13(tied) Water exports 4.47 24 3.73 28 Chlorinated water 4.37 . 25 3.81 26 Sewage treatment 4.35 26 4.06 24 Construction of logging roads 4.31 27** 5.50 4 Septic systems 3.96 28 4.19 21 Introduced aquatic species 3.88 29* 4.81 17 . Motor boating 3.66 30 3.38 31 Selective logging 3.42 31 3.63 29 Sport fishing 2.66 32* 3.38 31 Canoeing, kayaking and rafting (paddle boating) 1.87 33 2.00 33 Note: Scale ranged from 1 (poses no risk) to 7 (poses great risk). Note. Items marked with asterisks had significant differences between the mean ratings of the lay and expert samples based on two-sided t-tests (***p < .001; **p < .01; *p < .05). Table V also presents the mean ratings provided by the expert sample on general riskiness and their relative ranking of items. There are a number of items that experts rated significantly different from the lay population. Experts rated "construction of logging roads", "introduction of aquatic species", "population growth", "sport fishing", and "urban development" as higher in risk than did the lay sample. In contrast, experts rated "acid rain", "acid rock drainage from mining", "drought", "flooding", 20  "ultra-violate radiation from ozone depletion", and "leachates from landfills" all as less risky than did the lay sample. Experts and lay respondents did not differ significantly in their perceptions of risk on about two-thirds of the risk items.  Table VI presents the mean ratings and relative rankings of the risk items on the need  to regulate  scale for the lay and expert samples. There is a great deal of  consistency between the two samples in terms of perceptions of need for regulation. In general, there was a strong tendency in both groups to recognize the need for strict regulation of most items. Only four items in each sample were rated at or below the scale midpoint (4.00). However, some notable differences can be observed. Experts rated "hydro-electric power (dams)", "the construction of logging roads", and "introduction of aquatic species" as needing more regulation than did the lay sample. The former two items topped the experts list in terms of requiring strict regulation, whereas they were 16th and 20th in the lay sample, respectively . In contrast, experts rated "acid rain", "air emissions from automobiles", and "motor boating" as needing less stringent regulation than did the lay respondents. Apparently, lay people perceive the ecological risk associated with certain hazards, as well as society's need to regulate some hazards, differently than experts.  21  Table V I . Lay and Expert Ratings of 33 Items Regarding Society's Need to Regulate the Items  •-  Lay Sample Expert Sample Item Mean Rank Mean Rank Air emissions from automobiles 6.56 1* 6.12 8 Clearcut logging 6.42 2.. 5.94 12 Disposal of liquid waste products in sewers 6.31 3 5.88 14 Effluents from pulp mills 6.23 4 6.33 3 Acid rock drainage from mines 6.16 5 5.81 18 Loss of fish habitat 6.09 6 6.13 7 Pesticide use 6.08 7 6.00 10 Acid rain 6.05 4.44 29 Water exports 5.98 9 6.25 4 Agricultural waste disposal 5.98 10* 5.81 18(tied) Leachates from landfills 5.97 11 5.81 18(tied) Commercial fishing 5.91 12 6.06 9 Alterations of shorelines for development 5.88 13 6.00 10 Urban development ,5.74 14 6.19 5 Sewage treatment 5.70 15 5.88 14 Hydro-power development (dams) 5.67 .16** 6.56 1 Fertilizer use 5.60 17 5.44 22 Urban runoff 5.51 18 5.94 12 Urban water consumption . 5.49 19 '5.25 . 23(tied) Construction of logging roads 5.47 20** 6.44 . '.2 Chlorinated water 5.34 21 4.75 27 Increased ultra-violet radiation due to ozone depletion 5.30 22 5.87 16 Water demands for irrigated agriculture 5.25 23 5.50 21 Introduced aquatic species 5.09 24** ' 6.19 5 Climate change (e.g., global warming) 4.98 25 , 5.00 26 Population growth 4.95 26 5.25 23(tied) Septic systems ; 4.95 27 5.56 , 20 Selective logging 4.73 28 5.19 25 Motor boating . 4.41 29** 3.38 31 Sport fishing 3.93 30 4.50 28 Flooding 3.53 31 4.00 30 Drought ... 3.31 32 2.83 32 Canoeing, kayaking and rafting (paddle boating) 2.63 33 2.56 33 Note. Scale ranged from 1 (poses no risk) to 7 (poses great risk). Note. Items marked with asterisks had significant differences between the mean ratings of the lay and expert samples based on two-sided t-tests (***p< .001; **p< .01; *p< .05).  Relationships Between Factors and Dependent Variables Next, we consider how the factors discussed earlier are associated with participant's perceptions of general  risk  as well as with need  four factors were used to predict responses for general  risk  to regulate  and need  the risks. The  to regulate  with a  multivariate regression analysis using simultaneous entry. Table VII presents the results.  •22  Virtually all of the variability in the mean perceptions of general accounted for by the four factors  by Human p <  Benefits  riskiness  risk.  Impact  were most  (Beta (regression coefficient) = .80, p < .001), followed  [Beta = -.51, p < .001] and Knowledge  .001). Perceptions regarding Controllability  of general  could be  = .96), although only three of the four had  coefficients statistically different from zero. Ratings of Ecological correlated with general  risk  of Impacts  (Beta = .24,  were not associated with perceptions  Thus, perceptions of the general riskiness of items increases as  perceptions of ecological impact increase, human benefits decrease, and knowledge of impacts increase. This pattern of associations replicates that found in the original study (McDaniels, et al., 1995). Table VII. Multiple Regression Models Predicting Perception of General. Riskiness and Need for Regulation Across 33 Hazards to Water Environments  Ecological impact H u m a n benefits K n o w l e d g e o f impacts Controllability R2  General riskiness Beta P< 0.80 .001 -0.51 . 0 0 1 0.24 .001 -0.01 ns  N e e d to regulate Beta P< 0.57 .001 -0.22 .01 0.15 .07 0.66 .001  0.96  0.80  The 4-factor model accounted for less variance for need to regulate contrast to general riskiness, Controllability need  to regulate  (R^ = .80). In  was the factor most predictive of ratings of  the hazard items. Specifically, hazards seen as more controllable were  also rated as needing stricter regulation. The participants appear to recognize that the ability to control a hazard is an important (although not essential) component in pursuing societal regulation of that hazard. This point must be tempered by recognizing that the participants' assumptions regarding the controllability of a hazard may not be accurate. For example, participants rated "climate change" and "increased U V radiation from ozone depletion" as uncontrollable. Yet, numerous interventions do exist in the  23  attempt to control these hazards, although the potential degree of control possible is unknown. Perceptions of Ecological  Impact  were also highly associated with a need to  regulate the hazards. Thus, items with higher impacts were likely to also be rated as in need of strict regulation. Perceptions of Human with need to regulate'and  Knowledge  Benefits  of Impacts  had a moderate association  was only marginally associated with  the perceptions of regulation.  2.4 D I S C U S S I O N A N D C O N C L U S I O N S Results from the present study provide strong empirical support for the factor structure found in the original ecological risk perception work (McDaniels, et al., 1995). Several factors appear to characterize lay perceptions of ecological risk in general, and risks to water resources in particular. The single most important factor found to underlie ecological risk perceptions is an assessment of the impacts to species that result from a hazard, which includes risks to non-human species and to humans.  Although impacts are a central component in the formation of risk perceptions, the concept of risk is clearly more multifaceted. Consistent with previous studies (Alhakami and Slovic, 1994; Gregory and Mendelsohn, 1993), people's perceptions of risk are influenced by their assessment of the benefits conversely associated with a risk item. Thus, higher benefits derived from a risk item tend to reduce the perception of general risk associated with the hazard. For example, "chlorinated water" was perceived by study participants as having a relatively high (9th highest) ecological impact. It also was rated as offering high human benefits (2nd highest). The high rating of benefits likely offset the high rating of ecological impact, resulting in a reduced level of general risk (25th). The extent to which a hazard is controllable has been identified as a relevant predictor of risk perception.  The factor structure also proved to be valuable in understanding lay participants' beliefs regarding regulation of the risk items. The four factors were able to explain about 80 percent of the variability in participants' ratings of the need to regulate Specifically, ratings oi Ecological  Impact,  Human  Benefits,  significantly predictive of ratings of need to regulate, they are in terms of general  risk.  risk items.  and Knowledge  were all  although to a lesser degree than  Of the four factors, ratings of Controllability  most highly associated with need to regulate.  were the  Thus, as one would expect, the extent to  which a risk item is believed to be controllable appears to be an important influence on judgements of society's need to regulate a hazard.  s•  These findings can have dramatic implications for risk management activities associated with water resources, particularly if the public's perception of  controllability  is less than accurate. For example, risks such as climate change, increased U V radiation, water demands for irrigated agriculture and population growth were rated as low in controllability  in the current study. The perception that they are not controllable  could  lead to communal inaction regarding the hazard, or possibly denial that the hazard exists, even if strategies for risk management have been identified. Recognizing the salience of perceptions of controllability  can aid risk managers in communicating effectively with  the public about tactics to minimize a risk, and in implementing risk management programs that require community participation.  The present study also indicates potential divergences in perceived risk between the views of lay people and experts that could lead to controversy over risk management priorities. These divergences also suggest areas where dialogue between experts and lay groups is needed. For example, Tables V and VI show some substantial differences in perceptions on specific items. Experts see construction of logging roads as high in general risk and need for regulation, while the lay public sees them as much less  25  so. The lay public perceives acid rain as very high in general risk and need for regulation, reflecting the high degree of attention and concern over acid rain in Eastern Canada and elsewhere. Yet, in the Lower Fraser Basin, which is the area of interest in this study, experts rank acid rain as low in general risk and need for regulation. Finally, experts see introduced species as relatively high in general risk and need for regulation, while the lay public does not. These differences may be one of the keys to anticipating conflicts on risk management priorities.  26  CHAPTER m. PERCEPTIONS OF ECOLOGICAL RISKS ASSOCIATED WITH EUTROPHICATION SOURCES IN T H E LOWER FRASER RIVER BASIN, BRITISH COLUMBIA  3.1 INTRODUCTION The focus of this chapter is on public perceptions associated with human practices that lead to cultural (human-induced) eutrophication of aquatic environments. Eutrophication is defined as the process  of increasing  nutrient  content  within  a  water  body over time (Wetzel, 1983). In conditions not influenced by human activity, the nutrient concentrations of most water bodies tend to fluctuate naturally over geological time frames. Under these natural circumstances, nutrient fluctuations impart a normal contribution to the trophic structure of any aquatic ecosystem. Yet, human activities that cause direct input of nutrients (particularly phosphorus) to fresh water, commonly accelerate the rate of eutrophication beyond that to which the existing ecosystem can reasonably adapt in the short term. Ample research has demonstrated that when nutrient input is excessive, the dynamic equilibrium of an aquatic ecosystem is drastically offset until another, often entirely different equilibrium is attained (Dillon, P.J., et al., 1986; Hart, W.C., et al, 1978,1984; Janus, L . L . , et al, 1981; Kortmann, R.W., 1988,1994; Rast, W.,et al., 1978; Reckhow, K . H . , et al., 1980,1980; Waller, D.H., et al, 1977,1981; Wetzel, 1983). Until the new equilibrium is reached, conditions can prevail that have profound effects on intraspecific population structure (Casselman and Harvey, 1975). Rapid eutrophy leads to greatly accelerated rates of photosynthetic activity (algae and near-shore weed production) and water column oxygen depletion. Extreme oxygen depletion (anoxia) occurs after blooms of floating algae sink to the lake bottom and begin decaying. Under these circumstances, extensive die-offs of fish and other aquatic organisms often occur (Barica, 1975).  27  Human uses of a water resource are also affected by the abrupt change in trophic structure. Water drawn from eutrophic lakes or streams becomes more problematic for use as domestic water supplies. The prolific concentrations of algae add to the cost of treatment and can significantly alter the taste and odour (Canadian Council of Resource and Environment Ministers (CCREM), 1987). Additionally, many taxonomic phyla of algae that favour eutrophic conditions (e.g., Cyanobacteria, Chrysophyta, and Pyrrhophyta) are toxic and can cause illness or even death to aquatic life, water fowl, terrestrial wildlife, livestock and humans (Rose, 1953; Snell, 1980; Billings, 1981; Carmichael, 1981; Sykora and Keleti, 1981; Carmichael et al., 1985). Recreational use of the lakes is also affected. Aesthetic water qualities become impaired as the colour and odour of the water often changes, beaches become infested with weeds, and parasites that cause skin, ear, and eye infections occur (CCREM, 1987). Extreme eutrophication also severely damages both recreational and commercial fisheries, especially for the salmonid species which require oxygen rich waters (Casselman and Harvey, 1975).  Cultural eutrophication is a particularly interesting topic for study in terms of ecological risk perception because it can stem from several causes (some familiar, some less so). Four human practices (a subset of the 33 items considered in the broader study reported in Chapter 2) are of interest here as potential contributors to the problem. These include: (a) agricultural waste (manure) disposal; (b) fertilizer use; (c) failing on-site liquid waste disposal systems (septic systems); and (d) sewage treatment.  Two specific research questions regarding these human activities are addressed in this chapter: (1) are there differences in lay perceptions of the ecological risk associated with these activities, and, if so, what are the underlying characteristics that contribute to these differences?; (2) are there differences in the views of lay people and experts regarding the overall ecological risk associated with these activities?  28  Understanding perceptions of these sources of eutrophication is important for two reasons. First, if there are substantial differences in how the various sources are viewed by the lay public, the impacts of some sources may be under-recognized, while others may be overemphasized. Such a pattern could contribute to disparities in policies for managing the various pollution sources. Second, differences in how experts and lay people view these hazards may indicate possible future controversy about risk management priorities, and the need for risk communication efforts.  3.2 METHODS See Section 2.2 for a full discussion of the survey development and participant recruitment process.  Before beginning the survey, the respondents were instructed to review brief definitions of the items, ensuring that each respondent had the same basic understanding of what was meant by each item. Definitions provided for the four items of interest here were:  Agricultural waste disposal - land disposal of by-products (animal wastes) from farming operations;  Fertilizer use - application of minerals to soils to induce greater plant production; Septic systems - small scale (household)  treatment of human waste;  Sewage treatment - large scale physical, chemical or biological treatment of human waste.  29  3.3 RESULTS Lay Perceptions of the Judgement Scales: Descriptive Analysis of Mean Scores Table VIII presents the mean scores of judgement scales #1-15 (see Table I) for the four items (practices) that are sources of cultural eutrophication, as discussed below. Table VIII. Mean Scores on Each Judgement Scale for Items that are Sources of Eutrophication  Agricultural waste disposal Fertilizer use Expert knowledge 4.65 4.41 Controllability 4.87 4.90 Scope 4.69 4.89 Observability 4.10 4.08 Relevance 4.81 4.97 Benefit 2.67 3.76 People affected 5.04 5.02 Species loss 4.88 4.90 Equity 3:50 3.77 Immediacy 3.29 3.38 Rights 5.37 5.32 Reversibility 4.32 4.02 Alternatives 5.14 5.25 . Predictability 4.95 4.90 Risk to human health 5.03 5.09 Note: Ratings are means on 1-7 scales, n= 183  Septic systems 4.48 4.69 4.31 3.75 4.47 4.56 4.43 3.65 4.36 4.20 4.22 4.29 4.48 4.67 3.81  Sewage treatment 4.65 5.15 5.05 4.24 5.42 5.42 5.24 3.89 4.56 3.78 4.46 4.16 4.25 5.00 4.02  Agricultural waste disposal Of the four human activities reviewed here, agricultural waste disposal is viewed as having the least benefit to society (M=2.67). The respondents also judge this activity as being the least equitable (M=3.50) in terms of the distribution of the benefits and costs that it provides to society. Respondents perceive this practice as having the potential to influence the lives of a relatively large number of people (M=5.04), a perception that is reflected in their view that agricultural waste poses somewhat of a risk to human health (M=5.03). The respondents feel that the current practice of disposing of agricultural waste seriously infringes on the rights of non-human species (M=5.37).  30  Accordingly, they discern that there exists a potential for species loss associated with this practice (M=4.88). Respondents maintain a view that deleterious impacts that could potentially occur would do so within a relatively short time frame (M=3.29). Although they feel that there is only moderate expert knowledge about the impacts of agricultural waste disposal on water environments (M=4.65), they are fairly certain that these impacts are predictable (M=4.95) and that alternatives for the practice do exist (M=5.14).  Fertilizer use The respondent's perceptions regarding fertilizer use generally mirror their perceptions of agricultural waste disposal. Although they view the use of fertilizers as having somewhat more benefit to society (M=3.76, as opposed to 2.67 for agricultural waste disposal), they perceive this activity as having comparable, fairly high impacts in terms of the number of people affected (M=5.02), risk to human health (M=5.09), species loss (M=4.90), and infringement on the rights of non-human species (M=5.32). Once again, respondents judge this activity as being relatively poor in terms of equity (M=3.77), implying that those who receive the benefits of fertilizer use are not those who incur the costs of the activity. Of the four sources of eutrophication reviewed here, the impacts of fertilizer use are considered to be the least reversible (M=4.02) while it is deemed to be the practice for which alternatives are most readily available (M=5.25).  Septic systems In general, respondents have moderate views regarding risks associated with septic systems (values only marginally range about the mid-point of the scale, e.g., 3.65 4.69). This pattern likely occurs because, of the four practices, septic systems are viewed as being the least observable in terms of the impacts they impose on water environments (M=3.75) and these impacts are judged to occur farthest into the future (M=4.20 versus  31  3.29 for agricultural waste disposal, 3.38 for fertilizer use, and 3.78 for sewage treatment). Furthermore, respondents perceive that septic systems have less effect in terms of the scope (total area) of water resources potentially affected (M=4.31 versus 5.05 for sewage treatment). This judgement reflects the perceptions that the potential for species loss (M=3.65), the number of people affected (M=4.43), and the risk to human health (M=3.81) are all moderate relative to the other three practices. Conversely, the impacts associated with septic systems are viewed as being the least controllable (M=4.69) and least predictable (M=4.67) of the sources of eutrophication.  Sewage treatment Of the four practices, sewage treatment easily surpasses the others in terms of the perceived benefit that it offers society (M=5.42 versus 4.56 for septic systems, 3.76 for fertilizer use, and 2.67 for agricultural waste disposal). Accordingly, it is viewed as being the most equitable of these activities (M=4.56). The respondents judge that there exist fewer alternatives for the disposal of human waste [sewage treatment (M=4.25) and septic (M=4.48)] than exist for agricultural waste disposal (M=5.14) and fertilizer use (M=5.25). These perceptions, in conjunction with the views that the potential impacts associated with sewage treatment are controllable (M=5.15) and predictable (M=5.00), appear to provide a rationale for a practice which the respondents recognize can have impacts that occur rapidly (M=3.78), affect large areas (M=5.05) and affect many people (M=5.24). By acknowledging that sewage treatment is highly beneficial while, at the same time potentially severe in terms of the scope of its impact, the respondents consider sewage treatment as highly relevant to their lives (M=5.42). As demonstrated in the overall study (McDaniels, et al., 1997) and in other work (Alakahmi and Slovic, 1995), there is an inverse relation between respondents' judgements of perceived benefit and perceived hazard, with greater benefit leading to lower risk perception all else being equal.  32  Comparison of Ratings of General Risk and Need to Regulate among Lay and Expert Samples As a first step in comparing the lay and expert samples for each item, a statistical analysis was conducted to determine if there was a general tendency for one group to perceive risk in general, and the need to regulate, more than the other. A t-test was used to test the hypothesis that there exists no difference between the means (of all 33 items) for lay and expert perceptions of 'general risk'. The same test was conducted for the judgement scale 'need to regulate'. The mean value (M=4.786) for all items of the lay sample for the scale 'general risk' was determined not to differ from the mean value (M=4.575) for all items of the expert sample (ot=0.05). The mean value (M=5.370) for all items of the lay sample for the scale 'need to regulate' was also determined not to differ from the mean value (M=5.421) for all items of the expert sample (a=0.05). Therefore, any difference detected between lay and expert perceptions for any one item can be attributable entirely to that item (as opposed to population differences).  General risk Table IX presents the mean ratings of the eutrophication sources on the 'general risk' scale for the lay and expert samples. While there were no statistically significant differences between the two groups, two trends can be observed. Both groups concur that agricultural waste disposal and fertilizer use pose a relatively high level of risk to the health and productivity of water environments in the Lower Fraser Basin. The standard deviation of means for these two items is slightly lower (not significantly) for the expert sample, implying that there is somewhat more agreement among the experts than the lay people.  Sewage treatment and septic systems are generally perceived by both groups to pose less threat to water environments than do agricultural waste disposal and fertilizer  33  use. On the other hand, the higher standard deviations (compared to those for agricultural waste disposal and fertilizer use) for the expert sample imply less agreement among experts regarding the severity of the impacts associated with septic and sewage treatment. The expert sample appears to view septic systems as marginally more risky, and sewage treatment as marginally less risky, than do the lay sample, although the differences are not statistically significant. Moreover, the general risk associated with septic systems is the only case for which there exists greater standard deviations (and by implication, less agreement) among the experts than among the lay sample. The implication is that some experts judge septic systems to be a significant risk while other experts do not. Also, the experts, on average, judge that septic systems are slightly more risky than sewage treatment.  Table IX. Mean Values of Lay and Expert Perceptions for the General Risk Associated with Each Event  Agricultural waste disposal Fertilizer use Septic systems Sewage treatment  Lay perceptions Expert perceptions ()=standard deviation ()=standard deviation 5.02 (1.46) 5.25 (1.00) 5.05 (1.44) 5.00(1.21) 3.96 (1.47) 4.19(1.56) 4.34(1.56) 4.06(1.53)  Note: These values are means on a scale of 1-7, where 1 represents 'poses no risk' and 7 represents 'poses great risk'  Need to regulate Table X presents the mean ratings of the eutrophication sources on the 'need to regulate' scale for the lay and expert samples. There is general consistency between the two samples in terms of their perceptions that stringent regulations are needed for human activities that contribute to eutrophication. One notable, statistically significant, difference can be observed. The expert sample rates septic systems as more in need of regulation than does the lay sample (rxO.10). This difference is understandable given the expert sample appears to view septic systems as somewhat more 'risky' than does the  34  lay sample. The standard deviations indicate that the expert sample is generally more consistent in terms of their judgement of the need for regulation than their judgements about general riskiness. Also, the expert sample exhibits more agreement about regulation for all four sources of eutrophication than does the lay sample. Table X . Mean Values of Lay and Expert Perceptions for the Need to Regulate Each of the Events  Agricultural waste disposal Fertilizer use Septic systems* Sewage treatment  Lay perceptions ()=standard deviation 5.98(1.20) 5.60(1.23) 4.94(1.55) 5.70 (1.32)  Expert perceptions ()=standard deviation 5.81 (0.91) 5.44(1.09) 5.56(1.09) 5.87 (0.96)  Note:  These values are means on a scale of 1-7, where 1 represents 'requires no regulation' and 7 represents 'requires strict regulation' Note. Item marked with asterisks had significant differences between the mean ratings of the lay and expert samples based on two-sided t-tests (p < 0.10)  3.4 DISCUSSION AND CONCLUSIONS Although this chapter elaborates on a specific component of the overall study (McDaniels, et al., 1997), many of the trends and conclusions established in that study are shown to be consistent in this context. The overall study demonstrated that the single most important factor underlying ecological risk perceptions is the respondents' assessment of the impacts to species that result from a hazard. Also, impacts to humans are considered when assessing risk levels and high negative impacts are strongly associated with high perceptions of risk (McDaniels, et al., 1997). Of the four practices discussed here, agricultural waste disposal and fertilizer use were considered to create higher potential for species loss, infringe more on the rights of species, and pose greater risk to human health than did septic systems and sewage treatment. Accordingly, the former two practices were considered more risky in general than the latter two.  The present study also reinforces the conclusion that respondents' perceptions of risk are conversely influenced by their assessment of the benefits associated with a risk  35  item (McDaniels, et al., 1997). The perceived high benefit derived from sewage treatment is likely a contributor of the reduced general perception of risk associated with this practice. Conversely, the perceived lack of benefit derived from agricultural waste disposal, potentially elevates the general perception of risk associated with this practice.  This chapter identifies divergences between the views of lay and expert respondents regarding perceptions of the "overall riskiness" of and the "need to regulate" the four eutrophication sources. In general, the two groups tend to agree. However, the experts view septic systems as posing slightly more risk, and sewage treatment as posing slightly less risk than does the lay sample. These divergences, in conjunction with wide variations in views among experts regarding the severity of impacts associated with septic systems, could lead to controversy over risk management priorities. Conceivably, the variation in expert judgements could be the result of a scarcity of documented information regarding the long-term, cumulative effects from this source. Of the four sources of nutrients discussed here, septic systems are the most diffuse, non-point source, affecting surface waters from sub-surface origins over extended time frames. Given these factors, septic nutrients are perhaps the most difficult to trace. Experts are more inclined than the lay sample to advocate greater regulation of septic systems.  In sum, this study is a first step in characterizing how lay people conceptualize and judge the risks associated with various sources of eutrophication, and where their views differ from those of experts. Ongoing research and monitoring of non-point pollution, along with greater risk communication efforts, will likely be important factors in promoting effective regulation of practices that contribute to this important problem.  36  CHAPTER IV. PERCEIVED ECOLOGICAL RISKS TO W A T E R ENVIRONMENTS FROM S E L E C T E D FOREST INDUSTRY ACTIVITIES  4.1 INTRODUCTION Judging from public debate about forest land-use conflicts in the Pacific Northwest, the forest industry is seen by many observers as a source of substantial ecological risk (Kimmins, 1992). Why this is the case is easy to understand at a superficial level: visual images of clearcuts and forest processing mills provide a stark contrast to pristine mature forests. Perceptions of forest industry activities as sources of ecological risk lead to calls for increased regulation of forest practices, such as the recent British Columbia Forest Practices Code (1993) and could eventually limit the forest industry's ability to operate in many areas. It is no overstatement to say that public perceptions are one of the most important external factors affecting the future of the forest industry in North America and elsewhere.  While interested parties have, at times, undertaken standard polling to clarify public attitudes toward forest industry activities, little if any research has been conducted about the underlying factors that shape public perceptions of ecological impacts from these activities. Nor have the differences between lay and expert views about these issues been explored. These gaps are significant. A clearer understanding of the fundamental influences shaping lay perceptions and differences between lay and expert views would be helpful in many contexts: predicting and diagnosing conflicts about forestry practices, designing risk communication efforts regarding ecological impacts associated with forestry, and clarifying the public values that should be considered in making ecological risk management decisions (Slovic, 1987; National Research Council 1992; McDaniels, et al, 1997).  37  This chapter draws on the data set reported in Chapter 2 and focuses on perceptions of ecological risk from six important forest industry activities (e.g., clearcut forest harvesting, building of logging roads). In addition, these lay perceptions are compared to additional data regarding expert judgements of the overall riskiness of, and need for regulation of, the selected activities. The study focuses on ecological risks to water environments, defined in the survey as the uncertain potential for harm to the animal and plant life existing in or around water, in the Lower Fraser Basin of British Columbia. The emphasis on risks associated with water resources reflects on the increasing concern in forestry regulation for protecting water resource uses. By adopting this emphasis, we make the judgement tasks in the survey more straightforward for the lay subjects because the area of interest and impacts of concern are more clearly defined. We also narrow the range of expertise to be called upon when surveying experts.  Commercial forestry, the industry that dominates the provincial economy, is the predominant land-use at the eastern end of the Lower Fraser Basin (Figure 1) and in the mountainous areas above the flat valley floor. Moreover, important water resources in the Lower Fraser Basin are highly influenced by extensive forest industry harvesting and processing elsewhere in the Fraser Basin watershed. For example, some 55% of all industrial waste discharges to the Fraser River come from pulp mills located upstream of the Lower Fraser region (British Columbia, 1993).  The next section of this chapter briefly outlines the study methods. The following section presents the results, beginning with a comparison of lay and expert views regarding overall ecological riskiness and need for regulation of the selected forest industry activities. These views are compared to perceptions of other human activities to place the perceived riskiness of forest industry activities in context. Then, detailed  38  judgements regarding each forestry activity are considered in more detail. Finally, the multivariate statistical results are presented that place the selected forest industry activities on a two dimensional grid characterizing underlying factors shaping risk perception. Section 4 reviews the overall patterns and their implications, while the conclusion in Section 5 discusses descriptive and prescriptive interpretations of the results.  4.2 METHODS See Section 2.2 for a full discussion of the survey development and participant recruitment process.  Definitions for each item were provided to ensure that each respondent had the same understanding of what the terms meant. Definitions provided for the six forestry items of concern in this article were as follows:  Clearcut logging- timber harvesting where forested  areas are entirely cleared;  Selective logging- timber harvesting in which only the target species or age class is removed and that does not involve clearing large areas;  Construction of logging road's- building roads to provide access for logging  trucks to  remove cut timber;  Effluent from pulp mills- water containing industrial wastes that are released from pulp and paper operations;  Fertilizer use - application of minerals to soils to induce greater plant production; Pesticide use - application of chemical or biological products to eliminate  unwanted  pests. Note that fertilizer use and pesticide use were not defined as specific to forest activities, but were deliberately cast broadly. This is because lay subjects were less likely to have awareness of the use of fertilizers or pesticides in commercial forestry.  39  Data Analysis As afirststep in organizing and analyzing the data for this chapter, the 7 point (1-7) scale responses were recoded to a range of -3 to +3 with a mid-point of zero. The reason was to simplify the interpretation of the results. Next, a table of mean responses was created for every item / scale response combination, for each of the four communities from which respondents were recruited. Then a series of t-tests were conducted to see if there were significant differences in response patterns among respondents from the four communities. These tests showed no significant differences in responses from these communities, with two minor exceptions concerned agriculture and urban development. Hence, we thought it appropriate to treat the four groups as one pooled sample. All data presented here are sample means over the total respondent set.  The next step was to prepare a table of the intercorrelations of the means for each item / scale combination. The matrix indicated many of the responses were highly correlated, suggesting there may be underlying factors that explain the variation in the response patterns. In keeping with previous psychometric risk perception analysis, we conducted a principle components factor analysis with varimax rotation, to see if underlying factors shaping the responses could be identified. These results are discussed in the next section.  40  4.3 RESULTS Comparison of Mean Scores Between Lay and Expert Samples for Two Scales To begin, we report mean scores for the two most aggregate scales (risk in general  and need to regulate) for the lay and expert samples. Results are reported for  the six forestry items, as well as the other 27 items from the survey, in order to show the relative perceived riskiness of forestry activities compared to a wide array of other human endeavors.  In order to compare the lay and expert samples for each forest industry activity, a statistical analysis was conducted to determine if there was a general tendency for one group to rate overall risk (or the need to regulate) more highly than the other. T-tests were used to test the hypothesis that there exists no difference between the means (across all 33 items) for lay and expert perceptions of "riskiness in general" and the "need to regulate". The mean value (M=0.786) across all items of the lay sample for "riskiness in general" was determined not to differ significantly from the mean value (M=0.575) across all items of the expert sample (p<0.05). The mean value (M=1.370) across all items of the lay sample for "need to regulate" was also determined not to differ from the mean value (M= 1.421) across all items of the expert sample (p<0.05). Therefore, any reliable difference detected between lay and expert perceptions for any one item can be attributable entirely to that item (as opposed to population differences).  Risk in general Table XI presents the mean ratings for 33 items (including the forest industry activities) on the scale concerning "riskiness in general" for the lay and expert samples. Lay perceptions of the general risk associated with the forestry items is quite variable relative to the other items. Pesticide use, clearcut logging, and effluents from pulp mills  41  rank high (6th, 7th, and 8th, respectively). However, construction of logging roads and selective logging rank low relative to the whole suite of items presented to the respondents (27th and 31st, respectively). There were some notable differences between the expert rankings and the lay rankings for the items, as discussed below.  While there was general agreement among the two groups regarding their perceptions of ecological risk associated with each of the forestry activities, two notable, statistically significant differences did arise. One concerned perceptions of ecological risk associated with construction of logging roads (p<0.01). Experts viewed this activity as posing a serious degree of risk to water environments (M=1.50, ranked 4th overall), while the lay sample viewed this activity as posing much less risk (M=0.31, ranked 27th overall). The second difference was that the mean scores differed with respect to perceptions of ecological risk associated with pesticide use. The lay sample considered pesticides to be more of a risk than did the expert sample (M=l,60 and 1.00 respectively; p<0.05). Both groups viewed clearcut logging and effluent from pulp mills as posing a relatively high level of risk to water environments in the Fraser Basin, although the experts had slightly less extreme views. Both groups rated selective logging as the least risky of the six activities, and near the bottom of the entire set of items in terms of "riskiness in general".  42  Table XI.  L a y and Expert Ratings of the 33 Items regarding General  Risk to  Water Environments of the Fraser Basin (presented in descending order of lay perceptions) (forestry items are in bold font) Lav Sample Mean Rank 1*** 2.10  Item Acidrain Air emissions from automobiles Increased ultra-violet radiation due to ozone depletion Disposal of liquid waste products in sewers Climate change (e.g., global warming) Pesticide use Clearcut logging Effluents from pulp mills Population growth Acid rock drainage from mines Loss of fish habitat Drought Leachates from landfills Urban development Fertilizer use Flooding Agricultural waste disposal Urban water consumption Urban runoff Commercial fishing Alterations of shorelines for development Water demands for irrigated agriculture Hydro-power development (dams) Water exports Chlorinated water Sewage treatment Construction of logging roads Septic systems Introducedaquauc species Motor boating Selective logging Sport fishing Canoeing, kayaking and rafting (paddle boating)  1.98 1.74 1.64 1.62 1.60 1.56 1.51 1.44 1.36 1.36 1.28 1.19 1.17 1.05 1.03 1.02 0.99 0.91 0.77 0.63 0.54 0.52 0.47 0.37 0.35 0.31 -0.04 -0.12 -0.44 •0.58 -1.44 -2.13  2** 3***  4 5 6* 7 8 9*  10 11 12* 13** 15 16* 17 18 19 20 21 22 23 24 25 26 27** 28 29* 30 31 32* 33  Expert Sample Mean Rank -0.44 30 L31 5(tied) 0.31 20 1.19 9(tied) 1.06 12 1.00 13(tied) 1.19 9(tied) 1.13 11 2.19 1 0.13 23 1.88 2(tied) -0.25 27 0.19 21 1.88 2(tied) 1.00 13 0.06 24 1.25 8 0.44 18 1.31 5(tied) 1.00 13(tied) 5(tied) 1.31 0.38 19 1.00 13(tied) -0.27 28 -0.1? 26 0.06 24 1.50 4 '. 0.19 21 0.81 17 -0.62 31 -0.37 29 -0.62 31 -2.00 33  Scores are sample means on a scale of -3 to +3, where -3 represents 'poses no risk' and +3 represents 'poses great risk'. Items marked with asterisks had significant differences between the mean ratings of the lay and expert samples based on two-sided t-tests (***p < .001; **p < .01; *p < .05).  43  Need to regulate Table XII presents the mean ratings of all items on the scale "need to regulate" for the lay and expert samples. Lay respondents' rankings of the need to regulate forestry items are consistent with the ratings in Table XI. Clearcut logging, effluent from pulp mills, and pesticide use all rank high in terms of the need to regulate (2nd, 4th, and 7th, respectively). Construction of logging roads and selective logging rank relatively low on this scale (20th and 28th, respectively). Again, there were some notable differences between the expert rankings and the lay rankings for the forest industry activities. Lay respondents judged clearcut logging to be in need of strict regulation (M=2.42, ranked 2nd overall), while the experts were not as extreme in their view (M=1.94, ranked 12th overall, mean scores differed, p<0.10). Conversely, the experts judged the practice of constructing logging roads highly in terms of the need for regulation (M=2.44, ranked 2nd overall), while the lay respondents did not hold as extreme a viewpoint (M=1.47, ranked 20th overall, mean scores differ, p<0.01). Both lay and expert respondents agreed that pulp mill effluents and pesticide use are in need of strict regulations. Finally, they agreed that strict regulation was not as necessary for fertilizer use and selective logging.  44  Table XII.  Lay and Expert Ratings of the 33 Items regarding Society's Need to  Regulate the Items (presented in descending order of lay perceptions) (forestry items are in bold font) Lay Sample Mean Rank i** 2.56  Item Air emissions from automobiles Clearcut logging Disposal of liquid waste products in sewers Effluents from pulp mills Acid rock drainage from mines Loss of fish habitat Pesticide use Acid rain Water exports Agricultural waste disposal Leachates from landfills Commercial fishing Alterations of shorelines for development Urban development Sewage treatment Hydro-power development (dams) Fertilizer use Urban runoff Urban water consumption Construction of logging roads Chlorinated water Increased ultra-violet radiation due to ozone depletion Water demands for irrigated agriculture Introduced aquatic species Climate change (e.g., global warming) Population growth Septic systems Selective logging Motor boating Sport fishing Flooding Drought Canoeing, kayaking and rafting (paddle boating)  2.42 2.31 2.23 2.16 2.09 2.08 2.05 1.98 1.98 1.97 1.91 1.88 1.74 1.70 1.67 1.60 1.51 1.49 1.47 1.34 1.30 1.25 1.09 0.98 0.95 0.95 0.73 0.41 -0.07 -0.47 -0.69 -1.37  2* 3 4 5 6 7 g#*** 9 10** 11 12 13 14 15 17 18 19 20*** 21 22 23 24***  25 26 27 28 29*** 30 31 32 33  Expert Sample Mean Rank 2.12 1.94 1.88 2.33 1.81 2.13 2.00 0.44 2.25 1.81 1.81 2.06 2.00 2.19 1.88 2.56 1.44 1.94 1.25 2.44 0.75 1.87 1.50 2.19 1.00 1.25 1.56 1.19 -0.62 0.50 0.00 -1.17 -1.44  8 12 14 3 18 7 10 29 4 18(tied) 18(tied) 9 10 5 14 1 22 12 23(tied) 2 27 16 21 5 26 23(tied) 20 25 31 28 30 32 33  Scores are sample means on a scale of -3 to +3, where -3 represents 'requires no regulation' and +3 represents 'requires strict regulation'. Items marked with asterisks had significant differences between the mean ratings of the lay and expert samples based on two-sided t-tests (****p < .001; ***p < .01; **p < .05; *<. 10).  45  Factor Analysis Results and Findings for Forestry Items Factor Analysis Results A correlation analysis conducted on the entire data set (Table II) indicated a substantial relationship between many pairs of scales. A factor analysis of the correlation matrix identified four factors with Eigenvalues greater than 1, and provides a plausible characterization for the variations in the lay responses in the data set. Table III presents a summary of the factor analysis results. The first factor Ecological 50% of the variance in the data; factor 2, Human by 13%, and 8% for factors 3 and 4 respectively  Benefit,  Impact,  accounted for  accounted for 18%, followed  (Controllability  and  Knowledge).  A  comparison of this structure with the five-factor structure found in the study "Characterizing Perception of Ecological Risk" (McDaniels, et al., 1995) reveals substantial consistency. Three of the factors in the present study (Factor 2, Benefits;  Factor 3,  Controllability,  and Factor 4, Knowledge  of Impacts  Human  ) contained a  set of individual scales as predicted by the original 1995 study. Factor 1 appears to be an amalgam of the  Impact  on Species  and  Thus, we refer to this factor as Ecological  Impact  on Humans  Impact  factors found in that study.  (consistent with the view that  humans are part of the larger ecological system).  Results for Six Forest Industry Activities Table XIII presents the mean scores averaged over all lay respondents for i5 judgement scales for each of the forest industry activities (omitted are "riskiness in general" and "need to regulate"). Statistically significant judgement scale differences reported in the following text were determined through two-tailed t-tests (testing the mean of an item's scale against the mean score of that scale for each of the other items). Also shown in Table XIII are the four constructed factor scores for each of the forestry activities using labels from Table IV to identify each factor. Factor scores for each item were computed by weighting the ratings on each risk scale proportionally to the scale's  46  importance in determining each factor and then summing across all scales. In essence, the factor scores can be seen as "boiling down" the results for the judgement scales, although the richness of data for the individual judgement scales is also of interest. Results for each of the six activities are discussed below.  >  Table XIII. Overall Mean Judgement Scale Scores and Factor Analysis Scores for the Six Forest Industry Activities Items Clearcut Selective Cohstn of Pulp mill Pesticide Fertilizer Judgement Scales logging logging logging effluents use use roads Knowledge 0.83 0.58 0.43 0.42 0.41 .i 0.53 Controllability , "• 1.28 1.36 1.30 0.92 1.05 0.90 Scope 1.03 -0.11 . 0.27. 0.67 1.14 0.89 Observability 1.59 -0.03 0.58 0.51 0.32 0.08 Relevance 1.08 -0.17 -0.04 0.81 1.18 0.97 Benefit -1.45 0.85 -0.50 -2.50 -0.75 -0.24 People affected 1.18 -0.07 0.02 0.86 1.27 1.02 Species loss 1.76 r0.42 0.61 L37, 1,49 0^90 Equity -0.71 0.36 -0.49 -1.02 -0.32 -0.23 Immediacy -1.28 0.42 -0.48 -1.01 -0.59 -0.62 Rights of non-human species 2.14 0.09 1.05 1.77 . 1.70 1.32 Reversibility -0.32 0.89 0.26 -0.29 -0.33 0.02 Alternatives 1.93 0.48 0.70 1.21 1.33 1:25 Predictability 1.67 0.92 0.93 0.99 0.97 0.90 . Human health risk 0.61 -0.89 -0.52 1.35 1.50 1.09 ;  Factor Scores Factor 1 - Ecological impact 0.09 -1.15 -0.96 -0.22 0.60 Factor 2 - Human benefits -0.76 1.16 -0.28 -1.35 -0.06 Factor 3 - Controllability 1.44 -0.02 0.34 0.63 0.87 Factor 4 - Knowledge 1.71 0.23 0.47 0.51 0.12 Judgement scale results are on a -3 to +3 scale, where -3 represents very low rating on the particular scale and +3 represents very high rating on the particular scale. Note: refer to Table I for the end-point terms for each judgement scale. The factor scores were calculated using regression procedures, as described iri the text.  Clearcut  0.18 0.11 0.59 0.12  logging  For several of the judgement scales, the respondents' mean perception of clearcut logging was statistically different than the mean ratings for each of the other forest industry activities. For example, the impacts of this activity were considered to be the most observable (M=1.59, p<0.001, two tailed t-test), and the most predictable (M=1.67, p<0.001) relative to each of the other activities. Clearcut logging was also perceived as  47  having the greatest number of alternatives (M=1.93, p<0.001). Its impacts were judged to infringe more on the rights of non-human species (M=2.14, p<0.01) than any other forest industry activity.  While clearcut logging did not differ statistically from each of the other activities for the following scales, the perceptions of these scales were, nonetheless, extreme for this activity. It is an activity that is perceived to create a high potential for species loss (M=1.76), affect a somewhat large number of people (M=1.18), impact upon water environments rapidly (M=-1.24), and impart little benefit to society (M=-1.45). Although not statistically significant, the respondents had the perception that experts have more knowledge about the impacts of clearcut logging on water environments than they do about the impacts of the other forest industry activities. This pattern is reflected in respondents' judgements that the impacts of clearcut logging are relatively controllable (M=1.28) and predictable (M=1.67). Although this activity ranked the highest in terms of the extent of expert knowledge, the respondents did not perceive that experts had a great deal of knowledge about impacts of clearcut logging (M=0.83 on a scale where 3 indicates great amount of knowledge). Clearcut logging is viewed as affecting a fairly large area (M=1.03). The respondents also regard this activity to be less than equitable (M=-0.71) in terms of the distribution of the benefits and costs that it imparts on society.  Turning to the factor scores, clearcut logging was perceived as being the most extreme in terms of its controllability  (Factor 3) as well as being the forest industry  activity for which there exists the most expert  knowledge  (Factor 4). Conversely, only  pesticide use exceeded this activity in terms of its rating on the lack of human (Factor 2).  48  benefit  Selective  logging  For the majority of the judgement scales, the respondents viewed this activity as the least risky of the selected activities. Respondent views of the ecological risks of selective logging are moderate (ranging either side of the midpoint from -0:89 to 1.36). Relative to the other activities, selective logging is considered to have the smallest impact on water environments in terms of the total area affected (M=-0.11, p<0.05), the potential for species loss (M=-0.42, p<0.001), its observability (M=-0.03), the number of people affected (M=-0.07), and the risk imposed on human health (M=-.089, p<0.05). Furthermore, the impacts of selective logging are perceived to be the most reversible (M=0.89, p<0.001) and the most controllable (M=1.36). Its impacts are considered to occur farthest in the future (M=0.42, p<0.001). Selective logging is regarded as providing more benefit to society (M=0.85, p<0.001), as well as being the most equitable activity in term of the distribution of these benefits (M=0.36, p<0.001). In fact, selective logging is the only activity which ranked on the positive side of the scale midpoint for these two judgement questions. Surprisingly, selective logging is perceived to be the activity with the fewest alternatives (M=0.48). The result contrasts with the perception of clearcut logging as having the greatest alternatives, as noted earlier.  Selective logging ranked at the extreme, relative to the other forest industry activities, for three of the four factors. It was perceived to be the least damaging in terms oi ecological  impact  (Factor 1), it ranked high for human  interestingly, it ranked low in terms of controllability  Construction  of logging  benefit  (Factor 2), and  (Factor 3).  roads  Respondents generally perceive that building logging roads is only nominally damaging to water environments. The mean overall scale scores ranged from -0.52 to 1.30. The impacts of logging roads are considered to be relatively controllable (M=1.30)  49  while posing only modest risk to human health (M=-0.52). Selective logging is the only other forest industry activity addressed in this study considered to be more controllable and less of ariskto human health. The respondents judged that this practice infringes somewhat on the rights of non-human species (M=l".05) and poses a slight potential for species loss (M=0.61). Only selective logging was deemed to be less detrimental to nonhuman species. The construction of logging roads and selective logging are two activities that are considered to be only nominally relevant to the lives of the respondents (M=-0.04 and -0.17, respectively). The remaining four activities scored on the "more relevant" side of the scale mid-point for this judgement scale. The respondents' perceptions of the remaining judgement scales for this activity were relatively neutral.  With the exception of Factor 1  (ecological  impact),  the construction of logging  roads was not perceived in an extreme manner for any other factor. Of the forest industry activities reviewed here, only selective logging was perceived to have less overall ecological impact.  Effluent from pulp mills Respondents viewed pulp mill effluent disposal as having a relatively high degree of ecological risk for water environments. Of the six activities addressed in the study, the disposal of pulp mill effluent was judged to be the least beneficial (M=-2.50) by a significant margin (p<0.001). This practice was also considered to be the least equitable, which is evident in the respondents perception that those who derive the benefits from the activity are not those who incur the cost of its impacts (M=-1.02). It is perceived as imposing a high degree of risk to human health (M=1.35) as well as creating a serious potential for species loss (M=1.37). Pulp mill effluents were considered to infringe severely on therightsof non-human species (M=1.77). The respondents believe that the  50  impacts of pulp mill effluent disposal are relatively controllable (M=0.92), predictable (M=0.99), and that it is a practice for which alternatives exist (M=1.21). These judgements, in conjunction with the belief that the impacts occur quickly (M=-1.01), likely influence the poor perception that respondents have of pulp mill effluent disposal.  In terms of the factor analysis, pulp mill effluents were perceived in an extreme manner for Factor 2 only (human  benefit).This  activity was considered the least  beneficial of the forest industry activities.  Fertilizer use The respondents' perceptions of ecological risk from fertilizer use were moderate relative to clearcut logging, pulp mill effluents and pesticide use. However, it is considered to be a practice that imposes relatively high impacts for some judgement scales. It is perceived to affect a relatively large number of people (M=l-.02)' by virtue of the risk it potentially poses to human health (M=l .09). The impacts of fertilizer use are judged to cause some species loss (M=0.90), and therefore, it is an activity that is deemed to infringe on the rights of non-human species (M=1.32). As with all of the activities discussed here (with the exception of selective logging), the respondents consider that fertilizer use is not equitable (M=-0.23). It is the practice considered to-be the least controllable (M=0.90) [although for all six activities, the respondents believed that the impacts can be controlled to some degree (with means ranging from 0.90 to 1.36)]. It is also judged to be a practice for which alternatives are readily available (M=1.25).  This activity did not generate any extreme factor analysis scores. However, relative to the six activities discussed here, only pesticide use was considered to have  51  greater ecological  impact  having greater human  (Factor 1), while only selective logging was percieved as  benefit  (Factor 2).  Pesticide use Pesticide use was considered to be the worst of the six activities in terms of risk to human health (M=1.50), the number of people affected by its impacts (M=i.27) and the scope of the area potentially influenced by these impacts (M=1.14). Considering these three judgements, it is not surprising that the respondents deemed that pesticide use is the activity which is most relevant to their lives (M=l. 18) (recall that this item was concerned with all pesticide use, not just in forestry). It was considered to be a practice that is likely to cause high species loss (M=1.49), and as such, it definitively violates the rights of non-human species (M=1.70). The respondents also perceive that the impacts of pesticide use are likely to be rapid (M=-0.59) and somewhat irreversible (M=-0.33). Although they perceive this to be the case, the respondents maintain the view that the impacts of pesticide use are fairly controllable (M=1.05).  In terms of the factor analysis, pesticide use ranked as the most extreme relative to the other forest industry activities for Factor 1 (ecological for Factor 3 (controllability).  impacts).  It also ranked high  Of the forest industry activities, only clearcut logging was  perceived to be more controllable. In fact, only 2 other items for the entire study were considered to be more controllable than pesticide use (urban development and water exports).  Risk Maps The factor scores for all items in the study, including the six forestry items, can be used to create a two-dimensional "risk map" that graphically summarizes and communicates all the lay judgements in the study (Slovic, 1987; McDaniels, et al., 1995).  52  Figure 3 shows a risk map with Factor 1 forming the y-axis and Factor 2 forming the x-axis. Items with positive values along the y-axis are considered to have greater ecological  than those with negative values along this axis (e.g., the  impact  respondents' perception is that climate change has much greater ecological  impact  than  paddle boating). Items with positive values along the x-axis are considered to have greater human  benefit  than those with negative values along this axis. Therefore, the  upper left quadrant is where items that are perceived to have great ecological and little human  benefit  will be located (or the least ideal scenario). Conversely, the  lower right quadrant is where items that are perceived to have little and great human  benefit  impact  ecological  impact  will be located (or the most ideal scenario). Of the forest  industry activities that are the focus of this article, clearcut logging and pesticide use are located in the upper left quadrant. Selective logging is located in the lower right quadrant. Fertilizer use is located near the origin yet within the upper right quadrant, meaning it is considered to be of some ecological  impact.  human  benefit  while causing some degree of  The item mill effluents, however, is located in the lower left quadrant  far along the negative aspect of the x-axis (very low  human  benefit)  and marginally  bellow the zero point of the y-axis (considered not to have noteworthy impacts).  ecological  The item logging roads is also located in the lower left quadrant, therefore it is  considered to have little  ecological  impact  and little human  benefit.  4.4 DISCUSSION AND CONCLUSIONS These results indicate that impacts on non-human species are a crucial influence on perceptions of the ecological risk from forestry activities. Impacts on human health are also considered when assessing risk levels, with perceptions of negative impacts strongly associated with high perceptions of ecological risk (McDaniels, et al., 1995). These results are generally in keeping with the aggregate study findings for an array of human activities (Chapter 2 or McDaniels, et al., 1997). Of the six practices discussed  53  here, clearcut logging, pulp mill effluents, and pesticide use were considered to create higher potential for species loss, infringe more on the rights of species, and pose greater risk to human health than did selective logging, construction of logging roads, and fertilizer use. Accordingly, the former three practices were considered more risky in general than the latter three. The present study also reinforces the conclusion that respondents' perceptions of risk are inversely related to their assessment of the benefits associated with a risk item (Alhakami and Slovic, 1994; McDaniels, et al., 1995). The perceived higher benefit derived from selective logging, fertilizer use, and the construction of logging roads likely reduces the general perception of risk associated with these practices. People likely associate building roads (including logging roads) as a service that confers greater freedom of mobility. Conversely, the perceived lack of benefit derived from pulp mill effluents and clearcut logging potentially elevates the general perception of risk associated with these two practices.  This chapter also identifies divergences between the views of lay and expert respondents regarding perceptions of the "riskiness in general" of and the "need to regulate" some activities associated with the forest industry. These divergences could lead to controversy over future risk management priorities. For example, public perceptions of clearcut harvesting as being a high risk practice strongly conflict with prescriptions by some forest ecologists that in certain locations, clearcuts simulate the effects of widespread forest fires that shaped regeneration processes before the onset of forest harvests (Kimmins, 1992). These differences in perceptions lead to conflicts about the role of clearcuts in forest land-use practices. Similarly, public perceptions that logging road construction is relatively benign, conflicts with scientific writing on the issue (Ottens and Rudd, 1977; Anderson and Potts, 1987; Bilby, et al, 1989; Rice and Lewis, 1991; Grayson, et al., 1992; Tschaplinski, 1992) and the views of experts surveyed here. One of the key implications of reducing the size of clearcuts is an  54  increase in the length of logging roads constructed and active in a given year. Yet this change seems to hold less salience for public perceptions. As a result, there are conflicts between commonly held public views and the judgements of some experts about the pros and cons of reducing the size of clearcuts and extending logging roads.  55  CHAPTER V . CONCLUSION  Although the scope of the three sets of studies differ, common findings are evident. These findings relate directly to two of the three objectives of this thesis. Recall that the first objective was to clarify how lay people perceive various kinds of activities, in terms of the risk that each may pose to the health and productivity of water environments. The results demonstrate that people tend to conceptualize risk in a multifaceted fashion. As such, four factors were identified that characterize lay perceptions of ecological risk to water resources. The factors, as outlined in detail in Chapters 2 and 4, are  Ecological  Impact,Human  Benefits,  Controllability  and  Knowledge.  These four  factors can therefore be identified as predictors of environmental risk perception with Ecological  Impact  providing the most influence. A strong inverse relationship exists  between Human  Benefits  than Ecological  Impact,  Controllability  and perceived environmental risk. Although less influential Knowledge  is a predictor of environmental risk. Finally,  is highly associated with perceptions of the need to regulate an activity.  A further common finding is related to the second objective of the thesis. That objective was that the study compare lay views to the views of experts on the matters of general risk and regulation requirements. It was clear throughout the three component studies that there existed several divergences between the views of lay and expert respondents in these regards. Even for the two studies which dealt with only small subsets of the data (Chapters 3 and 4), lay respondents exhibited significantly differing views than those of the experts. This was strikingly evident in Chapter 2, which provided an analysis of the overall data set. The general observation that can be drawn from that chapter is that the lay respondents tended to associate activities for which there has existed some degree of national or even international media coverage with higher risk. The experts, on the other hand, had the perception that some of the activities for which  56  there exists little media coverage are, in fact, posing greater risk than others that receive this form of attention.  As per thefinalobjective of the thesis, thefindingsof this study have considerable policy and planning implications. Perhaps the most significant policy implication would be in the area of risk communication. Information drawn from studies of this nature could be helpful for designing risk communication materials that attempt to better inform the public about scientific understanding of the ecological risks and benefits of various human activities. Government agencies that are in the process of developing policies might find information of this sort useful for predicting and diagnosing conflicts between lay people and experts. Furthermore, these and other studies of this nature can be helpful tools to characterize the range of values and objectives important to lay people and experts. A clear understanding of these values and objectives would enhance the planning, design and evaluation phases of land and water resource management policies.  There are several advantages associated with studies of this nature. If the notion that a person's perceptions have bearings on the actions they take is a true reflection of reality, then it is important to accurately identify and clarify the public's perceptions and attitudes when developing policies that will influence how people are expected to act. A well constructed perception study will provide clues as to public preferences relating to the management of resources. The study format applied here removes some of the biases inherent in other techniques that attempt to assess public perception, such as a reliance on statements of interest groups, opinion polls and public hearings. Typically, these techniques reveal the perceptions and attitudes of select groups. Further to this, studies of this nature expand on the oversimplified means of gaining a feel for the public perception through mass referenda.  57  A further advantage of public perception studies conducted in the format presented here, is that information about a large suite of activities can be drawn from an entire data set and then refined to derive information about perceptions of related activities by utilizing a subset of the overall data. Policy development for specific environmental risk management concerns or, in other words, ones that focus on related activities, could be derived from the refined studies. However, it might be wise to conduct further research that focuses solely on perceptions of related activities to determine if it is reasonable to draw conclusions about those same activities when the information is derived from a study of larger scope, as was the case in this thesis. Additional further research might address perceptions of risks to other ecosystem components such as air, soil, ground water, and marine environments.  When reviewing these or any other risk perception results, it is crucial to bear in mind the distinction between descriptive and prescriptive interpretations of the findings. The present investigation is, in descriptive terms, a study in the vein of other social science research: it attempts to describe the processes shaping a social phenomenon. It indicates how the respondent sample (and by implication, the population from which it was drawn) view the relative ecological risks of human activities, given their current level of understanding, when asked to make quick, simple judgements in an established methodological framework. However, these descriptive findings should by no means be interpreted as having direct prescriptive significance. In other words, and by example, a finding that the public perceives clearcuts as highly risky, and that the construction of logging roads is not particularly risky, should not be interpreted as a vote that forest managers should eliminate clearcuts and ignore the impacts of road building; Decisions regarding the design of harvest patterns, or the merits of extending logging roads, or any other environmental management decision, require more careful attention to specifying objectives, alternatives, impacts and tradeoffs than is possible in surveys of this nature.  58  REFERENCES Alhakami, A and P. Slovic. 1994. "A Psychological Study of the Inverse Relationship Between Perceived Risk and Perceived Benefit." Risk Anal., 14(6): 1084-1096. Anderson, B. and D.F. Potts. 1987. "Suspended Sediment and Turbidity Following Road Construction and Logging in Western Montana." Water Res. Bull.,23(4): 681-690. Barica, J. 1975. "Collapses of Algal Blooms in Prairie Pothole Lakes: Their Mechanism and Ecological Impact." Verh. Int. Verein. Limnol., 19: 606-615. (Cited in C C R E M , 1987). Bilby, R.E., K. Sullivan and S. Duncan. 1989. "The Generation and Fate of Road Surface Sediment in Forested Watersheds in Southwestern Washington." For. Sci., 35(2): 453-468. Billings, W.H. 1981. "Water Associated Human Illness in Northeast Pennsylvania and its Suspected Association with Blue-Green Algae Blooms." In The Water Environment: Algal Toxins and Health. Ed. W.W. Carmichael Plenum Press, New York. (Cited in C C R E M , 1987). British Columbia. 1993. "Forest Practices Code." British Columbia Ministry of Forests, Victoria, British Columbia. British Columbia. 1993. "State of the Environment Report for British Columbia." British Columbia Ministry of Environment, Lands and Parks, Public Affairs Branch, Victoria, British Columbia. Canadian Council of Resource and Environment Ministers (CCREM). 1987. "Canadian Water Quality Guidelines." Water Quality Objectives Division, Water Quality Branch, Inland Waters Directorate, Environment Canada, Ottawa, Ontario. Carmichael, W.W. 1981. "Freshwater Blue-Green Algae (Cyanobacteria) Toxins - A Review." In The Water Environment: Algal Toxins and Health. Ed. W.W. Carmichael Plenum Press, New York. (Cited in C C R E M , 1987). Carmichael, W.W., C.L.A. Jones, N.A. Mahmood and W.C. Theiss. 1985. "Algal Toxins and Water Based Diseases." CRC Crit. Rev. Environ. Control, 15: 275-313. (Cited in C C R E M , 1987). Casselman, J.M. and H.H. Harvey. 1975. "Selective Fish Mortality Resulting from Low Winter Oxygen." Verh. Int. Verein. Limnol, 19: 2418-2429. (Cited in C C R E M , 1987). Dillon, P.J., K . H . Nicholls, W.A. Scheider, N.D. Yan, and D.S. Jeffries. 1986. "Lakeshore Capacity Study, Trophic Status." Ontario Ministry of Municipal Affairs, 90 pp. Fischhoff, B., S.R. Watson and C. Hope. 1986. "Defining Risk." Pol. Sci., 17:123-139.  59  Grayson, R.B., S.R. Haydon, M.D.A. Jayasuriya and B.L. Finlayson. 1992. "Water Quality in Mountain Ash Forests - Separating the Impacts of Roads from those of Logging Operations." J. Hydrol., 150(2/4): 459-480. Gregory, R. and R. Mendelsohn. 1993. "Perceived Risk, Dread, and Benefits," Risk Anal., 13: 259-264. Hart, W.C., R.S. Scott, and J.G. Ogden III. 1978. "A Phosphorus Loading Model for Lakes in the Shubenacadie Headwaters." Technical Report 2, 34 pp. Hart, W.C., and D.H. Waller. 1984. "Application of a Phosphorus Loading Model to Water Quality Management in the Shubenacadie River Basin." Contribution No. 8418. CWRS, TUNS. 11pp. Janus, L . L . , and R.A. Vollenweider. 1981. 'The OECD Co-operative Programme on Eutrophication. Summary Report." Canadian Contribution, Canada Center for Inland Waters, Burlington. Keeney, R.L. 1992. Value-focused thinking: A path to creative decision making. Cambridge, Massachusetts. Harvard University Press. Kimmins, H. 1992. Balancing Act: Environmental Issues in Forestry. Vancouver, British Columbia. University of British Columbia Press. Kortmann, R.W. 1988. "Septic Systems and How They Affect Lakes (or all you never knew about septic tanks and the environment)."La£e Line, N. Amer. Lake Manage. Soc, 8(6): 8-11,18-19. Kortmann, R.W., and P.H. Rich. 1994. "Lake Ecosystem Energetics: The Missing Management Link." Lake Reserv. Manage., 8 (2): 77-97. McDaniels, T.L., L.J. Axelrod, and N. Cavanagh, 1998. "Public Perceptions and Attitudes Regarding Water Quality in the Lower Fraser Basin of British Columbia." In Press with Water Resour. Res. McDaniels, T.L., L.J. Axelrod, N. Cavanagh and P. Slovic. 1997. "Characterizing Perception of Ecological Risk to Water Environments." Risk Anal,. 17(3): 341-352. McDaniels, T.L. 1997. "Ten Propositions for Untangling Descriptive and Prescriptive Lessons in Risk Perception Findings." J. Syst. Saf. Reliab.. In Press. McDaniels, T.L., L.J. Axelrod, and P. Slovic. 1995. "Characterizing Perception of Ecological Risk." Risk Anal., 15(5): 575-588. National Research Council (NRC). 1988. "Improving Risk Communication." Washington, DC. National Academy Press.  60  O'Connor, R.E., R.J. Bord., & A . Fisher. 1994. "Perceptions of Fresh Water." Nat. Geo. Res. Exp., 10 (3): 318-341. Ottens, J. and J. Rudd. 1977. "Environmental Protection Costs in Logging Road Design and Construction to Prevent Increased Sedimentation in the Carnation Creek Watershed (Vancouver Island)." Pacific Forestry Centre, Canada. Report No. BC-X155. Rast, W., and G.F. Lee. 1978. "Summary Analysis of the North American (U.S. portion) O E C D Eutrophication project: Nutrient Loading Lake Response Relationships and Trophic State Indices." Corvallis Environmental Research Lab, Corvallis, Oregon. EPA 600/3-78-008. 454 pp. Reckhow, K . H . , M.N. Beaulac, and J.T. Simpson. 1980. "Modeling Phosphorus Loading and Lake Response under Uncertainty: a Manual and Compilation of Export Coefficients." EPA 440/5-80-011. U.S.E.P.A. 214 pp. Reckhow, K . H . , and J.T. Simpson. 1980. "A Procedure Using Modeling and Error Analysis for the Prediction of Lake Phosphorus Concentration from Land Use Information." Can. J. Fish. Aquat. Sci., 37: 1439-1448. Rice, R.M. and J. Lewis. 1991. "Estimating Erosion Risks Associated with Logging and Forest Roads in Northwest California." Water Resour. Bull., 27(5): 809-817. Rose, E.T. 1953. 'Toxic Algae in Iowa Lakes." Proc. Iowa Acad. Sci. 60: 738-745. (Cited in C C R E M , 1987). Sewell, W.R.D. 1971a. "Integrating Public Views in Planning and Policy Making." In Perceptions and Attitudes in Resource Management. Ed. W.R.D. Sewell and I. Burton. Ottawa, Information Canada, p, 125. Sewell, W.R.D. 1971b. "Environmental Perceptions and Attitudes of Engineers and Public Health Officials." Environ, and Behavior, 3(1): 23-59. Sewell, W.R.D. 1974. "Perceptions, Attitudes and Public Participation in Countryside Management in Scotland." J. Environ. Mngmt, 2: 235-257. Slovic, P., B. Fischhoff, & S. Lichtenstein. 1979. "Rating the Risks." Environment, pp. 14-20, 36-39. Slovic, P. 1987. "Perceptions of Risk." Science, 236, 280-285. Slovic, P. 1992. "Perceptions of Risk: Reflections on the Psychometric Paradigm." In Social Theories of Risk. Ed. Krimsky, S. and D. Golding. New York, Praeger, pp. 89107. Snell, T.W. 1980. "Blue-Green Algae and Selection in Rotifer Populations." Oecologia (Berlin), 46: 343-346. (Cited in C C R E M , 1987).  61  Stehr, N., and H. von Storch. 1994. 'The Social Construct of Climate and Climate Change." Max-Planck-Institut fur Meteorologie, Report No. 137. Hamburg, Germany. Sykora, J.L. and G. Keleti. 1981. "Cyanobacteria and Endotoxins in Drinking Water Supplies." In The Water Environment: Algal Toxins and Health. Ed. W.W. Carmichael Plenum Press, New York. (Cited in C C R E M , 1987). Tschaplinski, P.J. 1992. "Effects of Roads on Freshwater Fish Habitats and Fish Production." British Columbia Ministry of Forests, Research Branch, Victoria, British Columbia. Waller, D.H. 1977. "Effects of Urbanization on Phosphorus Rows in a Residential System." In UNESCO IAHS-AISH Publn. 123:52-58. Waller, D.H., and Z.Novak. 1981. "Pollution Loading to the Great Lakes from Municipal Sources in Ontario." J. Water Pollut. Control Fed., 53 (3): 387-395. Wetzel, R.G. 1983. Limnology. 2nd ed. Saunders College Publishing, Toronto, pp. 4; 284-295.  62  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088884/manifest

Comment

Related Items