Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Culture and the forested landscape : inter and intra-cultural perceptions of modified forest landscapes Lewis, John Llewellyn 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2006-198933.pdf [ 24.22MB ]
JSON: 831-1.0075027.json
JSON-LD: 831-1.0075027-ld.json
RDF/XML (Pretty): 831-1.0075027-rdf.xml
RDF/JSON: 831-1.0075027-rdf.json
Turtle: 831-1.0075027-turtle.txt
N-Triples: 831-1.0075027-rdf-ntriples.txt
Original Record: 831-1.0075027-source.json
Full Text

Full Text

CULTURE AND THE FORESTED LANDSCAPE: Inter and Intra-Cultural Perceptions of Modified Forest Landscapes by John Llewellyn Lewis B A . (Honours), Queen's University, 1990 M.PI., Queen's University, 1996 M . S c , University of British Columbia, 2000 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Forestry) THE UNIVERSITY OF BRITISH COLUMBIA February, 2006 © John Llewellyn Lewis, 2006 Abstract A key feature of the current policy environment in which decisions about landscape management are made is its increasing complexity. In the past, relatively few interests commanded attention in decisions about forest management, today a host of values demand consideration in decision-making. There is a need to consider how this increasingly varied spectrum of interests can be taken into account, particularly where these interests are unfamiliar to land managers such as the preferences of First Nations communities. Two central questions underlie this aspect of preference: • What is the range of dimensions that local stakeholders consider in their evaluation of modified forest landscapes? • How and why do preferences for modified forest landscapes differ between and among First Nations and Euro-Canadians? A sample of First Nations and Euro-Canadian residents of the upper Skeena Valley in Northwest British Columbia were interviewed using a photo-elicitation technique for landscape preference evaluation. Photo-realistic simulations of alternative landscape changes in the upper Skeena Valley were presented to the participants, who ranked the landscape treatments and commented at length on the rationale for their rankings. Based on the pattern of results from the participant interviews, I have reached four main conclusions: • Conclusion 1: Landscape preference evaluations are based on a complex and simultaneous weighting of alternative perceived consequences, the degree to which disturbances demonstrate care for natural processes and future needs, as well as physical site characteristics. Conclusion 2: Ethnic and other sociodemographic characteristics only partially account for differences in landscape preference. People tend to evaluate the landscape from a user perspective and, with a few exceptions; forest-based uses often transcend distinctions of ethnicity, age and gender. Conclusion 3 : People are willing to accept modified or human-influenced landscapes, but the degree to which the treatment visibly demonstrates 'care' is a salient factor in the participants' preference evaluations. Conclusion 4 : Environmental value orientations can be conceptualized in terms of a two-dimensional matrix with the New Environmental Paradigm-Dominant Social Paradigm scale configured along one axis, and an additional axis that I have termed the 'hierarchy-interdependence' dimension. iii Table of Contents Abstract ii List of Tables vii List of Figures viii Chapter 1 - Introduction 1 1.1 Background and Research Questions 1 1.2 Research Justification and Assumptions 6 1.2.1 Dimensions of Acceptable Forest Management 6 1.2.2 Inter Cultural Landscape Preference Differences 12 1.3 Organisation and Content of the Dissertation 16 Chapter 2 - Literature Review 19 2.1 Introduction 19 2.2 Landscape Preference Theories 21 2.2.1 Biological Theories of Preference 21 2.2.2 Social Theories: Environmental Value Orientations 26 2.3 Determinants of Landscape Preference 31 2.3.1 Socio-demographic Characteristics 31 Age and Generational Influences 32 The Role of Gender 33 Formal Education 34 Landscape Familiarity 35 Urban versus Rural Residence 36 Culture and Ethnicity 37 2.3.2 The Social Construction of Landscape Preferences 39 2.3.3 Differing Human-Place Attachments 43 2.4 Chapter Summary 53 Chapter 3 - Environmental and Social Context 54 3.1 Introduction 54 3.2 The Physical and Ecological Landscape of the Upper Skeena Valley 55 3.3 The Pre-Colonial Culture and Landscape 64 3.3.1 Resource Procurement and Gitksan Social Organisation 64 3.3.2 The Question of Indigenous Cultivation 67 3.4 Post Contact Settlement and Landscape Change 76 3.4.1 The Decline of Traditional Indigenous Cultivation 76 3.4.2 Landscape Change Under the 'New Resource Economy' 86 3.5 Chapter Summary 90 Chapter 4 - Research Design & Methodology 92 4.1 Introduction 92 4.2 Research Paradigms and Methodological Assumptions 92 iv 4.3 Research Design 96 4.3.1 Sampling: Rationale tor Community Selection 97 4.3.2 Sampling: Rationale for Participant Selection 99 4.4 Stimulus Materials 107 4.4.1 Photo-Elicitation as a Qualitative Research Technique 107 4.4.2 Field Reconnaissance 110 4.4.3 Viewpoint Selection 112 4.4.4 Image Production Methods 120 Condition 1: Pre-lndustrial or Aboriginal Forest Management. 122 Condition 2: Industrial Timber Management 127 Condition 3: Natural or Pristine Forest 131 Condition 4: Multi-Use Forest Management 135 4.5 Data Collection Procedures 139 4.5.1 The Semi-Structured Interviews 141 4.6 Data Analysis 146 4.6.1 Research Question 1 147 4.6.2 Research Question 2 149 4.7 Criteria for Evaluating Qualitative Research 153 4.7.1 Internal Validity 154 4.7.2 External Validity or Generalisability 156 4.7.3 Reliability 158 4.8 Chapter Summary 159 Chapter 5 - Interpretation and Analysis 160 5.1 Introduction 160 5.2 Question 1: The Dimensions of Landscape Evaluation 161 5.2.1 The Perceived Balance of Impacts 161 a) Clearcut Patterns: Industrial and Multi-Use Scenarios 162 b) Aboriginal Burning: Multi-Use and Pre-lndustrial Scenarios... 169 c) Natural Condition 176 5.2.2 Perceptions of Care 181 5.2.3 Site Characteristics 187 5.3 Question 2: Inter and Intra Cultural Landscape Preferences 191 5.3.1 Individual and Group Preference Profiles 193 a) Participant Cluster One 193 b) Participant Cluster Two 199 c) Participant Cluster Three 206 d) Participant Cluster Four 214 5.3.2 Pattern of Findings: Research Question 2 221 a) The Value Orientation Matrix 221 b) Inter and Intra-Ethnic Dimensions of Landscape Preference. 228 5.4 Chapter Summary 240 Chapter 6 - Synthesis and Conclusions 241 6.1 Introduction 241 6.2 Summary of the Main Conclusions 242 v 6.3 Research Question 1: Dimensions of Preferred Landscape Change 244 6.3.1 Main Findings 244 6.3.2 Implications 248 Theoretical Implications 248 Practical Implications 249 6.3.3 Directions for Future Research 249 6.4 Research Question 2: Inter-Ethnic Determinants of Preference 250 6.4.1 Main Findings 250 6.4.2 Implications 255 Theoretical Implications 255 Practical Implications 256 6.4.3 Directions for Future Research 256 6.5 Conclusion 3: 'Cues to Care' and Perceptions of Human Influence 258 6.5.1 Main Findings 258 6.5.2 Implications 261 Theoretical Implications 261 Practical Implications 263 6.5.3 Directions for Future Research 264 6.6 Conclusion 4: The DSP-NEP Continuum Revisited 265 6.6.1 Main Findings 265 6.6.2 Implications 269 Theoretical Implications 269 Practical Implications 271 6.6.3 Directions for Future Research 272 6.7 Concluding Remarks 273 References 275 Appendix A1: Letter of Introduction 303 Appendix A2: Interview Consent Form 304 Appendix A3: Interview Guide 305 Appendix A4: Letter of Thanks 308 Appendix B1: Simulations, Pre-lndustrial Condition 309 Appendix B2: Simulations, Industrial Condition 310 Appendix B3: Simulations, Multi-Use Condition 311 Appendix B4: Simulations, Natural Condition 312 Appendix C1: Technical Appendix, Multidimensional Scalogram Analysis 313 Appendix C2: MDS Data Matrix 319 vi List of Tab les Table 5.1 Balance of Impacts 162 Table 5.2 Perceptions of Care 182 Table 5.3 Site Characteristics 187 Table 5.4 Summary of Cluster Preferences 235 vii List of Figures Figure 1.1 Study Area, Upper Skeena Watershed, British Columbia 3 Figure 2.1 Chapter Organisation 20 Figure 2.2 Landscape Preference Matrix 25 Figure 2.3 NEP-DSP Continuum 29 Figure 2.4 Place Dependence Landscape Preference Model 52 Figure 3.1 Upper Skeena Valley Communities 56 Figure 3.2 Skeena River Flood, 1936 58 Figure 3.3 The Skeena-Bulkley Valley Lowlands 59 Figure 3.4 Post Forest Practices Code Cutblock Patterns 62 Figure 3.5 Luu Skayansit, Traditional Berry Site Cultivated and Maintained by Fire... 71 Figure 3.6 Extent of Clearcut Harvesting in the Mid 1980s 85 Figure 3.7 Ksan - Gitksan Village Museum 89 Figure 4.1 Participant Selection Criteria 101 Figure 4.2 Participant Sample by Ethnicity, Age and Gender 104 Figure 4.3 Viewpoint Selection Criteria 114 Figure 4.4 Viewpoint 1 115 Figure 4.5 Viewpoint 2 116 Figure 4.6 Viewpoint 3 119 Figure 4.7 Gitksan Berry Patches, Kispiox Valley, 1948 124 Figure 4.8 Condition 1: Pre-lndustrial or Aboriginal Forest Management 126 Figure 4.9 Condition 2: Industrial 130 Figure 4.10 Condition 3: Natural or Pristine Forest 134 Figure 4.11 Condition 4: Multi-Use Forest Management 137 Figure 4.12 MDS Data Matrix 150 Figure 4.13 Multidimensional Scalogram Output 151 Figure 5.1 Multidimensional Scalogram (MDS) Analysis Results 192 Figure 5.2 MDS Results, Cluster 1 194 Figure 5.3 MDS Results, Cluster 2 200 Figure 5.4 MDS Results, Cluster 3 207 Figure 5.5 MDS Results, Cluster 4 215 Figure 5.6 Environmental Values Matrix 222 Figure 5.7 Participant Clusters by Ethnicity, Age and Gender 230-2 Figure 5.8 Revised Landscape Preference Model 238 Figure 6.1 Summary of Research Findings 242 Figure 6.2 Environmental Values Matrix 268 viii 1 Introduction & Overview 1.1 Background and Research Questions As government and industrial resource interests across North America strive to implement new, more sustainable approaches to land and resource management, stories accumulate of 'state-of-the-art' management frameworks that have been thwarted due to the inability of land managers to incorporate the knowledge, perceptions and preferences of local stakeholders (Armour, 1991; McCormack, 1998; Redmond, 1996; Satterfield, 2002). Consistent with the emerging landscape ecology principle that people and the landscape are mutually dependant parts of ecosystem patterns and processes, there is now the realization that we cannot prescribe what future changes should be made to environments without consulting local stakeholders' conceptions of preferred or acceptable 1 change (Brunson, 1993; Peterson, 1995; Stankey, 1996). People have affected most places in the world, to a greater or lesser extent, and many landscapes are the embodiment of some practical human choice. Recent 'interpretivist' approaches to landscape perception provide some insight into why different groups clash over the meaning of environmental change, and the potential human consequences of that change (Hull, et al., 2001; Satterfield, 2002). What scientists or resource managers may construe as a simple modification 1 The discussion assumes that the terms 'acceptability' and 'preference' are not synonymous, but are closely related. For clarification, I am defining acceptability as the outcome of a judgment process whereby individuals (1) compare a set of alternatives with one another as well as with a perceived 'reality' or status quo condition; and (2) decide among the alternatives which condition or conditions are the most 'superior' or favourable (Brunson, 1996: 9). This conception of acceptability, as a relative term, is similar to the standard conception of 'preference' which the Kaplans define as a process of categorization based on the relative extent to which a person 'likes' a given setting or landscape condition (1989: 41), based on its compatibility with human needs and purposes (1989: 10). In both conceptions a relative assessment of superiority and choice is made. However, in the latter case, the choice is largely driven by an assessment of personal needs and requirements. 1 to the natural environment, may be perceived as a threat to the fundamental meaning of a community's way-of-life. Alternatively, what professional specialists define as undesirable landscape alterations - the intentional burning of underbrush, for example - may represent the persistence of stewardship practices rooted in traditional subsistence or other material requirements. Thus, resource managers and related environmental professionals caught in the crossfire of land management controversy need to know how to respond in a manner that respects the atachments that local stakeholders including First Nations have for their landscape. Examining local preferences and developing a fuler understanding of the range of dimensions or factors that local stakeholders use to inform their assessment of landscape condition, is therefore an essential part of developing a forest management framework that has broad local support. An empirical approach to this requirement can be stated through the first research question: Research Question 1: What is the range of dimensions that local stakeholders consider in their evaluation of modified forest landscapes? For forest managers who have the technical capacity to alter the living landscape, designing change that delivers commercialy marketable products while sustaining local livelihoods and the quality of environmental experience for a wide range of users, including traditional subsistence cultures, remains a significant chalenge. A key feature of the current policy environment in which decisions about forest management are made is its increasing complexity. Whereas in the past, relatively few interests commanded attention in decisions about forest management, today a host of values and perspectives demand access and consideration in 2 decision-making. For instance, in the study area of the upper Skeena Valley used in this dissertation, (Figure 1.1), Crown forests have directly and indirectly contributed to the traditional way-of-life of Gitksan residents, as well as provided employment and solitude for First Nations and Euro-Canad ian 2 residents alike. A s such, forest management decisions can have wide ranging effects. Figure 1 . 1 : Study Area, Upper Skeena Watershed, British Columbia Digital map produced by John L. Lewis using ArcGIS rel. 8. I am treating the term 'Euro-Canadian' as any person with a predominantly European or Caucasian heritage. Excluded from this group are people of mixed European and Native American, Asian, or African descent. 3 A major, yet often overlooked impediment to socially acceptable resource management is the fact that environmental stakeholders may interpret management proposals quite differently from one another. This plurality of perspectives, combined with the tendency on the part of different stakeholders to believe that their point-of-view is the most legitimate, may result in myriad problems, including divergent problem definitions, misunderstanding, prolonged public-agency mistrust, and the possible breakdown of decision-making processes. There is a need to consider how an increasingly varied spectrum of interests can be taken into account, particularly in cases where First Nations and Euro-Canadian interests have historically conflicted over the meaning of preferred landscape change (Glavin, 1990), or where local values are unfamiliar to professional managers and decision-makers (e.g. First Nations land-based spiritual values, see Lewis and Sheppard, 2005). T h e central issue underlying this aspect of preference is - in what ways, and why, do landscape preference evaluations or judgments of acceptability vary among stakeholders? This is addressed in the second research question: Research Question 2: How and why do preferences for modified forest landscapes differ between and among First Nations and Euro-Canadians? For the upper Skeena Valley and other regions of British Columbia, the contemporary relevance of both research questions is clear. Forests harvested in the Kispiox and Bulkley forest districts sustained a substantial wood products industry for many years; however unsustainable rates of harvest and questionable harvesting practices combined with shifting global markets effectively closed the industry throughout much of the Northwest in the late 1990s (Weetman et al., 1990). 4 Throughout the economic cycles of the timber products sector, aboriginal subsistence activities have remained an integral part of the social, cultural and economic fabric of the upper Skeena region. For the Gitksan and Wetsuweten residents of the upper Skeena valley, participation in subsistence activities continues to sustain aboriginal traditions and customs, maintains community and family ties, supplements individual and family income, and provides a healthy source of traditional foods. Following the closure of the region's major sawmills in the late 1990s, renewed interest has developed among forest managers in the development of small scale, value-added, sustainable timber management, with the objective of minimizing effects on recreational and tourist uses of the forest, as well as mitigating the region's vulnerability to periodic declines in the forest products sector. Along with traditional modes of subsistence forest use, recreation and tourism, alternative approaches to timber harvesting will likely be part of a new approach to landscape management for communities that are transitioning from conventional modes of forest dependence. The perceived consequences of changing patterns of forest use, as well as how the dominant ethic communities in the upper Skeena differ (or potentially agree) on the preferred direction for landscape change, are important as well as interesting questions. In the following section, I clarify some assumptions that are implicit in both research questions, as well as present arguments that demonstrate the relevance and need for the following research. A consistent theme for both research questions is the prevalence of theoretical and empirical work over thirty years that has investigated both the determinants of socially acceptable forest management and 5 inter-individual differences in landscape preference evaluation. However, significant gaps and questionable assumptions exist in this considerable body of research, and these are identified as the main justifications for the present investigation. 1.2 Research Justification and Assumptions 1.2.1 Dimensions of Acceptable Forest Management The issue of socially acceptable forest management is of increasing concern to environmental policy makers and land management professionals as they recognise that implementation success depends on practices that respect the values of people living near managed forest landscapes (Brunson, 1996; Hansis, 1995). Over the last three decades, researchers across several disciplines have studied the acceptability of forest management practices, ranging from site-specific studies of alternative silvicultural and landscape design techniques (e.g. Arthur, 1977), to more broad-based examinations of public preferences for multiple-use management frameworks at the landscape scale (e.g. Behan, 1990). A shortcoming of this body of work has been the production of results and recommendations that lack the specificity needed by managers to make workable decisions (i.e. management guidelines, design standards, etc.). A s an example, Schindler and Reed (1996) conducted a site-scale investigation of local support for prescribed fire and mechanical thinning in the Blue Mountains of eastern Oregon and Washington. In addition to demonstrating clear public support for thinning over prescribed fire -likely due to the perceived increase in economic benefits (i.e. employment) and diminished public risk - the authors provided a general set of conclusions which indicated that credible and comprehensible information is essential to garnering 6 public support, and that people need to see management practices with their own eyes; the authors, however, provide no specific management direction. Similarly, Robert Ribe (1994) contributed to a multidisciplinary investigation of landscape management alternatives for a demonstration forest in the U.S . Pacific Northwest. He developed and implemented a multiphase survey of Oregon, Washington and national publics to explore perceptions of scenic effects, conceptions of 'natural' forest conditions, and preferences for recreational uses. From participants' ratings of acceptable forest conditions, Ribe concluded that managers need to be attuned to differing recreation preferences (active versus passive), consumptive versus amenity uses of the forest, the degree of stakeholder knowledge about forest landscapes, as well as the sociodemographic characteristics of stakeholders (i.e. rural or urban residency and income) in their formulation of management objectives, but did not link these to specific management practices. Notwithstanding the general nature of its conclusions, a study that has come to be regarded as seminal in its findings and methods (as well as highly relevant to the present dissertation in terms of its purpose and geographic location) was the work conducted in 1994 on the acceptability of alternative harvest practices on the Tongass National Forest in southeast Alaska (Schindler, et al., 1996). The study was an exploratory investigation to identify issues and areas of concern regarding alternative forest practices, and was designed more to identify potential research questions than it was intended to answer them. A limited number of small group meetings and individual interviews were conducted with residents of the Stikine region (i.e. Petersburg, Kupreanof, and Kake) where participants were presented 7 with photographs of alternative harvest treatments taken throughout southeast Alaska. Participants articulated a wide range of values and uses of the forest, including scenic values and attitudes towards clearcuts, timber production, recreational uses, habitat and biodiversity, subsistence, cultural resources, education, and spiritual values. Perhaps owing to its breadth and exploratory design, the Stikine study demonstrated that a wide range of factors affect people's perceptions of timber harvesting. From a scenic quality standpoint, most people preferred conditions resulting from alternative silvicultural techniques to those resulting from clearcuts, logging slash was generally perceived as a negative element, as was the visible presence of road works. Less tangible concerns were expressed regarding the cost effectiveness of timber operations, the viability of alternative practices, the effects on timber industry employment, and the manner in which alternative practices would affect fish and wildlife populations. The study also demonstrated that people tend to prefer mature forests to newly established stands, natural looking over managed stands, and partial cutting over clearcut harvesting. This investigation summarised the considerations that guide the preferences of southeast Alaskans as follows: People are more likely to find a practice acceptable if they can visualize how it will look, understand its effects on sustaining the natural characteristics of the surrounding forest, believe in the information they have received, feel that the practice will benefit the local community, and that they have had a chance to interact in the planning process (Schindler, etal., 1996). Among other recent investigations (e.g. Brunson and Schindler, 2004; Schindler, et al., 2002), the Stikine study rounds off nearly three decades of social acceptability research (Allen et al., 1998; Arthur, 1977; Benson and Ullrich, 1981; 8 Brunson, 1991; Gobster, 1996; Hansis, 1995; Magill, 1992; McBeth and Foster, 1994; Ribe, 1982, 1989, 1999; Steel et al., 1994; W o o d , 1988). Despite this significant volume of work, important questions remain unanswered or insufficiently addressed in the acceptability literature. For instance, much of this research has attempted to uncover stakeholders' conceptions of acceptable timber management. However, commercial timber management, and the various patterns that it can assume operationally (e.g. clearcut, patch cut, shelterwood, etc.), is one form of forest modification among a broad range of potential forest uses that has the capacity to visibly alter a forested landscape. For communities such as those in the upper Skeena Valley that are transitioning from a near exclusive reliance on industrial forest management, preferences for or the acceptability of forest landscapes that are modified for alternative uses such as agriculture, nature preservation, subsistence and/or timber remains largely uninvestigated. 3 More specifically, are the dimensions of social acceptability that have been described by Brunson (1996) and Schindler, et al. (1996; 2002) conceptually similar when the scope of forest management is broadened to encompass landscape modifications that include cultural/subsistence uses as well as commercial timber management? By avoiding the term managed and instead using the more generic term modified, the wording for Question 1 is intended to encompass any kind of purposive landscape alteration, and thereby avoid the 'timber bias' that is often implicit in the use of the former term. The focus here is on the investigation of preferences for change at the landscape scale. While there are several studies (Arthur, 1977; Benson and Ulrich, 1981; Brunson, 1991; Hansis, 1995; Magill, 1992) that have investigated landscape preferences in foreground and site level conditions, there are comparatively few studies that examine preferences for alternative and, in particular, unconventional landscape scale alterations (e.g. aboriginal subsistence uses). 9 In addition, the scope of social acceptability research has conventionally been quite broad, addressing a range of potential issues that may impinge on the acceptability of landscape management decisions - e.g. the clarity of planning information, the equity of decision-making processes, the technical capacity of planning professionals and stakeholders, risk perception, preferences for alternative silvicultural techniques, etc. In part, this breadth may account for the lack of specifiable planning guidelines discussed earlier. To address this issue, my dissertation research has been focused on presenting and examining stakeholder evaluations of alternative forest landscape conditions (see Chapter 4, 4.4: Stimulus Materials). While broader social issues such as information clarity and process equity are vital components of effective land-use planning (Lewis, 1996; Lewis, 2000), my interest through Question 1 is concentrated on understanding stakeholder perceptions of landscape conditions (i.e. in the absence of extraneous social factors such as the fairness of planning processes, etc), the dimensions or factors that stakeholders consider in their evaluation of landscape conditions, and on the development of landscape design guidelines or standards that may be of practical benefit to professional planners (i.e. foresters, landscape architects, etc.). 4 A s such, the focus of Question 1 is more consistent with landscape preference research (see the definition of 'preference' in footnote 1, page 1) than with a more broad-based social acceptability study (Kaplan and Kaplan, 1989): 4 I acknowledge that the social factors that have been central to social acceptability research are important determinants of socially acceptable or preferred landscape management. Consistent with the preceding body of social acceptability research, I am interested in and will not preclude discussion of both the social and physical dimensions that the research participants use in their evaluations of landscape conditions. However, the focus of Question 1 remains on how local stakeholders respond to physical landscape conditions, and on the examination of the dimensions or factors that are important to the research participants in their evaluations. 10 A few additional semantic assumptions regarding Question 1 need to be highlighted. Question 1 avoids focusing on what forest conditions local stakeholders prefer. While this kind of information is interesting and arises in this study during discussions with research participants (see Chapter 4, Section 4.3.3 - Stimulus Materials), of greater interest is understanding the conceptual basis or reasoning that underlies preference judgements - i.e. the why of landscape preference. Stankey (1996: 101) argues that public acceptability judgments are inherently "wicked" or messy. In effect, landscape evaluations are typically mediated by the simultaneous consideration of multiple consequences and tradeoffs (e.g. aesthetic judgments, wildlife impacts, economic opportunities, etc.), many of which may come into play in the consideration of a single forest management scenario. Moreover, these factors can be situationally dependent to the extent that their relative salience may vary over time and from person to person as environmental and social circumstances change. Therefore, understanding what landscapes are preferred may be less important than understanding why or, in other words, what factors contribute to a preference evaluation by a given person at a given point in time, particularly when unfamiliar or unconventional traditional practices (i.e. cultural subsistence uses) are involved. The question of acceptability has an alluring simplicity to it. However, as Schindler et al (1996) concluded, uncovering socially acceptable landscape management practices is primarily a matter of uncovering and "working through" complex issues with individual communities to identify locally relevant solutions. Question 1 has been formulated with this understanding in mind. 11 1.2.2 Inter-Cultural Landscape Preference Differences Across several disciplines (environmental psychology, environmental sociology, cultural geography, etc.), there is one question that underlies most inquiries into human conceptions of nature: are people from different backgrounds, placed in the same environmental context, likely to differ in their interpretations of the setting? More specifically, will they differ in their preferences for (or against) the setting or conditions, and why should this be the case? Advocates of two extreme theoretical positions are not hard to find, from those who insist, on the one hand, that there is nothing that is not socially constructed (Evernden, 1992) to those, on the other, who contend that every facet of human cognition is written into our genetic constitution (Wilson, 1999; discussed at length in Chapter 2). Those who espouse the latter position contend that all people prefer the same type of landscape -usually natural, savannah-like settings - and that there are inherited mechanisms responsible for these preferences. The accumulation of evidence documenting similar perceptual and preferential responses by subjects of different ethnic backgrounds has favoured the development of a 'consensus assumption,' i.e. the assumption that similarities in landscape evaluations generally outweigh differences across individuals, groups and cultures (Balling and Falk, 1982; Kaplan, 1987; Kellert, 1993). The consensus assumption has been used as an argument for the development of general models for predicting and explaining landscape preferences, often in evolutionary terms (Daniel and Boster, 1976; Kaplan and Kaplan, 1982). Despite a substantial body of empirical evidence, the consensus assumption has not gone unchallenged. Several authors have argued that much of the evidence 12 that supports the consensus assumption has been accumulated under circumstances with high potential for inter-subject agreement (Kahn, 2001; Van den Berg and Vlek, 1998; Yu, 1995). Studies that have employed more heterogeneous samples of subjects and environments suggest that there may indeed be important individual differences in environmental preference evaluations, and that these differences are often rooted in the unique experiences and purposes that individuals bring to their assessment of a landscape. For instance, Hull and Robertson (2000) refer to a number of studies in which they investigated how residents of rural Virginia evaluated forested landscapes in terms of concepts such as ecological health and integrity. According to Hull, these concepts are socially acquired to the extent that their meaning is a function of the intersubjective knowledge and life experience of the perceiver. A core assumption of this dissertation is that a basic cause underlying differences in environmental preference is differences in peoples' basic environmental motives and needs. In effect, the conditions that people prefer in a landscape are fundamentally driven by their material or experiential requirements from the setting - i.e. their needs (see section 2.3.3, Chapter 2). However, human motivations and purposes are largely ignored in experimental designs that take subjects out of their normal day-to-day existence and require them to evaluate unfamiliar landscapes for basic preference or beauty using a Likert scale. Moreover, research designs are often developed using socio-demographic factors (e.g. gender, age, level of formal education, etc.) as independent variables, implicitly assuming that individuals formulate preference evaluations in direct response to their socio-13 demographic characteristics. The validity of evaluations made under these conditions is doubtful, and has been recognized in the environmental preference literature as a potential problem that deserves more rigorous examination (see Chapter 2, 2.3.2: The Social Construction of Landscape Preferences). Returning to the question that was posed at the outset of this subsection - i.e. how and why do individuals differ in their preferences for the same environmental setting? - I anticipate that there will be some level of agreement among persons of different ethnic and cultural backgrounds in their landscape evaluations. This is in part based on my own intuitive sense that landscape preference judgements are not completely idiosyncratic or personal expressions of 'taste' (Kaplan and Kaplan, 1989), and in part based on a substantial body of empirical evidence which suggests that humans tend to be remarkably consistent in their preference evaluations (Kaplan and Herbert, 1987; Kaplan and Talbot, 1988; Kellert, 1993; Nasar, 1984; Orians and Heerwagen, 1992; Ulrich, 1993). Nonetheless, I expect that significant preference differences do exist and, more importantly, that some of these can be accounted for by differences in peoples' basic environmental motives and needs. T o state it simply, people approach the landscape with different personally or culturally driven motivations and requirements (e.g. subsistence, recreation, employment, solitude, ritual). Expectations of and preferences for landscape condition may ultimately be a function of who the user is in terms of their particular environmental requirements, and the landscape conditions that are capable of fulfilling or affording a desirable experience (Gibson, 1986). 14 The significance of Question 2 is perhaps more academic than practical. A n analysis df the environmental preference and perception literature have revealed that in many 'cross-cultural' studies, culture is commonly defined superficially to refer either to ethnicity or national identity (e.g. Tips and Savasdisara, 1987). A s the concept of culture is so all encompassing in everyday discourse, operationalising it for use in empirical studies has proven to be rather difficult, and resulted in some questionable definitions in the environmental preference literature. For instance, it is worth noting that the colloquial conception of culture invites us to use the concepts culture, society and nation interchangeably. However, comparing two nations, as is often done, does not necessarily imply that we are doing a cross-cultural comparison because the nations may be closely related historically and/or ethnically - e.g. Australians and Americans (Kaplan and Talbot, 1988; Orians and Heerwagen, 1992; Zube and Pitt, 1981). Moreover, a common methodological flaw in many preference studies stems from the fact that, although participants are selected from different countries (e.g. Korea and the United States), most use urban or suburban residents as subjects (Yang and Kaplan, 1990; Y u , 1995). These subjects are likely to have similar experiences with forested landscapes; experiences that are limited to a few camping trips, scenic drives, television documentaries, and coffee table picture books. T h e s e subjects are likely to evaluate such landscapes with recreation in mind rather than the purpose of timber extraction or plant gathering. Thus, the limited between-person variance reported in the bulk of landscape preference literature may not be due to an inherited scenic perception schema, but to research 15 designs that use an inadequate or inappropriate conception of culture in selecting subjects that share certain broad cultural similarities. In a more practical or applied sense, the value and significance of cross-cultural preference research lies in its relationship with specific planning and resource management objectives. It should provide environmental planners and resource managers with a better understanding of the evaluative processes of local publics. Matters of concern to professional land managers should include the range of resource and land-use values held by local communities, the relative strength of those values, and the preferred direction for landscape management and change. For instance, do aboriginal subsistence users, tourism operators and agriculturalists agree on the identification of lands to be protected because of their scenic value, the presence of rare plant and wildlife species, or the appropriateness of adjacent land-uses? The significance of such information rests in the insights and understanding that they provide for the design of landscape management policies and practices. To paraphrase Firey's hypothesis (1960), for resource policies and plans to be anything other than opportunistic ventures, they must be ecologically possible, economically gainful and culturally adoptable. 1.3 Organization and Content of the Dissertation The remainder of this dissertation is organised into the following chapters: Chapter 2: a review of the literature addressing the contingent nature of human environmental preferences. Part one of the chapter describes the competing biological and social constructionist paradigms in preference research, and attempts to forge a conceptual synthesis by suggesting that both biologically as well as 16 socially based mechanisms may function in the formation of environmental perceptions and preferences. The remainder of the chapter addresses the role of human values in preference evaluations, how environmental value orientations structure preference evaluations, and how differing conceptions of and experience with 'place' (i.e. place identity and place dependence) may account for inter-individual and inter-group variations in landscape appraisal. Chapter 3: this chapter provides an historical overview of the landscapes and cultural context of the upper Skeena valley. In addition to providing a regional history and physical context for the reader, the discussion examines the history of landscape change in the upper Skeena both prior and subsequent to European contact with the Gitksan people. From a period of subsistence-dominated landscapes that were largely maintained through the purposive use of fire, through agriculture and the intensive forest management that was the hallmark of the late twentieth century, this chapter argues that large-scale forest modification has a lengthy history in the Gitksan traditional territories and is not exclusively the legacy of European colonization. As well, the landscape patterns and forest disturbances that are described in chapter three provide the framework for the management scenarios that are presented in the following methods chapter. Chapter 4: This chapter describes the overall research strategy and methodology that was used in the dissertation to obtain landscape management preferences and supporting information from the Gitksan and Euro-Canadian residents of the upper Skeena valley. A procedural account of the methods that were used for data collection and analysis is provided, including the use of visualizations in interviews to 17 solicit preferences for patterns of landscape modification, accompanied by a theoretical rationale for each step taken. The chapter concludes with a description of the measures followed to ensure the validity and reliability of the findings, and how the ethical principles of social science research were followed. Chapter 5: This chapter outlines the main study results through a detailed analysis of interview transcripts and personal observations using insights gleaned from the literature review, observations from several visits to the Hazeltons and nearby communities, and conversations with key individuals within the upper Skeena region. The responses of interview participants to a series of visualizations of alternative modified forest scenarios are presented using both qualitative and quantitative data analysis and presentation techniques. Chapter 6: This chapter concludes the dissertation with a synthesis of the major findings from the analysis chapter and a discussion of their implications for landscape planning and research. Chapter 6 reintroduces the research questions and summarises the main findings for each. In turn, each finding is discussed in terms of its research and policy implications, with a focus on addressing the broad recommendations that have come from social acceptability research to date. Particular attention will be paid to the issue of visible cues or recognisable patterns of human intention and care in landscape alterations, and how the form that these cues may take varies between stakeholder groups. The chapter concludes with a consideration of the unanswered questions and opportunities for future research that result from this investigation. 18 2 Literature Review 2.1 Introduction Many people find it important to protect and enhance the natural values of their environment (Kellert, 1985). T o achieve this, several kinds of landscape management strategies may be applied, ranging from active strategies that guide natural processes by means of regulative activities, to more passive strategies that encourage the development of more spontaneous or uncontrolled natural processes by minimizing human activities in an area (Christensen et al., 1996). Decisions to apply a particular kind of landscape management strategy are typically based on a mixture of economic, ecological and social criteria. With respect to ecological and economic criteria, there is some scientific consensus about the relative advantages and disadvantages of active versus passive landscape management. However, with respect to social criteria, there is increasing evidence that people will differ in their relative preferences for landscapes with high or low degrees of human influence (Balling and Falk, 1982; Daniel and Boster, 1976; Dearden, 1984; Strumse, 1996). People living in rural settings who are dependent on the local environment for their livelihood generally appear to favour landscapes with a high degree of human influence (Hull, et al., 2001), while members of environmental organizations, younger people, people with high socioeconomic status, and exurban residents of wilderness areas have been found generally to favour landscapes with a low degree of human influence (Balling and Falk, 1982; Dearden, 1984; Y u , 1995). While evidence for inter-individual and inter-group differences in preferences for managed landscapes continues to accumulate, the hypothesised biological, 19 social and ecological mechanisms underlying these differences remain poorly understood and, from a conceptual standpoint, inadequately integrated. The present chapter atempts to shed more light on this issue by exploring the conceptual frameworks for understanding preference judgments, as wel as the bases for inter-group and inter-individual differences in landscape preference. Conceptualy, the chapter is organised using the framework shown in Figure 2.1: Figure 2.1: Chapter 2 Organisation Landscape Preference Theories Section 2.2 Biological Theories Section 2.2.1 Social Theories Section 2.2.2 Determinants of Preference Section 2.3 Sociodemographic Theories Section 2.3.1 Social Constructivism Section 2.3.2 Place Theories Section 2.3.3 First, the chapter wil criticaly discuss the hypothesised biological and social determinants of landscape preference (Section 2.2). From this review, it wil be argued that the major biological and social theories are not mutualy exclusive, and that a conceptual synthesis exists which links both biological and social factors as distinct but integrated elements of landscape evaluation. Section 2.2.2 wil expand on the social paradigm introduced and address the common (mis)conception that 20 preference evaluations are idiosyncratic judgments rooted in highly subjective or personal expressions of taste. A common assertion by proponents of biological preference theories is that social experiences vary to such an extent between individuals that it is difficult, if not impossible, to make generalisable statements about group preferences and their determinants (Brown et al., 1988). Using theoretical perspectives from the field of environmental sociology, this section will argue that human environmental preferences can be understood and studied methodically within a value orientations framework. T h e second half of the chapter (2.3 Determinants of Preference) moves beyond theoretical perspectives that describe inter-individual and inter-group variations in preference judgements and presents some of the factors that are commonly hypothesised to influence preference evaluations - i.e. sociodemographic characteristics (section 2.3.1), social processes and intersubjective learning (section 2.3.2), and the influence of the physical environment itself in conditioning preferences (section 2.3.3). 2.2 Landscape Preference Theories 2.2.1 Biological Theories of Preference During the last 25 years, a large number of empirical studies of affective and cognitive evaluations of landscapes have been published (see reviews by Bourassa, 1990; Kahn, 2001; Kaplan and Kaplan, 1989; Zube et al., 1975). O n e consistent finding has been that people from various ethnic and national backgrounds prefer (i.e. like) natural environments better than built or otherwise human-influenced environments (Hodgson and Thayer, 1980; Kaplan and Kaplan, 1989; Hull, et al., 2001), and that park or savannah-like landscapes often receive the highest ratings 21 (Orians and Heerwagen, 1992; Ulrich, 1993). Human influence may be accepted, however, provided that it is perceived as being in balance with natural elements -e.g. nature scenes containing old structures such as stonewalls or bridges (Simpson et al., 1976; Strumse, 1994). In combination with findings that demonstrate that natural scenes contribute to restoration from stress (Ulrich et al., 1991), these studies have been interpreted as supporting evolutionary theories of human landscape preference. Moreover, several researchers have concluded that similarities in evaluations of natural scenes far outweigh the differences across cultures or smaller groups (Kaplan and Talbot, 1988; Tips and Savasdisara, 1986; Ulrich, 1993; Yang and Brown, 1992). Most environmental psychologists agree that landscape evaluations are driven by peoples' basic motives and needs (Kaplan and Kaplan, 1989; Orians, 1980; Ulrich, 1983). Since certain environments provide better chances for survival than others, it seems likely that biologically driven mechanisms have evolved that compel people to seek out and prefer the most adaptive environments (Appleton, 1975; Kaplan, 1987). For instance, Stephen Kaplan (1987) has suggested that people today are still attracted to landscapes that enabled earlier primates to construct and exploit the cognitive maps they relied on for survival. In a similar vein, Ulrich (1983) has posited that adaptive behaviour in natural environments is directed by initial affective reactions elicited by environmental characteristics that triggered important survival cues during the early stages of human evolution. Since all humans are presumed to have evolved in the same savannah-like settings (Appleton, 1975; Wilson, 1984), it is often inferred that each individual's motivational 22 system is adapted to similar kinds of environmental information, thereby yielding similar kinds of landscape preferences. A n important as well as problematic implication of this view is that biologically driven mechanisms of human perception cannot be used to explain inter-individual variation in landscape preferences, because the strength of these mechanisms is assumed to be similar across individuals. T o address this problem, biological theorists posit that genetic variation is a key facet of evolution, and that such variation may facilitate human adaptation to different or changing environments. For example, Tooby and Cosmides (1990) have argued that while humans as a species have evolved complex adaptive mechanisms to promote their survival, the degree to which each individual possesses or utilises these mechanisms may vary (Kaplan and Kaplan, 1982; Lyons, 1983; Ulrich, 1983; Wohlwill, 1983). However, Bourassa (1988, 1990) and Hartig (1993) have worked towards more integrative theoretical frameworks, and argue that both biological and cultural factors are important determinants of environmental appraisals. Hartig (1993), for instance, describes biological and cultural factors as different mechanisms for human environmental adaptation that reflects our gradual transition from natural to culturally modified environments. In general, integrative theorists (e.g. Staats and Hartig, 2004) suggest that biology does not function as the sole or even primary causal factor in environmental preference, and that long-term exposure to different cultural, economic or other social experiences may give rise to differences in the extent to which different people rely on innate or genetically programmed perceptual frameworks. In other words, landscape preference may 23 operate at two levels of awareness (Kaplan, 1987) - i.e. innately driven, tacit awareness and experientially produced, conscious awareness. Empirical evidence to support the biological substructure-social superstructure conception of landscape preference comes from a number of sources. In a child development study by Balling and Falk (1982, summarised by S. Kaplan, 1992 and Orians and Heerwagen, 1992), a direct test of biological theories of landscape preference was conducted. Research participants from the eastern United States were asked to rate five different environment types representing rain forest, mixed hardwood forest, boreal forest, East African savannah, and desert. Based on Orians and Heerwagen's (1992) savannah hypothesis, Balling and Falk hypothesised that the younger children in the participant sample (eight and seven year olds) would prefer the savannah to the other four environments. In turn, since familiarity with an environment was hypothesised to be a factor in environmental preference, and since increasing age would correspond to children's increasing familiarity with the surrounding hardwood landscape, it was also hypothesised that the older children would equally prefer the hardwood forest and the savannah, and prefer both environments over the other three. Both hypotheses were confirmed, and Balling and Falk concluded that, in the absence of direct experience, young children prefer savannah environments, thus demonstrating that biology may be the dominant factor in early childhood environmental preferences. However, preferences for savannah-like environments are not immutable to the extent that the effect of biology appears to diminish as subjects accumulate experiences and knowledge of other environments as they mature. 24 A n additional description of how landscape preference may operate at two levels of awareness has been offered by Kaplan and Kaplan (1989), who suggested that at one level, landscape preferences are driven by unconscious and universal needs to explore (i.e. complexity and mystery) and understand (i.e. coherence and legibility) the environment (Figure 2.2). At a more conscious level, however, landscape preferences may be expressed in terms of personally salient and highly Figure 2.2: Landscape Preference Matrix Understanding Exploration Immediate Inferred Coherence Complexity Legibility Mystery Kaplan and Kaplan, 1989: 53 idiosyncratic spiritual or psychological needs, such as the desire to be away, to meditate, to be fascinated by nature, or to be spiritually connected to the natural context (see also section 2.3.3; Kellert, 1996; Lewis, 2000). Thus , when applied to landscape evaluation, even if it is assumed that people have developed a general adaptive preference for savannah-like landscapes that fulfill a common profile of basic needs, a particular individual may nonetheless prefer to experience a 25 coniferous forest because his or her personal experience has influenced the relative weighting of their basic environmental needs. Accordingly, it is reasonable to expect that both cultural as well as biological factors may contribute to evaluative responses to landscape types (Bourassa, 1990; Hartig, 1993; Ittelson, 1973; Staats and Hartig, 2004; Hartig, 2004 - personal communication). It has been shown that landscape preferences may change with age (Balling and Falk, 1982; Lyons, 1983; Zube et al., 1983), and that they may differ across demographic groups (Gonzalez-Bernaldez and Parra, 1979; Hull and Revell, 1989; Purcell et al., 1994). Yu (1995) demonstrated that living environment (urban versus rural), education level, and occupational interests influence the landscape preferences of Chinese participants. In a Dutch study (van den Berg et al., 1998), farmers' preference ratings for various landscapes differed from the ratings of visitors and non-farming residents. However, even though landscape evaluations may be affected by factors that are salient to particular individuals or groups, they are not as idiosyncratic or haphazard as some may suggest (Brown et al., 1988; Hull and Stewart, 1992; Zajonc, 1980). A s the following discussion posits, an environmental values perspective may provide the conceptual framework needed to understand how individuals (or groups) differ or are consistent in their landscape preference evaluations. 2.2.2 Social Theories: Environmental Value Orientations Our affective or aesthetic preference for landscapes is not an isolated mental process (Kaplan, 1982). Rather, it is associated with the affective and cognitive frameworks that each one of us possesses regarding our relationship with society as 26 well as our physical surroundings. According to social adaptation theory (Homer and Kahle, 1988), values are a form of cognition 1 that facilitates adaptation to the environment. In general terms, values are general and important beliefs, life goals or standards that serve as guiding principles in our lives (Golledge and Stimson, 1987; Rokeach, 1973; Schwartz, 1994), and they are thought to condition attitudes and behaviours towards more specific aspects of our environment. More specifically, Rokeach defines values as: A n enduring belief that a specific mode of conduct is personally or socially preferable to an opposite mode of conduct or end state of existence (1973: 5). In addition to more general values, we may have values that are oriented towards a more restricted part of the environment called basic beliefs or value orientations (Fulton et al., 1996). For instance, most people evaluate environmental issues such as forest management or trans-boundary pollution in a particular manner, thereby expressing their environmental value orientation. Attempts to measure aspects of environmental value orientations were made by Dunlap and V a n Liere (1978) and Dunlap et al. (2000). Their instrument, the New Environmental (Ecological) Paradigm (NEP) Scale, has received wide use. According to Dunlap and Van Liere, ideas such as 'limits to growth' and the importance of 'preserving the balance of nature' represented a challenge to previously held beliefs that the physical environment could support unlimited growth (Dunlap and V a n Liere, 1978; Bengston, 1994). The socially conservative view of the human-nature relationship is one in which there is a belief in economic growth, material abundance, and humans Cognition here is defined as the mental process of knowing or being aware of thoughts or perceptions, including understanding and reasoning. 27 are regarded as being above the rest of nature. Coined the "dominant social paradigm" (DSP) by Pirages and Ehrlich (1974), such views represented society's anti-environmentalist thrust. Dunlap and Van Liere argued that with the upsurge of environmental awareness in the developed world (i.e. mostly in Europe and North America) in the 1970s, a new set of ideas emerged which challenged the D S P . They set out to measure this developing set of ideas and developed the N E P concept to describe this value orientation. In an effort to refine the N E P - D S P framework, Stern and Dietz (1994) distinguished between the egoistic value orientation (environmental problems may harm the individual), the social-altruistic value orientation (problems may harm other people), and the biocentric value orientation (nature has intrinsic rights that exist independent of human interests). Thompson and Barton (1994) proposed a modification of this schema whereby the egoistic and social-altruistic orientations would be combined into what they termed the anthropocentric orientation, essentially because both original orientations are expressions of human interests in relation to the environment. According to Eckersley (1992), an anthropocentric orientation is an expression of the dominant social paradigm or a "human centred orientation toward the non-human world." Moreover, it assumes that the non-human part of the environment is "material to be used by humans as they see fit" (Scherer and Attig, 1983), whereby the natural world is "reduced to a storehouse of resources and is considered to have instrumental value only" (Eckersley, 1992). Thompson and Barton (1994) recognised the biocentric orientation originally proposed by Stern and Dietz (1994) but renamed it the ecocentric orientation. A s 28 originally conceived, it elevates the requirements of all organisms and natural processes to center stage and, in some versions, makes the earth or nature as a whole the focus of moral consideration. In contrast with the anthropocentric orientation, ecocentrism is a reflection of the New Environmental Paradigm to the extent that it asserts that human economic uses and benefits are not necessarily the most important uses of the environment. There are non-economic values associated with the natural environment, and these values are just as distinctive and, in some cases, are more valuable than economic values. Figure 2.3: NEP-DSP Continuum Ecocentric V^*-^__ Most People Anthropocentric Nature has intrinsic Nature is a storehouse rights independent of of resources with human interests. instrumental human value. Dunlap and Van Liere, 1978; Johnson, etal., 2004. For several years, the anthropocentric-ecocentric (or D S P - N E P ) value dimension has been applied to the investigation of values and attitudes towards forest management (Coufal, 1989; McQuillan, 1990; Rolston and Coufal, 1991; Schindler et al., 1993; Steel et al., 1994), and has been used to account for differences between European and Native North American environmental worldviews (Martinez, 1998). Biocentric and anthropocentric value orientations, however, are not mutually exclusive. Rather, these value orientations can be arrayed along a continuum with biocentric viewpoints on one end and anthropocentric orientations on the other (Figure 2.3). The midpoint of this scale 29 represents a mixture of the two extremes and reflects the value orientations held by the greater part of the public. Research conducted in Oregon (Schindler era/ . , 1993; Steel et al., 1994) and Colorado (Vaske and Donnelly, 1999) largely supports this conceptual continuum. Coming back to research on landscape preference, a connection can be drawn between the study of environmental value orientations and landscape preference. Many people may express a preference for a particular landscape, but for different reasons. S o m e may value and express a preference for a particular landscape condition because it serves or reflects anthropocentric needs, while others may emphasise ecocentric values. For example, a 'pristine,' forested landscape may be preferred because it can be harvested, and thus serve instrumental human needs (an anthropocentric orientation) or, alternatively, because it has intrinsic value as an intact ecosystem (an ecocentric orientation). Moreover, when individual or inter-group differences in landscape preferences are found, they may be closely aligned with underlying differences in values and attitudes towards specific environmental issues. For instance, when Chinese farmers responded negatively to water dominated and misty rocky scenes (Yu, 1995), or when Dutch farmers responded negatively to development plans involving restoration of natural wetlands and forests (van den Berg et al., 1998), part of the explanation was attributed to the farmers' relatively anthropocentric environmental orientations. However, empirical research on environmental values and preference ought to go beyond the mere determination of degrees of consensus in landscape preference appraisals. A s suggested at the outset of this chapter, the detection of 30 inter-individual or inter-group differences in itself is not a sufficient basis to reject biological explanations, or to adopt a completely social-deterministic framework. In order to be of theoretical and practical relevance, empirical research should not only provide information on the relative occurrence of individual differences in environmental values and landscape appraisals, but also on the determinants of these differences in terms of individual and social background variables, landscape characteristics, and characteristics of the judgement context. Possible causal factors underlying individual and group differences in landscape preference are explored in the following section. 2.3 Determinants of Landscape Preference From the environmental perception and preference literature, three general factors have been gleaned that account for individual and group differences in landscape preference: (a) sociodemographic differences, (b) social or group influences that condition normative values and beliefs regarding the environment, and (c) 'place-based' theories which posit that the physical environment itself structures human interactions with and knowledge of the landscape, and thereby creates differing human-place bonds. These are presented and critically evaluated in the following three subsections. 2.3.1 Sociodemographic Characteristics Group-based social attributes, often defined as sociodemographic or census based characteristics (e.g. age, income, education, gender), have long been an important feature of research concerning environmental values and preferences (Dunlap and Catton, 1983; Milbrath, 1984; Van Liere and Dunlap, 1980, 1981), and 31 a variety of different characteristics are thought to influence landscape perceptions and preferences. Various studies have demonstrated that sociodemographic characteristics such as age, income, place of residence (i.e. urban versus rural), and having a 'green' political preference are significant moderators of the preferred balance between natural and human-influenced patterns in landscapes (Balling and Falk, 1982; Dearden, 1984; Lyons, 1983; van den Berg era/ . , 1998). Age and Generational Influences Hypothesised age related preferences can be seen in several studies. A s discussed earlier, Balling and Falk (1982) found significant age-related changes in preferences for different landscapes, and that underlying preferences change with experience across an individual's life span. Abello and Bernaldez (1986) argued that preference for landscapes that demonstrate complexity, unpredictability and incongruence in natural or photographed scenes is a function of age, with a greater tolerance of complexity, uncertainty and surprise coming with greater maturity. Zube et al. (1983) suggest that some hierarchy of functional importance exists to the extent that young people are highly sensitive to the most elementary functional aspects of landscapes, particularly in terms of the environment's ability to facilitate play or 'fun' experiences. With cognitive development and experience, other more complex, information-based elements of coherence, complexity and mystery emerge. They further suggest that there may be a cohort effect associated with the scenic assessments of older adults, which may result from their socialisation prior to the advent of the environmental movement of the 1960s and 70s. Such a hypothesis ties in closely with theories of change in social values which posit, in 32 effect, that citizens in Western countries born after the Second World War are considered to be more likely than their parents' generation to focus on environmental concerns (Dalton, 1988; Inglehart, 1990; Inglehart and Flanagan, 1987). Consequently, age is often regarded as an important factor in any environmental preference study. The Role of Gender Numerous studies hypothesise that there may be a link between environmental values, preferences and gender (Tindall, 2003). However, research that uses gender as a predictor of environmental preferences and value orientations demonstrates a mixed pattern of findings. Van Liere and Dunlap's (1980) review of 21 studies from the 1960s and 1970s, for example, suggests that while some research has observed a correlation between sex and preservationist environmental values, other investigations have not reported a significant relationship. In a study supporting sex differences, Steger and Witt (1989) found that women are more likely than men to support pro-environmental value orientations and hold more preservationist attitudes. Other reviews (Mohai, 1992) come to a similar conclusion, but suggest that women are more concerned with local, as opposed to national environmental issues than men. Several researchers (Blocker and Eckberg, 1989; Davidson and Freudenberg, 1996; Van Liere and Dunlap, 1980) hypothesise that women are socialised to perceive moral dilemmas in terms of interpersonal relationships, and to resolve them by an ethic of care and cooperation. Males, by contrast, are socialised to perceive moral dilemmas in terms of more impersonal features of situations and to 33 resolve them in an emotionally detached manner, largely through appeals to objective rules of justice and rights, or through the inherently male traits of competitiveness and dominance seeking behaviour. Such differences in socialisation may lead women to take a more protective or preservationist (i.e. ecocentric) view of nature (Mohai, 1992; Steger and Witt, 1989), with men assuming a more detached, utilitarian and anthropocentric perspective. Taken together, however, while the empirical evidence is inconclusive, most researchers conclude that females are somewhat more environmentally oriented than males, particularly when the focus is on local landscape management issues (Mohai, 1992). Formal Education Educational attainment is generally included as an independent variable in environmental preference studies because it is broadly associated with having a strong impact on environmental values (Inglehart, 1990, Milbrath, 1984). However, research examining the relationship between education and environmental values also demonstrates a mixed pattern of findings. Steel et al. (1994), for example, found education to be a significant predictor of the biocentric-anthropocentric value orientation in their Oregon sample, but not in their national sample. Among the Oregon respondents, those with more education were more biocentric. In contrast, other research has reported an inverse relationship between education and biocentrism (Grendstadt and Wollebaek, 1998). However, these latter findings contradict the bulk of the literature, which demonstrates that higher education is associated with biocentric value orientations. According to most research findings, individuals with higher levels of education are significantly more likely to have value 34 orientations that are sympathetic to environmental concerns, and are more inclined to prefer landscapes that show no evidence of human impact or use than those with less formal education (Dearden, 1981; Harvey, 1995). According to Howell and Laska, this relationship is not surprising because: ...the evidence on both sides of an environmental issue frequently address a very complex etiology of causes comprehended more easily by the better educated (1992: 141). Landscape Familiarity Familiarity is noted as a factor affecting the perception of landscapes, and several researchers have looked at this factor with differing results (Kaplan and Kaplan, 1989). Prior experience with a landscape has a "profound influence" on human perception and preference, according to Balling and Falk (1982), who state that landscape preference is undoubtedly not just a function of some innate or biologically driven preference framework. Purcell (1992) comments that humans experience each new or previously encountered landscape within the context of mental models of prior landscape experience. Lyons (1983) found that preferences were highest for the most familiar environment, thus supporting the hypothesis that a person's landscape preference is strongly influenced by their residential experience in different environments. Penning-Roswell and Lowenthal (1986) note that perception and cognition are inseparable from experience, but speculate how familiarity creates affection for or indifference to particular landscapes. The study by Wellman and Buhyoff (1980) demonstrates that their participants did not show greater preference for a particular familiar environment and that participants drawn from widely disparate regions evaluated landscapes, in terms 35 of preference, in essentially the same manner. This result is supported by the work of Tips and Savasdisara (1986) who found that travel experience has virtually no influence on landscape preference; therefore suggesting that exposure to a wide variety of landscapes, hitherto unknown, has no discernable effect on preference. Ulrich (1993) states that the findings of the last few decades have led to the conclusion that the similarities in response to natural scenes usually far outweighs differences between individuals and groups with differing experiences or from different national backgrounds. Urban versus Rural Residence Place of residence, typically expressed as urban versus rural, suggests that urban populations are more likely to have pro-environmental values due to their better access to information and education opportunities, and because they "are more likely to experience environmental problems firsthand due to industrial activities and high concentrations of people" (Howell and Laska, 1992: 141). Moreover, when compared with the urban public of B C ' s Lower Mainland, citizens of rural British Columbia may have a somewhat different value orientation towards forests due to identification with the natural resource extraction culture and industry. O n e broadly held view suggests that much of the concern over preserving old growth forests originated with urban based environmentalists and policy makers (Wilson, 1998). A s Egan has commented: Environmentalists have learned that taking their case to a larger audience may be the best strategy for preserving forests (1991: 26-7). These findings have also been extended to exurban residents of rural communities. Length of residence in a community may influence a person's general 36 environmental value orientation and specific attitudes about acceptable landscape management practices (McCool and Martin, 1994; Rudzitis, 1999). In rural communities, 'culture clashes' between newcomers and long-term residents are a main form of growth-related conflict (Blahna, 1990). Newcomers often believe that the rural landscape should be preserved whenever possible, while long-time residents prefer land management strategies that balance protection and use (Bengston, 1994; McCoo l and Martin, 1994). For long-time residents this means that traditional environmental value orientations linked to agriculture, forestry, or ranching are challenged by new residents (Rudzitis, 1999). Culture and Ethnicity With regard to ethnicity and national identity, Guha's (1989) assessment of environmental preservation, particularly among aboriginal residents of Third World countries, provides insight into the cultural and ethnic basis of environmental preference. According to G u h a , landscape preservation for the sake of biotic communities can be disadvantageous for the indigenous populations of some Third World countries because traditional communities may be displaced through the designation of ecological preserves. The designation of these areas in the Western sense of nonhuman habitation effectively transfers land rights from the peasantry to wealthy power agents. Parajuli (2001) also writes that the nature/culture dichotomy that is so prevalent in Western conceptions of the environment is unfamiliar in traditional or indigenous societies throughout the world. For these "ecological ethnicities" (i.e. ethnic or other societal groups that practice sustainable forms of harvesting or resource management), there exist no discrete units of territory called 37 'wilderness' or 'nature' that exist in contrast with places where people dwell (Ingold, 2001). Western (2001) raises similar concerns about wild game preserves in Africa, arguing that the idea is foreign to indigenous groups (see also Southgate and Clark, 1993). Inglehart's (1990) postmaterialist thesis is often used as an explanation for the examination of indigenous and non-indigenous conceptions of landscape. According to Inglehart (1990, 1995), the emergence of scarcity-free, postmaterial societies in the latter half of the twentieth century allows individuals in those societies to adopt more egalitarian attitudes and express values that are more inclusive of others in society. Since the material needs of many industrialised societies and non-indigenous, Euro-Canadian communities are satisfied, more people are able to concentrate on issues and concerns that transcend their most fundamental needs - e.g. racial equality, animal rights, and environmental protection. Nash (1982) refers to this as the "full stomach phenomenon." Again, because many indigenous communities have yet to realise postmaterialist conditions on a par with non-indigenous peoples, Inglehart (1990, 1995) would predict that indigenous people would be less attuned to urban, European conceptions of landscapes (Altieri and Masera, 1993). Several years prior to the publication of Inglehart's thesis, Van Arsdol, Sabagh and Alexander (1964) studied the perception and reality of environmental hazards in ethnic communities of Southern California. A m o n g other things they found that people who were most exposed to air pollution and noise (urban non-white residents) tended to be least aware and concerned about these things, while 38 people who were least exposed (suburban white residents) tended to be the most aware and concerned. Th e authors concluded: It may be that, while nonwhites are concentrated in smog areas, the intrusion of social hazards in such areas and the preoccupation of such populations with other social problems may have obscured their perception of environmental hazards (1964: 49). Perhaps adaptation and numbing have a role. Whether such differences in environmental concern extend to landscapes is an empirical question, as landscape management may be a more esoteric concern to socially and ecologically disadvantaged communities than other environmental issues. Moreover, it needs to be acknowledged that as cultural groups, neither indigenous, immigrant, nor native-born North Americans are monolithic communities. Wide variations exist in culture within ethnic communities (see Chapter 1, section 1.1.2; Lewis, 2000), even communities coming from the same nation or region of the world. Nevertheless, by examining racial/ethnic minority responses to questions about an array of landscape attitudes and preferences, we gain some indication of the values attached to these lands by important constituent groups. 2.3.2 The Social Construction of Landscape Preferences The difficulty with sociodemographic explanations of environmental preference is that, while social characteristics may reflect the environmental orientations and preferences of a group at a given point in time, they are insufficient causal explanations of the environmental values and preferences held by different groups. In effect, people do not form value orientations or preferences in direct response to their sociodemographic characteristics (Kruger, 2004). While there may be some association between certain socio-demographic characteristics and 39 preference, the causal relationship is sometimes uncertain or variable. Indeed, the environmental value orientations of women and ethnic minorities, for instance, may change over time as their respective roles and positions in society shift. However, implicit in many value orientation and preference studies that structure sample groups according to gender or ethnicity is the notion that it is the social processes which underlie sociodemographic characteristics that determine how values and preferences take shape (Erickson, 1988; Mohai, 1992; Steger and Witt, 1989). Social constructionists debate the extent to which the natural environment can be known independently of the social context that shapes the purpose and process of knowing (Bird, 1987; Evernden, 1992; Escobar, 1999; Proctor, 1998). Greider and Garkovich (1994) explained that landscapes are symbolic environments used by people to define themselves. Therefore, the diversity of environmental preferences reflects the diversity of cultures, values, beliefs, and purposes of the people doing the defining. For example, the real estate developer looks across an open field and sees the potential for a housing subdivision, a farmer envisions a field of wheat, and a hunter sees deer feeding patterns (Steele, 1986). T he conceptual foundation for the social constructionist paradigm can be found in several sources, but among the most influential has been the work of Clifford Geertz in an article published in the mid 1960s on "the impact of the concept of culture on the concept of man" (1964). According to Geertz, human perception of the world is constructed according to a certain order, which ultimately comes from an intersubjectively defined set of categories and symbols that are imposed on the flux of experience. For any one individual, the range of symbols and conceptual 4 0 categories that can be elicited to make sense of the world is more or less fixed by the community into which the perceiver is born. Without the guidance of culturally constructed frameworks of meaning, humans would essentially be lost, unable to establish their bearings in the world. Unlike many other organisms that depend more on innate or genetically determined mechanisms in order to function within an environment, humans depend on a substantial amount of supplementary information that is learned: Undirected by cultural patterns . . . man's behaviour would be virtually ungovernable, a mere chaos of pointless acts and exploding emotions, his experience virtually shapeless (Geertz, 1973: 46). Thus , to return to the basic question that drives much of landscape perception research, the reason why two individuals perceive the same environment in different ways (Ingold, 2000), would be that each has brought a different symbolic system from distinct communities to bear in organising the same sensory input. In turn, what people like in a landscape, particularly in terms of whether or not that environment conforms to notions of 'pristine' or 'natural,' for instance, will ultimately be determined by socially constructed categories of naturalness. For instance, empirical investigations of how different stakeholder groups construe 'nature' or 'natural' have found that these constructs vary according to definitions or frameworks that have been socially acquired (i.e. learned). Two studies seem especially relevant to landscape managers. Peterson (1995) found that the conflicts between environmental regulators and the landowners being regulated were attributable to differences between the groups' respective conceptions of 'natural' landscape conditions. In addition, Richardson et al. (1996) 41 found that natural resource professionals' understandings and expectations of the landscapes that they managed varied according to their respective disciplinary training and agency affiliation. Further examples of how preferences for landscape condition are affected by dominant social networks come from studies that demonstrate that people living in or near 'natural' or rural landscapes tend to view evidence of human use as appropriate, acceptable, and compatible with the landscape. In contrast, seasonal tourists, visiting recreationists, and other people 'from away' tend to view the same visible markers of human influence in the landscape as a degradation of valued natural qualities (DuPuis and Vandergeest, 1996; Weaver, 1996). For instance, Senecah (1996) found differing statements of preference for the future of the Adirondack Park in upstate New York. O n e group, mostly local residents that wished to continue living in the area, were vilified as 'greedy' developers, 'taming' the wilderness, 'marring' vistas, and otherwise threatening a 'vulnerable' nature. Those wishing to exclude or minimise development in the area were construed as 'nature Nazis' or 'forest faggots' wishing to create a 'scenic gulag' that did not value the local history and culture of the region. Similarly, the side of the debate composed mostly of seasonal visitors and people living outside the park boundaries sought to minimise evidence of human use of the landscape. They characterised the area as a 'spiritual' retreat, a place of 'timeless' splendour, and the 'forever wild' jewel that ought to remain a biological 'treasure chest.' T o one side of the debate, human use and the landscape can coexist and even complement one another. Those on the other side of the debate believed that human presence could only 42 degrade the environment. These different perspectives can be interpreted as an expression of the old and overly simplistic human-nature dichotomy wherein the natural and the human exist as polarised, oppositional categories. Two important factors that would obviously affect environmental value orientations, particularly in relation to forest landscapes, are connections to the forest industry and membership in an environmental organisation. In effect, an individual's value orientation will be affected by where they stand in relation to the productive arrangements of society (Douglas and Wildavsky, 1982; Ribe 2002). Persons who rely on the timber industry for their economic well being, for instance, are more likely to consider commodity interests in their assessment of environmental condition. By contrast, environmentalists may tend to view forests more in terms of a public good and work to promote the preservation of forest landscapes (Dennis and Zube, 1988). S o m e observers have identified environmentalists and other contemporary social movements as engaging in 'elite challenging' strategies due to the grassroots nature of most movements and the use of unconventional forms of political participation to achieve their policy objectives (Inglehart, 1990). 2.3.3 Differing Human-Place Attachments C o m m o n to more current research into environmental perception and preference is a three-component view that weaves together the physical environment, human behaviours and social and/or psychological processes (Brandenburg and Carroll, 1995; Canter, 1977; Relph, 1997; Stedman, 2002). A major shortcoming of empirical research, particularly in the social constructionist paradigm, has been the absence of theoretical frameworks which posit that 43 preferences may be formed in response to the physical environment, focusing instead on environmental preferences as products of socially constructed behaviours and cultural processes. Recent research on the significance of 'place' attempts to address this disconnect, suggesting that the physical environment itself contributes to environmental perception and preference through specifiable mechanisms (Brandenburg and Carroll, 1995; Hashisaki, 1988; Stedman, 2002). Although social constructions are important, they hardly arise out of thin air. In effect, the local environment, in conjunction with social processes, sets bounds and gives form to what people know about a landscape and the conditions that they prefer. There has been a surge of interest in recent years into the investigation of human-place bonds. Most of this work has fallen under the study of 'place attachment' and related concepts such as 'sense of place' (Hay, 1998a, 1998b; Stedman, 2002), 'place identity' (Proshansky et al., 1983), 'place dependence' (Stokols and Shumaker, 1981), 'rootedness' (Tuan, 1980) and 'insideness' (Relph, 1976). Manzo (2003: 47) noted, however, that a limitation of this research is that it: . . .does not fully embrace all of the important dimensions of people's emotional relationships to places suggested by their definitions. In particular, she cited the predominance of work that has examined individuals' attachment to built environments. While she acknowledges the recent contributions of authors who have published works in the forest resources and leisure studies literatures, she suggested that much more needs to be learned about the role and meaning of places beyond the realm of urban environments. However, despite this criticism, a great deal can be gleaned from the broader literature on place, particularly in terms of how people differ in their evaluations of 44 and preferences for an environment. The concept of place used here draws from the work of Schreyer, Jacob and White (1981) and Williams and Rogenbuck (1989) who considered the human-place bond in terms of two components - place identity and place dependence (see also Low and Altman, 1992; Moore and Graefe, 1994; Proshansky et al., 1983; Ryden, 1993; Williams et al., 1992). Proshansky defined place identity in terms of the cognitive connection between people and their environment or: Those dimensions of self that define an individual's personal identity in relation to the physical environment by means of a complex pattern of conscious and unconscious ideals, beliefs, preferences, feelings, values, goals, and behavioural tendencies and skills relative to the environment (1978: 155). Identity asserts a person's place in the world, and a sense of cultural identity creates a secure base from which a person can attempt to deal with the pressures of the surrounding world. Identity is a fundamental and dynamic factor in the cultural configuration of a group that it allows for historical reconstructions as well for future projections. Conceptually, identity establishes a link between different ethnic communities belonging to the same "racial" stock even though dissimilar ethnic traditions have led to important cultural differences (Tremblay in Simon, J . etal., 1984: 65). In its conventional use in the place literature, identity is partially formed and takes strength from the symbolic messages that can be gleaned or read from the natural environment (Basso, 1994; Lewis, 2000). In effect, places or features of the landscape often serve as tangible symbols or emblems connecting people to historical figures and events. When these symbols are shared, places have the power to link people to one another to form a common identity that is rooted in a shared history and symbolic association. 45 Place dependence, on the other hand, is less abstract to the extent that it concerns how well a setting serves goal achievement given an existing range of alternatives (Jorgensen and Stedman, 2001). Stokols and Shumaker (1981) indicated that a place could be considered important to an individual because of its functional or instrumental value (see also Fried, 1982; Herting and Guest, 1983; Guest and Lee, 1983; Mesch and Manor, 1998; St. John, Austin and Baba, 1986). Dependence is viewed as "the utilitarian value [of a place] to meet certain basic needs" (Guest and Lee, 1983: 234) ranging from sociability to services to physical characteristics. For instance, in the context of recreational settings, users of specific resources can also be dependent on the place because of its unique ability to facilitate or 'afford' desired experiences (Gibson, 1986; Gregory, 1969; Kaplan S., 1975). Ryden (1993: 37-8) adds that: A place . . . takes in the meanings which people assign to that landscape through the process of living in it. Shumaker and Taylor (1983) suggest that the physical amenities of a setting provide attributes that satisfy certain needs. Sack (1997) links nature, culture and social interactions in the creation of place, and notes that some places are richer in natural elements than others (i.e. the attributes found in the landscape are foundations of attachment and satisfaction). These authors suggest that sense of place is not intrinsic to the physical setting itself, but resides in human interpretations and knowledge of the setting, which are constructed both through social interactions as well as experience with the setting itself. In effect, landscapes become 'places' as they become imbued with information and meaning acquired through intersubjective learning and embodied experience (Ingold, 2000; Tuan, 1977). 46 Several studies examining the relative effect of place identity and place dependence on preferences for landscape management scenarios in recreational contexts have shown that place identity and place dependence do not always act uniformly in spite of being related (Vaske and Kobrin, 2001; Williams and Vaske, 2003). For example, Kyle et al. (2003) examined place attachment's (i.e. place identity and place dependence) effect on visitor spending preferences for revenue collected from user fees on U.S . Forest Service lands. Their findings indicated that, overall, participants scoring high in their expression of place identity (i.e. symbolic attachment to a landscape) were more inclined towards expenditures directed toward the preservation and restoration of the landscape. By contrast, participants that scored high on place dependence were more disposed towards supporting expenditures that provided new and expanded facilities. T h e s e and related studies (e.g. Bricker and Kerstetter, 2000) indicate that generalised interpretations of place attachment and its effect on preferences for landscape management alternatives can be misleading. Th e dimensions of place attachment, while related to one another in the sense that they are facets of the same construct, can represent different elements of the human-place relationship and result in different landscape preference profiles. T o the same extent that place identity and place dependence may lead to differing land management preferences, there may be discrete facets of place dependence that will ultimately result in differing and potentially conflicting expressions of landscape preference. For instance, several authors have raised the possibility that individual variation in landscape preference is fundamentally 47 motivational or purposive in nature, or driven by personally and/or socially salient needs (Canter, 1983; Gibson, 1986; Kaltenborn and Bjerke, 2002; Knopf, 1983; Scott, 2003; Stankey and Schreyer, 1987; Zube era/ . , 1987). In effect, what people know about a landscape and the conditions that they prefer is fundamentally driven by their material or experiential requirements from the setting - i.e. their needs. Although this research has produced long lists of reasons why people like to visit natural areas, the fundamental question of which set of place dependence needs underlie these reasons has remained unanswered. At a more general level of analysis, other theorists have proposed that human needs can be classified into a limited number of ordered categories (Alderfer, 1972; Lewis, 2000; Simon, 1984, 1991). These general categories may serve as a useful starting point for deconstructing and identifying basic place dependence needs. A s a preliminary classification scheme, the three-fold taxonomy described by Alderfer (1972) is particularly useful: • Existence Needs: which refer to material and physiological requirements. • Relatedness Needs: which refer to requirements to relate to significant individuals and groups. • Growth Needs: refer to desires to fully develop psychologically and/or spiritually. Following Alderfer, it is assumed here that the three categories of place dependence needs are active in every human being, but that the relative strength of these needs differ depending on the individual's personal and social circumstances. Applied to the domain of landscape experiences, existence needs may be defined as all the various forms of material and physical needs relevant to human-nature interactions, such as the need for food and water, the need for shelter, and 48 the need to earn a livelihood. Relatedness needs, as a category of place dependence, may be defined as needs that represent goals to interact with groups of friends, community members or family members in an environment. Feelings of acceptance, confirmation, understanding and reciprocity are typical satisfiers of relatedness needs. Finally, growth needs may be defined as all needs that compel a person to fully develop his or her capabilities. With respect to place dependence, growth needs may be an important but particularly difficult to understand category of needs, due in part to their abstract nature, but also because the satisfaction of growth needs is dependent on the availability of stimulating environments (Alderfer, 1972; Lewis, 2000). In general, natural environments offer many opportunities for the satisfaction of growth needs. For instance, natural landscapes offer opportunities to learn new things about plants and nature, and they enable people to think about their lives in quiet and different surroundings. Consistent with the idea that growth needs are an important category of nature experience needs, many frequently observed nature experience needs, such as the desire to meditate or the desire to be relieved of pressure and tension through recreational activities, may be classified as growth needs. Conceptually, individuals with differing needs or place dependence profiles may be expected to differ in their preferences for human influenced patterns in the natural landscape. First, individuals with relatively strong existence needs may be expected to favour natural landscapes with a high degree of human influence, because these landscapes are generally less threatening, and yield more material benefits than landscapes with a low degree of human influence. Second, individuals 4 9 with relatively strong relatedness needs may also be expected to prefer landscapes with a high degree of human influence, because these landscapes are generally more accessible, symbolically reflect the community's identity, and have more facilities than wild, unmanaged natural landscapes, which makes them more suitable for social interactions. Finally, individuals with relatively strong growth needs may be expected to favour natural landscapes with a low degree of human influence, because these landscapes are very different from one's everyday surroundings, which may stimulate feelings of being completely away from culture and other people. Moreover, natural landscapes with a low degree of human influence may provide greater challenges for exploration and for learning new things than more human modified environments, because of the absence of man-made facilities and the occurrence of special types of plants and animals. The aforementioned classification of place dependence requirements as a basis for landscape preferences should be regarded as preliminary, especially since there have been very few empirical tests of the validity of this classification scheme with respect to landscape preferences in general, and the assumptions about the relationship between needs and landscape preferences in particular. However, several studies have examined the relationship between place dependence and preferences for natural landscapes in an exploratory manner (e.g. Stankey and Schreyer, 1987; Lewis, 2000). Unfortunately, most of these studies have compared need profiles among users of only a narrow range of wilderness settings. In one of the few studies that examined users of natural settings with varying degrees of human influence, Knopf, Peterson and Leatherberry (1983) found indications that 50 users of wild and undisturbed riparian settings were indeed strongly motivated by growth needs, such as a desire to escape civilization, while users of more developed and crowded rivers were strongly motivated by relatedness needs, such as a desire to spend time with their families. Similarly, in a recent examination of the significance of place to a British Columbia aboriginal community (Lewis, 2000; Lewis and Sheppard, 2005), the authors found that culturally based conceptions of place embodied several layers of meaning which included an appreciation for the landscape as a material provider and source of physical sustenance (i.e. existence needs), a catalyst for social relationships (i.e. relatedness needs), as well as the physical embodiment of their cultural and personal identities and spirituality (i.e. growth needs). Taken together, these limited findings suggest that the theoretical framework proposed by Alderfer and others may possess some empirical value. If it is indeed correct, the conclusion that place dependence needs can be predictive of stated preferences may demonstrate that preference evaluations of natural landscapes can be connected to people's purposive activities, as is often assumed (Brown et al., 1988; Hull and Stewart, 1992; Zajonc, 1980). Perhaps the most compelling facet of place dependence theories is the premise that, at the root of landscape preference evaluations, people actively or purposively seek information from the setting and intersubjectively from others (Anderson, 1981) within their social network which indicates that the landscape is capable of fulfilling basic needs (i.e. existence needs) or desired experiences (i.e. growth or relatedness needs, see Figure 2.4). This premise is consistent with a growing body of work that posits that landscapes 51 communicate, and a fundamental aspect of this mode of communication is that people seek information about the landscape and about others when they encounter a setting (Hull and Robertson, 2000; Nassauer, 1995; Sheppard, 2001). T h e degree to which people seek information about others and how they are caring for a Figure 2.4: Place Dependence Landscape Preference Model Perceiving Subject Environmental Needs & Motivations • Existence Needs • Relatedness Needs • Growth Needs...? Intersubjective Information • Cultural, Historic, Economic Environmental Information • Physical conditions Information concerning the motivation or need fulfilling potential of a landscape is derived 1) intersubjectively as people receive information from others in their social network and 2) from the setting itself. In effect, through physical patterns and conditions, the landscape communicates to the perceiver the extent to which it can afford desirable experiences or fulfil human needs. landscape is discussed in some detail in Chapter 5. However, what appears to be missing from these recent perspectives is the reflexive nature of landscape evaluation, or how people are motivated to seek information from a setting, not just about others or about abstract concerns for ecological sustainability, but about their own (material, spiritual, psychological) requirements and how the setting can fulfill those needs. Further thoughts on the motivational (or purposive) and reflexive nature of landscape evaluation are provided in Chapter 6. 52 2.4 Chapter Summary From this review of landscape preference theories, I believe that there are two important themes or perspectives that can be distilled from the literature: • Preferences are neither completely biologically nor culturally determined. Both biological (i.e. innate or genetic) and cultural factors may contribute to evaluative landscape responses by an individual, and operate at different levels of conscious awareness. Moreover, the relative strength of both factors may vary with the life stage of the perceiver as the effects of biology diminish with the accumulation of cultural and environmental experiences as individuals mature. • According to place dependence theories, preference formation may be a three part process whereby the physical environment in conjunction with intersubjective social processes and personal (or socially) based environmental needs and motivations set bounds to what people know about a landscape and the conditions that they prefer. In effect, preferences are formed in response to information that is gleaned from a landscape and intersubjectively from others regarding its need fulfilling potential. Removed from the arena of academic jargon - needs, attributes, motivations, etc. -landscape preference evaluation can be seen as a commonplace occurrence, as an ordinary way of engaging one's surroundings and finding significance or satisfaction in them. Writing about place theory, Albert C a m u s stated that "sense of place is not something that people know and feel, it is something people do" (Camus, 1955: 88). That idea brings the whole concept firmly down to earth, which is where I believe that notions such as preference and sense of place belong. 5 3 3 Environmental and Social Context 3.1 Introduction In social scientific research, the description of events and encounters is always from the perspective of the investigator, who highlights certain incidents and observations and draws out their implications for research. However, the characteristics of the research setting (i.e. the community, environmental context, local history, etc.) itself may colour the study's results, particularly when the subject of the research is the relationship between people and their landscape (Schensul et al., 1999). Therefore, it is essential to document the geographic and historical context of the people and culture involved in the research, particularly if such contextual information affects or at least illuminates the results (Goulet, 1998; Schensul et al., 1999). In effect, participation within and observation of other people's lives can generate useful insights, but it must also be recognised that the perspectives and concepts that are a product of social science research are a joint creation of the historical experiences of a community with their social and physical environment (Goulet, 1998). As such, this chapter documents the social, historic and geographic context of the Gitksan people. The essential objective of this chapter is to provide a descriptive account of the Gitksan in terms of the contemporary landscape that they inhabit (Section 3.2 - The Physiographic Setting) as well as the landscape history of the upper Skeena valley (Section 3.3 - The Pre-Colonial Landscape). A theme that runs throughout much of this chapter is that the recent history of European settlement in the Skeena Valley, with the associated influx of agricultural development, industrial 54 forest management and urbanisation, is only the latest stage in a long history of landscape modification that precedes contact with the First Nations of the region. The final section of the chapter (3.4 - Post Contact Settlement and Changing Landscape Patterns) addresses the changes to the Skeena Valley landscape that have resulted from European settlement. Physical changes that can be linked to social processes of assimilation, depopulation, and resource exploitation have altered if not nearly eliminated traditional Gitksan patterns of land stewardship and, in turn, affected the appearance and ecology of the landscape. However, as patterns of resource use and economic activity continue to evolve, I speculate in the final section (3.4.2 - Landscape Conflict Under the 'New Resource Economy') on the potential for new landscape patterns in the region's future, as well as new sources of land-use conflict. 3.2 The Physical and Ecological Landscape of the Upper Skeena Valley The landscapes of the Skeena valley are exceptionally diverse. There are noticeable differences in precipitation, temperature, landforms, vegetation, and soils as you travel from the coast to the interior of the province and, of these, the most important controlling factors are topography and precipitation. Together these two factors influence most other environmental attributes, as well as much of the human experience in the Northwest. The remarkable vegetation contrast between Prince Rupert and Terrace in the lower Skeena, and Hazelton and Smithers along the upper reaches of the Skeena and Bulkey valleys is due in large part to the shape of the land and the long-term influence of weather. First Nations and recent Euro-55 Canadian economic activities, settlement patterns, and transportation routes also reflect these natural influences. Along the entire length of British Columbia's west coast, only three major rivers empty into the Pacific (the Fraser, Skeena and Nass), while three more traverse Canadian territory before seeking their outlet in Alaska (the Stikine and Iskut) and mainland United States (the Columbia). Ranking second in importance among the three Canadian rivers is the Skeena, currently the road to the northern half of British Columbia (Figure 3.1). The Tsimshian and Gitksan First Nations who I Figure 3 . 1 : Upper Skeena Valley Communities Context map from Anderson and Halpin, 2002: 2. 5 6 have traditionally resided along its shores called the river K-shian or K-san, which translates roughly as "river of mists," and is generally regarded as the source of the contemporary name for the river. The Skeena is approximately 560 kilometres long, drains a watershed that encompasses 4,000,000 hectares, and is the third largest river in British Columbia. Historically, the Gitksan people controlled the upper extent of the Skeena watershed, along with portions of the Nass River system (Sterritt, et al., 1998), while their Tsimshian neighbours claim the lower and westernmost portion of the watershed as their traditional territory. The Skeena's source is in the Gunanoot Mountains, and begins in close proximity to the headwaters of the Nass and Stikine rivers, and flows in a southerly direction for 320 kilometres to its junction with its main tributary, the Watsonquah or Bulkley River. From that point it swings in a westward direction and empties into the Pacific at Prince Rupert. Like most mountain rivers, the Skeena is unpredictable in its behaviour and, in the early summer, it is subject to sudden and sometimes severe flooding. Under normal circumstances, the melt-water that flows out of the Coast Range Mountains is controlled by the cold chill of spring nights. However, in an occasional year, the heat of the summer comes early and lasts throughout the night. At such times, the water can rage down the relatively short course of the river, eroding banks, tearing up trees and inundating the land generally. Such an event occurred in 1936 when whole villages along the Skeena's course were rendered homeless, totem poles were uprooted and washed away, and transportation and telegraphic communication 57 networks we re s e v e r e d (Roth , 2 0 0 1 ; F igure 3.2). H o w e v e r , s u c h y e a r s a re infrequent, a n d genera l l y the river r ema ins within its b a n k s . Figure 3.2: Skeena River Flood, 1936 Photo courtesy of the Village of Hazelton. T h e l a n d s c a p e through wh ich m u c h of the S k e e n a p a s s e s is very moun ta inous , and only in the immed ia te vicinity of the river is there any level land . Moun ta in s l o p e s a n d va l ley f loors a re heav i ly t imbered , a n d on ly o n the cent ra l p la teau of the eas te rn tr ibutar ies (i.e. the K isp iox and Bu lk ley r ivers) near Haze l ton is there any o p e n g raz ing country . A sma l l - s ca l e ove rv iew of the upper S k e e n a ' s topography revea ls that the cha in of mounta in r a n g e s and va l l eys fo rms a large t r iangle. It l ies in a vert ical nor th-south or ientat ion, a c r o s s the S k e e n a and N a s s r ivers, a n d e n c o m p a s s e s m u c h of the agr icul tural land of the reg ion . T h e wes te rn 58 leg of the triangle begins in the Nass valley and extends south along the Cranberry River to the Gitksan communities of Gitanyow and Kitwanga. The eastern leg begins in the Nass valley and extends south along the Kispiox River to the village of Kispiox and ends at the confluence of the Skeena and Bulkley rivers, where the Gitksan village of Gitanmaax is situated adjacent to the predominantly Euro-Canadian communities of Hazelton, New Hazelton and South Hazelton. The base of the triangle is made by the Skeena itself, which turns west from the Forks and flows towards the village of Gitsegukla (pronounced git-see-ook-la) and back to Kitwanga. During the last ice age, these valleys were filled with glacial ice and functioned as spillways for the runoff from the retreating ice flows. The pulverized surfaces of the surrounding mountains have been deposited as sand and gravel outwash terraces visible along the Skeena valley floor from Terrace to Kitwanga. Figure 3 . 3 : The Skeena-Bulkley Valley Lowlands Photo by the author. Both historic and contemporary settlements have been concentrated along the southern extent of the Gitksan territories at or near major river junctions with the Skeena. Stretching from Kitwanga in the west to Hazelton in the east, the Skeena-Bulkley valley lowland is the site of small-scale agriculture (mostly grazing and hay 59 production); vast deciduous forests of cottonwood, poplar and aspen; and village and town settlements (Figure 3.3). This is the heart of the Gitksan territories, as well as the confluence of the Skeena and Bulkley rivers (locally known as 'The Forks'), where the natural endowments of access to the coast and interior by way of two major river corridors and a relatively mild climate has made this part of the Northwest a natural choice for habitation, as well as a vital nexus for trade among coastal and interior First Nations peoples. Much of the river valley would be a coniferous forest, largely dominated by Interior Cedar and Hemlock were it not for the intensive cultivation and historic First Nations practice of clearing land with fire (See Section 3.3.2). The Wetsuweten First Nation, an interior or Carrier group bearing some cultural but no linguistic relationship with the Gitksan 6 , has two large villages both within and in close proximity to the Gitksan territories. Hagwilget (pronounced haggle-gyet) is the larger of the two villages and is situated at the end of the Bulkley Canyon near Gitanmaax, which in addition to being a prolific location during the salmon harvest was the site of great gatherings of coastal and interior peoples during the autumn months. The second village is only a short distance up the Bulkley River outside the southeastern corner of Gitksan territory and situated beside a waterfall, where the salmon were easily obtained during their struggles to surmount the obstacle. The village now bears the name of Moricetown, in memory The Wetsuweten and Gitksan, despite being interior and coastal peoples respectively, share several cultural practices (e.g. the feast system) and institutions (e.g. hereditary chiefs and matrilineal house groups). However, they belong to separate linguistic families and, therefore, possess mutually unintelligible languages (Mills, 1994). 60 of the catholic missionary (Father A . G . Morice) who worked with the Wetsuweten and other interior peoples. The Skeena Range forms the eastern margin of the Coast Mountains and ends abruptly at the confluence of the Skeena and Bulkley rivers, forming a highly varied landscape of steep mountains and deeply cut canyons such as the Bulkley canyon adjacent to Hagwilget. The Skeena Range holds scattered veins of precious metals that attracted multitudes of prospectors to the area and north into the Omineca district in the nineteenth century. Dense forests of Interior Hemlock and Cedar cover the hillsides, nurtured by abundant precipitation in the spring, early summer and autumn. Large scale timber harvesting began in this region as recently as the mid 1980s, and evidence of past as well as current clear-cutting becomes more prevalent as you move away from the more settled areas along the major river courses. Harvesting displays an unnatural, regular geometry in the back road portions of the Gitksan territories, while the patterns in the front country reflect some attempt to soften the visual imprint by emulating natural disturbance patterns through irregular cutblock patterns (Figure 3.4). As the traveller explores the Skeena Range from west to east, dramatic isolated peaks punctuate the landscape (e.g. Blunt Mountain, Stekyoden or Roche de Boule), and end abruptly with the foothills and undulating topography of the interior plateau. These are the eastern most summits of the Coast Range, the highest of which, Hudson Bay Mountain, looms 1,500 metres (5,000 feet) above Smithers. Its bowl shaped summit holds a glacier that displays a vibrant blue 61 l ead ing e d g e dur ing the s u m m e r mon ths , and attracts s c o r e s of tourists to its b a s e for sk i ing , vaca t ion co t tages , and other recreat ion faci l i t ies on the moun ta in ' s s l o p e s . Figure 3.4: Post Forest Practices Code Cutblock Patterns Photo by the author. Throughou t the C o a s t R a n g e , r iparian z o n e s , fo res ts and a lp ine a r e a s feature quite dist inct env i ronments . M a r i n e air mov ing in land distr ibutes precipi tat ion uneven ly , with tempera te ' ra inforests ' of C o a s t a l C e d a r and Mounta in H e m l o c k on wes t - fac ing s l o p e s , and relat ively dry, Interior C e d a r and H e m l o c k domina ted r a i n s h a d o w s l o p e s on the leeward s i d e s of the moun ta ins (Haeuss le r , 1980 ; H a e u s s l e r , et al., 1985 ; L e P a g e , 1995). M o r e o v e r , the co ld and dry air of the interior p la teau m a n a g e s to f ind its w a y through narrow va l ley p a s s a g e s s u c h a s the Bu lk ley and into the S k e e n a and K isp iox V a l l e y s to c rea te a temperature invers ion effect, 62 where the dense colder air of the interior sits on the valley floors beneath the lighter humid air of the coast. Thus, within the Gitksan territory alone, annual precipitation can vary more than 100 inches, creating a 'transition zone' between the Coast and Interior ecoregions (Haeussler et al., 1985). In addition, landslides, glaciers, wind, and naturally occurring fires have produced a diverse assortment of open spaces and biotic communities at different locations. The Skeena River ranks as one of the great arteries and lifelines of the Pacific Northwest. Along its often-turbulent course, First Nations settled, explorers sought routes to the rich trapping territories of the interior, and settlers immigrated to establish small farms, harvest the region's forest or mineral wealth, or convert and preach to the Native population. Today, Highway 16 links the interior of the province with the port of Prince Rupert and Asia, along with a railway, pipeline and communication facilities that run parallel to part of its path. With the Nass, the Skeena River has been a major gateway through the Coast Range for northern First Nations, a rendezvous point for Coastal and Interior peoples. First Nations from both sides of the mountains met at the Forks to trade the products of their very different environments. Shellfish, cedar bark baskets, eulachon grease, and shell beads from the Tsimshian and Haida peoples on the coast were traded for berry leather, mountain caribou, goat and moose meat, and precious stones from the inland areas. The place known as the Hazeltons today was among the most important contact points between the northern Coast and Plateau First Nations. 63 3.3 The Pre-Colonial Culture and Landscape 3.3.1 Resource Procurement and Gitksan Social Organisation The social institutions and ecological knowledge of the Coast and Plateau First Nations of the upper Skeena Valley remain subjects of considerable interest to anthropologists. However, since the nineteenth century, the relationship between the First Nations of the Northwest Coast and their landscape has received passing mention in the literatures of anthropology, historical ecology and other disciplines. Indigenous cultures were of interest to explorers, travellers and contemporary anthropologists, but they did not address aboriginal peoples' use of their landscape as a central theme. Most of them - following the lead of prominent Northwest Coast anthropologist Franz Boas - described human uses of the marine and forest environments on the basis of brief visits and post hoc analyses of indigenous oral traditions (Boas, 1895). Impressed by the region's abundant aquatic and terrestrial resources, most writers concluded that food and other material resources were plentiful at all times and in all places. Thus, in books such as Ruth Benedict's Patterns of Culture, resource gathering was depicted as a simple, even leisurely practice. Referring to the Kwakiutl First Nation of the Central Coast, Benedict wrote: They were a people of great possessions as primitive peoples go. Their civilization was built upon an ample supply of goods, inexhaustible, and obtained without excessive expenditure of labour (1934: 173-4). However, if the Northwest was a place of inexhaustible abundance, how then do we account for the elaborate systems of land tenure found along much of the coast and into the interior, and why were long and brutal wars fought among adjacent First 64 Nations over particular territories and resource sites (Sterritt, et al., 1998)? Moreover, why did displacement from traditional resource areas, as well as the prohibition of particular resource stewardship techniques such as burning, result in such material hardship for the First Nations of the region? The answer to these questions, and the possible key to material success in pre-contact Northwest society, lies in a consideration of the ecological heterogeneity of the Skeena Valley landscape and the deliberate management or 'improvement' of the land to enhance the yield of culturally significant products. In effect, although the natural resources in the Northwest were in many respects prolific, they did not pervade the landscape. They were unevenly distributed temporally and spatially, and First Nations' valuation and use of the landscape varied accordingly. Indeed, the pre-contact history of Northwest Coast land-use appears to have involved peoples' efforts to maximize output from, and assert control over, a finite and highly diverse range of resource-rich sites. Since the Gitksan territories are situated in a transitional forest ecosystem (see section 3.2), this part of the Northwest is characterised by a variety of 'patchy,' small-scale environments. Small concentrations of forest, rock outcrops, sand and gravel terraces, grassy clearings and sub-alpine meadows, mixed coniferous-deciduous forests, and so on tend to predominate. Animals feed and breed, and plants grow in a finite range of habitats, and their spatial distributions tend to be irregular as well. Without modern technologies for transporting food over long distances, First Nations who depended on the landscape for their material existence tended to locate near places where food plants and animals were relatively 65 abundant and easily transportable, and they jealously controlled access to these resource procurement sites to the extent that trespass was punishable by death. The availability of staple plants and animals in the Northwest varied both seasonally and geographically. For instance, streams with large populations of salmon or eulachon could be tens of miles apart. Elk or mountain caribou - now conspicuously absent throughout most of the region - were relatively scarce outside of the forest clearings where they grazed. Plant foods were also unevenly distributed. On the coast many berry producing plants that were a vital source of Vitamin C - e.g. Black Huckleberry (Vaccinium membranaceum) - grow poorly under the dense forest canopy that blankets much of the zone, but tend to be more prolific towards the interior where canopy openings are more common. Even the Western Red Cedar from which peoples of the coast crafted their homes, watercraft, totemic art, clothing and other goods, is irregularly distributed throughout the coast and into the interior. This discontinuous distribution of natural resources presented a host of challenges, both logistically and in terms of social organization to the people of the Northwest. From the standpoint of societal organization, what we refer to today as 'nations' or 'tribes' had no formal existence prior to contact with Europeans. Linguistic, ceremonial, symbolic, economic and kinship ties may have bound communities together, but the Gitksan, Tsimshian and Nisgaa were not 'nations' but ethnolinguistic groups. The Gitksan villages described briefly in section 3.2 were the primary political and social units. Each was largely autonomous in matters of politics, war and trade and each appears to have formed alliances with other 66 communities based on mutual, inter-village interests. Individual house groups (i.e. wilps), essentially extended kin groups that were controlled by a hereditary chief and a cadre of successors (i.e. wing chiefs), were the fundamental social units in ceremony, resource procurement and ownership. House groups were largely independent economic units that controlled a subset of territorially defined resource sites within a larger, usually village-owned resource territory. This social structure was sustained by a number of distinctive cultural institutions. Perhaps the most important among these at the time of European contact was the feast, a ritual exchange of goods that accompanied significant events such as marriage, title acquisition, death, puberty or personal shame. Through these exchanges, hereditary chiefs and other societal elites repaid or generated debts, mediated interpersonal disputes, affirmed authority, as well as enhanced their personal status as well as that of their wilp. Although the emergence of the feast has been difficult to establish archaeologically, there is some evidence that its foundations in Northwest Coast aboriginal society extend 3,500 years into the past (Adams, 1973; Carlson and Hobler, 1993; Halpin and Anderson, 2000; Mills, 1994). 3.3.2 The Question of Indigenous Cultivation The central role that the feast played in Northwest First Nations society, particularly in terms of elite legitimation, greatly influenced Gitksan systems of resource use and landscape patterns. Resource procurement was required not only to meet basic dietary or subsistence needs, but also to produce a surplus that could be exchanged to maintain social status and political structures. In effect, to maintain 67 or enhance their position, chiefs and others with hereditary names had strong incentives to acquire wealth and enhance the productivity of resources within their wilp territories. Social institutions and personal status, in other words, required access to surplus resource wealth (i.e. beyond that required for subsistence). By most accounts, the First Nations peoples of the Northwest did not improve, cultivate or enhance the productivity of their landscapes prior to European contact (Benedict, 1934; Flanagan, 2000). There is, however, a growing body of evidence suggesting that pre-contact people of the region did in fact engage in some form of cultivation. In large part, these practices expanded the amount and productivity of land within a given wilp territory on which coveted plants could grow. In the process, people altered the ecology of the Northwest landscape while significantly augmenting their diet and material wealth. Scholars who challenge Euro-centric notions of indigenous land-use now posit that First Nations were "emergent agriculturalists" who often merely augmented hunting, fishing and plant gathering activities with managed food production, thus increasing the abundance and predictability of their food supply (Butzer, 1990; Denevan, 1992; Doolittle, 1992, 2000; Sauer, 1936). Accordingly, cultivation has been redefined to include a continuum of activities that involve the repeated and purposeful manipulation of plants and landscapes to enhance productivity. In many indigenous societies, an indication of the presence of cultivation is what has generally been referred to as "agro-ecosystems," or human modified environments such as the small swiddens of indigenous slash and burn agriculture, canals and 68 other improvements designed to enhance the yield and concentration of the land's produce (Wiersum, 1997). Although the Northwest Coast possesses some of the highest levels of terrestrial and aquatic productivity in the world, there are obstacles to the growth of nutritionally valuable plants in this zone, most notably the low levels of sunlight beneath dense forest canopies as well as the region's leached soils. The First Nations of the Northwest Coast overcame both solar and nutritional deficits by modifying existing plant communities - in effect, creating agro-ecosystems - to provide nearby villages with staple food plants that might otherwise have been dispersed too widely to prevent efficient surplus accumulation (Gottesfeld-Johnson, 1994). Among other plants, scholars have noted that the peoples of the Northwest managed two plant species on a large scale: tobacco and berries (Johnson, 1997; Turner and Taylor, 1972; Turner, 1991). Tobacco cultivation was easily recognisable to early European explorers. In 1787, 13 years after the first documented contact between Northwest coast Natives and Europeans, Archibald Menzies found both the Haida and Tlingit cultivating a variety of tobacco in patches of ground near their primary settlements (Menzies, 1923). Tobacco was likely native to the drier, interior portions of the Northwest rather than to the moist coastal zone, as evidence by its rapid disappearance from the region after indigenous cultivation of the plant ceased. Genetic and archaeological evidence, as well as the presence of tobacco in indigenous oral traditions suggests that the pre-contact cultivation of 69 introduced tobacco on the coast was a practice with a considerable history (Denevan, 1970; Deur, 2000). Nonetheless, the extent to which tobacco cultivation modified terrestrial environments to any large extent is debatable. It is likely that the practice was confined to small patches near village settlements with little, if any, discernable effect on the surrounding landscape. However, despite its poor representation in the historical ecology literature, there is clear evidence along much of the coast and interior of British Columbia (Gottesfeld-Johnson, 1994; Turner, 1999); Alberta (Lewis H. , 1988); Washington, Oregon and California (Agee, 1993; Lewis H. , 1973) for indigenous cultivation of black huckleberry and other fruit bearing shrubs - e.g. Low-bush Blueberry (Vaccinium caespitosum), Thimbleberry (Rubus parviflorus), Soap Berry (Sheperdia canadensis) - through the controlled but large-scale use of fire. Many researchers have examined the role of fire as an integral part of human culture, primarily since the early part of this century. People have used fire since prehistoric times for warmth, cooking, hunting, tool making, in cultural rites and ceremonies, and have also worshipped it (Hough, 1926). In the context of landscape history, the significance of anthropogenic fire is in its use by indigenous peoples to manipulate vegetative communities. Anthropologist Omer Stewart (1956) considered such activities to be universal: In ancient times, as in recent ones, thick forests and dense jungles of brush offered very little use to the hunter or collector. On the contrary, narrow trails through tangled and heavy growth were dangerous because of the concealment provided to human, as well as animal, enemies. Whenever possible, aboriginals have set fire to jungles and thick woods in order to 'open them up.' Widely spaced trees and clear meadows and plains offer better and safer hunting (1956: 148). 70 That the First Nations of the Pacific Northwest carried out prescribed burning for a variety of purposes has long been recognised, as shown by this extract from botanist David Douglas' journal for 1826 describing the Willamette region of Oregon: Most parts of the country burned; only on little patches in the valleys and on the flats near the low hills that verdure is to be seen. Some of the natives tell me it is done for the purpose of urging the deer to frequent certain parts, to feed, which they leave unburned, and of course they are easily killed. Others say that it is done in order that they might better find wild honey and grasshoppers, which doth serve as articles of winter food (Douglas, 1914:43). Figure 3.5: Luu Skayansit, Traditional Berry Site Cultivated and Maintained by Fire The relic burnt portion of the mountainside (see outline) is situated mid-slope, below a ridge-line and to the left of a cutblock harvested in the mid 1970s. Photo by the author. Indigenous burning not only helped maintain grassland and woodland meadow ecosystems, but it also was the process that promoted, expanded and, in some areas, formed them. The extent to which this may be true is still hotly contested (Baker, 2000; Lepofsky er al., 2003). However, there seems to be a growing consensus among fire ecologists and ethnoecologists that, prior to European settlement and the advent of fire suppression policies, repeated burning did maintain and expand certain ecosystems (Agee, 1993; Anderson, 1996; Bonnicksen, 2000; Boyd, 1999; Lewis, 1973; Lewis and Ferguson, 1999; Johnson, 71 1999 ; P e a c o c k a n d Turner , 2 0 0 0 ; F igu re 3.5). T h e difficulty l ies in f inding abso lu te e v i d e n c e b e c a u s e m a n y of t hese f i res w e r e f requent , low sever i ty f i res wh ich do not l eave the s a m e s igna tu res a s l ess f requent , more in tense f i res ( A g e e , 1993 : 354) . M o r e o v e r , in the G i t k s a n commun i t i es , the spec ia l i s t s w h o w e r e o n c e respons ib l e for uti l ising fire to m a n a g e vegeta t ive commun i t i es have long s i nce p a s s e d a w a y , leav ing a t r e m e n d o u s g a p of skil l a n d k n o w l e d g e regard ing whe re , h o w regular ly a n d to what spat ia l extent burn ing w a s pract iced (see sec t ion 3.4.1). W h i l e it is difficult to d i sce rn how prevalent f i res w e r e d u e ei ther to l ightning or h u m a n ignit ion, a n d h o w they a l tered the a p p e a r a n c e of the l a n d s c a p e , the e thnoeco log i ca l l i terature s u g g e s t s that a s imp le c o m p a r i s o n of known u s e a r e a s with a r e a s without e v i d e n c e of h u m a n u s e demons t ra tes that there are often c lear pat terns of i n c r e a s e d fire f r equenc ies in a r e a s a s s o c i a t e d with h u m a n habi tat ion (Wal l in e r a / . , 1996) . T h e r e is a m p l e e v i d e n c e that burn ing w a s cer ta in ly car r ied out on a large s c a l e throughout the interior of the cont inent w h e r e a natural ly dr ier a n d more fire-p rone c l imate cont r ibuted to the creat ion of g r a s s l a n d s hund reds or t h o u s a n d s of s q u a r e mi les in extent. Hen ry L e w i s r e s e a r c h e d abor ig ina l burn ing p rac t i ces in the borea l fo res ts of nor thern A lber ta a n d found that there w e r e a n u m b e r of r e a s o n s for the u s e of fire in t hose e c o s y s t e m s , inc lud ing (Lewis , H. , 1988) : • M a i n t e n a n c e of m e a d o w s and g r a s s l a n d s for ho rse pas tu re or b r o w s e for favoured g a m e s p e c i e s . Wi thout per iod ic f ire, s h r u b s and t rees wou ld i nvade an th ropogen ic g r a s s l a n d s . • S u p p r e s s i o n of unders tory plant mater ia ls to faci l i tate ove r l and travel and reduce fuel l oads . • Improvement of the l a n d s c a p e ' s aes the t i c qual i ty. • Habi tat e n h a n c e m e n t for fu r -bearers a long r ipar ian z o n e s . • Ki l l ing t rembl ing a s p e n and co t tonwood to c rea te f i rewood. • M a i n t e n a n c e of berry pa t ches , part icular ly wild raspber r i es , b lueber r ies a n d s t rawber r ies . 72 • Elimination of deadfall as well as maintenance of village and camp sites and trail networks. The southern Tutchone of the Yukon are known to have burned to enhance wildlife habitat and facilitate the harvest of Alpine Sweet-Broom roots {Hedysarum alpinum) by thawing the ground in early spring (Hawkes, 1983). However, berry production appears to have been one of main intended outcomes from the purposive use of fire. If prescribed fire was used by the Natives of Coastal Washington and Oregon, as well as by the interior First Nations of British Columbia, Alberta and the Yukon, it seems safe to assume that it was also used in the Northwest Coastal region of British Columbia. First hand written accounts of burning to enhance plant production in the Gitksan territories are exceptionally rare and only indirectly associate areas of burnt forest with the deliberate use of fire by First Nations. For instance, during his geological survey of the Skeena Valley for the Canadian Pacific Railway in 1879, George Dawson suggests that the residents near present day Cedarvale had used fire to clear the terraces adjacent to the river: Terrace flats or benches are even more prominent than before and in some places spread widely. One flat in particular, about ten miles below Kitwanga, having been burnt over and grown up with grass and coppice presented a very attractive appearance. . . . Higher mountains appear at some distance. A prominent group of these to the south-east being called Ish-gan-isht by the Indians. Most of the hillsides on this part of the river have been burnt over, and are largely, in some cases for considerable areas, exclusively covered with aspen second growth (Dawson, 1881: 15). Similarly, during his Crown land survey of New Caledonia in 1891, A . L . Poudrier notes the extensive evidence of burning near the Forks, but does not directly attribute the disturbance to the First Nations inhabiting the area: 73 The fire has formerly destroyed the high timber, and these terraces are now covered with a short growth, where aspen poplar predominates (P. tremuloides); it is one of the favourite spots where the Indians prepare their crop of small fruits for the winter use. ... The soil is rich, the water plentiful; it would cost very little to clear the land, and the foot-hills of the mountains where the timber has been destroyed by fire, offers some first class grazing (Poudrier, 1891: 361). The only comprehensive and detailed studies of aboriginal burning in British Columbia have been carried out and reported upon relatively recently. Nancy Turner (University of Victoria) and Leslie Main Johnson (Athabaska University) have been the primary investigators in BC and, by combining their geographical areas of interest, what is currently known about the prescribed burning practices of the First Nations of British Columbia can be summarised (Gottesfeld-Johnson, 1994; Johnson, 1997, 1999; Turner, 1991, 1999; Turner and Efrat, 1982; Peacock and Turner, 2000). The primary focus of their studies has been prescribed burning for the management of food plants, but it is recognised that the practice provided a variety of benefits including the creation and maintenance of wildlife habitat patches, travel corridors, medicinal plant gathering sites, and fuel load reduction around village settlements. Burning was done at specific times of the year and day, under certain weather conditions and required specialised knowledge of fire behaviour and vegetation response. Often burns were carried out towards the beginning or end of the rainy season to limit its intensity and spread, and there is some anecdotal evidence that burning was done towards the end of the day as warm air currents from the valley-bottom ascended the mountainsides, thus reducing the risk of escapement and destruction of human settlements. In the case of black huckleberry, burns were carried out every five to ten years to suppress forest 74 succession processes, renew the plants' productivity, and replenish soil nutrients. The practice of rotational burning kept a larger landscape in constant production, with several patches of a few thousand square metres being maintained in varying stages of production within a larger mosaic of several hectares. The preferred locations for burning were close to villages or seasonal camps to facilitate the processing (i.e. drying) of the berry harvest and its transport, with most of the patches being located at higher mountain elevations and on a southerly aspect to maximize sun exposure and warmth. Berries and tobacco were not the only plants to be cultivated by Northwest First Nations. In fact, coastal peoples also cultivated several other plant species using similar methods - e.g. Camas (Camassia quamash), Springbank Clover (Trifolium wormskioldii), and Pacific Silverweed (Potentilla anserine spp. pacifica). However, these plant communities and the cultivation methods that sustained them were alien to the European eye and were therefore consistently overlooked in travel accounts and other descriptions of the region. Although tobacco and other plants were likely imported from the interior through trade networks, the planting practices used in their cultivation were very likely long-standing and possibly endemic. If the Northwest Coast was not 'agricultural,' it was at the very least a place of considerable plant tending and cultivation. Although the First Nations of this region generally lived well, notions of resource superabundance oversimplifies if not entirely obscures First Nations patterns of active land-use and management. 75 3.4 Post Contact Settlement and Landscape Change 3.4.1 The Decline of Traditional Indigenous Cultivation Settlements in Northwest British Columbia are location bound. That is, they were founded and populated in specific places because of what the natural environment around them had to offer. This relationship between human and physical geography becomes apparent when a map that details the physical environment is compared with one that shows the distribution of the Northwest's population. East of the Rockies, uniformity of climate, soil fertility, and transportation potential has led to settlements that are evenly spaced and graduated in size. In contrast, the Pacific Northwest, particularly through the Coast Range, is far from homogenous. Its terrain, soils, transportation links and the occurrence of basic resources are widely scattered and differ markedly throughout the region. Thus, in the Skeena Valley, as elsewhere throughout the region, the physical environment has played a major determining role in the location of human settlements. Landscape planners refer to the natural affordances of a location for particular land-uses as intrinsic suitability. Euro-Canadian settlers, like the First Nations that preceded them, gravitated to places on the land where they could make a living. Historically, these were sites where water, game, materials for shelter, transportation and other essentials of existence were relatively plentiful. First Nations located primarily near rich estuaries, the confluence of major river systems or along river cataracts where salmon could be netted. The first European explorers, mainly under the commission on the Hudson's Bay Company, sought out places where fur-bearing animals thrived - sea otters along the coast of the Queen Charlotte Islands 76 or along the mainland coasts of British Columbia and Alaska, and beavers in the interior of the continent. New trapping and fur trading centres (e.g. Fort St. James, Fort Fraser) sprang up where abundant animal populations and their luxurious pelts could be collected and shipped to distant markets. However, other European settlements were situated near long established First Nations communities (e.g. Hazelton, Kispiox), in part reflecting the natural advantages of these sites in terms of proximity to resources and major transportation routes to the Coast and Interior. In the early to mid nineteenth century, settlements were established as trading centres for the exchange of furs or, where the supply of this commodity was relatively meagre, to provision trading centres further inland with staples from the coast. In this way, the fur trade founded as well as maintained Euro-Canadian communities along the Skeena Valley and into the heart of the interior. In. the mid-nineteenth century, a new source of material wealth spurred colonial settlement in the Northwest as mineral strikes brought throngs of prospectors in search of gold, silver and other valuable ores. Settlements from Port Edward at the mouth of the Skeena to Hazelton grew into bustling itinerant tent communities overnight as miners moved into the district from the United States and southern B C bound for the Omineca gold fields in the northern reaches of the Skeena watershed. New gold strikes sparked the wanderlust not only of countless miners, but also of attendant entrepreneurs determined to relieve the prospectors of their newfound wealth. Merchants, bankers, venture capitalists, hotelkeepers, and bandits quickly moved into the region to exploit the latest strike. 77 However, the surge of European settlement in the region during the late nineteenth century also coincided with dramatic changes to First Nations' patterns of cultivation and land-use. These changes were associated with three major factors. First, prior to the arrival of the first Europeans in the Pacific Northwest, diseases spread throughout the region and, as mortality rates soared, population pressures on local resources diminished, as did the incentive and skill to enhance the productivity of resource sites. As populations plummeted throughout British Columbia, survivors regrouped with survivors from other villages in new multi-house or, in some instances, multi-nation communities. Although traditional foods continued to be eaten by Northwest communities, labour-intensive food production techniques diminished extensively as skilled practitioners perished, survivors were severely weakened and unable to maintain traditional resource procurement sites, and sites were largely abandoned in the course of relocation (Johnson, 1997). The second factor that contributed to the decline of traditional cultivation practices was the introduction of alternative staple foods. As Native communities consolidated or new ones were founded near Christian missions (e.g. Meanskinisht or Cedarvale), canneries, sawmills, and trade depots, traditional foods were gradually supplemented or replaced by foodstuffs obtained from European settlers. The abandonment of traditional cultivation practices was also paralleled the simultaneous decline of the fur trade. As fur-bearing animals were depleted, First Nations looked for new items to trade with Europeans arriving by ship. Potatoes were imported and began to be cultivated widely, replacing the most endemic root crops in the Native diet. This transition was partly accelerated by missionaries who, 78 not recognizing localised forms of agriculture in the unfamiliar plants and cultivation systems, promoted the aboriginal adoption of European forms of agriculture as a requisite step in the progress from 'savagery' to 'civilization' (Flanagan, 2000; Tomlinson, 1993). Since few other European crops did as well in the cold, moist soils of the Northwest, potatoes became the new staple crop. Third, it was not until the 1940s when the provincial Forest Service put a stop to most traditional landscape burning practices that the pattern of aboriginal burning virtually disappeared from the British Columbia landscape (Parminter, 1995). Aboriginal burning is still carried out in BC, albeit on a much-reduced scale since reserves are within federal jurisdiction and, therefore, are not subject to provincial regulations (Gottesfeld-Johnson, 1994). Yet, for the most part, the gathering of traditional fire-managed food plants by the Gitksan and several other BC First Nations has been limited to areas affected by wildfire or logging. In the words of a Secwepmc elder interviewed by Nancy Turner (1991): If you got to burn then you get into trouble because the white men want to grow trees. Because they changed our ways for us and we eat the food that the white men use. Then we forget the food of our earliest forefathers. ... Now they have all disappeared because the hills grew weedy and no one seems to tend them, no one clears them as our forefathers did so thoroughly. In general, through the loss of knowledge and skill brought by disease, the introduction of alternative agricultural methods and foods, the rising dependence of First Nations upon the fur trade and, eventually, the commercial economy, and the prohibition of aboriginal burning by the provincial government, only a portion of traditional procurement methods - such as salmon fishing, berry picking in logged or naturally burnt over areas, and hunting - persist among the majority of First Nations 79 peoples. A small but declining segment of contemporary Gitksan people, as well as more remote First Nations communities further north in the Stikine and Tahltan territories, continue to use endemic food plants as well as practice the controlled use of fire in vegetation management (Gitksan Informant - Personal Communication). Notwithstanding the declining use of fire by First Nations, fire as an agent of landscape change remained significant throughout the nineteenth and early twentieth centuries. It is believed that accidental fires were responsible for affecting even larger portions of the landscape than Native burning activity. However, the evidence is somewhat inconclusive as some early travellers through B C remarked that First Nations were 'careless' with fire, while others observed just the opposite (Lutz, 1959). Nevertheless, there is some evidence that unextinguished Native campfires were responsible for starting large wildfires. Omer Stewart was of the opinion that First Nations in general did not put out their campfires in order to maintain a source of fire and, in addition, they saw no real harm from the burning or clearing out of flammable vegetation (Stewart, 1963). However, more conclusive evidence points to prospectors, lumbermen and careless travellers as the major causes of these fires. There are as many accounts of non-Native wildfires, which often resulted from campfires escaping as well as from firing the forest to create a supply of dry fuel wood, ward off mosquitoes and other insects, drive moose in the direction of hunters, and to promote the growth of grass for livestock forage. Fire was also used routinely in mining operations and for clearing land, with mixed results. The apparent increase in the number of forest fires brought about by 80 European settlers likely had a major impact on the landscape. George Dawson remarked upon this in the mid 1880s: It is often stated that the Indians are responsible for this [forest] destruction . . . They would not, however, willingly destroy their own hunting grounds, and the best evidence of their care is found in the fact that, while along the north Kootanie [sic] Pass (which so far has been scarcely used, except by the Indians), the woods are generally unburnt, those in the vicinity of the parallel Crow [sic] Nest Pass, which has now been for a few years a route used by the whites, are entirely destroyed and represented only by bleaching or blackened trunks (Dawson, 1881: 61). While prescribed burning by Native peoples was probably constant and sometimes quite local, the ecological impact of non-Natives and their use of fire was more abrupt and widespread. Given the size of British Columbia and its limited population in the late 1800s, there is no doubt that the vast majority of forest fires passed almost unnoticed. With the fledgling province's economy based largely on fur and minerals rather than timber, which was seen to be inexhaustible, little concern was expressed regarding forest fires. Furs and mineral wealth were not the only forms of wealth to be taken from the region that the colonial government at the time referred to as New Caledonia. The vast woodlands of the Coast Range and interior regions became valuable sources of fuel to power steamships and locomotives. Mills sawed raw timber into lumber to shore up mining excavations, to construct burgeoning towns, and later to expand and fuel the expanding web of the railroad. Communities such as Terrace, Kitselas, South Hazelton and Smithers owe their genesis, in part, to the construction of the railway through the Skeena and Bukley valleys to connect Prince Rupert with Edmonton. However, changes to the region's forests were not only brought about 81 by expanding infrastructure and construction demands, but the job of feeding the Northwest's expanding population also depended on the natural environment. Agriculture and ranching began as primarily subsistence activities on tiny plots along the river, mainly to supplement provisions that were difficult and costly to transport from the southern reaches of the colony. With population growth, however, the need for food pushed more farmers and ranchers into the region with the associated clearing of greater tracts of lowland forests, particularly near the Forks and in the Bulkley valley. Indeed the vast woodlands of the Coast Range became not only impediments to settlement, prospecting or the development of agriculture, but valuable sources of fuel to power steam driven sternwheelers and locomotives at the beginning of the twentieth century. Local mills at the beginning of the twentieth century were small and did not adequately supply the needs of the region, and local requirements for construction materials were filled largely by lumber imported from the south coast. However, with the coming of the railroad before the First World War, a new use for timber was introduced - the production of poles and ties. Trees suitable for this purpose were found on the upper Skeena, and a fledgling timber industry soon thrived with mills strung along the Skeena and Bulkley rivers from Kitwanga to Houston. Two factors combined over a period of thirty years following the Second World War to bring industrial forestry to the forefront, not only in terms of its economic contribution to the region, but in terms of landscape and ecological change as well. First, throughout most of the nineteenth and early twentieth 82 centuries, logging technology based on horse teams was capable only of tapping stands within two miles of streams large enough to float logs to local mills. Railroads during the interwar period changed all this by opening large stretches of previously inaccessible stands to loggers. The impact was greatest through the main stretch of the Skeena Valley itself where the low-lying terraces between Terrace and Hazelton were within easy access of the railway. After the Second World War, logging trucks supplanted the railway as the dominant means of transportation, opening remote backcountry stands that railroad loggers had previously been unable to reach. Second, during the postwar period, demand for Crown timber resources exploded in British Columbia due in large part to the postwar baby boom, the associated housing boom, sustained economic growth, a westward shift in the national population, and the expansion of transportation networks. However, the Skeena region remained a relatively minor producer of pulp and timber products during the during the 1950s and '60s, because much of this demand could be satisfied by the timberlands in Washington and Oregon, as well as from British Columbia's south coast, all of which were closer to main consumer markets and shipping ports. This pattern of resource production changed in the 1970s with the advent of environmental protection legislation in the United States, which restricted the flow of timber products from the American Pacific Northwest. Combined with the entry of Japan onto the world stage as a major economic engine and consumer of North American raw materials, the demand for wood products from the more remote corners of British Columbia began to mount, and regions such as the Skeena Valley became commercially practicable. Ultimately, to satisfy this rising demand and 'get 83 the cut out' of Crown forests, forestry officials in the Northwest shifted timber management policy from the more conservative selective harvesting model that was characteristic of the immediate post-war period, to a more aggressive and 'efficient' approach based on clearcutting (Weetman etal., 1990). During this period, forest policy and practices were designed to remove commercially valuable, overmature stands and replace them with homogenous immature plantations in the most efficient manner possible - i.e. through large-scale clearcutting and artificial regeneration. The forest industry's justification for clearcutting practices in the region were generally based on: • Its relative efficiency as a logging practice, • The use of planting to satisfy promptly legal reforestation obligations, • The decay, decadence, and windthrow hazards associated with the first harvest of many overmature, unmanaged old growth forests. According to the BC Forest Service, some early attempts at partial cutting in the region during the 1960s unsuccessfully attempted to stimulate natural regeneration, and other poorly regulated 'high-grading' practices on moist sites produced a legacy of poorly stocked hardwood-conifer stands. These obvious 'failures' led to a widespread discrediting of any partial cutting practices among foresters in the Skeena region (Weetman etal., 1990). The legacy of this management philosophy has produced large and often highly visible clearcut areas on mountainsides, lakeshores and roadsides (Figure 3.6). The consequence has been that forest management has been among the most visible and hotly contested environmental issues in the Northwest during the past two decades. Throughout the '80s and early 1990s, protests from First Nations 84 and the local non-Na t i ve publ ic e x p r e s s e d conce rn relat ing to a range of r esou rce i s s u e s inc lud ing wildl i fe habitat p reserva t ion , w a t e r s h e d v a l u e s , forest sustainabi l i ty , tour ism impac ts and loca l recreat iona l oppor tuni t ies. Industrial t imber product ion , with its large c learcu ts and impac ts to aquat ic , r ipar ian and terrestr ial habitat, e n e r g i s e d a local ly dr iven forestry reform m o v e m e n t in the Nor thwest in the 1980s and 1 9 9 0 s . At the s a m e t ime, the rapid l iquidat ion of o ld-growth fores ts in the region led to a n internat ional ly dr iven preservat ion m o v e m e n t that resul ted in s igni f icant reduc t ions of logging on provincia l ly regulated C r o w n lands . Figure 3.6: Extent of Clearcut Harvesting in the Mid 1980s Photo by the author. 85 In addition to dramatic changes to the physical and regulatory environment, the post-war period of large-scale, industrial management of the landscape had produced a second set of consequences. Early Euro-Canadian settlement in the Skeena Valley, as in the West at large, had been driven by a vision of limitless abundance. The forests seemed endless, the land in need of 'improvement,' and the land's resources generally were available for the taking. Even though forest management in the upper Skeena region started out conservatively, it quickly gained a momentum that seemed to overwhelm the good sense of the local foresters. Perhaps the most surprising thing about this environmental history is that in the 1970s and '80s, regional foresters set up plans knowing that harvests would drop precipitously as the Skeena's stocks of merchantable timber could not keep pace with demands set by the province's allowable annual cut, leading to probable mill closure in the late 1990s (Anonymous B C Forest Service Informant - personal communication). Unfortunately, this is what happened. Harvests collapsed at the end of the '90s, not because of environmentalists or beetle infestations, but because planners set it up that way, perhaps assuming that it was a reasonable price to pay for getting forests regulated as fast as possible. 3.4.2 Landscape Change Under the 'New Resource E c o n o m y ' The common thread shared by the explorers, trappers, farmers, ranchers, steamship companies, railroad corporations, town founders, and forest companies was a reliance on the natural landscapes of the Skeena district. Often the entire economy of a major community depended on a single natural resource such as salmon or timber. So dominant was a particular industry at many times in the history 86 of the region that they were occasionally referred to as "company towns." The economy of the Northwest historically and into the present is measured according to what can be extracted from the surrounding landscape. Unfortunately for the stability of the region's communities, most of these commodities have been subject to the whims of the market or the finitude of the resource. When the vein of gold ran out, felt hats fell out of vogue, prime quality saw timber was logged out, or the soil was exhausted, the raison d'etre for a community soon came into question. Similarly, a change in technology such as the transition from steamships to locomotives may render a town redundant or unnecessary. Just as quickly as resource booms commence, they often collapse with equal rapidity and many towns go bust in the process. This has been a common theme throughout much of the economic history of the Northwest. The boom and bust cycle, however, may change with the advent and development of a different economic sector, an economy that does not extract resources in the traditional sense. Although it is intimately tied to the land, it does not require cutting trees or digging up the rocks, turning the cows loose, or ploughing a prairie. This new sector in the Northwest's economy finds intrinsic, marketable value in the landscapes of the Skeena, Bulkley and Nass valleys. Lofty mountains, majestic views, sprawling forests, fish-filled streams, wildlife and unhurried towns have recently become valuable both to exurban migrants and roving tourists alike. Former boomtowns in the Northwest possess these amenities in abundance. Interestingly, many of these newfound assets were not so long ago regarded as liabilities. Historically, settlers judged mountains as too steep to farm and too 87 difficult to push transportation routes through. Whitewater rivers were less a source of exhilaration and more a navigation hazard or impediment to be overcome. Wild animals were a threat to people and livestock, and thick forests were forbidding and required taming. Today these hazards have become the attractions. The same landscapes that were once regarded as difficult to conquer are now valued for hiking, skiing, rafting, hunting and fishing, or simply living "close to nature" (Chipeniuk, 2004). What were once considered inadequately equipped little towns are now valued as bucolic villages without the congestion, pollution and crime of larger centres, but with the added attraction sometimes of romanticised and sterilised representations of First Nations and pioneer heritage (Figure 3.7). This recent appreciation is not completely new, but increasingly, large numbers of tourists, retirees and "back-to-the-landers" are choosing the towns of northwestern British Columbia as destinations. The "pristine" landscapes around them have become the "places to be." The Northwest's new sources of economic and demographic have a host of attendant concerns. The influx of new residents and visitors to the region come with expectations that impact the landscape. Although their expectations are embodied in very different uses of the land, often the results are equally transforming. At a general level, the expectations of tourists and newcomers alike are for a landscape that is 'natural,' or in other words, untrammelled by the kinds of land-uses that were characteristic of the past and which involved the clearing of large tracts of forestland. Such expectations have pitted recent migrants against long-time residents. At one level, natural landscapes are unproductive to the extent that they fail to provide the 88 k inds of mater ia l or e c o n o m i c benef i ts that h a v e histor ical ly sus ta ined Nor thwest commun i t i es . But at another leve l , logg ing roads b e c o m e mounta in b ike trails or a c c e s s roads for recreat iona l hunters, a n d farm f ie lds b e c o m e cot tage proper t ies for w e e k e n d vaca t ione rs . It m a y be te rmed 'min ing the v iew ' or 'se l l ing the scenery , ' but the growth and t ransformat ion that is beg inn ing to o c c u r in the Nor thwest is c a u s i n g m a n y to worry about a n e w sui te of impac ts - l oss of l ive l ihood, ove rc rowd ing , habitat reduct ion, etc. Figure 3.7: Ksan - Gitksan Village Museum H u m a n re l iance on the natural env i ronment will con t inue to evo lve in the Nor thwest . T h e l a n d s c a p e is still the pre-eminent def in ing character is t ic of the 89 region and its residents. Management of the landscape and the extraction of natural resources has sustained First Nations for generations and spurred Euro-Canadian exploration and settlement for two centuries. However, the deleterious effects of over-exploitation of the region's forest resources are a legacy only of the past twenty or thirty years. Expectations of what the landscape will provide are beginning to change, but now these expectations are rooted in a desire for the landscapes of the Northwest to sustain local residents in new, and perhaps more benign, but still consumptive ways. The ultimate question is, whether these new relationships with the forest landscape will create a future that is more or less sustainable, both from an ecological as well as a social perspective. 3.5 Chapter Summary The story of the changes that were made to the Skeena Valley landscape is, in its simplest version, a story of the land's transformation into a set of resources that could be removed from one landscape and transported to another. First Nations had certainly altered the landscape in a cyclical, modestly scaled and apparently sustainable regime of vegetation management. However, when Euro-Canadian settlers arrived they set into motion changes that far outpaced the previous period of transformation. The critical difference was that places like the Skeena Valley were becoming a source of commodity wealth - timber, minerals, and fish. Before Europeans came, the Skeena Valley was certainly connected to markets outside the region. Local villages had an extensive set of ties to trading networks that spread west to the Coast and east into the Interior Plateau. First Nations did extract elements from the local ecosystem, and in the process they changed the local 90 ecology to meet their needs. However, the process of transformation throughout the period of contact and colonisation by Euro-Canadian settlers not only resulted in greater and more destructive changes to the ecology of the region, but also led to the undermining of the resource base on which the economy of the region had come to depend. Future demands to tread lightly on the environment of the Northwest, although relatively advantageous to the extent that forest ecosystems will be preserved for the sake of animal habitat, water quality or scenery, nonetheless situate human values and expectations as the driving force that will shape the pattern and function of the region's landscapes. 91 4 Research Design & Methodology 4.1 Introduction It has been suggested that the selection of a particular research strategy should be determined by how the problem is shaped, by the questions that are posed, and by the type of end product that is desired (Kirby and M c K e n n a , 1989). Yet the reasons why a certain approach is taken occasionally go unmentioned in research reports. In this chapter, I provide an overview of some of the methodological approaches and assumptions that have been documented in related research on landscape preference. This is done in an effort to provide a theoretical rationale for the methods chosen, and to outline some of the common methodological postulates found in social science research that addresses landscape preference. In the second half of the chapter, I provide a procedural account of the methods that will be used for data collection and analysis, as well as comment on the measures that will be taken to ensure the integrity of the research findings. 4.2 Research Paradigms & Methodological Assumptions This dissertation poses two interrelated questions, the first of which addresses the dimensions or consequences that are salient in local stakeholders' evaluations of forest change. At a more detailed level of investigation, the second question analyses the participant sample in terms of First Nations and Euro-Canadian sub-groups, to explore the cultural or ethnic basis of landscape evaluation. T o address both questions, a primarily qualitative research paradigm has been adopted, due mainly to the differing goal orientations that characterize the 92 quantitative and qualitative research paradigms. A key difference between these broader paradigms is that, in qualitative research, it is argued that we must first discover what constructs are important in peoples' thinking and the reasons that they give for their perceptions and actions before undertaking more broadly based quantitative studies or programs of action (Schensul, era / . , 1999; Spradley, 1979). The focus of such research then is on how research participants define their world, what is important to them, and what contextual factors influence the thoughts and conceptions that they have of their environment. Thus, the most striking point of contrast between the qualitative and quantitative paradigms is the way in which each paradigm treats its analytic categories. The quantitative goal is to isolate and define categories as precisely as possible before the study is undertaken, and then determine the relationship between them with great precision and the power to generalize. The qualitative goal, on the other hand, is often to isolate and define constructs during the process of research. Thus , for the quantitative paradigm, well-defined categories are the means of research, particularly when they are well known in advance, while for the qualitative paradigm they are the goals of research. While question two has a more particular ethnic focus, the basic objective of both questions is to uncover the factors that people consider in their assessment of landscape condition. A s such, it is essential that the method will allow for an open and unbiased discovery of the constructs as the participants define them, in other words, largely unconstrained or unmediated by the researcher's own preconceptions. In general, qualitative research is concerned with the ability of the method chosen to illuminate concepts or ideas that are relevant to the research 93 participants and, by extension, grounded in the data that is obtained from them and other sources. A second and related justification for using qualitative research is its emphasis on discovery, particularly where the field of investigation chosen by the researcher can be described as relatively new or unexplored territory. Although it is not possible to characterise cross-cultural landscape preference research as a new field of inquiry, the potential contribution that research questions one and two can make to this area may be described as unique. For instance, Chapter 1 briefly outlined thirty years of social acceptability research that has addressed conceptions of acceptable forest management, but argued that there has been a near exclusive focus on the determinants of acceptability for commercial timber harvesting. While leading-edge research in this field is currently examining the social acceptability of alternatives to conventional timber harvesting (i.e. clearcutting compared with variants of partial cutting, see Kruger, 2004), absent from this body of work is a consideration of alternative and highly visible forest uses such as agriculture, recreation or subsistence. T o paraphrase the argument that was made in the introductory chapter, for communities that are transitioning from traditional modes of forest dependence, an expanded consideration of the acceptability of alternative forest uses is potentially interesting as well as unique in its approach. There is an equally lengthy history of cross-cultural environmental preference research that has investigated Asian (Pierce-Colfer, 2001; Tips and Savasdisara, 1986; Yang and Kaplan, 1990), European (Habron, 1998; Oresczyn and Lane, 2000; Van den Berg and Vlek, 1998) and Euro-American (Burger, 1998; Kaplan and 94 Talbot, 1988; Zube and Pitt, 1981; Zube and Sell, 1986) perceptions of and preferences for modified, managed or designed landscapes. However, missing from this body of work has been a systematic comparison of the landscape preferences of Asian, North American or European populations with those of aboriginal communities, particularly from C a n a d a or the United States. This is not meant to suggest that indigenous or First Nations environmental preference research is untrammelled territory. Indeed, there is a wealth of scholarship in anthropology and environmental sociology that addresses the meta-level values and worldviews that underlie indigenous environmental perceptions, particularly as they pertain to forested landscapes and how indigenous values are affected by landscape change (e.g. Basso, 1994; Booth and Jacobs, 1990; Kingsley, 1995; Mauze, 1998; see also Pinkerton, 1998 for an examination of Gitksan environmental values). However, there are few (if any) systematic or controlled comparisons of First Nations and European (i.e. Euro-Canadian or Euro-American) environmental preferences as they are expressed in relation to a defined set of forest landscape conditions. More exceptional yet are studies that have the practical objective of seeking to identify culturally salient landscape preferences with the ultimate purpose of informing forest management and landscape design practice. In sum, the need for information that is grounded in the participants' view of the world, coupled with a distinctive approach to the cross-cultural investigation of landscape preference, coincides well with the emphasis on discovery and localised meaning that is the hallmark of qualitative research. 95 4.3 Research Design A hybrid research design using elements of both naturalistic inquiry (or grounded theory) and conventional hypothesis testing was chosen as the basic research method (Glaser and Strauss, 1967; Glaser, 1992). A s Glaser and Strauss originally conceived it, naturalistic inquiry is predicated on an attempt to close a perceived gap between social scientific theory and empirical research. Consistent with the broader goals of qualitative research described in the preceding section, the focus of naturalistic inquiry is on the emergence of concepts or theories from the data collected, rather than on verifications of existing theories found in the scholarly literature. Due to its exploratory and inductive nature, pure or typical naturalistic inquiry is an iterative process where data collection is guided by the findings of the research process and theory is generated from the data as the investigation proceeds. Despite naturalistic inquiry's core emphasis on allowing constructs to emerge from the research context, most forms of social scientific investigation do not operate in a conceptual vacuum independent of the researcher's own hunches and cognitive baggage (Miles and Huberman, 1994). For instance, Miles and Huberman (2004: 8) note that grounded theory approaches to qualitative research, despite the emphasis on naturalistic inquiry, do not preclude the use or refinement of theories or conceptual frameworks. For instance, the theories presented in Chapter 2 that account for inter-individual variations in environmental preference will structure or affect in some manner the interpretation of data. Moreover, the research questions themselves may influence how data collection is conducted before the process 96 begins. Since one objective of this research is to contrast ethnic groups in their evaluations of modified forest landscapes, some consideration has to be given to the issue of comparability. This is a basic issue for question 2 as both the validity of the comparisons made, and the degree of confidence with which any observed differences can be attributed to inter-ethnic variation, can be cast into doubt without some prior consideration of the comparison groups and comparison dimensions. Thus, the process of participant selection raises issues that are discussed in detail below. 4.3.1 Sampling: Rationale for Community Selection This research involved two sampling decisions - the choice of a subject community, and the individual participants within that community who would contribute to the investigation. The sampling strategy used to select both the community and the research participants can best be described as purposive (Henderson, 1991; Johnson, 1990). Commonly used in qualitative research, purposive sampling is based on the notion that in order to gain the most insight into a particular phenomenon, a sample needs to be selected that can provide the most relevant and insightful information. In this study, there are three main rationales for selecting the sample community. I was familiar with an organization called the Gitksan Treaty Office ( G T O ) , an association of lawyers and resource and community planning specialists who, among other objectives, are working to make the resource management processes in their region of northwestern British Columbia more responsive to Gitksan needs and interests (rationale 1). After meeting with personal contacts with the G T O and 97 reviewing a number of their documents, I learned that they were engaged in several watershed-based planning exercises that would assist Gitksan community members in the articulation of culturally acceptable notions of forest management. It is also generally well known that the Gitksan are leaders among the First Nations of British Columbia in conducting and supporting ecological research, and possess a substantial repository of GIS and related land-use data. Moreover, as described in Chapter 3, fire has traditionally been a significant management tool for the Gitksan people, particularly for the cultivation of berry plants and other subsistence resources. The availability of digital landscape data, as well as historic fire-use information (i.e. photography and narrative accounts), made the Gitksan an attractive and sensible partner for this research (rationale 2). In addition to the research needs and interests that were expressed by the Gitksan, Chapter 3 described the challenges that have been brought about by thirty years of intensive forest management in the Northwest, and the subsequent collapse of large-scale commercial forestry throughout much of the region. These changes have created a greater need to understand First Nations conceptions of acceptable forest management, particularly for alternative land management practices that respect or accommodate cultural and in particular subsistence forest uses. However, this research need is as significant for the Euro-Canadian population that shares the landscape with their Gitksan neighbours, and for whom the economic changes of the last five to ten years have been nearly as disruptive. Therefore, both the need for an enhanced understanding of social acceptability, as well as the inter-cultural salience of forestry issues in the upper Skeena valley are clearly 98 relevant to research questions one and two (rationale 3). In this respect, the upper Skeena region provided what Goetz and LeCompte (1984: 82) would characterise as a "unique case selection." In effect, the congruence of my dissertation objectives with the social and economic circumstances of the upper Skeena valley was the most central consideration for selecting this region as the basis for my research. 4.3.2 Sampling: Rationale for Participant Selection T he decision to work with upper Skeena valley residents generally was not a unilateral one. A s a requirement of naturalistic inquiry, the first step of the research process involved establishing a formal relationship with community members through informal conversations, as well as more formal open-ended interviews to evaluate the feasibility of my dissertation project. Moreover, I needed to assess whom to approach within the local communities (i.e. the Hazeltons, Kispiox, Kitwanga and Cedarvale) for interviews and how to obtain access to documentary materials maintained by the Treaty Office, provincial agencies, and local municipalities. Establishing contacts with Euro-Canadian residents was a relatively straightforward process of meeting community members through introductions by key informants, as well as informally by introducing myself at various public gatherings and social venues. T o establish connections with prospective Gitksan participants, my initial meetings with G T O representatives left me with the impression that the organization would be willing to participate and assist in the identification of suitable Gitksan research participants. However, several G T O staff members advised me that the willingness of Gitksan community members could not be guaranteed or assured solely by the Treaty Office's endorsement. Ultimately, I 99 would need to pursue traditional Gitksan channels of authority and seek the permission of selected wilp (i.e. House) chiefs both to conduct field research in their territories, as well as to approach their respective wilp members and request their participation in the data collection process. For their part, the G T O would provide their written consent to conduct the research (see Appendix A5), as well as provide access to their archival materials and electronic databases. During the spring and summer of 2003, I spent considerable time attending meetings, introducing myself to members of the community, examined archival photographs and collected periodical materials that I borrowed from the G T O and local community college libraries (i.e. Northwest Community College in Smithers, Hazelton and Terrace). This was to familiarize myself with the Gitksan people, the history of Euro-Canadian settlement in the area, and the changing patterns of land-use over the last century and a half. However, the most valuable information came from the people with whom I had developed a rapport. These sources, both Gitksan and Euro-Canadian, provided me with a sense of who the most active and knowledgeable community members were, and who may be approached as key informants or as participants in the data collection process. In addition to building the trust of and gathering information from local residents, I needed to establish criteria that would make the participant selection process more focused or appropriate, systematic and logistically manageable. Criteria for the selection of participants were of two general types (Figure 4.1). T h e first of these (Criteria 1) refers to the innate abilities or characteristics of the participants. The second group of criteria refers to the theoretical qualifications of 100 the participants. I use the term theoretical qualifications for Criteria 2 to underscore the purposive role of theory in guiding the selection of participants in terms of such factors as their ethnicity and sociodemographic characteristics (e.g. age, gender, etc.) that are believed to influence landscape preference. While Criteria 2 are narrowly specified and attempt to draw a sample that is theoretically representative, Criteria 1 concerning innate abilities is more a function of personality and the ease with which the participant can communicate with the researcher. Figure 4.1: Participant Selection Criteria Criteria 1: Innate Qualities of the participant. e.g. willingness to co-operate, communication abilities. Criteria 2: Theoretical Characteristics of the participant. e.g. sociodemographic determinants of preference - age, gender, etc. T o satisfy the requirements of research questions 1 and 2, the participant selection factors for Criteria 1 were relatively straightforward and consistent for both research questions. In essence, I sought participants who, in the estimation of the key informants, have a demonstrated willingness to co-operate with researchers and possessed good communication abilities. By these, I had hoped to garner participants who already had some familiarity or experience in contributing to academic or other forms of research, and would not be intimidated by the structured or formalistic nature of the research enterprise. Implicit in both considerations was my desire to select participants who would be at ease with the research experience, who would be comfortable in articulating their perceptions of the world, and do so in a manner that honestly reflects their ideas and beliefs. Encompassed within Criteria 101 1 are Spradley's (1979) conceptions of a good participant as someone who not only knows their culture well, but also is willing and non-analytical in communicating about his or her world. By contrast, the factors defined for Criteria 2 were developed in a way that was specific to the data requirements of each research question. T o satisfy the requirements for question one (i.e. What is the range of consequences or dimensions that local stakeholders consider in their evaluation of modified forest landscapes?), I personally sought and asked key informants to recommend participants who have a recognised interest in forest management. By 'recognised,' I am referring to people who are known by members of their community to be concerned about the current and future condition of the region's forests, and who may play an active role in expressing their concerns through formal channels (e.g. public meetings, citizen committees, newspapers) as well as informal channels of communication (e.g. coffee klatches, social clubs). Although participants with a high level of interest and engagement in local planning issues may not be statistically representative of the larger population, it can be argued that they are influential in local planning processes and, therefore, provide an appropriate and illustrative case study that will shed light on the dimensions of landscape preference within this and similar populations. In effect, purposive samples of this kind can yield a small number of participants who provide a representative picture of aspects of information or knowledge distributed within a population. Those participants who are most interested in a given subject matter also tend to be the most accurate and reliable in terms of their reporting of issues and perceptions. This is consistent with the 102 conventional wisdom in ethnographic research that 'key' or active people make good participants (Spradley, 1979). Satisfying the requirements of Question 2 (i.e. How and why do preferences for modified forest landscapes differ between First Nations and Euro-Canadians?) appears relatively straightforward to the extent that it requires selecting a roughly equal number of participants drawn from both the local First Nations and Euro-Canadian communities. However, there are additional theoretically motivated factors that need to be addressed particularly in light of the discussion in Chapter 2, which posits that certain sociodemographic characteristics may influence landscape evaluations. While the empirical evidence for education and landscape familiarity is inconclusive or contradictory, there is relatively strong evidence that age (or generation) and gender may be strong determinants of preference. Thus, to control for the effects of age and gender, thereby isolating ethnicity as the independent variable, both the First Nations and Euro-Canadian samples were subdivided by generation with both male and female participants represented in each cohort. T h e final sample of research participants by ethnicity, age cohort and gender is show in Figure 4.2. A n additional hypothesised determinant of preference that is not reflected in the participant sample but was accommodated with some difficulty is occupation and, in particular, involvement in the forest industry. It is argued that 'culture clashes' occur in communities where a segment of the population that has been historically dependent on a single resource sector for employment, and thereby favours development, is challenged by newcomers who value the same landscape for its natural properties and believe that the setting should be preserved wherever 103 possible (Blahna, 1990; Bengston, 1994). However, factoring forest sector employment into the participant sample was complicated by a number of considerations. First, the communities in the upper Skeena have not received the influx of exurban back-to-the-land migrants that has been typical of larger B . C . centres such as Smithers, or the more trendy artists' communities of Nelson and Castlegar in the East Kootenays. I did learn that there are a handful of individuals in the upper Kispiox Valley who are self-described 'urban refugees' or draft dodgers that came to the area forty years ago and, moreover, have had no direct or indirect (i.e. through family members) connection with the forest industry. However, their meagre representation in the community combined with the fact that they have no analogue among the Gitksan precluded them from serious consideration as an additional sub-category in the participant sample. Second, and perhaps reflecting the degree of forest dependence in the Figure 4.2: Participant Sample A g e Cohort First Nation Euro -Canad ian Elder (Male) 4 2 Elder (Female) 1 2 Elder Total 5 4 Baby Boomer (Male) 6 7 Baby Boomer (Female) 2 2 Baby Boomer Total 8 9 Generation X (Male) 2 1 Generation X (Female) 0 1 Generation X Total 2 2 Total N = 15 15 Age Cohort Assumptions: Elder: 61 > Baby Boomer: 40-60 Yong Adult: 20-39 104 region, the employment history of upper Skeena residents is fairly homogenous to the extent that most people have been employed in the forest industry at some point in their lives (Key Informants - Personal Communication). Indeed, locating people without forest sector experience proved to be difficult, and was limited to people working within specialised occupations such as banking or the hospitality industry. For the majority of the community, although the exact nature of their forestry experience varies between individuals (e.g. logger, mill worker, cruiser, spacer), forestry work has been a significant source of livelihood across gender, ethnic groups and generations. O n e approach to representing forest sector employment in the participant sample would have involved breaking the groups (defined by ethnicity x age cohort x gender) into further subcategories according to type of forestry employment (e.g. logger, mill worker, cruiser, spacer, etc.). This was rejected because it would have produced a large and complicated number of subcategories, making the ultimate goal of conducting an inter-ethnic comparison of landscape preferences needlessly awkward and time consuming. Ultimately, controlling for the effects of forest industry employment on preference was achieved by precluding residents with no forest sector experience from the sample, thereby ensuring that identification with the resource extraction culture through the forest industry is nearly similar across ethnic (i.e. in both Gitksan and Euro-Canadian communities), gender and age cohorts. In addition to building a participant sample through key informants and spontaneous introductions, an additional strategy that expanded and possibly 105 strengthened the participant sample was the use of informal networks or the "snowball" technique as recommended by Goetz and LeCompte (1984) and Werner and Schoepfle (1987). Claiming that most qualitative sampling is opportunistic or less than explicit, both groups of authors suggest building on or expanding systematically obtained samples through the use of social networks. Based on an understanding of these social networks, the 'well-informed' participant should emerge. T o this end, every participant that was identified through the criteria identified above was asked for names of other knowledgeable or active community members who could provide helpful perspectives in an interview. Within the Gitksan community, wilp chiefs were asked to provide names and contact information for individuals that satisfied criteria 1 and 2. Based on the criteria-based key informant lists, personal introductions and network referrals, thirty people were invited to participate in the research. Austin (1994) recommends a sample size of eight to ten participants per comparison group when using semi-structured or structured interviews that employ some form of 'systematic data collection' method (Miles and Huberman, 1994). 1 According to McCracken , the first principle of participant selection in qualitative research is that "less is more" (1988: 35). In effect, it is important to work longer, and with greater care, with a few people than more superficially with many of them. While many quantitatively oriented social scientists may be contemptuous of samples that are so small, it should be remembered that qualitative participant samples are not chosen to represent some part of the larger world (i.e. for the purpose of statistical Q In this case, the main sub-groups that are being compared are defined in terms of ethnicity, or First Nations and Euro-Canadian participants. 106 generalization). Rather, qualitative samples are driven by theoretical or purposive selection criteria, which are designed to maximize the opportunities for uncovering the phenomenon of interest, and to probe deeply into the character and nuances of a culture. It should only be after the researcher has taken advantage of a qualitative investigation that he or she should be prepared to determine the distribution, frequency and generalisability of the cultural phenomena that have come to light. 4.4 Stimulus Materials 4.4.1 Photo-Elicitation as a Qualitative Research Technique Qualitative research designs commonly employ interview or observational techniques without the aid of extraneous stimulus materials. However, stimulus materials in the form of visual aids can be useful during semi-structured interviews in that they give research participants the opportunity to pull together and discuss their perceptions of items, sometimes leading them to observations that may be difficult for the participants to articulate through a more conventional interview format (i.e. discussion in the absence of stimuli). Use of the materials allows participants to focus on and visualize several concepts or items at a single moment, allowing for discussion of more complex interrelationships and issues. In other words, it allows participants to express themselves in a more systematic fashion. Since its first documented use in the late 1950s (Collier and Collier, 1957), photo-elicitation has remained a straightforward method to understand and employ. It involves using photographs to invoke comments, memory and discussion in the course of a semi-structured interview. The strength of the technique rests in the power of images to invite people to take the lead in an inquiry, to render full use of 1 0 7 their expertise. For instance, interviews can often become stilted exercises as the interviewer probes for detailed information that the participant seems reluctant or incapable of articulating, but visual information can invite open expression while maintaining concrete and explicit reference points. Psychologically, the stimulus materials perform as a third party in the interviews, shifting the focus of attention and pressure away from the participant as the subject of questioning. In effect, both the interviewer and the participant engage the photographs together, and the role of the participant shifts from research subject to that of a co-investigator in discovering answers to questions posed in relation to the photographs. The participant's role can be seen as one of expert guide leading the researcher through the content of the pictures. The incentive to take on the role of expert guide often comes from the fact that people have a fascination with settings or visual information that they know intimately and a common compulsion to express that knowledge. The need for stimulus materials as part of the research design for this dissertation was based on three considerations. First, because the basic objective was to gauge public preferences for different modified landscapes, it was felt that the presentation of a carefully chosen set of images would be a more efficient method of gathering this information than taking 30 participants into the field to discuss forest settings that are in various stages of disturbance, and separated by large distances. In addition, it would minimize interpersonal differences in imagining future conditions. Second, the 'grounded theory' or reconnaissance interviews conducted in the summer of 2003 revealed the difficulty that many people encounter in their effort to describe landscape conditions in the abstract (i.e. without the benefit of 108 some tangible or visual referent), as well as their thoughts and feelings about them. This situation was particularly problematic for the Gitksan participants who were asked to describe in some detail the characteristics (i.e. spatial extent, location and frequency of disturbance) of traditional aboriginal burning practices for berry crops. O n several occasions, the information was difficult for the participants to express, either because the memories for these vanishing landscape features have faded, or they simply lacked the means to articulate the information, or both. In either case, it seemed worthwhile to use photo-elicitation as a means of jogging memories or simply providing reference points for information that participants cannot easily express. Third, this study provides a unique opportunity to use visual simulation techniques (described further in subsection 4.4.4) to depict and elicit preferences for modified landscape conditions that no longer exist, such as the burning practices described in Chapter 3, as well as future conditions that are not yet part of the upper Skeena landscape (e.g. 'new forestry' harvesting methods or preservationist approaches to landscape management). Moreover, the power of landscape simulation to turn the clock back and represent historical landscape conditions was identified as one means of capturing information about traditional land management practices that are quickly disappearing from the Gitksan community. In a qualitative study of this nature, the selection of images for presentation to interview participants is driven by the same purposive logic used in the selection of the participants, that is, to maximize the opportunities to uncover and elicit information pertaining to the subject of interest - i.e. preferences for alternative modified landscapes. A s such, some basic decisions and criteria needed to be 109 established about the content of the images and, with regard to computer-based simulations, how they were to be produced. At this juncture, the defensibility of the stimulus materials needs to be established by providing an explicit description and rationale for the following factors: • Landscape conditions: see section 4.4.2 Field Reconnaissance; • Landscape settings: see section 4.4.3 Viewpoint Selection; • Image production: see section 4.4.4 Image Production Methods. 4.4.2 Field Reconnaissance The first step in developing the stimulus materials involved a scoping exercise that required several lengthy visits to the Northwest during the first field season from April to August of 2003. In addition to developing a rapport with the community, meeting key individuals, and assembling documentary materials, the purpose of these visits was to familiarize myself with the upper Skeena landscape and the history of human-induced forest disturbance in the region. By documenting examples of historic and recent forest modifications, it was my intention to use these as templates or prototypes for the conditions that would be presented in the semi-structured interviews to be held the following year. The scoping exercise was approached in three ways. First, a number of informal meetings were held with local community members including G T O representatives to address logistical issues pertaining to the research, as well as to discuss the history of forest use in the study area. These discussions were intended to elicit information about the purpose, location and extent of aboriginal burning practices; and the history of industrial logging, particularly in terms of its scale and distribution throughout the region, as well as the harvest and silvicultural methods 110 employed. These and a range of other topics were addressed through a series of relatively formal but open-ended interviews conducted throughout the summer of 2003 with 14 Gitksan and non-Gitksan community members, and 3 local forest ecosystem experts. Second, a cartographic profile was created to provide a visual inventory of recent timber harvesting activity throughout the study area. Constructed from GIS-based forest cover datasets obtained from the B C Ministry of Forests as well as Gitksan sources, the maps served to illustrate the size, density and timing of industrial logging operations, particularly as they proliferated during the 1980s and early '90s throughout the upper Kispiox and Suskwa watersheds. Third, Gitksan informants and forest ecology experts accompanied me on tours of the landscape to scrutinize the region and help me to develop some familiarity with local forest types and biogeoclimatic classifications. More specifically, the tours were intended to visit and photo-document areas that were burnt by the indigenous population to sustain culturally important food plant communities (e.g. black huckleberry) and wildlife habitat (e.g. mountain caribou corridors). T h e sites visited have been documented as aboriginal burn areas by other researchers (Johnson, 1997; Trusler, 2002), largely from evidence provided by Gitksan cultural experts, archaeological evidence (e.g. bentwood boxes or drying racks), ethnographic reports and fire history data. In combination, the information gleaned from formal interviews and informal conversations, GIS mapping data, field visits and ethnographic reports was consolidated and reviewed to identify easily accessible examples of aboriginal burn m sites, as well as cutblocks from the 1970s to the late 1990s. I subsequently visited these locales to take photographs, and performed the inventory according to a few technical specifications that are standard practice for visual reconnaissance exercises (Sheppard, 1989): • Panoramic photography ( 1 8 0 ° ) was taken with a 35 mm camera equipped with a 50 mm lens, using colour slide film and mounted on a level camera tripod. Panoramic photos with a 50 mm lens provides the closest approximation to the average human field of vision, and colour slide film is considered to be superior to analog print and most digital camera formats for providing accurate colour representation. • Photograph locations were recorded in Universal Transverse Mercator (UTM) co-ordinates using a standard, handheld Trimble® Geographic Positioning System (GPS) unit. In addition to creating a digital record of the camera positions, manual field notes were compiled to record the height of the tripod, compass bearing, panorama sequences (number and degree intervals), and lighting and weather conditions. • The pictures were taken between the months of June and August during the height of the growing season, and immediately prior to the onset of fall colouration, which typically comes early in the Skeena region during mid to late September. Photographs were taken during ideal lighting (i.e. non-overcast) conditions, with the landscape or feature of interest suitably front lit. • Wherever possible, the photos were taken without foreground plants or topographic features to provide a clear, unobstructed view of the feature of interest. Photo locations were also selected to be representative of the various landscape and forest disturbance types that are evident throughout the Gitksan territories. • 36 panoramic photographs were taken from 30 viewpoint locations, and reflect a typology of human-induced forest modifications in the upper Skeena valley, as described in section 4.4.4 (Image Production Methods). 4.4.3 Viewpoint Selection A key advantage of computer-based landscape simulation is its ability to depict alternative landscape conditions from a single location in order to represent 112 historic conditions, natural processes of ecological succession, as well as alternative patterns of landscape management and design. This allows the researcher to represent a suite of landscape changes from the same viewpoint, in effect, controlling for the effects of variables (e.g. lighting, season, differences in place significance or familiarity, presence or absence of water, etc.) that may spuriously affect or bias the viewer's responses if the setting is allowed to change. T he field reconnaissance exercise and resultant photographic inventory helped to identify the range and types of forest disturbances that are present on the Gitksan landscape. These photos were key in the identification of possible disturbance conditions for the interview stimuli; however, a number of additional considerations needed to be addressed regarding the presentation format. The primary issue involved determining how many viewpoint locations would be chosen and, more importantly, the criteria that would be used to select the viewpoints. The criteria were intended to minimise viewpoint selection as a potential source of bias in participant responses, as well as to ensure the credibility of the conditions that were created. Figure 4 . 3 presents the viewpoint selection criteria that were used. 113 Figure 4.3: Viewpoint Selection Criteria Criterion 1 Accessibility - viewpoints should be locally accessible and familiar to a wide segment of the community. Criterion 2 Visibility- viewpoints should present a full view of the landscape and be relatively free of occluding or framing elements. Criterion 3 Credibility/Defensibility of Forest Modification Types - demonstration that the viewpoint chosen, both historically and into the future, can represent a range of forest modification types. la Criterion 1 stemmed from the need to ensure that the images used in the interviews would depict settings that are equally accessible and familiar to all segments of the participant sample. Since several of the photos had been taken in relatively remote, back country locations, it was conceivable that participants who are more familiar with such areas would be more sensitive to particular forms of forest disturbance. Thus, to eliminate familiarity as a potential source of bias, it was decided that the photographic inventory would be culled to produce two or three viewpoint locations that are likely to be accessible and familiar across the participant sample, regardless of lifestyle or other factors that may generate biased results. Criterion 2 omitted locations from which it was difficult to see the full extent of the landscape, either because foreground elements obscured a significant portion of the mid or background features thus minimising their visual significance, or where landscape elements such as trees and nearby ridge lines framed the view within a vista thereby accentuating the visual significance of mid or background portions of the landscape. Finally, Criterion 3 specified that ideal viewpoint locations should be able to represent a range of potential or historic forest modifications (e.g. aboriginal burning for berry production or intensive timber harvesting). In effect, I wished to 114 a v o i d a n in te rv iew s i tua t ion w h e r e a pa r t i c ipan t w o u l d d i sc red i t t he s i m u l a t e d l a n d s c a p e c o n d i t i o n s b e c a u s e 'that c o n d i t i o n w o u l d n e v e r h a v e h a p p e n e d t he re ' o r ' that cond i t i on c o u l d n e v e r h a p p e n the re . ' In o the r w o r d s , I w a n t e d to e n s u r e that t he v i ewpo in t l o c a t i o n s w o u l d not u n d e r m i n e the cred ib i l i t y of the m o d i f i e d fo res t c o n d i t i o n s . Figure 4.4: Viewpoint 1 Photo by the author. V i e w p o i n t 1 (F i gu re 4.4) is s i t ua ted o n a s c e n i c ou t l ook f a c i n g w e s t o n the K i s p i o x V a l l e y R o a d , a n d o v e r l o o k i n g the t o w n of H a z e l t o n at the c o n f l u e n c e of the B u l k l e y a n d S k e e n a r ivers . S i t u a t e d a s it is a l o n g a h e a v i l y t r ave l l ed co r r i do r b e t w e e n the c o m m u n i t i e s of H a z e l t o n a n d K i s p i o x , th is l oca t i on w a s s e l e c t e d for its o b v i o u s a c c e s s i b i l i t y a n d fami l ia r i ty to t he l oca l p o p u l a t i o n . H o w e v e r , th is l oca t i on is i dea l l y s u i t e d a s a v i ewpo in t b e c a u s e of the r a n g e of a n t h r o p o g e n i c a n d na tu ra l d i s t u r b a n c e pa t t e rns that a r e v i s i b l e in the foo th i l l s s u r r o u n d i n g the c o m m u n i t i e s of n e w a n d o ld H a z e l t o n . T rad i t i ona l l y , m u c h of t he va l l ey bo t t om w a s burnt a n d m a n a g e d by t he G i t k s a n a n d the i r W e t s u w e t e n n e i g h b o u r s in H a g w i l g e t for l o w - b u s h b l u e b e r r y a n d s o a p b e r r y p r o d u c t i o n . E c o l o g i c a l l y , t he flat te r ra in a n d xe r i c s o i l s that a r e c o m m o n t h roughou t the v a l l e y bo t t om a r e e s s e n t i a l for the cu l t i va t ion of t h e s e cu l tu ra l l y s ign i f i can t p lan t s p e c i e s . E v i d e n c e that th is a r e a h a s b e e n the sub jec t of 115 frequent and extensive aboriginal burning is substantiated by the dominance of deciduous vegetation throughout the valley, as well as oral histories and early surveyor's records that describe the terrain surrounding Hazelton as a "savannah-like" landscape filled with a patchwork of grasslands, berry patches and forest bosques (Dawson, 1881, Poudrier, 1893). Like so many other landscapes throughout the region, proximity and visual Figure 4.5: Viewpoint 2 prominence has done little to preserve the area from the effects of industrial logging, the remains of which are still visible on the cedar-hemlock dominated foothills northwest of Hazelton. Thus, owing to a combination of this landscape's familiarity and accessibility to both aboriginal and non-aboriginal residents, its rich history of active management by the Gitksan, as well as the visible effects of current industrial and natural disturbances, Viewpoint 1 was selected as an ideal setting from which perceptions of modified forests could be discussed with aboriginal and non-aboriginal participants. Viewpoint 2 (Figure 4.5) shows the western edge of the Roche de Boule range, a prominent local feature that can be seen from the community of Gitsegukla 116 on Highway 16 as well as the Gitsegukla Forest Service Road (FSR). The feature's distance from the highway, as well as the presence of occluding foreground elements, were among the major considerations for excluding the highway or village as potential alternative viewpoint locations. Local residents (i.e. First Nations and non-First Nations) use the F S R extensively during the summer and autumn to access berry and mushroom picking grounds, as well as for hunting and fishing, despite its primary function a backcountry corridor for logging and forest service vehicles. In addition to providing a clear and unobstructed view of a major landscape feature, there is some evidence that the forests along the western slope of the range have been modified extensively over the last several decades, and perhaps centuries (Gitksan Informant, Personal Communication; Trusler, 2002). Most recently, the imprint of timber harvesting is visible along a bench that extends across the mid-slope portion of the range. Two large clearcut openings in the middle of the photo have matured or 'greened-up' significantly since they were harvested in the mid 1970s, and industrial logging continues to occur on a smaller scale, and with a less geometric signature to the immediate right on the southern portion of the range. In addition to ongoing forestry activity, a subtle but extensive feature of the forest cover below the bench is a mixed canopy of deciduous and coniferous forest (i.e. aspen, birch and mountain hemlock) that is visible in part because of its sharp upper edge and lighter hue in relation to the surrounding cover. From several conversations with Gitksan wilp chiefs and researchers, I have learned that this portion of the mountainside is traditionally known as Luu Skayansit, and refers to a 117 locally significant berry harvesting and processing site that has ostensibly been used by the local First Nations (i.e. both Gitksan and Wetsuweten) communities for generations (Gitksan Informant, Personal Communication). Since the site is situated on Crown Land, human-induced fires have not occurred on Luu Skayansit since the Forest Service's prohibition on prescribed burning in the mid-1940s. A s such, any assertion that the Gitksan traditionally managed the site is fairly circumstantial. According the ethnobotanical literature (Johnson, 1997; Trusler, 2002), the site is a sensible location for berry management given its proximity to the communities of Gitsegukla and Gitanmaax, and its vertical elevation (i.e. position on the mid-slope portion of the range) is ideal for the cultivation of black huckleberry. Archaeological work in the area has produced numerous bent-wood boxes, traditionally used by the Gitksan in the storage and transport of dried berry products, and evidence of drying racks and temporary habitation has turned up on the bench immediately above the burnt portion of the range, suggesting that the site was once the focal point of significant berry production activity. Consistent with the selection criteria, Luu Skayansit was chosen as an ideal viewpoint location due to the combination of the viewpoint's accessibility, visibility and breadth of historic, current and potential forest modifications. T he landscape at Viewpoint 3 (Figure 4.6) is the south-facing aspect of Blunt Mountain within the Suskwa valley 10 kilometres east of Hazelton on Highway 16; this is the major transportation corridor that connects the coastal community of Prince Rupert with Prince George in the interior, as well as the towns of Terrace, Hazelton and Smithers in between. In addition to its visually prominent situation 118 Figure 4 . 6 : Viewpoint 3 Photo by the author. along a busy travel route, it is visible to a loosely scattered residential and farming population that inhabits much of the valley bottom north of the highway and the Skeena River. Conversations with key informants also attest to the area's lengthy and continuing legacy of forest management. Given its proximity to the aboriginal villages of Hagwilget and Gitanmaax, Blunt Mountain's southerly aspect and temperate mountainous terrain makes it ideally situated for black huckleberry (Burton, 1998) and both oral histories and archaeological evidence point to the concentrated use of much of the valley corridor for berry production (Dawson, 1881; Key Informant, Personal Communication). More recently, the timber resources of the valley have been the focus of large scale commercial harvesting. The 80-hectare cutblock shown in the viewpoint photo reflects the scale and nature of timber removal that was prevalent throughout the Gitksan territories during the 1980s. As recently as the late 1990s, adjacent areas within the watershed continue to be harvested using clearcut silvicultural practices despite the area's visual prominence, although the scale and configuration of the cutblocks reflect the different forest 119 management paradigm that was brought about with the introduction of the Forest Practices C o d e in the mid 1990s (See Figure 3.4). 4.4.1 Image Production Methods Altering the viewpoint photographs so that they would depict different forest modifications required a combination of computer and photo simulation techniques. The simulation procedure chosen consisted of three-dimensional surface modelling (Lewis, et al., 2005), an approach that represents a common method of relatively accurate and precise landscape simulation that is employed by forestry and environmental planning practitioners throughout North America (McGaughey, 1998; Sheppard, et al., 2004). A s such, my decision to use this method is representative of current practice in landscape simulation, and represents an industry tested and affordable method. Generically, surface modelling software consists of workstation based computer programs that create and display perspective, full-colour images of landscapes ranging in size from a forest harvest unit to the multiple watershed level (i.e. 10 to 8000 ha). Most surface modelling packages are capable of displaying ground models, or terrain represented by a digital terrain model (DTM), forest stands, and generic line information representing roads, harvest unit boundaries, and so on. The surface-modelling package selected to construct the simulations for this research, Visual Nature Studio ( V N S ® ) , is currently the preferred landscape or surface-modelling package used by the B C Ministry of Forests and several landscape planning consultants throughout British Columbia. The major factor underlying its status as the preferred modelling software by government and industry, and use for this research, is the software's ability to manipulate tree stands 120 and simulate various operational or landscape planning alternatives to a high level of accuracy and photographic realism, which in turn draws on the following capabilities: • Accurate ground surfaces can be constructed from digital topographic files obtained from the B C Ministry of Sustainable Resource Management. • Forest stand data can be imported directly from B C Ministry of Forests stand inventories. Visible stand characteristics such as species composition, stand height and stand density can also be built into the simulation from B C M o F data. • Tree libraries can be built from photographs of trees and understory vegetation to replicate vegetation characteristics such as colour, composition and texture. • Ephemeral conditions such as sunlight intensity, ambient light, atmospheric blur, fog and cloud cover can be calibrated to site photographs. • Disturbance conditions such as harvest cutblocks, fire patches and plantings can be imported as polygon boundaries from a geographic information system (GIS). Using Visual Nature Studio, four modified forest conditions were created for each viewpoint location (i.e. 4 conditions x 3 viewpoints - 12 images). The simulations are representative to the extent that they reflect two of the dominant forest conditions that have been characteristic of large (i.e. landscape) human-induced forest disturbances in the Skeena region over the past century and a half -i.e. aboriginal burning for berry cultivation and industrial timber harvesting (as described in Chapter 3). T h e remaining simulated conditions reflect potential or future forest conditions that may exist in the upper Skeena landscape given some assumptions derived from conversations with local residents and forest management and forest ecology experts. In effect, the simulations reflect a temporal sequence of forest change from the relatively recent past (i.e. from the early to late twentieth 121 century) to a reasonable approximation of future landscape conditions. Despite a considerable amount of inference and assumptions in the design of the simulations, information used in the design of the modified forest conditions was derived from a variety of documented data sources including field reconnaissance photography and notes, archival ground and aerial photographs, forest development plans, forest inventories, 19 t h century Canadian Pacific Railway and Hudson's Bay Company surveyors' diaries, and open-ended interviews which, in part, were intended to elicit detailed information regarding the physical characteristics of forest openings created by aboriginal burning. The following discussion introduces the simulated conditions with a general description of the conditions portrayed by each. Condition 1: Pre-lndustrial or Aboriginal Forest Management T h e Pre-lndustrial simulations were the most difficult to produce in large part because aboriginal burning for food production has not been practiced on a large scale in the Gitksan territories for nearly a century. The last of the Gitksan fire experts who practiced this form of management first-hand passed away during the 1980s and early 1990s (Leslie Johnson, Personal Communication 2003), and most of the direct visual accounts of how and where berry management was practiced are provided by elders who witnessed the practices but whose memories were often imprecise. 2 The most precise source of information that was used to size and obtain some sense of the geometry of Gitksan berry patches came from photographs, mostly taken in the first three decades of the twentieth century, and maintained by the B C archives in Victoria and the Northwest Community College campus in Q While there are residents of the Gitksan communities who burn small patches of reserve land for berries or brush management, large-scale berry management using traditional rotation techniques is largely a vestige of the past. 122 Terrace (Figure 4.7). High quality prints of the photos were obtained from both archives and shown to Gitksan plant and berry management experts to verify that the forest disturbances shown in the photos were berry patches that had traditionally been burnt by Gitksan people, and not some other form of natural or anthropogenic disturbance - e.g. wildfire or clearings made by mineral prospectors. These conversations, and other documented sources (Gottesfeld-Johnson, 1994; Trusler, 2002), were instrumental in deriving the physical parameters or assumptions for simulating traditional burn sites, as follows: • The spatial scale or size of sites that were repeatedly burnt varied from year to year and among Gitksan wilps. How large the berry patch should be was traditionally based on a House chief's estimation of the quantity of berries that would be required to sustain their wilp members. In effect, since wilp populations varied, so too did the size of the patches that were burnt. However, recent ethnobotanical studies in the upper Skeena region suggest that the 'typical' berry patch mosaic varied between 80 to 100 hectares (Trusler, 2002). • Burning occurred within relatively small sub-patches within an overall mosaic of productive berry growing grounds (see Chapter 3). In effect, berry management consisted of a shifting mosaic of recently burnt, fallow and productive patches. Patches were burnt every five to ten years to renew areas that were passing peak productivity, and to ensure that there would be a continual flow of berries from adjacent productive patches. • Perhaps the most conclusive piece of evidence about the physical dimensions of Gitksan berry patches pertains to how they were sited. According to Gitksan informants and recent ethnobotanical research (Burton, 1998; Gottesfeld-Johsnon, 1994; Trussler, 2002) the Gitksan chose south and southwest mountain aspects to maximise solar exposure, and typically created openings in the forest canopy on mid to upper elevations where cooler temperatures and abundant precipitation were particularly favourable for huckleberry cultivation. This pattern of patch location seems to be consistent with the photos that were verified as genuine traditional berry sites, as well as from the aspect and elevation of known patches such as Luu Skayansit (Viewpoint 2). 123 • Burning was used primarily for berry management, but it was also useful for other land-uses such as grassland and range management, particularly during the first half of the twentieth century. Photographic evidence, local informants (both Gitksan and Euro-Canadian) and surveyors diaries (Poudrier, 1881) suggest that extensive portions of the Skeena valley lowlands were cleared for agriculture and range management. In addition, the prevalence of steam ships on the Skeena prior to the advent of rail transportation often meant that timber was cleared or harvested from the river's banks to provide fuel for the boats. This pattern of land-use may account for Poudrier's (1881) description of the region in the late nineteenth century as "savannah-like." Figure 4.7: Gitksan Berry Patches, Kispiox Valley, 1948 Image courtesy of the BC Archives, Photo # 1-21896. Although lacking in measurement detail, these assumptions were factored into the design of the pre-industrial condition for all three viewpoints (Figure 4.8). Viewpoint 1 is consistent with the latter description of the Skeena Valley lowlands 124 during the early twentieth century as an open expanse of range and pasture, maintained in part by aboriginal burning. A simulated recent burn can be seen in the middle of the simulation where the ground shows the mottled black, charcoal and light brown pattern that is typical of low-intensity grass and shrub fires. In addition, I have shown small patches cleared in the forest cover adjacent to the Skeena River in an attempt to depict the pattern of riparian clearing that ostensibly resulted from fuel wood gathering by the Skeena river steamers (Bennett, 1997). The hills that dominate the background portion of the Viewpoint 1 simulation are covered with an unmodified stand of trees. I could find no evidence from local Gitksan informants that these hillsides had ever been altered for berry production, or for timber management during the early twentieth century. T he simulations for Viewpoints 2 and 3 both depict traditional berry management at Luu Skayansit and the Suskwa Valley, respectively, in different stages of productivity and as they may have been seen in the early 20 t h century. The burn patches for both viewpoints are approximately 80 hectares in size and, within this larger patch, there are a number of smaller (20-30 hectare) rotational patches of recent low intensity burns, and older productive berry patches. Two points are worth emphasising in relation to these smaller patches. First, while the size and location of the larger patch matrix is derived from relatively substantial evidence, the size, shape and adjacency relationships of the smaller rotational patches are based on a greater degree of conjecture. In effect, I was unable to find any information (written or oral) that would allow me to replicate the actual Gitksan practice of patch rotation with any degree of accuracy. 125 Figure 4 .8 : Landscape Simulations, Pre-lndustrial Condition Viewpoint 1 126 A s the interview data in Chapter 5 attests, some more experienced elders (to the extent that they had witnessed berry management practices as children) had some difficulty with the simulated patch sizes and challenged my assumptions based on their own indirect experience. Second, the colouration for the patches was derived from B C Ministry of Forests aerial and ground level photography of low intensity shrub fires. Portions of these aerial photos were cropped using Adobe P h o t o s h o p ® and draped onto the V N S digital elevation model as 'texture maps' in the areas that corresponded with recently burnt patches. Photographs of live shrubs in the upper Skeena valley were used to create the productive huckleberry patches. A s with the burn surface textures, these photographs were cropped to isolate the huckleberry plant, which was then imported as a set of flat images into the simulation software. These flat images were in turn placed throughout the surface model within 20-30 hectare polygons that I had defined in V N S as mature berry patches. Condition 2: Industrial Timber Management Stepping forward in time, the Industrial condition reflects the advent of large-scale timber management in the upper Skeena Valley during the 1970s and early 1990s. A s described in Chapter 3, the prohibition of aboriginal burning in the 1940s allowed the region's forests to return to a more or less natural condition, as old berry producing sites were abandoned to natural processes of succession, and post war timber demands were limited to the selective removal of single stems by small scale sawmill operators and pole companies. However, by the 1970s, clearcut harvesting began to dominate the upper Skeena landscape, particularly through the Suskwa, 1 2 7 Kispiox and Bulkley valleys as demand for 'virgin' timber pushed industrial scale harvesting further into the Northwest (Weetman et al., 1990). A s a consequence, the forest landscape was once again subject to extensive disturbance, but this time the general pattern of clearcut harvesting left large, unnatural or geometric openings throughout the landscape. Simulating this period of landscape history was a relatively straightforward task of replicating the current forest condition and imposing on the surface model the cutblock boundaries that were designed twenty to thirty years ago, as recorded digitally in forest cover datasets obtained from the B C Ministry of Forests. In addition to spatial data, the datasets contained detailed attribute data on the area harvested, the silvicultural technique used, and the harvest year. From this information, I could produce a series of maps that illustrated where the harvest activity occurred throughout the 'industrial' period, as well as derive statistics that reflected the range and average cutblock size. However, the most challenging decision involved selecting a reference year on which the simulations should be based. In other words, what point in the twenty-year period of industrial harvesting should the simulations depict? Analysis of the forest cover dataset revealed that large clearcuts (i.e. average of 65 hectares, with a range of 55-80 hectares) began to appear in the upper Skeena region in the early 1970s between 1972 and 1975. Clearcut openings larger than 55 hectares appeared at all three viewpoint locations in 1975. With the exception of Viewpoint 2 where there was no timber management activity after 1975, a new block was clearcut with near perfect regularity every five years at Viewpoints 1 and 3 until 1990. If 1990 had been selected, or some point in 128 between as the reference year (e.g. 1980 or 1985), it is possible that the profusion of large geometric openings on the landscape would have 'stacked-the-deck' against clearcut forest conditions, in effect showing the landscape at its absolute worst and thereby eliciting preference responses that would almost invariably be negative. A s such, 1975 was chosen as the reference year for the Industrial condition to control for this potential source of response bias. From the Ministry of Forests GIS dataset, polygon boundaries for the 1975-clearcut openings at Viewpoints 1 through 3 were imported and draped onto the V N S surface model. The Industrial Condition simulations shown in Figure 4.9 therefore correspond with the actual harvest pattern and block geometry on the upper Skeena landscape as it appeared in 1975. Colouration for the cutblock ground surfaces was derived from an aerial photograph taken of a clearcut in southeastern British Columbia and archived by the Collaborative for Advanced Landscape Planning (UBC). A portion of the photo was cropped and draped onto the surface model as a texture map within areas defined in the V N S model as clearcut openings. Some colour adjustment and calibration was performed in Adobe Photoshop to bring the texture map as close as possible to the colouration of cutblock surfaces that were photographed in the study area. 129 Viewpoint 2 Condition 3: Natural or Pristine Forest For obvious reasons, a natural forest condition cannot be described as a modified landscape. However, I have chosen to include a natural landscape among the simulations largely because it would be interesting to elicit preference responses for a condition that is readily conceptualised by the public but has rarely existed in the front country of the upper Skeena valley. In effect, I found it difficult to find an area of the upper Skeena Valley that has not been disturbed at some point in the lengthy history of human habitation in the region. Whether for traditional subsistence uses, settlement, agriculture or timber management, purposeful modification of the region's forests has been a persistent facet of the Northwest's ecological history, such that a natural forest condition is almost an anomaly. Moreover, a natural condition reflects human purpose as well. In effect, it embodies the deliberate modern human intention to refrain from using landscape resources, or use them as little as possible, and allow natural processes to dictate the course of landscape succession and patterns. Given the inescapable human relationship with and dependence on the landscape (i.e. materially, symbolically and spiritually), a natural landscape condition is in many respects a product of human culture and (biocentric) values. How the research participants evaluate a condition that is inconsistent with the region's ecological history as well as with the reality of human existence and material dependence on landscape resources is, in many ways, an interesting question to pursue. It is conceivable, given a continuing economic decline in the region's timber industry, a growing preservationist ethos, and an 131 expanded ecotourism industry, that a natural condition may come to represent the upper Skeena's future. Simulating a natural condition is achieved only in part by 'erasing' evidence of past timber harvesting activity (i.e. recent cutblocks, and cutblocks that are in the process of greening-up). It is also important to consider to what extent has Gitksan berry management, particularly if it has been practiced over centuries, left visible patterns on the region's landscape ecology. Given the lengthy history of aboriginal burning in the Northwest, I sought the guidance of a local landscape ecologist (Dr. Phil Burton) to develop a better understanding of what visible effects, if any, Gitksan berry management has left on the contemporary landscape. Dr. Burton indicated that the Gitksan have left a lasting imprint on the landscape as evidenced by the profusion of deciduous forests (i.e. birch, aspen and cottonwood stands), both on the Skeena and Bulkley Valley bottoms, as well as on the mid slopes of several mountains near major Gitksan settlements (e.g. Luu Skayansit, as in Figure 3.6). According to the biogeoclimatic mapping for this part of the Northwest, the forest throughout the lowland and mid-slope portion of the landscape should be a cover of Interior Cedar-Hemlock and Mountain Hemlock, respectively (Haeussler, et al., 1995). The fact that these species do not predominate is a result of generations of aboriginal burning 3 , from which has accumulated a thick horizon of organic soils that are difficult for most xeric and mesic loving coniferous species to colonise (Haeussler, et al., 1995; LePage , 1995). According to Dr. Burton, coniferous 1 0 According to Dr. Burton, large stand replacing fires are of such rare occurrence in the Northwest (i.e. every 200-300 years, see also LePage, 1995), that the deciduous cover in the Skeena-Bulkley valley lowlands, as well as at Luu Skayansit, cannot be accounted for by wildfires. Further, fire scar sampling within these deciduous stands demonstrates that the fire return interval in these areas has been every 5 to 10 years, or consistent with the Gitksan fire rotation interval (Trussler, 2002). 132 species will eventually succeed and come to dominate these portions of the landscape, but this process may take another century or more. Thus, a pristine condition where natural processes control visible landscape patterns should reflect the biogeoclimatic classifications for the area and represent a cedar-hemlock and mountain hemlock dominated landscape. The natural simulations for all three viewpoints assume that the Upper Skeena forests have been left untouched for over a century and have thereby been left subject to natural processes of succession (Figure 4.10). Viewpoint 1 is where the effects of aboriginal burning are most pronounced. In the natural simulation for this viewpoint, the deciduous forest cover along the Skeena-Bulkley Valley lowlands has been replaced, based on biogeoclimatic mapping, with an Interior Cedar-Hemlock forest. Similarly, along the mid-slope portion of Viewpoint 2, the light green deciduous forest cover within Luu Skayansit has been replaced by a mature stand of Mountain Hemlock. Finally, evidence of past timber harvesting along the face of Blunt Mountain at Viewpoint 3 has been removed, and the deciduous stands along two terraces at the base of the mountain are covered with a mature stand of Interior C e d a r and Hemlock. 133 Figure 4.10: Landscape Simulations, Natural Condition Viewpoint 1 134 Condition 4 - Multi-Use Forest Management Th e last of the four landscape conditions reflects an alternative vision for the future of the upper Skeena's forests. It is based on the assumption that the population and economy of the Skeena-Bulkley valley will grow over the next century. In fact, there is some evidence that within the next ten years, communities such as Smithers (-30 minutes away from the Hazeltons) will receive a large influx of new residents as new mining and tourism opportunities open the region to an expanding population seeking employment and a "bucolic" rural lifestyle (Chipeniuk, 2004). It is conceivable that the timber economy in adjacent areas such as Kispiox and the Hazeltons may become reinvigorated as new construction augments the demand for local forest products. However, it could also be assumed that, under such a condition, local residents may not settle for the status quo approach to timber management, and may even react strongly to any form of timber harvesting that reflects the management style of the 1970s and '80s. Moreover, as discussed in the concluding section of Chapter 3, the future of forest management in the Northwest may have to contend with the simultaneous demand for timber products, aesthetic interests, and forest uses that are increasingly varied (i.e. subsistence users, timber management, agriculture, recreationists and tourists), and reflect the trend towards increased First Nations control of the landscape with an emphasis on traditional land-uses. In effect, a simulated multiple-use condition was both a plausible and potentially interesting future landscape condition to present to the research participants. 135 T he multi-use condition consists of three modified landscape elements (Figure 4.11): • A visually sensitive approach to timber harvesting, • The reintroduction of aboriginal fire management for berries, • Range management on the Skeena-Bulkley Valley lowlands. The timber management component of the multi-use condition was first approached by analysing the GIS dataset to identify timber stands that are operable as well as merchantable, based on the Ministry of Forests' definition of these two terms ( B C M o F , 2002). The objective was to identify potential areas for siting cutblocks. Subsequent to this analysis for the three viewpoint locations, cutblocks were designed based on the following standards: • Block openings are smaller than the 60-hectare clearcuts of the 1970s and 1980s. The simulations for Viewpoints 1 through 3 depict a block size that varies between 35 and 40 hectares, reflecting the recommended cutblock size for coastal forests within the former B C Forest Practices C o d e ( B C M o F , 1995a, b). • Cutblock edges were designed with an irregular or sinuous pattern, reflecting recent attempts by local operators to emulate natural disturbance patterns (see Figure 3.4 and Figure 4.11). • The cutblock edges have also been feathered to the extent that a thinned margin of timber extends 25 metres into the block, further lending a rough or more natural appearance to the openings, in recognition of the visual sensitivity of the sites to public viewpoints • S e e d tree retention patches and riparian leave strips (25 metre buffers) have been left within the blocks largely to enhance the natural appearance of the blocks, as well as to suggest a deliberate effort to mitigate ecological impacts (consistent with current moves toward variable retention harvesting in the industry). 136 Figure 4.11: Multi-Use Condition 137 In addition to timber management, the simulated conditions depict aboriginal berry cultivation and range management. In the former case, the assumptions regarding patch size and location, colouration and rotation presented for Condition 2 have been repeated for Condition 4. The only discrepancy between the two conditions is the relative location of patches that have recently been burnt and those that are in a mature or productive state. In other words, to avoid participant confusion, as well as to suggest that Conditions 2 and 4 reflect different moments in time, the patches that are depicted as recently burnt in Condition 2 are shown in a state of relative maturity in Condition 4, and vice versa. Range management has historically occurred in the lowland portions of the Skeena and Bulkley Valleys, and is represented by Condition 1 (i.e. the Pre-lndustrial Condition) at Viewpoint 1 as an open expanse of grassland punctuated by a few remnant bosques and exposed riparian edges. For Condition 4, it has been assumed that future agricultural uses in the lowlands may not be as extensive as they were historically, and the aesthetic and environmental sensibilities of the contemporary public (both First Nations and Euro-Canadian) may not permit the kind of poor riparian management that was prevalent a century ago. As such, substantial riparian corridors (25-50 metres in width) are simulated along all river and stream channels on the valley bottom, and the entire length of the Skeena River has been enclosed with a fifty metre buffer of mixed deciduous and coniferous vegetation. In addition, there are more bosques or patches of remnant forest scattered throughout the valley bottom in Condition 4 than there are in Condition 1. Overall, the timber, agricultural and berry management modifications in Condition 4 occur simultaneously at each viewpoint, reflecting a 138 multi-use approach to land-use, as well as provide for smaller openings, more irregular and naturally formed edges, and remnant vegetation within disturbed areas. If the assessment presented in Chapter 3 of the future of landscape management in the upper Skeena region is plausible, then Condition 4 may accurately reflect a future where the competing demands of resource development, traditional use, aesthetics and recreational use necessitate a lighter, more natural appearing (i.e. designed) imprint on the land. In general, I believe that the time spent studying and photographing the Gitksan landscape, interviewing individuals who are familiar with the forests' changing patterns, careful review of documentary materials that provide a detailed visual and written account of historical forest conditions, and analysis of the GIS forest cover dataset has resulted in a set of landscape conditions and simulations that are credible as well as defensible (Sheppard, 1989, 2001a). 4.5 Data Collection Procedures Like most qualitative research designs, this dissertation has involved multiple data collection methods. Over the course of my contact with the G T O , I have been able to: • Review documentary materials, both internal (e.g. newsletters, memoranda, court transcripts and archival materials) and external (published articles, consultants' and government reports). • Observe community meetings and ceremonial gatherings. • Conduct open-ended (June-July, 2003) and semi-structured (October, 2004) interviews with community members. The use of multiple data collection methods in the same study is referred to as triangulation. According to Denzin (1970: 308), "The rationale for this strategy is 139 that the flaws of one method are often the strengths of another, and by combining methods, observers can achieve the best of each while overcoming their unique deficiencies." This helps to ensure the validity of the data by checking the information found in documentary materials against data that emerges from the interviews. T he use of documentary materials provided a means of obtaining background information on the Gitksan's history and interaction with the forest management sector. Combined with the open-ended interviews, the information obtained from the document review helped focus and ask relevant questions about the information provided by the participants in the semi-structured interviews, conducted the following autumn. These data were gathered over the course of several trips to the G T O library in which I read all of the information that is pertinent to this dissertation available in their files. My observations of community meetings (e.g. conferences) and ceremonial functions (e.g. traditional feasts) were relatively informal and impressionistic as less attention was given to taking extensive field notes than to establishing relationships with the people with whom I was working. The bulk of my data comes from the two sets of interviews, the most important of which being the semi-structured series held in the autumn of 2004. I decided that the participants' experience and perceptions should take precedence in the study as the latter series of interviews was structured specifically to address the dissertation questions. The following discussion is, therefore, a detailed review of the semi-structured interview process. 140 4.5.1 The Semi-Structured Interviews T he interviews can best be described as guided conversations that took an hour and a half to two hours to complete. I favoured the semi-structured format because, while the interviews were guided by a formal interview schedule, neither the exact wording nor, for the most part, the order in which the interview questions were asked needed to be rigidly constrained by the schedule (Merriam, 1988). In this way, I was permitted to respond to unanticipated comments and insights provided by the participants and to allow the information to emerge largely independent of any prior assumptions or theoretical frameworks that I may have brought to the interviews (McCracken, 1988). The interview phase of the data collection consisted of a three-part process. The first stage was to prepare and submit a letter of introduction prior to the interviews to all potential participants (Appendix A1). The letter describes briefly the objectives of the research, the individuals involved in conducting and supervising the work, an assurance of confidentiality, and a statement expressing my wish to contact that individual to establish an interview date and time. T he second part of the interview process was to review a statement of consent with the participants prior to commencing the interviews (Appendix A2). This statement essentially fulfilled two purposes, both dealing with the ethical considerations of the research. First, it obtains the participant's consent for the researcher to tape record and transcribe their statements for quotation in the dissertation. Second, the statement sets out explicitly what the participants can expect from the researcher in terms of my obligations to them. In brief, the 141 statement assures the participants that I will not disclose their identities through the transcribed materials or the dissertation text, that they may review copies of their respective transcripts to verify or amend statements, that they will receive a copy of the research report (i.e. the analysis and conclusion chapters), and that they may discontinue their participation in the research at any time. The final part consists of the interview questions themselves. The questions were compiled and refined from the preceding series of open-ended interviews, from theoretical insights gleaned from a review of the landscape preference and ethnographic literature, and from informal discussions with key informants and anthropologists experienced in working with the Gitksan or their Tsimshian neighbours. A draft interview procedure and set of interview questions had been prepared prior to embarking on the final phase of interviews, which were pre-tested with 6 individuals (3 First Nation and 3 Euro-Canadian). The pre-test interviews were instrumental in two respects. First, they helped assess which questions from the original draft list would be understood and most effective in eliciting environmental preference information from the interview participants. Second, the dry-run interviews also served to gauge how many questions could be asked and tolerated by the participants within the maximum 2 hours allotted for each interview. The final list of questions derived from the pre-tests is presented in Appendix A 3 and summarised next. In Chapters 1 and 2 , it was hypothesized that landscape preferences are rooted in the purpose-based, intersubjective and embodied experiences that an individual accumulates from interactions with other members of their community or 142 cultural group and with their environment. Thus, the factors that are important to an individual in their evaluation of a landscape condition (Research Question 1) and to different individuals with unique ethnic identities and, perhaps, ways-of-life (Research Question 2), require a better understanding of the participant and his/her experience with the environment - i.e. beyond the information provided by the key informants and determined by the participant selection criteria. Thus, Part A of the interview schedule began with a set of biographical questions, designed primarily to explore the participants' personal background - their employment history, history of residence in the area, their level of interest in and understanding of forest management issues, etc. More generally, these questions are designed to fulfill two purposes. First, they are intended to put the participants at ease by giving them a chance to discuss their experiences and opinions in a relatively loose and informal manner. Second, they provided a body of data that described in greater depth the participant's forest-related knowledge and experience that, ultimately, will be used to explain or account for the preference information provided in the subsequent interview phases. The second part of the interview schedule was intended to assess the participants' preferences for the modified forest conditions using both qualitative and quantitative methods, respectively. For Part B, the participants were presented with the simulated conditions corresponding with Viewpoint 1. The simulated conditions were printed separately on four 60-inch wide sheets.4 With some minor variation, the questioning during the interviews generally followed a consistent sequence: 1 1 It was assumed that the participants would be directly handling the images and examining them roughly at an arm's length distance (i.e. 36"). Sheppard (1989: 185) published a formula to determine 143 • The participants were first asked to carefully examine each of the images, and describe their initial impressions of the landscape conditions that they could see in the simulations. • The simulated conditions were explained in detail to the participants, and they were asked to indicate which of the conditions they preferred, which they preferred the least and, if applicable, which conditions they 'moderately' preferred (i.e. the simulations were close to preferable but the participant still felt some discomfort with the conditions). A s the participants stated their preferences, they were asked to sort the images into separate piles by preference. • Additional 'probing' questions were asked to elicit more detail on the participants' preference judgements - e.g. 'why do you prefer this condition?,' 'why do you prefer this condition the least?,' 'how does this condition make you feel?,' 'what significance/importance does this area have for you?', etc. The images in each pile were recorded and discussed and, when it appeared that the participants could provide no additional information in relation to the conditions for Viewpoint 1, these were substituted for simulations depicting the same set of modified landscape conditions at Viewpoint 2. The line of questioning and procedure was repeated for the second set of images and the discussion progressed until the participants had exhausted their perspectives and comments. Finally, the participants were presented with the Viewpoint 3 images and questioned in relation to the simulated conditions. For Viewpoints 2 and 3, it is acknowledged that this adaptation of the pile sort method would make it possible for participants to transfer their knowledge of the four conditions from the review of Viewpoint 1 to their evaluations of the conditions at Viewpoints 2 and 3 . 5 the correct image size for matching real-world image size, based on assumptions of a fixed viewing distance and field of view - Correct Image Size = 2*Viewing Distance*tan of 0.5 Viewing Angle. The simulated images use a viewing angle of 80°. Using the above formula, the values can be substituted into the equation to derive a printed image size of approximately 60" (y=2*36*tan 40°=60.42"). 12 For this reason, the independent effect of additional information on preference has not been analysed. 144 The interviews were conducted at the GTO offices, in local restaurants and cafes, or in the participants' homes for their convenience. Generally, I had developed some familiarity with the participants through the time I had spent in the upper Skeena during the preceding summer. However, for those participants that I had not previously met, I typically began the interviews with a lengthy personal introduction and review of the research objectives. During all of the interviews, I tried to let the participants determine the direction of the conversation, only stepping in with prompts or to redirect if necessary. Throughout this process it was quite challenging to find a balance between allowing the participant to speak freely and obtaining the particular information sought for the research. At the close of each interview, all participants were thanked verbally and tangibly with a gift of tea and coffee. For the Gitksan participants, this gesture was not only considered an expression of good etiquette, but a necessary part of the social transaction involved in the participants' expenditure of time, effort and knowledge through the interviews. Subsequent to the interviews, the tapes were reviewed, transcribed verbatim and sent back to the participants for review. Through a cover letter, I asked the participants to clarify, change or omit any information from the interviews if they deemed it necessary to do so. This step served as a second, self-administered interview with the same questions, only this time the participants could take time to mull over their answers. I included this step for two reasons. First, it was to ensure that the data were as accurate and as representative of the participants' views as possible by having them check the transcripts for errors and clarify points. Commonly known as "member checks," this method is one of the key means of 145 ensuring the validity of the findings (Lincoln and Guba, 1985: 236). Second, I hoped that it would maximize participant satisfaction with the finished product by protecting them from misrepresentation. I also felt that it provided the participants with a degree of control over the data and thus gave them an active role in the research. 4.6 Data Analysis Glaser's approach to grounded theory was used in the analysis of transcribed data (Glaser, 1992, 1994), whereby the constant comparison technique is utilised as the basis for qualitative theory building and, wherever possible, the data is allowed to tell its own story. Within grounded theory, data collection and analysis take place concurrently, rather than consecutively by collecting the data and then analyzing it. As with most forms of naturalistic inquiry, the starting point is a general perspective or set of questions; however, some effort is made to avoid biasing initial interpretations and decisions regarding the data by predetermined theories. As the data is collected, concepts are generated and themes emerge, and working notes or 'memos' are maintained to keep track of the assumptions used to arrange and compare constructs. Working on concepts that emerge from the data means that developing theories are firmly rooted in the participants' statements rather than being influenced in an a priori manner by existing theories or my own preconceptions. It also generates theories that are representative and relevant to the participants involved in the research, which is ultimately crucial to establishing the validity of any interpretations gleaned from the data. 1 4 6 4.6.1 Research Question 1 The data analysis strategy used for Question 1 was based on two considerations. First, due to the exploratory nature of the question, it was felt that an inductive process of analysis that allows for themes and patterns to emerge from the data, and thereby permit theory building, would be the most appropriate strategy for this question. Second, consistent with the particular objectives of question one, the images were analysed in terms of the dimensions or consequences that the participants considered salient in their evaluations of the modified landscape conditions. Given the largely qualitative nature of Question 1, a content analysis of the interview transcripts was performed. Consistent with the broader goals of grounded theory (Glaser, 1992; Strauss and Corbin, 1990), the data was approached with few presumptions about what themes would emerge other than the expectation, developed in part from pre-test interviews and prior research with First Nations communities, that repeated references to particular consequences or impacts (e.g. cultural, ecological, economic) may reflect key attributes that the participants use in their preference evaluations of landscape conditions. The review and thematic sorting of data began early in the research process as interviews were conducted and transcribed. Individual pieces of data or blocks of text were extracted from each interview and assessed in terms of similarities, differences, consistency and inconsistency within and across participants. As themes began to "emerge from the data" (Strauss and Corbin, 1990: 73), they were identified by codes and used to guide subsequent analysis. Miles and Huberman (1994) distinguish between first and second level codes. The first level codes 147 applied to the data were descriptive as labels are attached to groups of words or blocks of text that have been culled from the transcripts. The second level coding is referred to as pattern coding and involves the grouping of initial descriptive codes together as common themes in the data are identified. It was particularly important at this juncture to keep a record of my thoughts and interpretation of the codes as the analysis proceeded. According to Lincoln and Guba (1985), these second level codes or "units of information" should be: Heuristic: in that the codes should be relevant to the study and encourage the reader to think beyond that individual piece of information, and Choate: interpretable without the addition of other pieces of information other than a general understanding of the context in which the inquiry is carried out. Thus, through a process of iterative review and comparison, each second level code becomes a conceptual unit that describes the particular dimensions used by the participants to evaluate the forest conditions depicted in the simulations. Although it may seem contrary to the aims of qualitative research, some effort has been made to identify what is inconsistent or missing from the participants' comments. In this respect, I have attempted to look for units of information that are anticipated from theoretical insights derived from the literature review, but which are either contrary to that body of theory or surprisingly missing from the interview data. This approach to content analysis may serve as a guide to question existing hypotheses and theories, as well as aid in the exploration of alternative theoretical perspectives. 148 4.6.2 Research Question 2 The primary objective of Question 2 is to uncover the similarities and/or differences between individual participants as reflected in their appraisals of the landscape simulations. Two widely used and complementary analytical techniques to identify similarities and differences between participants are multidimensional scaling (MDS) and qualitative content analysis (Canter, 1997; Weller and Romney, 1988). In order to explain the way that M D S works and how it is used in qualitative research, the following overview is presented of how the M D S dataset was derived for this research. 6 In order to illustrate the use of M D S , this example will consider only six participants from the research sample, which form the 'items' for this analysis. T h e first key piece of M D S input are the 'variables,' in this case the participants' sorting of the four landscape conditions for the three viewpoints on a three point ordinal scale - preferred=2, moderately/somewhat preferred=1, not/least preferred=0. At this stage the participants and their ordinal ratings are entered into a spreadsheet using commercially available statistical analysis software, in this case S P S S ® , to produce a 'data matrix' (Figure 4.12). T h e matrix lists each participant (i.e. item) as a row of data, and each numeric rating of a landscape condition (i.e. 'variable') as a column. 13 References used to provide additional technical background for this procedure included, Cox and Cox (1994), Kruskal (1964), Lingoes (1968), Togo, etal. (2000). Canter, etal. (1985) and Weller and Romney (1988) provide an overview of the use and interpretation of MDS data in conjunction with pile sort procedures. 149 Figure 4.12: MDS Data Matrix Vpt l_Pre Ind Vpt l_Ind Vpt 1_MU Vpt l_Nat Vpt2_ EC 13 0 EC 14 2 EC 15 2 FNOl 0 FN02 0 FN03 0 0 0 0 0 0 0 1 1 1 0 0 1 2 2 2 1 1 1 Thus, reading down the column corresponding with the pre-industrial condition at Viewpoint 1, it is possible to see that two Euro-Canadian participants have classified this as their most preferred condition, while three First Nations and one Euro-Canadian participant has classified it as their least preferred condition. However, we are looking for similarities and dissimilarities within the participant sample, so reading along a row of data it is possible to see a profile of ratings that summarises the preferences that each participant expressed. For instance, a visual comparison of participants E C 1 4 (2012) and E C 1 5 (2012) shows that they have rated the simulated conditions exactly the same for Viewpoint 1 and are therefore identical in their preference profiles. Participants E C 1 3 (0012) and E C 1 4 (2012), however, provided identical ratings for three of the four conditions and, although they are not identical in their preference profiles, they are quite similar in the ratings that they used. Now compare the ratings made by E C 1 4 (2012) and FN01 (0001). The two profiles show that the participants have different ratings on three of the four variables/conditions. Using this kind of 'eyeball' analysis of the data matrix, it is possible to identify relatively similar and dissimilar preference profiles. However, when this task is complicated by large numbers of participants, the data matrix can be represented 150 graphically where the participants are plotted as points in two-dimensional geometric space. Generally, programs such as S P S S use Euclidian distance equations (e.g. the distance dy points i and j is defined as df = [I(x / a - xja)2]V2) to model similarity/dissimilarity for ordinal, interval and ratio level data (see Togo, era/. (2000) for an overview of the algorithms used to calculate Euclidian distance for ordinal Figure 4.13: Multidimensional Scalogram (MDS) Output Object Points - Common Space V 1-The number adjacent to each data point indicates the number of participants represented by each data point. data). S P S S plots each item (in this case a participant) as a point in geometric space (see Figure 4.13)7 1 4 The MDS plot is not read in the same way that a graph would be interpreted in terms of having an x and y axis with a graduated scale of numeric values along each axis. The MDS plot confines the geometric space within a 'box,' which is merely a graphic device for defining the extent of the geometric space. Quantitative 'values' within the MDS plot are defined in terms of the relative proximity or distance between any two points within the space. This does not mean, however, that MDS results cannot be interpreted or conceptualised in terms of a matrix, which is simply a set of quantities arranged and displayed in a rectangular array. Other rotations of the data and ways of segmenting the clusters are possible, as long as the key consideration is distance from nearest neighbours and the qualitative interpretations of the clusters. 151 It attempts to find a configuration of points so that the plot can be divided into clear regions, which distinguish the items on the basis of patterns of response to the variables. In effect, this means that the more qualities two items (i.e. participants) have in common, the closer together they are in the plot. Conversely, points that are distant from one another indicate the simulations have been given different ratings by the participants and are viewed as conceptually dissimilar in some way (a detailed discussion of the algebraic calculations for M D S output is provided in Appendix C1). T o use the example provided in the preceding paragraph, participants E C 1 4 (2012) and E C 1 5 (2012) have identical preference profiles for Viewpoint 1 and will therefore be represented in the M D S plot by a single data point. A s participants become increasingly dissimilar in their preference ratings geometric distance between the participants will increase the M D S plot. Thus , participants E C 1 3 (0012) and E C 1 4 (2012) will be displayed in closer proximity than participants E C 1 4 (2012) and FN01 (0001). T he first task in M D S analysis is to examine the output visually for clusters or groupings of participants. However, for the groupings to be conceptually meaningful, and to identify inter-ethnic similarities and/or differences in landscape evaluation, the participant groupings were examined in terms of three factors: • Their relative proximity to other clusters within the M D S plot, • The qualitative explanations given by participants within the cluster for their ratings, and • Socio-demographic similarities and/or differences among the cluster participants that may shed light on their preference evaluations. T o do this, the interview transcripts pertaining to the simulations were combed for individual pieces of data or 'bibbits' (Kirby and McKenna , 1989: 135) that describe 152 the common aspects described by each participant as explanations for their simulation ratings. A s a form of content analysis, these bits of information can be extracted from each interview and analyzed in terms of similarities and differences, consistencies and inconsistencies both within and across participant clusters (Goetz and LeCompte , 1984; Glaser and Strauss, 1967). In addition to probing deeper into the schemata that the participants use to evaluate the simulations, and the possible ethnic/cultural basis for their evaluations, the outcome of this analysis provides a key means of validating the M D S results through data triangulation (LeCompte and Schensul, 1999). 4.7 Criteria for Evaluating Qualitative Research All research is concerned with producing valid and reliable knowledge, and qualitative research is no exception. In fact, owing to the largely qualitative nature of this dissertation, such concerns may be more significant than in experimental designs where reliability and validity are accounted for from the start. This discussion addresses issues of research quality as they apply to the presentation and interpretation of research data. Three common evaluation criteria were used, paying particular attention to how these tests are applied throughout the subsequent analysis of the research data and their synthesis into findings. Respectively, these tests include (Merriam, 1988: 163-77; Yin, 1984: 40-1): Internal Validity: Ensuring that the study's evidence and the interpretations made of that evidence reflects 'reality' as the study's participants perceive it. External Validity: Also referred to as Generalisability, it refers to the degree to which a study's findings can be extended to or compared across groups. 153 Reliability: Demonstrating that the operations of a study - such as the data collection procedures - can be repeated with the same results. 4.7.1 Internal Validity Internal validity deals with the question of how faithfully research findings match with the participants' conceptions of reality. Ultimately, the question which the researcher needs to address is - what is being observed in the research, and how does one assess the validity of those observations (Merriam, 1988)? What is being observed are peoples' perceptions of reality, how they understand the world: T h e [researcher] constantly attempts to capture and portray the world as it appears to the people in it. In a sense, for the [researcher], what seems to be true is more important than what is true. For the [researcher] . . . the internal judgements made by those he [sic] studies, or those who are close to the situation, are often more significant than the judgements of outsiders (Walker, 1981: 45). Thus , in qualitative research, internal validity involves understanding the perspectives of those involved in the phenomenon of interest, to uncover the complexity of their conceptions of the world and portray, as accurately as possible, an interpretation of the participants' perspectives. A s with many social science methods, a research design that is largely reliant on interview data is based upon participants being willing and able to give accurate and complete answers to the questions that are posed. Yet, for various reasons, participants may be motivated to distort their answers. They may distrust the researcher and, even if they wish to cooperate, they may be unable to answer accurately because they cannot remember the details requested or because they cannot understand the questions. S o m e attempt has been made to overcome these potential difficulties by constructing the interview schedule in order to provide 154 evidence of consistency across responses (see Appendix A3 , Schensul , et al., 1999). Having a pattern of questions that probes for the same information, but is expressed or worded differently, offers one way of assessing the consistency of participant responses and, in turn, the validity of the data. Thus , by restating questions in different segments of the interview, I have been able to determine if the participants were being inconsistent with their answers. From this determination, the questioning can be extended using additional probing questions to clarify the source of the inconsistency - confusion, poor comprehension, a deliberate attempt to mislead, etc. - or exclude that part of the participant's data from the analysis. O n c e interviews are completed, and the data has been transcribed and analysed, a researcher will invariably encounter a critic who asks - "How can I trust that what you as the researcher see in the interviews are the same things that your interview subjects and other people would see?" It generally helps to discuss at length in the presentation of the data (Chapter 5) explicit interpretations of key interview statements. Wherever possible, the objective is to put the reader in a position to question or re-examine interpretations made of the data. Thus , throughout Chapter 5, interview data has been presented using the following model: 'Here is an example of my qualitative data, here is how I interpret it, here are possible rival or competing interpretations, and this is what I conclude.' A n additional key strategy that has been used to ensure the validity of the findings is data triangulation. In effect, greater confidence in the researcher's understanding and interpretation of an event can be established if the study demonstrates that information from multiple sources coincide. Thus , in the present 155 study, participant interviews, documentary and archival sources and field notes have been employed to strengthen assertions that the data presented and interpretations made are robust and valid. The final strategy has involved the use of member checks, or taking evidence and interpretations back to the people from whom they were derived and asking them if the interpretations are plausible and realistic. Lincoln and G u b a (1981) suggest doing this throughout the study. For this dissertation, interview participants have been asked to review their own interview transcripts to ensure that their testimonies were recorded accurately and to allow them to double-check their claims and assertions. In addition, to prevent misinterpretations of evidence from entering the dissertation, I made a formal request of each participant to review draft copies of the chapters. 4.7.2 External Validity or Generalisability Questions of external validity in qualitative research commonly address two issues. First, generalization can be expressed in terms of statistical generalization, or the extent to which the results obtained in one study also hold true for other populations or, in other words, will related or similar populations in other regions of the province or country display patterns of landscape preference that are similar to the population being investigated. Second, a conceptual approach to generalization in qualitative studies posits that analytic generalization may be established where the theoretical frames, definitions and concepts used in one study can be applied by other researchers to comparable projects (Miles and Huberman, 1994: 27). 156 In terms of any attempt to establish the statistical generalisability of this research, a generalized application of the research findings to other First Nation or non-aboriginal communities across the province would be difficult to substantiate because of the unique historical, political, and economic circumstances experienced by different communities. Political, economic and social encounters are important factors to consider but more precisely aboriginal as well as non-aboriginal communities across British Columbia have experienced varying degrees of urbanisation, resource exploitation, acculturation and other historical factors that make it difficult to establish the relevance of any particular set of findings established among the communities of the Skeena valley to the communities of the Fraser Valley, the Kootenays or the Peace River District, for instance. Thus , to avoid making incorrect or inappropriate generalisations from this dissertation, I have determined that assessments of generalisability should be left up to the reader or, more appropriately, to communities who may wish to apply the findings to their own situations (Merriam, 1988: 173). They may do so with a full understanding of how the results were obtained and the conceptual assumptions that informed the research. Analytic generalization is a very different matter. Qualitative research in general is used to put flesh on the bones of theoretical constructs and their relationships, with the ultimate objective of understanding generic processes. Given that the main questions of this dissertation are driven by an interest in how landscape preference evaluations work within and across cultures, the focus of generalization in this context is not to all First Nations, or even to all members of the 157 Gitksan community broadly, but to existing or new theories of how environmental preference works. This of course relates back to the sampling decisions that have been made which, in qualitative research designs, are theoretically driven (Glaser and Strauss, 1967). In effect, choices of participants are driven by a conceptual question and with a concern for maximizing opportunities to uncover the phenomenon of interest, rather than by an interest in demographic representativeness. T o get to the construct, the objective in this research is to see an instance of it working with a particular set of people at a particular moment in time. Thus, the main concern is with the conditions under which the theory or construct operates, not with the generalization of the findings to other settings. More will be said about the nature and theoretical relevance - i.e. analytic generalisability -of the findings made in this research to 'established' theories of landscape preference in the concluding chapter. 4.7.3 Reliability The touchstone of scientific research is the replicability of the investigation. However, when the focus of the research is either social values or perceptions, it cannot be guaranteed that the research design and its results can be repeated (Lincoln and G u b a , 1985). Social reality is not a stable, fixed phenomenon. Since there are so many rival interpretations of what is happening in the world around us, there is no single benchmark from which a set of repeated measures can be taken to establish reliability in the traditional scientific sense. However, qualitative analyses ought to be open to verification. A clear and detailed description of how the data were obtained should be provided which is clear enough for someone else to repeat 158 what has been done and to verify or, at the very least, support the conclusions reached by the original researcher. Since the conventional concept of reliability is problematic when applied to social science research, Lincoln and G u b a (1985) suggest thinking about the dependability or consistency of the results obtained from the data. In essence, rather than demanding that other researchers derive the same results from the data, one would wish to concur that, given the data collected, the results make sense -i.e. they are consistent and dependable. O n c e again, the primary technique the researcher can use to ensure that the results are dependable is data triangulation (Yin, 1984; Merriam, 1988). 4.8 Chapter Summary The purpose of this chapter has been to provide the reader with sufficient information to interpret the research, as well as to demonstrate the rigour with which the research was conducted. Care has been taken to follow a research design appropriate to a study of this nature. My own philosophical approach to research has been made explicit, and each step has been accounted for and justified by supporting methodological literature. A s with all research, however, there are some limitations to be noted. These limitations as well as opportunities for future research are discussed in the conclusion of Chapter 6. 159 5 Interpretation & Analysis 5.1 Introduction This chapter is organised primarily according to the research questions that guided this study and, secondarily, according to the salient themes within the narratives provided by the upper Skeena participants. The first section (5.2: The Perceived Dimensions of Landscape Change) describes the factors that the participants considered in their evaluations of the landscape conditions. The essential argument made from the data that runs throughout this section is that, while landscape evaluations may be inherently 'wicked' exercises, the discovery of preferred or socially acceptable landscape management need not be complicated by the daunting array of considerations that stakeholders bring to their evaluations. Rather, as discussed in sections 5.2.2 and 5.2.3, judgments of acceptability may be associated with the degree to which management demonstrates care or respect for the landscape or cultural uses of the land. The final section (5.3: Inter and Intra Cultural Landscape Preferences) addresses the second research question and provides evidence, which supports in a limited sense the ethnic basis of interindividual differences in landscape preference, while also noting key intra-cultural differences. Beyond this, the second half of the chapter presents results that behavioural or motivational factors (i.e. purpose-based uses of the forest landscape) also appear to influence judgements of landscape acceptability. Moreover, these behavioural influences transcend distinctions of ethnicity, age and gender and may account for the significant degree of inter-ethnic agreement and intra-cultural 160 difference in landscape preferences described in the environmental perception and preference literature. 5.2 Question 1: The Dimensions of Landscape Evaluation The participants that examined the twelve images provided lengthy commentaries on the considerations affecting their judgements of the four landscape conditions. These commentaries have been organised around three overarching categories and dimensions identified by the participants as affecting their judgements of acceptability: 1) Perceived Balance of Impacts (see Table 5.1, Section 5.2.1), 2) Perceptions of Care (see Table 5.2, Section 5.2.2), and 3) Site Characteristics (see Table 5.3, Section 5.2.3). 5.2.1 The Perceived Balance of Impacts The first dimension is the concept of balance, or the tradeoffs that individuals contemplate when making a judgment on a particular landscape condition. In effect, the participants would concurrently weigh different possible or perceived consequences of the landscape conditions, demonstrating that people make complex assessments when evaluating landscape change. A s the comments summarised below in Table 5.1, and discussed at length in the following sections demonstrate, the participants generally considered how a change appears from a scenic quality standpoint, but they are also fundamentally concerned about what is right for natural (e.g. erosion potential, habitat protection) and human communities (e.g. material livelihood, resource waste, public relations). 161 a) Clearcut Patterns: Industrial and Multi-Use Conditions^5 Table 5 . 1 : Balance of Impacts Issues Dimensions Balance of Impacts Material Livelihood Clearcut Patterns Scenic Quality Habitat Erosion Resource Waste Scale Public Relations Aboriginal Burns Scale Cultural Identity Multiple Uses Competence Material Need Resource Waste Care Natural Condition Scenic Quality Habitat Ecological Restoration Subsistence Uses Spiritual Restoration Resource Waste Illustrative Comments If that's a trapline, and if it's your trapline, you'd be pretty well heartbroken.'ve got a larger visual impact because of the clearcuts. doesn't provide the wildlife with the shelter they require. ...when it rains a lot of it slides, and a lot of the sand and mud goes into the rivers. see nice stuff just chucked away because they can't be bothered ... to clean up. These I didn't like because the clearcuts were too big. There's less visual impact, it's better public relations. This burn is not too bad, but this is huge's cultural in that the Gitksan ... have through the generations made use of non-timber resources. least there's another forest use. It's not just the big companies who are benefiting from it. ...if it was managed and not just some Joe Blow who decides that he's going to go out and burn... I know they used to do it for berries in the past, but do they really need to do it today? ...there's enough value in that timber that it would be stupid to burn it... would be nice to see that these areas are being cared for. ...I think it's just much nicer to look at. This is what I like. It's better for wildlife... This is the way it should be done, left to regrow like that. That's what we look for when we go picking our Indian medicine. ...this is the kind of place where I'd like to go when something's troubling me, to find healing.'s a waste because a lot of our timber here's over mature. 15 It was common for the logging patterns in both the Industrial and Multi-use conditions to be evaluated simultaneously by the participants. As such, the harvest treatments for both conditions are presented together in this subsection. Similarly the participants' simultaneous consideration of the burn treatments in the Multi-Use and Pre-lndustrial simulations warrants their combined discussion in the following subsection. 162 T he complexity of arriving at a preferred strategy for landscape management is highlighted by the comments made generally by the upper Skeena participants. W h e n examining the four conditions, people weighed several factors simultaneously and would often reconsider their comments as they processed additional information provided by the interviewer. The participants were able to embrace cognitively the idea of tradeoffs and would frequently articulate multiple issues within the context of a single response: F N 14: That's what strikes me first about this picture. A n d it's not very cosmetic, but they've left the sides of the mountain alone. . . . There would be massive seepage into the creeks, and stuff like that. I've seen a lot of grazelands up the Kispiox Valley where there's absolutely no instream vegetation and it's really ugly. The cows go through there, screws up the substrate of the creeks. E C 11: Well to me, you've got a larger visual impact because of the clearcuts, and then for wildlife, if you have smaller clearcuts, it's better for wildlife. They usually only use the edge, and just for public relations too. Y o u know, they used to use 60-hectare clearcuts, and to me, you should keep them in the 20s. E C 8: it's all a matter of, if you're looking at it just like fibre, fibre, fibre, then that's no good. But I mean there's, what if you want to go shoot a moose, or what if you just want to walk through a nice forest? Yeah , there's other values there besides the fibre. For many participants, the industrial harvest patterns depicted by the industrial and multi-use simulations reflected the management style that predominated in the region during the last thirty years and, from their perspective, failed to balance important forest attributes. Although most participants acknowledged that the treatments represented an immediate economic benefit to the local communities, there were other negative characteristics that outweighed the positive economic consequences. In general, comments on the acceptability of the 163. industrial condition and, to some extent, the multi-use condition reflected a common tension between conflicting material uses of the forest - i.e. the forest as a source of commercial timber wealth, subsistence or non-timber products - and values which emphasised either the scenic quality of the landscape or its intrinsic importance for wildlife and aquatic habitat (i.e. biocentric values): FN 9: Wel number one, when it rains, a lot of it slides, and a lot of the sand and mud goes into the rivers. It takes longer to grow back, and then they wipe out everything. FN 1: Oh yeah, it's just terrible today. You start getting sick looking at it. We went to Babine yesterday, my son and I we got one moose, and where we used to walk around there was just nothing, it's just mud. FN 4: ..we'd only pick out what we were going to need. We don't just take it all. And that's thinking about tomorrow, you have to leave enough for somebody else. EC 11: These I didn't like because the clearcuts were too big. In my opinion they should have gone to smaler clearcuts. ... to me, you got larger visual impact because of the clearcuts, and then for wildlife, if you have smaler clearcuts it's better for wildlife ... and just for public relations too. EC 8: ...this is going to have all kinds of drainage issues. Like road building, you talk to the guys that are building roads the ones that go fal everything ahead, and you get a couple of days like this, they're down for a long time because they've got nothing sucking the water up. ... And it's the same thing here, you get this and where the hell's the water going to go? You're going to have all kinds of erosion and shit. EC 10: yeah, wel I mean what do the people of Gitsegukla think of that? I guess that's more to the point. ... I think what pisses people of  is when they see this but you see nice stuf  just chucked away because they can't be bothered going after that nice log and they can't be bothered to clean up ... I've got no problem with these things just because they're clearcuts, as long as you get the most out of it that you can. While reactions of this sort were anticipated for the industrial condition, participant evaluations of the multi-use harvest treatments were more surprising. 164 Despite the irregular cutblock configurations, several participants focused on the large openings and effectively dismissed the attempt to soften the visual imprint through patch retention, smaller openings (40 ha and less), as well as irregular and feathered edges arguing, in effect, that "a clearcut's a clearcut" (FN 7): F N 10: I'm a little more comfortable with this, but not much. These are still clearcuts. F N 12: There's not much difference, what is the difference between these two? . . . Well these two here (i.e. industrial and multi-use conditions), there's so much disturbance. F N 1: The young people say it's beautiful, but with me . . . you see something like this, well that breaks me up inside.... E C 1: Yeah , it looks like they've done an awful lot of cutting in this area. T o o big, I'd say. The openings are way too big. But yeah, I think it's better here with the cover. Could be more of it, but some is better than nothing. More substantively, a group of First Nations and Euro-Canadian participants who depend on the forest for income from trapping, argued that clearcutting in any form - i.e. using geometric or naturalistic patterns - represents a threat to their material livelihood. In effect, the industrial and multi-use treatments provide insufficient cover for small fur-bearing animals, and the degree to which the conditions destroy habitat or render valued species vulnerable to higher levels of predation from other animals or hunters, clearly affected their judgments: F N 14: It's definitely exposed it quite a bit especially in terms of the wildlife. You know, it doesn't provide the wildlife with the shelter they require. It's a little bit more exposed to hunters and whatnot. F N 4: I don't like it. The government says they're clearcutting trees and they're coming back in better shape, but where are these animals going to go around there. A hunter will go in there and as soon as an animal shows its face in a clearcut, bang. 165 F N 1: If that's a trapline, and if it's your trapline, you'd be pretty well heartbroken. Like I say, the animals don't go in anything that's charcoaled or if there's nothing there to feed on. However, most criticisms of the harvest treatments reflected a generalised concern for the integrity of wildlife and aquatic habitat. In most cases, habitat was something that the participants appear to value intrinsically. O n occasion, this was expressed explicitly through comments that affirmed the importance of a high quality natural environment, or from statements that spoke in relatively abstract terms about the need for more ecologically sustainable forms of forest management: F N 15: It's created a lot of erosion, it's dirtied up the streams that were once salmon bearing streams, it's warmed up the creeks and the fish don't like warm water too much. F N 13: This one would be totally unacceptable. You're dealing with the runoff heading into the water and impacting the salmon. And I would go along the lines of . . . select harvesting for many a purpose and some of it would be to have sustainable logging rather than just clearing out the area. E C 8: ...if you do this you're just asking for trouble. This isn't going to stay anyways. It's all just going to get washed out. Human beings eat fish and this is one of the places where they come up to lay their eggs. T he final criticism brought against the multi-use and industrial harvest patterns focused on the degree to which the reserve patches and riparian leave strips represented in the multi-use simulations would be viable over the long term. In other words, they were dismissed as a largely superficial attempt by the forest service or logging operators to mitigate wildlife and/or visual impacts, which would eventually be eliminated through windthrow. S o m e participants chose to characterise the retention treatments as "sacrificial" and potentially hazardous to wildlife and forestry workers: 166 F N 11: . . .you always get the snags right around there that create the blowdown, and for wildlife it really creates a bit of a problem for them to manoeuvre around. F N 9: I don't like them, because the patches, a lot of them end up to be blow down. . . . Because there's no protection, eh? So a lot of the patches, you could see them, they're just blowing down because they're up rooted and the winds are too heavy. E C 13: Well they leave those little islands in there and soon there gone. F N 8: I've worked in this area for tree spacing and stuff like that, you can see the amount of trees that are blown over after it's logged off like that. These trees aren't going too long if you log it off like that. F N 7: ...why do you leave these sacrificial trees in the middle there, they're going to get blown down anyway? W h y bother? That's just stupid. While the participants were virtually unanimous in their disapproval of the industrial condition, a small and vocal minority expressed their qualified or outright approval of the harvest patterns in the multi-use condition. Comprised largely of loggers and actively employed mill workers, this group characterised the condition as a reasonable effort to balance the paramount economic needs of the community with effective visual management: E C 11: When we go out hunting and fishing we don't want to look at clearcuts scarring the landscape, that's why I prefer smaller ones. There's less visual impact, it's better public relations. E C 8: Yeah , well you see the other thing is everybody wants to get going on tourism here and that's (Multi-use Condition) a lot nicer to look at than that (Industrial Condition). Y o u know, a lot more interesting to people I think. F N 5: it doesn't bother me because the animals can come through this timber (i.e. reserve patches and corridors) . . . they've got that area they want to get to. E C 9: me, I like this one. 167 JL: why's that? E C 9: ...I think partially PR, you know, if there's more wildlife in there. There's more edge this way, I think you'd have less wildlife impact in there. Moreover, as the latter two comments suggest, the multi-use condition is more benign in its effects on wildlife. A s the participants argue, the irregular cutblock boundaries produce longer edges and, along with the reserve patches located in the centre of the openings, there is sufficient cover and refuge for wild game. From the comments that have been reviewed to this point, there is a clear and largely anticipated tension between the participants in terms of those who value the forest largely as a material and economic asset, and those who value it for its natural, habitat providing functions. However, this synopsis may oversimplify the distinction. Of the participants who rejected the industrial and multi-use conditions due to their potential to damage wildlife and aquatic habitat, a similar distinction can be drawn between those who value the forest either intrinsically or instrumentally. A s the preceding discussion argues, the former group contend that the forest, and the wildlife communities that it supports, possesses inherent value that is independent of human needs and uses. However, from the comments made by the latter group, the need for intact and healthy wildlife habitat is bound up with their use of the forest as a source of livelihood from subsistence hunting or commercial trapping. Although more will be said to clarify this in the second half of the chapter, preferences for or against particular landscape conditions can cut across value orientations to the extent that participants who are largely biocentric or anthropocentric can express similar preferences. A s argued in chapter 2, simple preference evaluations that ask participants to evaluate or rate landscape treatments 168 without clarifying the factors that underpin their assessments may obscure information that is potentially more interesting and important. While this may be due to an incomplete understanding of the participants' value orientations, it may also point to a potential weakness in the anthropocentric-biocentric framework. This is not intended to suggest that the framework is incorrect, but rather that it may need to be modified to better account for similarities and differences in landscape appraisals. The evidence and arguments in section 5.3 provide some preliminary reflections on how the framework may be amended. b) Aboriginal Burning: Multi-Use and Pre-lndustrial Conditions T he burn treatments depicted in the multi-use and pre-industrial conditions tended to elicit some confusion. Perhaps owing to the absence of anthropogenic fires in the Skeena valley, or because the approximation of the historical disturbance pattern was incorrect, the openings were often mistaken for slash-burned clearcuts. More will be said about the accuracy of the simulated burn patterns in section 5.2.3 (Contextual Aspects of Landscape Change), but until the purpose and nature of the burn treatments was clarified, many of the same concerns that were articulated in relation to the Industrial and Multi-Use clearcuts were expressed by the participants. In sum, comments on the initial acceptability of the treatment reflected the tension described above between timber production and worries about higher anticipated levels of erosion, and long-term wildlife and fish habitat productivity: F N 1: Most of the elder people as myself, it hurts you when you see that. It hurts you because it's all charcoal. Y o u see the animals feed on these roots and everything else. Y o u won't see a track up there if you go up there after they burn them. 169 E C 1: There's just too many problems. Not only with erosion but there's no protection for animals, and it's a big area right? F N 8 :'ve got to remember when you burn all that floor, when you get a big rainstorm, that all washes down to the river. That's what we don't want. It kills the fish. E C 9: this one I don't like because the clearcuts were too big. In my opinion they should have gone to smaller clearcuts. . . . T o me, you've got a larger visual impact because of the clearcuts, and then for wildlife, if you have smaller clearcuts it's better for wildlife. Quite often the participants would question their own interpretations of the images and ask for clarification. However, invariably, I would correct their misinterpretation of the simulations and inform the participants that the simulated burns reflected a best approximation of historical fires set by First Nations to cultivate berry crops (i.e. the Pre-lndustrial condition) or, in the case of the multi-use condition, future efforts to reintroduce the practice. With this information, assessments of the treatments diverged considerably with comments ranging from a focus on material considerations or use values, to expressions of concern or support that were rooted in competing notions of cultural identity. In terms of comments that favoured the burn treatments, comments could be grouped into two categories. The first emphasised the material benefits associated with the management of non-timber products, underscoring the dietary and potential pecuniary benefits that the community may realise from the cultivation of berries and other forest-based food plants (e.g. wild mushrooms). In some cases, preference for the burn patterns was qualified by reflections on the scale of the disturbances and the practical necessity of large berry patches for small communities: F N 7: This burn is not too bad, but this is huge. How close to the settlement is this, it would make sense. If you compared it to the 170 population that this is trying to feed it would make sense. Then it would be ok. F N 1: It would be nice to pick huckleberries there (Pre-lndustrial condition), wouldn't it? S a m e with that (Multi-Use condition). E C 4: Well it's a human use, at that end of it, it's a positive thing. It's not a human use in terms of a negative thing, it's pretty positive I think. You're feeding people, and it's local and it's high quality . . . But boy it looks like a big burnt area. It changes your view of it, at least my view of it if it's being used for food directly, as opposed to just straight profit. E C 2: I like the fact that it's being managed for many uses, for other uses, as well as just forestry, for wood, for money, for industry. EC1 14: Wow, I didn't know that they (First Nations) used to do that. . . . It would love to be able to hike up there and pick some berries. F N 14: . . .you need something to generate the economy in this area. A n d you have to keep a lot of the Indians happy, and as long as the berry areas are accessible . . . I think it would work to our advantage. The second group of comments that favoured the burn treatments lends some support to the idea that ethnicity may mediate environmental preference, at least to the extent that the landscape condition supports or reflects expressions of cultural identity. For some of the First Nations participants, the reintroduction of berry cultivation practices would not only be conducive to a healthier lifestyle, but would also reinforce traditional patterns of land-use and relationships with the landscape: F N 15:'re taking one resource that may not be of value to you and you're creating resources that you require to live in the area. A n d one of them is berries, and that can go culturally, but if you look back a little bit more with the harsh elements that we're faced with nine months of the year, you definitely require the berries... . F N 12: Well it's another, at least there's another forest use . . . it's cultural and it's cultural in that the Gitksan, and in other areas the Wetsuweten, have through the generations made use of non-timber forest resources. I think it's important that we continue to have that kind of access to those values, to those activities, to those areas within our own traditional lands. 1 7 1 F N 14: I like this one because it kind of shows the rotational burns. Yeah , I know they moved around a lot when they were burning for berries. And it kind of looks here like they've done that . . . that they've tried to do it using traditional methods. In part, the reintroduction of aboriginal burning to the upper Skeena valley may serve as a marker of First Nations culture through the visible imprint that traditional land uses may leave on the landscape. O n a more abstract level, activities themselves (i.e. the harvesting of traditional food plants) are capable of forging intimate bonds between settings and people who possess a unique ethnic identity (Lewis, 2000). In effect, people who have historically been rooted to and derived their livelihood from the land will often characterise their relationship with the landscape in terms of "the earth is part of us" or "we are the land, and the land is us." This may be a factor in the preceding set of comments, however more will be said about ethnic identity as a possible basis for environmental preferences in the second half of the chapter. In the same way that preferences for the burn treatments were rooted in expressions of use value and identity, these same factors were prevalent among the participants who opposed the conditions. The most vocal statements of opposition came from the participants who are actively employed in the forest industry. Initially, most of these participants easily recognised the scorched colouration displayed by the images as an indication of some form of burning and concluded, as did most other participants, that these were clearcuts which had subsequently been burnt. From these individuals, assessments of slash burning were generally positive: 172 E C 11: Well I actually think broadcast burning's a good option in this country because of the brush and stuff, and the fuel loading. I don't have a problem with it, with the burning. E C 9: I think it's an acceptable management tool. Fire has always been a natural part of forest disturbance up here anyway, so. F N 8 : Fire's ok for timber management, but you have to be careful how you do it. You don't want too much run off going into creeks and streams killing fish. E C 8 : Burning's ok. W e need to bring back that type of management in this area. It's good for the soil and makes it easier to replant. However, these assessments changed with the knowledge that the burn patterns were intended for aboriginal food plant cultivation rather than timber management. The handful of Euro-Canadian participants who had formerly supported the use of fire as a silvicultural tool, appeared to base their concerns on the appropriateness of aboriginal burning and, in particular, on the competence of First Nations as fire managers: E C 8 : I think that it's not two hundred years ago now, that we know enough and there's enough value in that timber that it would be stupid to burn it, just to make some huckleberries. . . . Y o u see I take issue with the idea that anything the Indians do is just automatically good because they're so in touch with the land because they're Indians. . . . A n d it's the same thing like the fishing, there's so many people that have the right to fish with nets, but they can't seem to get it through their heads that they don't need to leave the nets in there after they've got enough fish. . . . There's a lot of people who just leave their nets in there because they're lazy, there's some of them that do that. So it's just, wanting to burn because they're Indians makes no sense to me. E C 11: seems to me that they don't do as much berry picking as they used to. Just from observations though. A n d I'm not wild about them torching big areas like that. I mean, if it was managed and not just some Joe Blow who just decides that he's going to go out and burn off a hillside for a berry patch. 173 E C 9: I'm not really too comfortable about the Natives burning that much land. I know they used to do it for berries in the past, but do they really need to do it today? Whether these statements are rooted in competing notions of cultural identity (i.e. they are our forests to manage and no longer belong to the 'Indians'), or simply expressions of concern regarding the reintroduction of forest management practices that have unknown risks and benefits are questions that were not explored in further depth. I am inclined to think that both factors may have played into this dramatic reversal of opinion by all of the Euro-Canadian participants quoted above. However, it is interesting to note that there were First Nations participants who changed their evaluations and, consistent with their Euro-Canadian counterparts, felt that burning for berries would be a waste of more valuable timber resources. In effect, despite the basis of fire management in traditional land management, the need for fire was questioned particularly when it was believed that mechanical stand harvesting could accomplish the same results: F N 8: . . .sure if it was logged off, we will go there. If it's accessible, we'll go there and pick berries. Y o u know . . . like you say if this was a burn and it was logged off, well then it should be replanted. A n d we'll pick berries there until it starts growing again. F N 9: There's a lot of berries in strip logging, in fact tonnes because, I don't know if you know that the berries grow on tree lines, by the tonnes. A n d so strip logging, you see, you don't need to wipe out a place and bum it. If you do strip logging, that's where the berries come back. The same sentiment was expressed by some of the Euro-Canadian participants who argued that a better approach to forestry would be to manage the landscape for the production of multiple resources rather than exclusively for timber or ethnobotanical foods: 174 E C 8: W h y not log it and then burn it? E C 10: Wouldn't it be better to try and kill two birds with one stone? I mean, couldn't this area be logged and then burnt for berries. That may keep both sides happy. In addition to appeasing those participants who felt that the burn treatments were intended to supplant more economically lucrative forms of timber management, multi-use forest management by adapting timber harvesting to accommodate an interim flow of non-timber resources was an idea that held broad support among the research participants: F N 12: Well it's another, at least there's another forest use. It's not just the big companies who are benefiting from it, it's actually the community that is accessing the different forest values or forest resources that are non-timber values and that's important. E C 2: this does sit well with me, very well, because it means you're not going to be depending on those loggers for the rest of your life. F N 15: I think you have to consider the botanical forest products as well as being a resource. It's a naturally producing resource that local people, and not just local we get a lot of outsiders coming in to pick the botanical forest products to sell, to assist their income. . . . Obviously there's a profit, but if you look at it from a community base level the pine mushrooms or morel mushrooms . . . that are harvested in this area are twenty times more valuable to the community than clearcut logs which get exported out of our community because we don't have any sawmills operating. In addition, several participants argued that berry management would provide an additional use which would indicate that the forest landscape is being utilised and cared for on a continuous basis: E C 14: T o me, it would be nice to see that these areas are being cared for. J L : How would you be able to see it. E C 14: Well when it's clearcut, companies just come in, take the logs and then just disappear. At least this way, with people on the land 175 harvesting and tending to the berry patches, people may actually start caring for these areas. F N 4: W e should do it the way the old people did it. W h e n we were out there harvesting berries for ourselves there was always somebody out on the land, or at least more than you see nowadays. Conventional timber management suffers from the public perception that, once the resource has been removed from the land, little (visible) interest or effort is expended into caring for the landscape to ensure that it will remain healthy and productive. A s one participant described it, forestry appears to be characterised by a "make a quick dollar and just grab it and run (FN 7)" style of land stewardship. Cultivating non-timber resources may be one means of providing a continuous stream of benefits from the landscape while the timber 'crop' grows, which ultimately places people on the land throughout the rotation (i.e. as opposed to just at the beginning and end of a timber rotation) and provides a visible human presence that, to many, reflects ongoing stewardship and care. This approach to "visible stewardship" (Sheppard, 2001) is discussed in more detail in the following section (5.2.2: Perceptions of Care: Sustaining Natural Forest Conditions), as well as in the concluding chapter. c) Natural Condition For the remaining condition, most participants identified the natural condition as the most favourable alternative. In general, the high aesthetic value of an untouched landscape combined with considerations of fish and wildlife habitat concerns over the long term were clear factors in the participants' evaluations: F N 1: That's the way it's supposed to look like. That way the animals can roam their own country out there. 176 F N 5: Yeah , I think it's a better deal when they do leave these areas. . . . I think there should be more study on animal habitat. E C 5: This is what I like. It's better for wildlife and I think it's just much nicer to look at. However, considerations of scenic beauty and the integrity of animal habitat were not the only factors that motivated participants to favour the natural condition. A s anticipated from the discussion in Chapter 2, and previous research findings with First Nations communities (Lewis, 2000; Lewis and Sheppard, 2005), evaluations were also affected by a consideration of conflicting material and human growth needs (i.e. psychological or spiritual). For instance, in terms of material considerations, some participants saw the natural condition as a necessary interim step that would allow the landscape to heal the damage of past forest practices, and permit the development of more culturally appropriate and ecologically benign harvesting methods: F N 9: See with this, we could do what we call our aboriginal assessment, what we call it around this area. A n d we could tell them where they might be able to log, and we could tell them areas where they shouldn't log. E C 8: Well, obviously I like that because there's a lot of opportunity for someone to go in there and do it right. F N 2: This is the way it should be done, left to regrow like that. This way we can go in and harvest the trees in a way that's better for the animals, more selectively. Invariably, comments of this sort were accompanied by a clear preference for landscape conditions where selection harvesting or individual stem removal is the dominant form of timber management (despite the absence of simulations representing this form of timber management). Working towards selection 177 harvesting through the restoration of natural forest conditions was preferred, in part, because forest management has been and remains an integral part of the region's economic foundation. However, it also reflected the need to balance multiple forest uses as well as the long-term economic needs of the communities: F N 7: But that's how I always felt about, after I got involved in all the blockades and trying to stop the logging so that they would do it selectively, I used to write poetry about it and try to think of ways that we could get together to stop it and have it in a different way because I could see that this is way, way, way back in the early 80s or something like that. There weren't as many clearcuts then, they were just starting to take off, and I could see that all the people who are here, they wouldn't have jobs eventually, because it makes sense where you see a land base and you can only clearcut it so much, then you're out of a job. But if you do it in a different way you'd have a job for a longer time. E C 13: I'd take out single stems. . . . There's some other values that are way more worth it. J L : What other values? E C 13: Well, the visual, wildlife, water quality. E C 2: . . .even the thinning where they do selective logging because of things like mushrooms and medicinal plants and all that kind of thing . . . and treat them with respect and say, well, this is a special patch of plants and maybe they'll flag it and say where these plants grow, leave that. The latter two comments reflect the range of multiple values that the participants addressed in their appraisals. Consistent with the material underpinnings of landscape preference, several of the First Nations participants identified the natural condition as clearly suited to the cultivation and harvesting of sacred and medicinal plants. For instance, two First Nations elders pointed to the intact forest ecosystems at the base of Blunt Mountain (i.e. viewpoint three) as ideal settings for the gathering of sacred medicines: 178 F N 1: This is pretty good, yeah. See the slides, they get soaked up. That's what we look for when we go picking our Indian medicine. J L : The slides? F N 1: Yeah , where there's no logging. F N 2: It's good habitat. F N 1: S o you go down to the foot of this, that where it grows the medicine. F N 7: That would be the pure water. You would hike up there and pack that water for your sacred medicine. Because that would be the water you would use. Unexpectedly, this sentiment was expressed by a handful of the Euro-Canadian participants: E C 2: ...that leaves lots of territory where you would find your medicinal plants probably. E C 6: I think this would be a better condition for things like medicines and other plants that the First Nations use. In terms of more abstract needs, several participants couched their preference for the natural condition in terms of the spiritual and restorative benefits of a landscape that is undamaged or, as one participant stated, an environment that is 'whole:' F N 7: This one I like the best because it feels whole. E C 2: All the good stuff is still there. . . . A n d I remember we used to go out to Kisegas, you could almost feel the people there, the old people. It was like, that's spiritual. And that's what I used to like about this place. F N 4: I like this one because this is the kind of place where I'd like to go when something's troubling me, to find healing. These comments were anticipated given that much of the environmental preference literature posits that people generally prefer landscapes where there is no visible imprint of human disturbance (Abello and Bernaldez, 1986; Benson and Ulrich, 179 1981; Kaplan and Kaplan, 1989). Moreover, pristine settings are generally regarded as being more conducive to personal and psychological restoration, where contact with nature has been empirically demonstrated to promote restoration from psychophysical stress (Ulrich, 1979, 1981; Ulrich et al., 1991) and mental fatigue (Hartig etal., 1991; Kaplan and Talbot, 1983). Less anticipated were the comments which opposed the natural condition, largely because unutilised or underutilised landscapes were deemed to be an 'unproductive' waste of resources: E C 8: Well what are you looking for? Are you looking for healthy trees or are you looking for forest management? None of this will be good because it's just a big weed patch . . . if you're not going to make use of it, well this is just great. F N 5: Well it's alright I guess, but it's not natural the way we live nowadays. You know, like I said before, you have to create employment. Y o u have to live. E C 11: well it's nice to see untouched forest but, to me, it's a waste because a lot of our timber here's over mature. I mean, if you don't log it, you're either going to have fire or bugs, so that's just part of the natural cycle here. A major theme throughout the interviews was the tension between managing the landscape for a single material use (i.e. timber resources) and competing material (i.e. non-timber resources), identity-based and spiritual values. In many ways, this finding underscores the argument by Stankey (1996) that the search for socially acceptable forest management is inherently wicked, that is, confounded by multiple and competing conceptions of appropriate land-use and for which there is no discrete, unequivocal solution. When confronted by the intractability of technical solutions in value-based conflicts, planners often conclude that equally complex 180 processes involving community-based negotiation, education and trial-and-error management are the most appropriate means of planning landscapes in the face of competing social values and meanings of forests (Armour, 1991; Coyle, 1998; Fisher and Ury, 1981). While I remain a proponent of community consultation and collaborative planning processes (Lewis, 1996, 2000; Lewis and Sheppard, 2005), I do not believe that technical and, in particular, environmental design solutions do not have a place in the search for (more or less) socially preferred forest management. In the following subsections (5.2.2 and 5.2.3), I provide illustration from the upper Skeena valley interviews which demonstrate that the participants sought physical indications or cues that the forest landscape was being managed carefully (5.2.2), or according to culturally appropriate conventions of land-use (5.2.3, see Table 5.2). In either case, value based landscape conflicts need not be construed as the province of social scientists and professional negotiators. Indeed, there may remain considerable opportunity for technically oriented environmental planners and landscape architects to assist in the search for workable solutions. 5.2.2 Perceptions of Care Despite the overall preference for unaltered landscapes, most participants demonstrated a pragmatism that affirmed the need for forest management largely to maintain the region's standard of living. In effect, without returning to the conditions of the last twenty to thirty years, there was an unambiguous sense that forest management will remain an important part of the upper Skeena's economic future: FN11: Y o u have to be realistic and think that in this area there's going to be interference or human disturbance of the land, the forestland area. 181 FN12: I have to be realistic and understand that not all of the forest areas in this forest region can be pristine. FN5: I would prefer this, but It's not natural the way we live nowadays. You know . . . you have to create employment. Y o u have to live. E C 2 : I think it's reality because we're getting more and more people. I think that it's still pretty devastating, but you do know that the trees will come back in the forest when they've been logging. For those participants who do not play an active or direct part in the management of forest resources, the timber harvest patterns in the industrial and multi-use conditions were criticised for the degree to which they failed to demonstrate proper care or attention to natural processes or the needs of future generations. The main categories of 'care' dimensions are listed in Table 5.2 and described below: Table 5.2: Perceptions of Care Issues Dimensions Illustrative Comments Perceptions of Care Respect for Nature ...treat them with respect and say, well, this is a special patch of plants... Cultural Identity ...they moved around a lot when they were burning for berries looks here live they've done that. Interim Uses ...when it's clearcut, companies just come in, take the logs and then just disappear. Selective Disturbance ...some of it would be to have sustainable logging rather than just clearing out the area. Sustained Presence I would like to see people out there more often tending to the forest, making sure it's healthy. E C 2 : I'm not totally against logging, and I know that there are some, sort of, economic basis for clearcuts. But I still think that they should be taking more care in how they do it. I mean that kind of thing is a terrible scar. And it is. It's just rotten and not just aesthetically it's, I imagine that it changes a lot of things. FN10: I think that all this multiple use stuff needs to be done with care because even the thinning where they do selective logging because of things like mushrooms and medicinal plants and all that kind of thing . . . and treat them with respect and say, well, this is a special patch of plants and maybe they'll flag it and say where these plants grow, leave that. 182 FN13: Because they don't really take, I'd like to see more care in how they design the blocks, the cutblocks, or how they take measures to protect riparian areas or areas which, this has wildlife. When pressed to explain how they would recognise a carefully managed forest, some First Nations participants pointed to the burn treatments in the multi-use and pre-industrial simulations to suggest that the smaller patchwork pattern of burning approximated the traditional practice of rotating burn treatments, and is analogous to the rotational or farm-fallow system of European agriculture: FN6: My mum used to tell me that when her family burnt an area they would only burn off a small area and the rest of the berry patch would be left alone. That way there would always be a good source of berries every year. T o me that's the way forest companies should do it, take out only what you need from a small area and leave areas that have been harvested in recent years to regrow. F N 4: W h e n the old people did this (i.e. burnt for berries) they used to burn one area off and leave the area nearby to grow and produce berries. Kind of like they do on a farm, you know, they would rotate the crop. W h y can't forestry do that? W h y do they have to remove everything and leave nothing behind for the other plants and animals? In effect, as described in Chapter 3, berry-growing areas were managed according to a rotating system of applying fire to a confined portion of a larger berry producing area (Trusler, 2002). Only after a smaller patch had passed its peak production capability would it be burned, destroying much of the plant growth aboveground and allowing the rhizome base beneath the soil to sprout more vigorous and productive shoots. While recently burned areas regenerate and reach a subsequent phase of peak production, a process that is estimated to take between three to five years, adjacent areas that were burned three to five years earlier have reached their productive potential and are actively harvested for their fruit crop. In this manner, 183 the land is maintained in a constant state of food production, and the process of periodic burning is known both scientifically and anecdotally to improve soil fertility. From the perspective of some of the First Nations participants, emulating this culturally based pattern of land management would reflect care to the extent that harvesting is done in smaller patches and, moreover, that the periodic burning of overmature vegetation returns nutrients to the soil. Moreover, a regular and more frequent presence on the landscape may suggest that the forest is being tended: E C 14: T o me, it would be nice to see that these areas are being cared for. J L : How would you be able to see it. E C 14: Well when it's clearcut, companies just come in, take the logs and then just disappear. At least this way, with people on the land harvesting and tending to the berry patches, people may actually start caring for these areas. For most of the First Nations and Euro-Canadian participants who expressed a need for careful forest management, the most common conception of care was based on a selective approach to timber harvesting: E C 2 : I was just looking at that one [multi-use condition], you can see the riparian zones and see that care has been taken in managing it. F N 14: A n d I would go along the lines of . . . select harvesting for many a purpose and some of it would be to have sustainable logging rather than just clearing out the area, leaving the roads to erode and ensuring there will be something for future generations to harvest as well. Selective harvesting suggests that less is being removed from the land (perhaps even in circumstances where the removal is spread over a larger area), and that the harvest has been conducted without the degree of industrial intensity that is commonly associated with clearcut methods. However, some Euro-Canadian participants expressed views that were consistent with the preceding set of 184 comments that suggested that a visible and sustained human presence on the landscape would represent active stewardship. A m o n g this group, suggestions for how to accomplish this varied considerably from erecting built structures (e.g. silviculture camps or growth monitoring stations) to direct community involvement in the restoration of damaged landscapes or the stewardship of recently harvested areas: E C 1 4 : I'm not sure, it might be good to see some kind of building or permanent camp set up in areas that have been disturbed. That way you could see that someone is looking after the area. E C 1 5 : Why not just get people out there looking after the land? I mean, if you have people involved in looking after their backyard wouldn't that go a long way towards making people feel better about how the forest is managed. E C 3 : I think we should be involved in the management of our region. I would like to see people out there more often tending to the forest, making sure it's healthy. A few observations can be made about the comments that have been reviewed so far. First, despite the overwhelming preference for natural looking landscapes, there is a clear recognition that humans are dependent on the natural environment and that some visible human alteration of the forest is to be expected and regarded as part of the natural condition. The fact that this has been observed among the interview participants may simply be a result of the rural resource culture described in Chapter 2. In effect, forest management is such an omnipresent facet of life in rural British Columbia that it is hardly surprising that many residents of the Northwest have come to regard a 'natural connection' to the land as a human-nature relationship that encompasses the commercial harvesting of trees. Indeed, as one participant remarked, upon coming to the region nearly forty years ago, the desire to 185 see untrammelled nature was tempered after several years by the realisation that, whether for timber, grazing lands or berries, human beings have in some way transformed the upper Skeena's forest landscapes: E C 1 5 : I guess when I first came here as a young guy I expected to see more of a wilderness, and it really got to me, you know, how the forest was being damaged. But I guess with age I've either just mellowed or come to the realisation that using the forest is what makes this area tick. Second, where forest management has occurred, some participants have expressed a need to see that it is being done according to a standard of care. Moreover, that standard of care may have its basis in culturally defined notions of appropriate land stewardship. For instance, for some First Nations participants, care is implicit in the traditional pattern of berry patch rotation. Whether there is a genuine understanding that customary forms of land cultivation were indeed better for the environment, or these participants are simply basing their views on an idealised conception of their customs and history is uncertain. What is significant is that the First Nations participants have predicated their conception of care on a culturally recognisable form, as has the small group of Euro-Canadians who identify care through agrarian lenses - i.e. built structures and people present on and actively tending the land. For both groups, the interval between harvests (or a forest rotation) is too long. A s one component of socially acceptable management, forestry ought to alter the "let's get it and run (FN7)" style of management to accommodate visible demonstrations of stewardship and care throughout the harvest interval. Suggestions on how to accomplish this are discussed in the concluding chapter. 186 5.2.3 Site Characteristics When responding to the acceptability of any of the given treatments, in addition to reflecting on such themes as the balance of impacts, perceptions of naturalness and visual indications of stewardship, participants frequently advanced other issues that revolved around the physical or spatial contexts in which the forest treatments were depicted. In the transition landscape of the upper Skeena valley, there is enough physiographic diversity to allow for slope position, angle of slope, or adjacent landscape features to influence participant evaluations of landscape conditions. Typically, participant reflections on the physical characteristics of the forest modifications focused on the scale and proximity of the disturbance to sensitive features, and an assessment of cumulative impacts (Table 5.3): E C 8 : This is extremely steep too. They're pretty big for such a steep slope. E C 1: you've got some pretty big cuts up in here, huge. That's a big, big cut. Although they're going about it half at a time, I'm not too sure what the timing was but there was a timing thing. FN1: Gitksan people don't go up and make a big square block like that. FN12: I just don't like it because the disturbance is so huge, and it's similar to the disturbances we can see on Blunt Mountain just down the valley towards Smithers. That's the same thing. So the disturbance is so huge, like the clearcut area on Blunt Mountain, I've always seen that as negative and unfavourable. Table 5.3: Site Characteristics Issues Dimensions 3. Site Characteristics Disturbance Scale Adjacent Features Cumulative Impacts Illustrative Comments I just don't like it because the disturbance is so huge. ...what about all the toxins coming off here and going into your food-source? When such a huge cutblock is right next to an area that has already been harvested, that is unacceptable. 187 In terms of harvest location and the scale of the operation, both of these considerations potentially encompass a broad range of issues that are salient to forest management - adjacent land use, cumulative effects, potential for ecological deterioration, economic feasibility and sustainability. A simple focus on stand level patterns may obscure the more pressing issues affecting people's judgements. For instance, the potential for cumulative effects (i.e. the combination of past and adjacent land-use histories) affects the evaluation of a particular treatment. It may be one matter to impose a relatively intensive modification in a previously undisturbed landscape, but the same treatment may have different aggregate effects in a situation where dramatic human presence already exists. Perhaps due to their prominent front-country location, most participants were acutely aware of the recent land-use histories for Viewpoints 1 and 3. In evaluating the landscape modifications depicted at these locations, participants were thinking broadly: E C 1 3 : If there are places that are a little less valuable or sensitive than others, that's where I think I could handle a little more activity, more intense development. E C 4 : When such a huge cutblock is right next to an area that has already been heavily harvested, that is unacceptable. Y o u get a bigger visual impact. E C 2 : That's what they did, didn't they. And they went in there and I said why are you making the clearcut bigger? A n d they said well it's already messed, it's already visually messed, so we feel that we can do the edges, and nip at the edges. Also, many areas within the upper Skeena valley have highly productive capacities for other valued resources such as fish, deer, moose or bear. A s discussed in the first section of this chapter, some locations are valued for their scenic character, and visitors may not appreciate harvest operations in full view of 188 these valued locations. In terms of wildlife, several participants noted the territorial behaviour of grizzly bears and attributed recent bear incursions into settlement areas to the destruction of bear habitat by clearcut harvesting: FN13: The issue is how high it is, I think, how close it is to the alpine. Because there are grizzly bears even though it's close to Hazelton . . . but it's too close to alpine I think. FN6: we've had grizzlies coming into town recently and, even though I have no problem with bears, they've given a few people quite a scare. J L : Why has that become such a problem? FN6: Because they're logging out their habitat. Grizzly bears have their own territories, and where are they supposed to go? E C 6 : Logging has encroached on bear habitat. I don't think they should be doing anything up there unless they're sure that it isn't displacing bears and moving them out of their browsing areas. T he placement of a treatment on a hillside can make a difference to some people, especially if an improper placement threatens another resource that people wish to maintain. With the attention that many First Nations people give to subsistence resources in the Northwest, harvest patterns, the intensity of harvests, and disturbance locations need to respect the integrity of culturally important food and medicinal plants: FN7: do you know what? O n e of the things that I know of, particularly the water, is that this is pure water, pure mountain water coming from here. It's even more so here where this peak is. . . . That would be the pure water. You would hike up there and pack that water for your sacred medicine, because that would be the water you would use. The further up you go, in our minds, the stronger the medicine is that you're going to make. FN6: If you have foodstuffs down there and you have a logging show up there [multi-use condition], what about all the toxins coming off here going into your food-source? It doesn't make any sense. Y o u can't have both and have pristine foodstuffs. 189 FN1: My dad used to work up at the Kuldo, that's after they clearcut and started putting roads in, he cried . . . I asked him why are you crying, and he talked to me and said he's just broken hearted about what's going on. He tells me where they picked huckleberries up there, where they picked their Indian medicine, it's all under clearcut, it's just ruined. . . . you can't harvest medicine, we have to go for miles now just to try and get some, pick medicine. . . . It has to be pure. This is the right weather when you want to in and get them because everything washes off. That's when we get them. Now we have to go for miles and miles. Thus , for the First Nations participants, the placement of forest modifications can defy common sense in terms of their proximity to sensitive features and land-uses. However, this common sense understanding of disturbance location appears to be rooted in a culturally based understanding of how the landscape was traditionally managed for subsistence resources. For instance, several participants who continue to harvest traditional food and medicinal plants pointed to potential siting and sizing problems in the simulation of traditional burn patterns: J L : would the burnt berry patches in the old days have looked anything like this? FN2: No, they would have been smaller. If they're going to pick huckleberries, but they've got to see the huckleberries first before they burn it. FN6: if these are going to be berry grounds, isn't it too far down. It would be more here [mid-elevation]. Yeah , you don't want to be climbing up . . . wouldn't you try to make it easy for yourself, to maximise your time and picking time as well rather than wasting energy running up? FN8: ...we got these berry patches up in the mountain in Salmon River, and they're quite high up in the alpine region of this area here, not down below here. Berries grow at a certain height, huckleberries at a certain height on the mountain. . . . that's where they used to burn berry patches. A special consideration for this latter group of First Nations participants was the appropriateness of the burn treatments depicted in the images, both in terms of 190 how they may be impacted by other forest uses - e.g. timber removal - as well as the intrinsic suitability of the simulated patches in terms of their size and topographic location. From what is known about these participants, their responses draw upon a specialised body of knowledge to the extent that they are personally familiar with the traditional practice of berry management using fire. In effect, the participants who provided such direct and highly informed comments on the appropriateness of the simulated burn treatments have either engaged in the practice of fire management, or have witnessed parents or grandparents use the technique. In all cases, the participants acknowledge that subsistence berry harvesting, either on specially managed (i.e. deliberately burned) reserve lands or on naturally occurring burn areas, remains an important part of their way-of-life. From this, it would be reasonable to assume that culturally based practices, as a source of environmental knowledge, may contribute to the acceptability of a landscape treatment. Moreover, to the extent that these practices are specialised both within and across ethnic communities, inter-individual variation in landscape preference may be mediated, in part, by embodied or inter-subjectively derived knowledge - i.e. knowledge that comes from the lived experience of the perceiver. This issue is central to the second research question, and is explored at length in the following section. 5.3 Question 2: Inter and Intra Ethnic Landscape Preference Differences T he second research question asked - How and why do preferences for modified forest landscapes differ between and among First Nations and Euro-Canadians? Comparison of the participants' preference ratings (i.e. 2=preferred, 1 =moderately/somewhat preferred, 0=not/least preferred, Appendix C2) with the 1 9 1 comments that were made in relation to the conditions yielded an interesting pattern of results. As described in Chapter 4, participant ratings were analysed quantitatively using Multidimensional Scalogram Analysis (MDS), which produced the scattergram plot shown in Figure 5.1. Each point in the MDS output represents an individual participant (or group of participants if their preference ratings correspond exactly), and their similarity or dissimilarity with any other individual in the MDS plot is measured by their respective proximity or distance from other points. Figure 5.1: Multidimensional Scalogram (MDS) Analysis Results Object Points - Common Space The number adjacent to each data point indicates the number of participants represented by each data point. In effect, the more two individuals concur in their rating or assessment of the forest conditions, the closer together they will be in the geometric space of the MDS output. Conversely, disagreement is represented by greater distance between 192 points. However, for the configuration of the M D S plot to be meaningful, each point or cluster of participants must be examined in relation to three factors: • Their relative proximity to other clusters within the M D S plot, • The qualitative explanations given by participants within the cluster for their ratings, and • Socio-demographic similarities and/or differences among the cluster participants that may shed light on their preference evaluations. T o this end, the following discussion combines an examination of the quantitative M D S output with a qualitative explanation of the M D S results. 5.3.1 Individual and Group Preference Profiles a) Participant Cluster One The first task in M D S analysis is to examine the scattergram output for clusters or groups of participants defined by relative similarity. T h e first cluster (cluster 1, Figure 5.2), and perhaps the simplest to address, is situated in the upper right corner of the M D S plot. With the exception of the other groups that are discussed below, cluster one is the only participant grouping that is clearly defined by ethnicity, age, gender, and occupation. In effect, all four members of the cluster are relatively young (i.e. baby boomers), Euro-Canadian males who are actively employed in the forest industry as contract loggers or small-scale sawmill operators. Moreover, from their biographical statements, a fairly narrow range of activities, namely employment and the occasional hunting or fishing trip, defines their relationship with the forest. A s such, it was anticipated prior to interviewing these individuals that their condition evaluations would be marked by a strong timber bias. However, even this apparently safe characterisation of forestry workers can be inaccurate. All four of 193 Figure 5.2: MDS Results, Cluster 1 Object Points ® 1 1 [ > ® ' Cluster 1 <8> ® The number adjacent to each data point indicates the number of participants represented by each data point. the cluster one participants expressed some discomfort with the industrial condition. While one participant rejected the condition outright, the remaining three rated it as moderately acceptable arguing that clearcut harvesting is sometimes a necessary evil that ought to be done at a smaller scale and with a more natural appearance: E C 1 1 : I just don't like the squarish looking boundary . . . just in terms of visual impact it looks kind of artificial. I prefer to see something like this where it's hard to tell if it's a cutblock or a natural burn (i.e. pre-industrial condition). E C 8 : Well there's just the straight aesthetic value . . . people don't want to be living in a big clearcut here. E C 1 0 : Yeah , I don't have anything that's blanket against anything, even clearcuts are good in certain places but it's, yeah what are the reasons for it. 194 E C 9 : I don't know, sometimes clearcuts are ok, it just depends on what your objectives are. But that doesn't mean I like looking at them. In terms of the pre-industrial condition, the irregular and naturalistic edges accounted for the moderate level of acceptance that two of the cluster one participants expressed. However, prescribed burning was regarded with some scepticism. From their perspective, burning as a silvicultural tool is of questionable value in maintaining or enriching forest soils: E C 8 : In my mind they've always been a little suspect these burns, because I haven't seen the benefits of doing them. I know they add a little bit of nitrogen to the soil but. Thus, for operational reasons, the value of burning is moderately persuasive. However, as discussed earlier, the use of fire as a means of creating and sustaining traditional food crops by First Nations was regarded as completely unacceptable, in large part because food production was seen as a waste of more valuable timber resources: E C 8 : I think that it's not two hundred years ago now, that we know enough and there's enough value in that timber that it would be stupid to burn it, just to make some huckleberries. . . . Y o u see I take issue with the idea that anything the Indians do is just automatically good because they're so in touch with the land because they're Indians. . . . A n d it's the same thing like the fishing, there's so many people that have the right to fish with nets, but they can't seem to get it through their heads that they don't need to leave the nets in there after they've got enough fish. . . . There's a lot of people who just leave their nets in there because they're lazy, there's some of them that do that. So it's just, wanting to burn because they're Indians makes no sense to me. E C 11: seems to me that they don't do as much berry picking as they used to. Just from observations though. A n d I'm not wild about them torching big areas like that. I mean, if it was managed and not just some Joe Blow who just decides that he's going to go out and burn off a hillside for a berry patch. 195 E C 10: Why would you want to ruin all that good timber just for the sake of a few berries? The Indians are just like the rest of us, they need jobs and probably could get away with making a few patches on their reserves. With the exception of one participant, the natural condition was rejected for similar reasons. In effect, setting aside large landscapes for recreation, scenery or wildlife habitat without some form of commercial timber management was described as unacceptable in its wastefulness: E C 1 1 : I wouldn't like it because I think it's kind of a waste of the land base. We're in tough economic times, you know, people kind of need the work. You know, I'd like to see it managed properly. Yeah , to me, I don't know why you'd want to leave the whole area there untouched. E C 1 0 : Well the second one (i.e. natural condition) I figure would be a waste because it's nice timber close to town, so it's relatively cheap . . . it's just good economics. E C 8 : Well what are you looking for? Are you looking for healthy trees or are you looking for forest management. None of this will be good because it's just a big weed patch . . . if you're not going to make use of it well this is just great. Yeah , like that's kind of a nonsense thing. E C 9 : That's the lifeblood of the communities up here. You're not going to make enough jobs to keep the people here employed just by making pretty scenery. T he one condition that was unanimously preferred as an acceptable forest condition was the multi-use condition. In general, the cluster one participants characterised it as an effective compromise between managing the forest for timber as well as the "other concerns" that impinge upon forest management. From their perspective, these other concerns were largely defined by the public image of the forest industry, and the irregular patterns of the multi-use condition created a landscape that is consistent with the natural disturbance regime of the upper Skeena: 196 EC11: Yeah, I think they're heading in the right direction. You go to irregular boundaries, it mimics fire and stuff. It would kind of blend into the landscape here.... EC10: When we go out hunting and fishing we don't want to look at clearcuts scarring the landscape, that's why I prefer smaller ones. There's less visual impact, it's better public relations.... EC9: This looks better. Yeah, I think it's much better for tourists and locals to look at, you know. As professional timber workers, the participants in cluster one qualified their appraisals of the multi-use condition by pointing to the operational challenges of working with irregular cutblock patterns. Nonetheless, their support for and characterisation of the condition as a 'win-win' solution was unambiguous. To summarise the mindset of this group, it appears that their assessments were driven to a large degree by a focus on the material and, in particular, the economic imperatives of timber management in the upper Skeena valley. Such a position is perhaps self-evident given the cluster one participants' reliance on commercial forestry for their personal livelihoods. However, their qualified acceptance of the industrial condition and unanimous support for the multi-use treatments points to a further set of factors that guide their assessments of forest management. As the first half of this chapter demonstrated, participant evaluations are guided by the consideration of how multiple impacts are balanced and, according to the participants in cluster one the imperative of commercial timber management is tempered by a consideration of forestry's visual impacts. The scope of visual impacts includes how forest management appears to tourists and others visiting the region, as well as their own desire to encounter natural looking landscapes when recreating. 197 To the extent that cluster one defines the value of forest resources primarily in material and economic terms and, in particular, how the forest is valued and utilised by humans, this group of participants appear to fit well within the anthropocentric category of human-nature values described in Chapter 2. Overall, this group of participants share the following perspectives: • Human material/economic uses of the forest are paramount in their assessment of forest related impacts - i.e. the forest is a resource for human use and benefit, • Visual impacts are an important but secondary concern, • Forest-related uses are limited to a narrow range of activities - i.e. employment and recreation. • A general pattern of landscape condition preferences: • Preferred: Multi-Use • Moderately Preferred: Industrial • Least Preferred: Pre-lndustrial, Natural However, I would like to make this categorization with one caveat. From what is known about these individuals through numerous personal conversations, I believe that it would be a misrepresentation of their mindsets to argue that they filter their perceptions of the landscape entirely in terms of immediate material or economic needs. Although less anthropocentric statements did not emerge from these individuals during interview discussions, I am inclined to think that most people would reject forms of environmental management that (are proven to) seriously degrade the environment even if they are profitable and present a favourable public image. Moreover, as one participant indicated, long-term dependence on the region's resources can force someone to think beyond immediate economic needs, particularly when you are raising children who may grow to depend on the same basket of resources: 198 EC8: starts to sink in that it's, you know, that you're responsible for what's going to happen here in fifty years and that your kids and grandkids will want to do something with it, well it just changes the whole thing. So you just starting taking the nice ones and leaving the other ones, and you've got a better, it's what you're leaving behind instead of what you're taking. However, statements such as this were more an exception than the norm for cluster one participants. The clear predisposition for these individuals was to evaluate the conditions from a characteristic anthropocentric orientation. This position contrasted sharply with the participants in cluster two who conceptualised human-environment relationships in oppositional terms and, in particular, as a largely damaging relationship. b) Participant Cluster Two Relative to cluster one, the small group of three participants that comprise cluster two are more heterogeneous in background (Figure 5.3). The main attribute that unites all three participants is that they have some experience in the forest industry as silviculture workers or mill employees. The one female in cluster two is Euro-Canadian, currently works as a teacher, is among the younger members of the entire participant sample (mid twenties to late thirties or 'Generation X') and describes her primary use of the forest in recreational terms, that is, hiking and the occasional fishing trip. The other participants, in addition to being male, both belong to the baby boomer generation, and are self-employed. However, they belong to different ethnic groups to the extent that one is Gitksan and the other is a Euro-Canadian resident. 199 Figure 5.3: MDS Results, Cluster 2 Object Points 3 ® ® Cluster 2 ® ® ® ® The number adjacent to each data point indicates the number of participants represented by each data point. • • • • • • • • • • • • • •HE ~ ! S 9 H I H m H H a H H B S ^ H B 3 i Notwithstanding the mixture of socio-demographic characteristics, and perhaps owing to the similar purposes for which they use a forest environment, all three participants were consistent in their appraisals of the forest conditions. The most unequivocal endorsement of the conditions was expressed in relation to the natural condition. Unlike their counterparts in cluster one who regarded the condition as a material waste, cluster two attributed their preference for an untouched, pristine or natural forest to an abstract conception that human management involved the destruction or disruption of something that is intact and in balance: FN 12: Al my life I've liked that sort of forest area and I realy don't know why. ... it might be, probably is something that I've regarded as 200 more natural, it's a natural forest condition. A n d there's been no attempts by humans to interfere with things, to do so-called management of the forest area. . . . I think I've always had the bias that what we call management is more interference, and it's interference done without sufficient thought and knowledge. J L : T h e bottom one sits well with you. E C 1 5 : Always. J L : Why's that? E C 1 5 : Just because it's all there. E C 1 4 : I like it because it looks natural and healthy. Like people haven't been in here and upset things. I don't know much about this subject, but I suspect that there would be more wildlife in here. Conversely, the industrial condition was viewed as representing the worst form of landscape management. In part, the cluster two participants were reacting to the scale and shape of the treatments, but their assessments were also clearly informed by a consideration of humanity's place in nature as an agent of change: FN12:'s so huge. I've always considered this unfavourable. . . . Without doing a lot of deep thinking about it, I just don't like it because the disturbance is so huge, and it's similar to the disturbances we can see on Blunt Mountain just down the valley towards Smithers. E C 1 5 : When you've got that kind of thing scarring the landscape, I don't think that it benefits anyone other than the bottom line of the big companies. T o me it just kind of reflects the destructive power that people have over nature. E C 1 4 : I don't like it, mostly because it's just so big. But I just get tired of seeing stuff like this that can't be anything but harmful to nature. That's to me what industrial scale management implies. Given the strong bias that the participants in cluster two have for a natural landscape, and the general perception that management equates to interference and destruction, it would be reasonable to conclude that these participants have expressed a value orientation that is consistent with biocentrism. However, a 'pragmatic' outlook that affirmed the need for some form of commercial forest 201 management tempered their assessments. Whether this broader outlook is due to their experience in forest management, or due simply to living in a region that is deeply reliant on the forest industry, is uncertain from the comments that were obtained and further research may be needed to clarify this question. O n e participant suggested that their current economic independence from the forest industry might account for a more critical stance towards forest practices: FN12: I realise that not all of the forest areas in this region can be pristine . . . I suppose I would say that, but I live in modern times and I don't entirely depend solely on those basic human needs. So I think that's why I see the pristine condition as way more favourable because I have the luxury, so to speak, of having a decent or reasonable standard of living. . . . However, I suppose if you were ask my grandfather that question, if you were to show my grandfather these in the 1950s, he would say this is the best one [pointing to the industrial condition]. Despite having a standard of living that comes from sources other than forestry, the other participants in cluster two affirmed that the well being of the broader community depends on an economically healthy forest sector: E C 1 4 : Even though I don't rely on the forest industry, I guess we need a certain amount of forestry. That's what these communities were built on right? E C 1 5 : I guess when I first came here as a young guy I expected to see more of a wilderness, and it really got to me, you know, how the forest was being damaged. But I guess with age I've either just mellowed or come to the realisation that using the forest is what makes this area tick. Perhaps reflecting their economic pragmatism, the participants in cluster two were moderately accepting of the multi-use condition. Predictably, the participants were critical of the large (35 hectare) cutblocks, and felt that the treatments would have a detrimental impact on wildlife: 202 F N 12: Well these two here, there's so much disturbance. . . . T h e issue is how high it is, I think, how close it is to the alpine. Because there are grizzly bears there even though it's close to Hazelton. . . . E C M : These seem pretty big. I can't believe that wildlife won't be affected by this. E C 1 5 : I don't think this is a good picture, just for the sake of the animals that live here. I mean, I'm not a hunter but I don't like to see anything that's going to damage sensitive wildlife populations. From their biographical statements, the participants in this cluster do not hunt recreationally or engage in subsistence fishing. When pressed to explain why the potential for habitat impacts is important to them, their responses appeared to reflect the kind of intrinsic human-nature values that reflect the biocentric orientation: E C 14: Isn't that why most people choose to live here in the north, just to be closer to nature? I like knowing that animals are out there, healthy and surviving. In effect, nature is valued for its own sake and people derive inner satisfaction from the knowledge or perception that an environment is intact and functioning well. From a utilitarian or use value perspective, however, two of the cluster two participants explained that they enjoy wildlife encounters during recreational uses of the forest: FN12: I just love the mountains and I love the solitude, the beauty. I love the aesthetic beauty of the mountain ranges, of being high up. . . . I think it's the solitude . . . I'm not a hunter. If I hunt I hunt with a camera. E C M : I really enjoy coming across wildlife when I'm out hiking with friends. It's great to see a moose or the odd bear. T he multi-use condition, however, moves closer to being acceptable to the extent that the treatments demonstrate care for ecosystems. Overall, the large riparian leave strips, remnant tree patches, irregular and feathered edges appeared 2 0 3 more natural and may benefit wildlife. Moreover, the participants were sensitive to the operational difficulties of the treatments and suggested that the condition reflected greater care and attention to natural habitat and topography: F N 12: It looks more natural because the trees, there's no sharp, like the cutblocks where they've done clearcutting, it's like this wall that they've done around the edge of the cutblock. But this one, the wall of trees, there's a sharp change right from the large tree stems to the clearcut area . . . they're not really being careful about how they design the blocks, the cutblocks, or how they take measures to protect riparian areas or areas which, this has wildlife. While this one here, it's more like a feathered edge. The trees extend into the open area, all around it, and it's a more irregular, more natural pattern. E C 1 4 : it looks like they've paid more attention to the lay of the land, you know, the topography and stuff like that. Isn't that why the edges of these openings are so wavy? E C 1 5 : I know from talking to forestry folks up here that these aren't easy to harvest. T o me it just says that they've been a little more cautious in how they've taken the trees out. T he pre-industrial condition garnered a surprising level of support from the cluster two participants but, from their explanations, there were striking differences among them. The irregular, naturalistic geometry of the treatments played some role in their assessments, but the participants were unanimous in their support for the objectives that motivated the treatment. For the Gitksan participant, the pre-industrial condition is a visible embodiment of his culture in the sense that it reintroduces activities and forms of land management that have been practiced by the Gitksan for generations. In a larger sense, the condition reflects ancestral practices and connections with the land and, in this way, has the potential to strengthen Gitksan cultural identity: FN12:'s cultural and it's cultural in that the Gitksan and, in other areas the Wetsuweten, have through the generations made use of 204 non-timber forest resources. I think it's important that we continue to have that kind of access to those values, to those activities, to those areas within our own traditional lands. So that's why. I don't see it as a negative to have areas where there have been intentional burns. However, for the Euro-Canadian participants, the appeal of the pre-industrial condition is less abstract. Both participants were anecdotally aware that the First Nations of the region had traditionally used fire to manage berry crops, but their assessment was predicated more on the condition's potential to provide browse for bears and other large mammals. In addition, they felt that berry patches in various parts of the forest would enhance the quality of their hiking experiences as well as provide a further inducement to enter the forest during their leisure time: E C 1 4 : I think it would be pretty cool to collect a few berries on a Sunday hike. E C 1 5 : we like harvesting berries when we head out into the bush, at least when you come across a good patch or two. It would be great to see more of them on trails or beside logging roads. Based on their initial reactions to the conditions, there is a clear sense that the cluster two participants value the integrity of natural ecosystems and generally equate human management with environmental damage. Under ideal circumstances, the landscape ought to be left alone and be allowed to function as a natural ecosystem, that is, bereft of human interference and use. Overall, the following factors appear to summarise the cluster two participants and their evaluations of the landscape conditions: • Wildlife and the integrity of natural ecosystems are paramount in their assessment of forest related impacts - i.e. the forest exists independent of human use and benefit, and its integrity ought to be preserved. 205 • Care and attention to natural processes is reflected in efforts to manage the landscape using irregular, naturalistic patterns. • Forest related uses are limited primarily to recreational and nature appreciation (e.g. hiking) activities. • A general pattern of landscape condition preferences: • Preferred: Natural • Moderately Preferred: Multi-Use, Pre-lndustrial • Least Preferred: Industrial T o the extent that the cluster two participants appear to elevate the requirements of the natural environment to center stage in their thinking, I have chosen to characterise them as biocentric. However, their value orientation is clearly complicated by the 'modern pragmatism' that compels them to recognise that communities cannot function without some form of resource development. Moreover, their own recreational needs or human-centered use values clearly played some part in their choice of the pre-industrial condition as a preferred treatment. The observation that the cluster two participants do not fit squarely within the biocentric grouping reflects the conclusions reached by various authors (Bengston, 1994; Schindler, et al., 1993) that people generally fit somewhere between the biocentric-anthropocentric continuum in the sense that most of us embody different mixtures of the two extremes. Thus, it is generally expected that individuals will be more or less oriented to either one of the two value-orientations. c) Participant Cluster Three For the participants that I have identified as belonging to cluster three, they are arrayed as a group at the bottom right corner of the M D S output (Figure 5.4) and, from their responses to the conditions, they appear to evaluate landscape change in terms of multiple and interrelated impacts, as well as describe themselves as emotionally and materially connected to the land. 206 Figure 5.4: MDS Results, Cluster 3 Object Points ( ® ® • ® Cluster 3 \. ® 3 T The number adjacent to each data point indicates the number of participants represented by each data point. For many of the cluster three participants, forest industry employment has been a significant part of their lives. However, the range of forest uses identified by the participants in this cluster as significant to their way of life is much broader than those described by the participants in clusters one and two. The following statements reflects the degree or breadth of forest dependence that is typical of this cluster: FN 1: I was raised on this country in the bush, hunting and trapping and doing things ... and when I was falling I used to tell our bosses keep clear of this creek, there's fish coming up here. 'Oh, don't worry about it,' they'd just run their skidders right over it. 207 Consistent with clusters two and four (described in the following subsection) is the socio-demographic heterogeneity of this group. Of the nine people that comprise the cluster, six participants are First Nations and three are Euro-Canadian and, in terms of age, six participants are belong to the elder cohort and three are baby boomers. Among the nine members, there is only one female member of cluster three. It is interesting to note that this individual has lived a remarkably varied life. In addition to being an elder and having witnessed first-hand the Gitksan practice of berry cultivation with fire, she has worked as a tree planter as well as a mill worker, and continues to harvest food and medicinal plants from the forest. Consistent with their female counterpart, the male members of cluster three have led a similarly varied lifestyle, with the minor exception that they are active subsistence hunters. In other words, hunting for these individuals is not an occasional activity that is undertaken for leisure, it is a vital part of their existence to the extent that it supplements their diet, as well as provides for an extended network of family members who may be incapable of providing (i.e. by hunting or purchase) meat and fish for themselves: F N 1 : Sure I still go out, even at my age with my son. W e need the meat for ourselves and the rest of our family, otherwise we can't afford to eat. F N 8: Hunting's an important part of our way-of-life. Whatever we bring in we share with our family, especially with the old people, they can't get out for themselves anymore you know. E C 1: Hunting's a huge part of the way people live up here, a huge part, for whites and Natives. How else are you going to feed a family if you've just been laid off or the mill's been shut down? 208 T he subsistence needs of the cluster three participants' were a dominant factor in their rejection of the industrial condition, as well as in their qualified acceptance of the multiple-use condition. In terms of the former condition, the participants felt that the industrial harvesting that had taken place over the previous twenty to thirty years had detrimentally affected their use of the forest as a food source: F N 8: The way forestry's done things here, I'm surprised we have anything left. W e have to go further and further out just to find good hunting. F N 1: I guess I went out hunting yesterday, my oldest son and I, and just looking at it it's, anything that's good has just been ploughed over with a bulldozer. F N 5: Forestry got into a lot of trouble for doing that, and then they went at it for almost twenty years doing that and it didn't work. . . . The thing is that up there it makes a hell of a difference if you're a hunter. A n d there's a lot of animals in there. Although they could clearly identify the intention or assumptions that were implicit in the multi-use condition, the cluster three participants characterised the reserve and riparian leave treatments as ineffectual, arguing that windthrow would virtually eliminate the reserve patches and produce a situation which is of little benefit for wildlife and loggers alike: F N 9: I don't like them, because the patches, a lot of them end up to be blow down. . . . Because there's no protection, eh? S o a lot of the patches, you could see them, you go out and see them, they're just blowing down because they're uprooted and the winds are too heavy. So it causes a lot of blow down, and there again it's too tough on our animals. F N 5: Most of this will just wind up falling over anyway. W h y would you want to waste all that timber? 209 In a similar manner, the participants' assessment of the natural condition was based largely on a simultaneous consideration of the condition's tradeoffs between wildlife habitat and economic potential. Whereas all of the participants in cluster three described the natural condition as a preferred condition for the region's forests, that condition was seen as a necessary but intermediate step to a more sustainable forest. In effect, allowing the forest to revert to a natural state was seen as tantamount to allowing the landscape to heal the damage of the last three decades, following which, a selective model of timber harvesting could be practiced that would provide wildlife habitat as well as economic benefits for the community: F N 2: This would be good, that way they could go in and do things better, a little more selectively. F N 9 : See with this . . . we could tell them where they might be able to log, and we can tell them areas where they shouldn't log. . . . But it's just beautiful like this, and it's the most perfect area for trapping. F N 5: And that's why we want to push it like that because, one of the main things now are the moose are coming back, they were disappearing for a while when all these big areas were logged off. But they're growing back in now so the moose population's coming back. Of the four conditions, however, the participants in cluster three provided the most interesting reactions to the pre-industrial treatment. Given the prevalence of elders and self-identified subsistence users in this cluster, it was expected that the burn treatments would elicit a clear expression of support. For one sub-group of cluster three participants (Figure 5.4, data point '6'), the condition represented the destruction of commercially valuable of timber resources for a single, albeit culturally important forest use. For these individuals, utilising 'traditional' burn sites or naturally burnt locations made more sense than burning new berry patches. 2 1 0 Moreover, they felt that greater value could be derived from the forest if timber management could be reconciled in some way with berry production: F N 9: I'll say no to that myself for one reason only . . . if it's accessible, we'll go there and pick the berries . . . [but] if this were a burn and it was logged off, well then it should be replanted. A n d we'll pick the berries there until it starts growing again. E C 7: I think a better way of doing this would be to log it and then manage for berries. W e need to stop managing our land base as though it should be producing a single crop. E C 3: Sure but I like the way this picture is going (i.e. multi-use condition). They should try and find a way of managing the forest for timber and other resources. That would probably be way better for the local economy. The salience of timber management to this sub-group of cluster three participants may be due to the fact that they remain actively employed in forestry. For the second sub-group (Figure 5.4, data point '3'), employment is less of a concern than the need to safeguard vital hunting and trapping locations. In addition, the First Nations elders in this sub-group had witnessed as children the Gitksan practice of burning for berries, and their concerns about the pre-industrial condition centered on the appropriateness of the physical and spatial attributes of the burn treatments (see also section 5.2.3). My efforts to approximate traditional burning practices were clearly deficient to the extent that I was told in remarkable detail how the 'old people' would actually use fire to create berry patches: F N 4: No they would have burnt down along these ridges like this here. They would never have burnt across the hillside like that. F N 1: No they would have been smaller, if they're going to pick any huckleberries, but they've got to see the huckleberries first before they burn it. 211 F N 5: I don't recall them doing it like this. They would have been higher up on the mountain and away from these areas in here where goats like to browse. This information is significant in itself to the extent that it has been documented only in a handful of other sources (Johnson, 1997; Trusler, 2002; Turner, 1991). However, it also points to the significance and relevance of intersubjectively acquired, culturally based knowledge as a factor in the evaluation of landscape change. Although most of the participants in cluster three continue to harvest berries and other ethnobotanical products from the forest, most rely on naturally occurring fires to create or maintain productive berry sites. With the exception of one participant who continues to maintain (i.e. regularly burn) his own berry patch, largely because it is located on reserve land, the understanding of how and where burning was practiced is mostly socially acquired and indirect. From what has been described, I have chosen to describe the participants in cluster three as 'ecologistic.' There is a strong undercurrent of human-nature interdependence in their evaluations of the landscape treatments, and this is articulated most clearly in their desire to see land management conducted in a way that is "respectful." Respect or the degree to which the conditions failed to reflect respectful land-use was a common refrain among the cluster three participants: F N 1: Y o u got to be respectful of the fish or the game you get . . . and we got to respect the trees. Now we moved out of that logging base where her dad had his logging camps, now it's all clearcut. Just nothing. E C 7: We've got to respect what we've got here, otherwise we'll have nothing left. What we do to the forest affects everything from timber to fish and hunting. 212 E C 3: ...treat them with respect and say, well, this is a special patch of plants and maybe they'll flag it and say where these plants grow, leave that. In effect, humans and nature exist in a relationship of mutual dependence, but it is the responsibility of the former to ensure that sufficient resources are left behind to provide for the needs of other living communities and future generations. This contrasts in some way with the dualistic views expressed by the participants in clusters one and two whereby humans and nature exist in a hierarchical relationship and, in some ways, nature is separable from human existence. For the anthropocentric cluster one participants, care and attention is given to the landscape as long as it sustains something we might need, even if only because it is aesthetically pleasing. For the cluster two participants, on the other hand, nature is not worthy of care because it benefits humanity, but because it is purer and stands above humanity. Although there remains a strong sense that humans stand to benefit materially from the landscape, the line between nature and culture for the cluster three participants is blurred. Moreover, because of this blurred distinction, cluster three participants are inclined to argue that nature should be treated by people as carefully as they do their home places, tending or taking from it as needed. From the preceding discussion, the participants in cluster three appear to share the following characteristics: • Wildlife and the integrity of natural ecosystems are important to the extent that they sustain a broad range of important human forest uses. • They are sceptical of efforts to display care and mimic natural processes using irregular landscape patterns. Care is demonstrated through the selective removal of timber, thus demonstrating 'respect' for other forest users, both human and non-human. 213 • Forest related uses are more extensive than the cluster one and two participants and include employment, subsistence hunting and plant collection, fishing and recreation. • A general pattern of landscape condition preferences: • Preferred: Natural 2 • Moderately Preferred: Multi-Use, Pre-lndustrial • Least Preferred: Industrial d) Participant Cluster Four T he participants that have been grouped into cluster four (Figure 5.5) are similar to the cluster three participants in that they utilise the forest for a variety of purposes. Hunting, plant collection and fishing are integral facets of their way-of-life and, much like their counterparts in cluster three, the value of these activities as a mode of subsistence is significant: F N 15: Yeah , I still go out hunting a lot. A few of us went out just the other day and came back with a couple of moose. I don't need that all for myself though. A lot of it I give to close family members who can't get out for themselves. F N 14: Well, hunting sure saves me and my family a lot of money. There's not too many hunters in my family, so the game that I get I usually bring it over to my grannies and we split the bounty. People usually help in with the butchering and the drying, so it's really important to go hunting for me. And , you know in Kitwanga, the community's 95 percent unemployed and it really helps them out. I usually get two maybe three moose a year, but that's split between three, well two families. E C 1: It's pretty darned important for most of the people that live up here, not only for myself. You'll see a lot of First Nations and non-Native people put in the bush at this time of year hunting, gathering mushrooms and generally providing for themselves that way. A n d it's 1 6 While the Natural condition was the most preferred of the four conditions, for this cluster is was generally seen as an interim step that would allow the current forest landscape to heal from past damage. Ultimately, the 'ideal' future condition for these participants would be a landscape where the forest is selectively harvested for timber and used for a range of additional purposes - e.g. subsistence, recreation, etc. Although this conception of the future is by definition multi-use, the actual multi-use condition that was presented to them was given qualified acceptance, largely because the condition did not go far in its approach to retention. Thus, I addition to the Natural condition, this group may have preferred a 'lighter touch' multi-use condition. 214 not just a way-of-life, it's a necessity when you're out of work, or not earning enough to feed your family. In terms of the cluster's socio-demographic composition, among the fourteen members of this group, there is an even balance of seven First Nations and Euro-Canadians; six females and eight males; as well as three elders, nine baby boomers and two Generation-x members. Thus, in addition to being the largest of the four Figure 5.5: MDS Results, Cluster 4 Object Points <8> ® ® ® Cluster 4 ® ® The number adjacent to each data point indicates the number of participants represented by each data point. clusters, this cluster is perhaps the most diverse and reflects the degree to which ethnicity, gender and age overlap in evaluations of landscape condition. Th e one aspect in which the cluster four participants differ from their counterparts in cluster three is their record of forest-based employment. While the latter indicated that they were currently engaged in forestry related work, or had 215 been employed by the forest sector throughout much of their lives, the cluster four participants indicated that forestry employment has been a relatively insignificant component of their employment histories. In this one aspect, the cluster four participants were more similar to cluster two, or the biocentric group, and departed from cluster three. Nonetheless, from their assessments of the industrial and multi-use conditions, the breadth of wildlife, aquatic, visual and economic impacts that they considered in their evaluations was similar to the cluster three participants and perhaps reflected their multifaceted relationship with the forest landscape: E C 2: I think they've been lucky that we didn't have a lot of flooding, erosion and that kind of stuff on this mountain. . . . I'm not totally against logging, and I know there's some sort of economic basis for clearcuts. But I still think they should be taking more care in how they do it. . . . It's just rotten and not just aesthetically it's, I imagine it changes a lot of things. F N 15: it's definitely exposed it quite a bit especially in terms of the wildlife. You know, it doesn't provide the wildlife with the shelter that they require. . . . It's created a lot of erosion, it's dirtied up the streams that were once salmon bearing streams, it's warmed up the creeks and the fish don't like the warm water too much. E C 13: Well they leave those little islands in there, and soon they're gone. There's the wildlife, and then it's just the visual thing. Leave sixty percent of the stems there and I'd start to like it. Based on their explicit preference for the natural condition, the cluster four participants are comparable to cluster two or the biocentric group. Although the cluster four participants discussed the importance and need for forest management in the upper Skeena landscape, generally expressed in terms of a common preference for selective forestry, their clear preference was for a 'naturalistic' landscape that would provide wildlife habitat and, in turn, support critical subsistence activities as well as economic development ventures such as eco-tourism: 216 E C 12: Well you like to think of it both ways, you like to think of it as a place where the animals can still survive, things are healthy. F N 15: For the tourism potential as well. . . . I think we need, the trees are, the trees and natural scenery are much more valuable to us in the tourism aspect that it would be for a few people to line their pockets off of it. F N 11: . . . it doesn't affect the wildlife . . . Like I got a couple of moose up there, we got one on Sunday and one on Tuesday. There wouldn't have been a chance of us getting them if it wasn't like this. While clusters two and four prefer natural conditions, the conceptual basis for these preferences is not identical. For the former group, a natural forest is desirable in large part because human management is equated with interference and destruction. For the latter group, however, people are viewed as part of the landscape, subsisting from it and ideally managing the environment in such a way that it provides a continuous flow of benefits over time: FN14: I would go along the lines of . . . single stem, select harvesting for many a purpose and some of it would be to have sustainable logging rather than just clearing out the area, leaving the roads to erode and ensuring there will be something for future generations to harvest as well. A s such, I have chosen to characterise this group as 'naturalistic' in order to underscore their preference for a landscape that is more or less intact and unspoiled, but to distinguish them from their biocentric counterparts to the extent that the environment is not regarded as something that is separate and above humanity. In effect, the cluster four participants regard the landscape as something that ought to be utilised, but used within the limits of natural patterns and processes. In a sense, therefore, both clusters three (i.e. 'ecologistic') and four ('naturalistic') concurrently value the landscape in terms of the intrinsic and use values described 217 in Chapter 2, however, the latter cluster is inclined to a greater degree towards intrinsic environmental values. There were, however, noteworthy preference differences among the participants in cluster four, particularly in relation to the pre-industrial condition. A m o n g the group that gravitated spatially in the M D S output towards cluster two (Figure 5.5, data point '9'), there was qualified support for the condition. Among the First Nations participants in this sub-group, restoring a traditional form of land-use appealed to their sense of identity and echoed the comments expressed by the First Nations participant in cluster two. However, for both the First Nations and the Euro-Canadian participants, berry management would provide an additional land-use that would potentially have tremendous health, cultural as well as commercial benefits for the local community: F N 7: This burn is not too bad . . . How close to the settlement was this? It would make sense. If you compared it to the population that this was trying to feed it would make sense. Then it probably would be ok. F N 15: I think you have to consider the botanical forest products as well as being a resource. It's a naturally producing resource that local people, and not just local we got a lot of outsiders coming in to pick the botanical forest products to sell, to assist their annual income. F N 10: . . . in terms of harvesting berries you don't see very many people of my age going out and harvesting berries, so we definitely got to keep that going to remember who we are and who our ancestors were and how they lived, rather than just read it out of a book. For this sub-group, berry management has ethnic as well as material or economic benefits, and it is perhaps the latter factor that serves to separate this sub-group from cluster two. 218 For the second sub-group, which is situated in closer proximity to cluster three (Figure 5.5, data point '5'), the value of berry management was acknowledged. However, their rejection of the condition was based upon two factors. First, the First Nations in this sub-group criticised the logic of siting and sizing the burn treatments in the manner that was represented by the simulations. Like many of their counterparts in cluster three, these participants could not draw on direct experience to inform their judgements of burn treatments, but they do appear to have a pool of knowledge that comes to them from elders and other community members that have engaged in the practice: F N 6: I'm not sure they would have done it this way. First off, it looks so big, but wouldn't they have burnt further up the mountain and have avoided these slide areas. I don't know. I'm just basing that on what I've been told by my parents. F N 13: Just from what I've heard from more knowledgeable people in the community, elders and whatnot, the burns would have been narrower and come down the mountain like this. However, the participants in this sub-group, both First Nations and Euro-Canadian, also addressed the need to consider the wildlife impacts that may result from burn treatments that are as large as those depicted by the simulations. Like their counterparts in cluster three, this sub-group within cluster four expressed their concerns for a resource that is a critical facet of their material livelihood and cultural way-of-life: E C 6: Y o u have to be careful in terms of how you do this. There's more to look out for in here than just berries, there's wildlife, fish habitat, and so on. F N 10: I could see this being good for bears and other animals that feed on these plants. But bears are pretty territorial, we have to respect their needs. 219 F N 11: I think that all this stuff needs to be done with care . . . because of things like mushrooms and medicinal plants and all that kind of thing. Cluster four is unique among the M D S groupings that have been described in that: • Wildlife and the integrity of natural ecosystems are important to the extent that they serve both natural (i.e. intrinsic) and human (i.e. use) values, but with a partiality towards natural processes. • They are sceptical of efforts to display care and mimic natural processes using irregular landscape openings. Care is demonstrated through the selective removal of timber, thus demonstrating 'respect' for other forest users, both human and non-human. • Forest related uses are extensive and include subsistence hunting, fishing, food and medicinal plant gathering, recreation, and to a limited extent, employment. • A general pattern of landscape condition preferences: • Preferred: Natural • Moderately Preferred: Multi-Use, Pre-lndustrial • Least Preferred: Industrial T o this point, however, I have not discussed how clusters three and four described their conceptions of visible care or stewardship in managed landscapes. While clusters one and two equated care with landscape patterns that emulated natural disturbance patterns, clusters three and four expressed the need to see patterns that reflected their conceptions of 'respect.' In the following discussion, I address the following issues based on the pattern of findings: • What respect encompasses for these groups and present some thoughts on how the N E P - D S P continuum may be revised to reflect the value orientations that have been described thus far. • The section concludes with a reflection on the heterogeneous (i.e. ethnic, age and gender) composition of the participant clusters and what this means for inter and intra ethnic dimensions of landscape preference. 220 5.3.2 Pattern of Findings: Research Question 2 a) The Value Orientation Matrix The N E P - D S P continuum was conceived as a framework that would explain human-environment relationships in terms of two opposing and prototypical categories - i.e. anthropocentrism and biocentrism. Most theorists contend that the majority of people fit somewhere in between, and that in the interstitial space along the continuum are orientations that are more or less representative of actual human environmental values. From the data that I have obtained from the upper Skeena Valley participants, I would like to posit that there are other value orientations in addition to the biocentric-anthropocentric framework and, moreover, that environmental value orientations may not be best arrayed along a single axis or spectrum as suggested by the N E P - D S P continuum. Based in part on the graphic results of the M D S analysis, but more importantly on a qualitative interpretation of the participants' comments, I believe that environmental values may be conceptualised in terms of a two-dimensional matrix with the anthropocentric-biocentric dimension configured along the x-axis, and an additional y-axis that I have termed the hierarchy-interdependence dimension (Figure 5.6). 221 Figure 5.6: Environmental Values Matrix Object Points Cluster 2 Biocentric ® ® ^ Cluster 1 Anthropocentric ® ® ® Cluster 4 Naturalistic ® ® Cluster 3 Ecologistic ® u c <D •o c (U a <D T3 a> c I o b. (0 k_ 0) to 'x < Cluster 3 Ecologistic Cluster 4 Naturalistic Cluster 1 Anthropocentric Cluster 2 Biocentric X-Axis: DSP-NEP The clusters in the original data plot (top) have been inverted so that Cluster 1 is displayed at the origin (bottom plot). The relative position of the clusters in the matrix has not changed. This bottom configuration better reflects the original DSP-NEP continuum shown in Figure 2.3. In terms of the x-axis, the anthropocentric and ecologistic participants conveyed a clear sense that natural resources exist for human benefit and use (i.e. a high human use value), and that landscapes are worth protecting insofar as they sustain something that we may need. This was in part expressed by the moderate preference expressed by these participants for the multi-use condition, as well as by comments where human needs and utilitarian values figured prominently in their assessments. For the naturalistic and biocentric participants, on the other hand, nature is better to the extent that the natural condition was clearly preferred by the participants in both clusters and by the suggestion that 'management' means letting the landscape find its own way, preferably with no human influence or, at the very least, with low impact. Yet, what seems to distinguish the ecologistic and naturalistic participants along the y-axis is their conception of humanity's embeddedness in the natural order. For the participants at the bottom of the y-axis, people and nature exist outside of one another. In effect, for the participants who were either strongly anthropocentric or biocentric, human society is something that is independent of and acts upon the natural world (either for its benefit or destruction), which exists 'out there.' At the top end of the y-axis, however, there is a stronger sense that ecologistic and naturalistic participants take a more complex view of the world and suggest that people exist within a network of human-environment relationships, rather than in a significance-based vertical hierarchy. There are several strands of evidence that I have drawn upon to make this observation. First, participants along the interdependence (top) end of the y-axis 223 described how they exist in a relationship of shared dependence with the environment and discussed both the tangible or physical consequences of land-use change, as well as the affective or emotional impacts that they may feel as a result of landscape alterations. Typically, interdependence participants talk of 'belonging to' the landscape and being 'part o f or 'connected to' something that transcends themselves: FN1: Most of the elder people as myself, it hurts you when you see that. It hurts you because it's all just charcoal. Y o u see the animals feed on these roots, and everything else. FN1: I was raised on this country in the bush, hunting and trapping and doing things. And you see something like this, well that breaks me up inside because if you go trapping in there, there would be nothing. F N 15: It definitely does make me feel bad, and especially with some of the other resources out there that have been disturbed, the cultural heritage resources and with the berry picking there... . FN7: It looks like a big festering sore, and it feels like your heart's ripped out. But that's how I always felt about, after I got involved in all the blockades and trying to stop the logging so that they would do it selectively. E C 2 : Well I just don't like the look of it. I guess if you look at things as if you're connected to it, I don't know if other people go around with these blinkers that maybe I have, is that if you're connected to everything you feel what that feels like. In addition to the prevalence of comments that described deep emotional and physical connections with the landscape, interdependence participants recognised the need for landscape management. However, perhaps owing to their conceptual framework that everything is interconnected, these participants often spoke of conducting forest management in a manner that is more 'respectful' of other users (e.g. present and future generations) as well as other species: 224 FN4: I don't like it because it just seems so disrespectful. See , the mountain is an entity, that's what people forget. This is alive, the same as you or I alive as a person, this mountain is an entity. E C 2 : I really think that all this multiple use stuff needs to be done with care because even the thinning where they do selective logging because of things like mushrooms and medicinal plants and all that kind of thing . . . and treat them with respect and say, well, this is a special patch of plants and maybe they'll flag it and say where these plants grow, leave that. FN1: but we got to respect the trees. Now we moved out of that logging base where her dad had his logging camps, now it's all clearcut. Just nothing. A n d yet within a few years, those small trees are growing big enough to recut again. While many of the hierarchically oriented participants were satisfied with landscape treatments that either mitigated visual impacts or emulated natural disturbance patterns, the interdependence participants sought reflections of care that were more substantial. For the latter group, retention patches such as those depicted in the multi-use condition were characterised as superficial exercises in visual management that provide no real benefit to wildlife and would ultimately result in the waste of commercially valuable timber: FN9: I don't like this because of the windfall . . . it's just harder on everybody. It's harder on the animals, I mean you're just ruining all the timber when you log like this. FN8: I've worked in this area for tree spacing and stuff like that you can see the amount of trees that are blown over after it's logged like that. These trees aren't going to last too long if you log it off like that. FN5: It's even worse up in the Suskwa, it's just decimated, one block after another. They say they'll leave a little leave strip, but it's not much of a leave strip. It's similar to what they've done here, that's not much of a leave strip. FN7: A n d my uncle . . . would say, why do you leave these sacrificial trees in the middle there, they're going to get blown down anyway? 225 Why bother? Because that's just stupid. That's what he would say because he was a logger as well. F N 15: ...these ones are just sacrificial trees, the wildlife corridors. E C 2 : I'm not totally against logging, and I know that there are some sort of economic basis for clearcuts. But I still think that they should be taking more care in how they do it. I mean that kind of thing is a terrible scar. A n d it is. It's just rotten and not just aesthetically it's, I imagine it changes a lot of things. However, among this group of participants, respectful timber management did mean harvesting trees more selectively - i.e. using a dispersed retention model - typically to provide sufficient cover for wildlife and, in effect, ensuring that enough of the landscape is left intact to ensure that other living communities can survive. Further, respectful management required paying attention to the location of timber harvesting in relation to sensitive landscape features and processes. In the case of culturally significant plant communities, this meant ensuring that sensitive features such as medicinal and food plant sites are not situated downstream from harvest areas: FN9: when it rains, a lot of it slides, and a lot of the sand and mud goes into the rivers. It takes longer to grow back, and then they wipe out everything, and then a lot of it is just turned into waste. FN2: That's what we look for when we go picking our Indian medicine . . . where there's no logging. So you go down into the foot of this, that's where it grows the medicine. FN15: You're dealing with runoff heading into the water and impacting the salmon. And I would go along the lines of single stem, select harvesting for many a purpose and some of it would be to have sustainable logging rather than just clearing out the area . . . and ensuring there will be something for future generations to harvest as well. E C 5 : ...this is all pretty drastic up in here. I really think that all this multiple use stuff needs to be done with care because even the thinning where they do selective logging, because of things like 226 mushrooms and medicinal plants and all that kind of thing, they're going to be impacted by the runoff coming from the logging up in here. FN7: If you have foodstuffs down there and you have logging show up there, what about all the toxins coming off here? Going into your food-source? It doesn't make any sense. You can't have both and have pristine foodstuffs. This more complex conception of human-environment relationships may be a product of the multiple uses through which the interdependence participants encounter the natural environment. A s discussed throughout this chapter, the anthropocentric and biocentric participants experience the landscape through a narrow range of activities. For the former, the forest is a source of employment and material livelihood, while the latter described their primary use of the environment in terms of recreational pursuits. The ecologistic and naturalistic participants, on the other hand, experience the landscape through a variety of lenses - i.e. as hunters, plant collectors, trappers, recreationists, forestry workers, etc. - and thus may be more attuned to the multiple benefits that the forest provides, as well as perhaps how landscape processes interact and are affected by change. T o borrow a concept from social network theory, the ecologistic and naturalistic participants are more 'multiplex' in their relationship with the landscape. T o the same extent that people's attitudes and behaviours are strongly shaped by the complex interaction of many simultaneous or 'multiplex' social influences (Gluckman, 1967), human environmental values may be shaped by the complex interaction of multiple sources of environmental information. By contrast, 'uniplex' human-environment relationships provide a limited range of experiences, and environmental values will likely reflect the character and range of those experiences. 227 Although it would be difficult to argue that the data presented conclusively demonstrates that this pattern of environmental value orientations exists in reality, the data does however suggest that it is plausible. While social network research has been conducted to explain differing environmental attitudes and beliefs, I believe that future research needs to be conducted that examines patterns or networks human-environment interaction as determinants of environmental values and preference. A more detailed discussion of this future research is provided in the concluding chapter. b) Inter and Intra-Ethnic Dimensions of Landscape Preference There are three observations that I would like to make about the ethnic dimensions of landscape preference in light of the discussion presented in section 5.3 so far. First, the results demonstrate the heterogeneous composition of the participant clusters, both in terms of ethnicity as well as other sociodemographic attributes - i.e. age and gender. What is clear from the preceding discussion is that, with the exception of cluster 1, which is exclusively comprised of Euro-Canadian males, clusters 2, 3 and 4 encompass both First Nations and Euro-Canadian participants. A simple graphic breakdown of the participant clusters by ethnicity underscores this observation (Figure 5.7a). However, it is interesting to note that Figures 5.7b and 5.7c demonstrate a similar heterogeneous pattern of results when the participant clusters are displayed by age and gender. First, to the extent that the elder participants cluster towards the ecologistic dimensions of the M D S plot, this may reflect an age-related preference for landscape condition that is based on a greater tolerance for change or visible modification, in large part due to this 228 generation's historic involvement in the forest industry. In contrast, the Generation X participants gravitate towards the naturalistic and biocentric dimensions of the M D S plot, perhaps due to their relative detachment from the forest industry for their material livelihood. In a similar manner, female participants cluster towards the naturalistic and biocentric dimensions. With the exception of the one female participant in the ecologistic cluster who has worked in the forest industry, this pattern of results may support the argument that preference is a product of perceived functional importance. In other words, elder male participants and one female participant appear to be sensitive to the functional or purposive aspects of landscapes to the extent that the landscape can facilitate direct material or employment needs. Returning to the more central question regarding ethnicity does the mixed distribution of First Nations and Euro-Canadians in the M D S plot demonstrate that ethnicity has little to do with landscape preference? It is interesting to note that 13 of the 14 First Nations participants (or 87% of the F N sample) gravitate towards the interdependence end of the M D S plot, while only 10 (or 66%) of the Euro-Canadian participants are similarly inclined. Perhaps this reflects a greater tendency on the part of First Nations, through culture and traditional patterns of resource use, to be connected to and dependent on the land in multiple ways (e.g. subsistence), which are less characteristic of Euro-Canadian society. However, it is difficult to conclude that ethnicity accounts for this difference, in part because of the small numeric disparity between the two groups (i.e. 13 and 10), as well as the other observations that I have made. 2 2 9 Figure 5.7a: Participant Clusters by Ethnicity EC (2), FN (1) EC (1) ST EC (1) EC (1) EC (4), FN (5) U S EC (2), FN (4) EC (3), FN (2) «r~| EC (1), FN (2) FN = First Nation, E C = Euro Canadian. Numbers reflect the number of participants in each ethnic sub-group. Figure 5.7b: Participant Clusters by Age E (2), B B (6), E ( 1 ) , B B (3), GX (1) E(3) GX = Generation X, BB = Baby Boomer, E = Elder. Numbers reflect the number of participants in each age sub-group. 231 Figure 5.7c: Participant Clusters by Gender F ( 1 ) , M ( 2 ) F (5), M (4) F ( 1 ) , M ( 4 ) M (1) M ( 1 ) M (1) M ( 1 ) z1 M (6) <gT~| F ( 1 ) , M ( 2 ) M = Male, F = Euro Canadian. Numbers reflect the number of participants in each gender sub-group. 232 My second observation posits that participant preferences transcend distinctions of ethnicity and, moreover, that this may be due to the pattern of human-environment relationships discussed in the preceding sub-section. In effect, whether an individual is First Nations or Euro-Canadian appears to have less influence on their landscape preferences than if: a. They engage or use the landscape in a multiplex or uniplex fashion (i.e. the scope of forest-based uses), and b. The kinds of benefits (e.g. material, recreation, identity, spiritual) that they derive from the landscape and, more importantly, the relative importance of forest-based uses and benefits. For instance, Table 5.4 summarises the socio-demographic composition of the participant clusters, how they interact with the forest, and their preferences for the landscape conditions. The Euro-Canadian and First Nations participants in clusters three and four (the ecologistic and naturalistic participants, respectively) were unified in their preference for the natural condition as a vital means of restoring or healing the landscape following three decades of destructive timber management practices. What appeared to unify the participants in these clusters was a common view of the forest as a source of subsistence from hunting, trapping, food and medicinal plant collection and fishing. Multiplex subsistence uses of the forest are common to both ethnic groups, and were important factors in their evaluation of the landscape conditions. Clusters three and four, however, differed in their assessment of the pre-industrial condition, and it is here that the relative importance of forest uses appears to affect preference evaluations. Given the greater importance of forest industry employment to the participants in cluster three, the pre-industrial condition was 233 viewed, much like their counterparts in cluster one, as a threat to valuable timber resources. Most participants in this cluster felt that rather than burning a portion of the landscape to produce a single product, berry management and other non-timber forest products could be reconciled with timber management to produce multiple economic and subsistence benefits for the local community. For the participants in cluster four, the pre-industrial condition was viewed as a preferred means of escaping from the region's traditional dependence on the timber economy, and expanding the range of subsistence and economic benefits drawn from the forest. The fact that most of these participants do not rely on the local timber economy for their material well-being may account for their 'non-timber orientation' and greater preference for the pre-industrial condition. 234 Table 5.4: Summary of Cluster Preferences Cluster (N) Socio-Demographic Forest-Related Uses Condition Preferences 1: Anthropocentric (4) Ethnicity: 4 EC Gender: 4 M Age: 4 BB Uniplex: Employment, Recreation. Preferred: Multi-Use Mod. Preferred: Industrial Least Preferred: Pre-lndustrial, Natural 2: Biocentric (3) Ethnicity: 2 EC, 1 FN Gender: 2 M, 1 F Age: 2 BB, 1 GX Uniplex: Recreation Preferred: Natural Mod. Preferred: Multi-Use, Pre-lndustrial Least Preferred: Industrial 3: Ecologistic (9) Ethnicity: 3 EC, 6 FN Gender: 8 M, 1 F Age: 6 E, 3 BB Multiplex: Employment, Subsistence Hunting, Plant Collection, Fishing, and Recreation. Preferred: Natural Mod. Preferred: Multi-Use, Pre-lndustrial Least Preferred: Industrial 4: Naturalistic (14) Ethnicity: 7 EC, 7 FN Gender: 8 M, 6 F Age: 3 E, 9 BB, 2 GX Multiplex: Subsistence Hunting, Plant Collection, Fishing, Recreation, and Employment. Preferred: Natural, Pre-lndustrial Mod. Preferred: Multi-Use Least Preferred: Industrial Sociodemographic Key: Ethnicity - EC (Euro-Canadian), FN (First Nations) Gender - M (Male), F (Female) Age - E (Elder), BB (Baby Boomer), GX (Generation X) Note: Forest-related uses are listed in the order of priority or importance expressed by the participants. r o c o c n Third, should these observations be interpreted to mean that ethnicity has nothing to do with landscape preference evaluations? I do believe that purpose-based, human-environment interactions significantly affect what we look for in a landscape particularly in terms of how that landscape can fulfill human needs. In this sense, I think that the pattern of findings provides some support for the Landscape Preference Model summarised in the place dependence discussion in Chapter 2 (see Figure 2.4, page 52). However, I also believe that ethnicity provides an additional set of referents that influence preference. For instance, among several of the First Nations participants their preference for the pre-industrial condition, or the burn component of the multi-use condition, was affected by their appreciation for the aboriginal berry management using controlled burns as a traditional and culturally-based form of land management. While the burn patterns depicted in the landscape simulations may not have been consistent with their understanding of how berry management was carried out by their ancestors, the use of the land for traditional (i.e. Gitksan) food products resonated with their sense of this practice as a part of their identity: F N 15:'re taking one resource that may not be of value to you and you're creating resources that you require to live in the area. A n d one of them is berries, and that can go culturally, but if you look back a little bit more with the harsh elements that we're faced with nine months of the year, you definitely require the berries... . F N 12: Well it's another, at least there's another forest use . . . it's cultural and it's cultural in that the Gitksan, and in other areas the Wetsuweten, have through the generations made use of non-timber forest resources. I think it's important that we continue to have that kind of access to those values, to those activities, to those areas within our own traditional lands. 236 F N 14: I like this one because it kind of shows the rotational burns. Yeah , I know they moved around a lot when they were burning for berries. And it kind of looks here like they've done that . . . that they've tried to do it using traditional methods. T o the extent that landscape patterns and uses are representations of ethnic or culturally particular land management practices, and that an ethnic group prefers these patterns of use, ethnicity does appear to play a role in landscape preference. Thus , a slight refinement can be made to the Place Dependence Landscape Preference Model (Figure 5.8). I have added psychological or identity needs to the list of environmental motivations largely to elevate place identity from what is generally regarded in the sense of place literature as an abstract, esoteric conception of human-environment relationships that is distinct from the more tangible and material expressions of place needs. Identity or finding expressions of cultural identity in the landscape may itself be a psychological need to the extent that stories, histories and visible markers of traditional resource use possess a form of power for people like the Gitksan because they are shared, they link people to one another to form a cultural identity that is rooted in a common history and symbolic association that is bound up in the landscape. Over time, these places, the stories and uses that are associated with them combine to become significant cultural talismans and markers of group identity (Lewis and Sheppard, 2005). N. Scott Momaday (1976: 1) suggests that finding identity and meaning in place features is a human impulse or need that has been present in First Nations societies such as the Gitksan for a very long time: From the time the Indian first set foot upon this continent, he centred his life in the natural world. He is deeply invested in the earth, committed to it both in his consciousness and in his instinct. The 2 3 7 sense of place is paramount. Only in reference to the earth can he persist in his identity. Figure 5.8: Place Dependence Landscape Preference Model (Revised) Perceiving Subject Environmental Needs & Motivations • Existence Needs • Relatedness Needs • Growth Needs • Psychological/Identity Needs Intersubjective Information • Cultural, Historic, Economic Environmental Information • Physical conditions Information concerning the motivation or need fulfilling potential of a landscape is derived 1) intersubjectively as people receive information from others in their social network and 2) from the setting itself. In effect, through physical patterns and conditions, the landscape communicates to the perceiver the extent to which it can afford desirable experiences or fulfil human needs. A connection has also been drawn in Figure 5 .8 between the Intersubjective and Environmental information components of the Landscape Preference Model to reflect the interaction between intersubjectively transmitted expressions of identity (e.g. stories, history, traditional uses) and visible landscape patterns that reflect or embody culturally based expressions of identity. In effect, people may read the landscape to find markers of group identity (i.e. environmental information), and their understanding of what these markers signify is largely supported by what the perceiver knows about the setting through intersubjectively transmitted and culturally-based information. Thus , these are complementary rather than discrete forms of information that satisfy the perceiving subject's psychological requirements. 238 I discuss future prospects for exploring the synergistic role of environmental and intersubjective information for landscape preference formation in Chapter 6. O n a final observation regarding identity, it is interesting to note how the participants in cluster one may have been using a conception of ethnic identity in their opposition to traditional berry management. The pre-industrial condition was moderately preferred by this group when it was understood to be a silvicultural technique to control brush and replenish soil nutrients. However, when it was made clear that First Nations traditionally used fire to cultivate berries and other food crops, the statements of opposition were nearly hostile in their tone: E C 8 : I think that it's not two hundred years ago now, that we know enough and there's enough value in that timber that it would be stupid to burn it, just to make some huckleberries. . . . Y o u see I take issue with the idea that anything the Indians do is just automatically good because they're so in touch with the land because they're Indians. . . . A n d it's the same thing like the fishing, there's so many people that have the right to fish with nets, but they can't seem to get it through their heads that they don't need to leave the nets in there after they've got enough fish. . . . There's a lot of people who just leave their nets in there because they're lazy, there's some of them that do that. So it's just, wanting to burn because they're Indians makes no sense to me. E C 11: seems to me that they don't do as much berry picking as they used to. Just from observations though. A n d I'm not wild about them torching big areas like that. I mean, if it was managed and not just some Joe Blow who just decides that he's going to go out and burn off a hillside for a berry patch. E C 10: W h y would you want to ruin all that good timber just for the sake of a few berries? The Indians are just like the rest of us, they need jobs and probably could get away with making a few patches on their reserves. In part, these statements are a response to the potential loss or mismanagement of a resource that this cluster values most, i.e. timber. However, perceptions of "Indians" as less than competent managers or completely acculturated and, 239 therefore, no longer dependent on non-timber products are implicit (explicit in the case of EC8) throughout these statements. In this sense, conceptions of ethnicity and ethnic stereotypes were a significant component of cluster one's evaluation and rejection of the pre-industrial condition. Thus , ethnicity factors into landscape preference evaluations to the extent that land-uses and landscape conditions symbolically represent or embody a group and their relationship with the environment. This can occur reflexively in the case of the Gitksan participants, or it can be imposed by one ethnic group to define appropriate environmental interactions for another group based on inaccurate and outmoded cultural assumptions. This and other findings presented in this chapter are further addressed in Chapter 6. 5.4 Chapter Summary The research findings in this chapter can are presented in the following chapter where a more comprehensive and concise connection can be made between the research questions discussed in Chapter 1, the findings and interpretation of results in Chapter 5, and the implications for theory, practice and future research discussed in Chapter 6. 240 6 Synthesis & Conclusions 6.1 Introduction T he general aim of this dissertation was to gain more insight into the factors that affect people's judgments of preferred landscape change. Two main questions guided this research. Th e first research question asked: "What is the range of dimensions that local stakeholders, including subsistence users, consider in their evaluation of modified forest landscapes?" The second research question sought to uncover the mechanisms underlying individual and cultural differences in landscape preference: "How and why do preferences for modified forest landscapes differ between and among First Nations and Euro-Canadians?" Chapter 2 provided a general overview of theoretical notions and empirical findings relevant to these research questions, while Chapter 5 dealt with findings related to these questions in more detail. This final chapter includes an overview and discussion of the main conclusions that can be drawn from the empirical component of the research. A few basic caveats need to be presented before discussing the findings. The sample of participants is small when compared to most quantitatively based studies, and was purposively selected. A s such, the results cannot necessarily be statistically generalized to other communities in British Columbia (or elsewhere) and should be replicated in another community setting. The visual stimuli used to depict landscape disturbance conditions were limited to three main viewpoints/locations and four disturbance types. Again, future studies should be replicated with a wider range of locations and scenarios. In addition, the possibility cannot be ruled out that, enabled by the semi-structured qualitative process, my explanations of the 241 landscape conditions and questioning may have inadvertently influenced the preference responses. However, given the measures taken to minimise this possibility and, as revealed by repeated examinations of the transcripts, the disclosure of interview questions, and lack of obvious bias in the questioning, this explanation for the preference profiles that emerged seems unlikely. 6.2 Summary of the Main Conclusions Figure 6.1: Summary of Research Findings Research Question 1: Landscape preference evaluations are inherently wicked exercises (Stankey, 1996) to the extent that they are based on a complex and simultaneous weighting of alternative perceived consequences (i.e. Balance of Impacts), the degree to which disturbances demonstrate care or attention to natural processes and future needs, as well as physical site characteristics. Research Question 2: Ethnic and other sociodemographic characteristics only partially account for differences in preferred landscape condition (e.g. age, gender, occupation). People tend to evaluate the landscape from a user perspective and, with a few exceptions; forest-based uses often transcend distinctions of race, age and gender Conclusion 3: People are generally willing to accept modified or human-influenced landscapes, but the degree to which the treatment visibly demonstrates care or attention to the vitality of the forest is a salient factor in the participants' preference evaluations. Conclusion 4: The New Environmental Paradigm-Dominant Social Paradigm continuum identified by Dunlap and VanLiere (1978) only partially accounts for the range of environmental value orientations expressed by the Upper Skeena Valley participants. T h e main contributions of the research presented in this dissertation can be summarised in four crosscutting conclusions (Figure 6.1). The first conclusion largely confirms the argument made by Stankey (1996) and others (Schindler, et al., 2002) that landscape preference evaluations are inherently 'wicked' exercises to the extent that they are based on a complex and simultaneous weighting of alternative perceived consequences. Although upper Skeena participants are generally concerned with how a change appears from a scenic quality standpoint, but they are also fundamentally concerned about what is right for natural (e.g. erosion potential, 242 habitat protection) and human communities (e.g. material livelihood, resource waste, public relations). While the Industrial or clearcut condition was clearly the least preferred of the four alternatives, the results presented in Chapter 5 do not provide conclusive evidence that any one type of treatment is most acceptable. Instead, the interests of upper Skeena residents are multifaceted, driven by potentially complementary but sometimes conflicting desires for vital communities and healthy landscapes. The second conclusion is that ethnic and other sociodemographic characteristics only partially account for differences in preferred landscape condition (e.g. age, gender, occupation). People tend to evaluate the landscape purposively or from a user perspective and, with a few exceptions, forest-based uses often transcend distinctions of race, age and gender. Forest-based behaviour seen in terms of how an individual uses the forest and the knowledge that results from purposive activity, appears to be strongly related to preference judgements. Ethnicity does, however, play a part in preference evaluations to the extent that landscape conditions reflect land-uses that are distinctive to a particular cultural or ethnic community. In this sense, the landscape is a tangible embodiment or reflection of an ethnic community's history and material way-of-life and, to that extent it can be a powerful stimulus for preference responses. W h e n