UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Grizzly bear (Ursus arctus) management and mortality distribution along the administrative boundaries… Dean, Carrie Saviers 1999

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

Item Metadata


831-ubc_1999-0494.pdf [ 19.6MB ]
JSON: 831-1.0099408.json
JSON-LD: 831-1.0099408-ld.json
RDF/XML (Pretty): 831-1.0099408-rdf.xml
RDF/JSON: 831-1.0099408-rdf.json
Turtle: 831-1.0099408-turtle.txt
N-Triples: 831-1.0099408-rdf-ntriples.txt
Original Record: 831-1.0099408-source.json
Full Text

Full Text

GRIZZLY BEAR (Ursus arctos) MANAGEMENT AND MORTALITY DISTRIBUTION ALONG THE ADMINISTRATIVE BOUNDARIES OF WATERTON-GLACLER INTERNATIONAL PEACE PARK by CARRIE SAVTERS DEAN B.A., Biology, Colgate University, 1996 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Resource Management and Environmental Studies Programme We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1999 © Carrie Saviers Dean, 1999 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I further agree that permission f o r extensive copying of t h i s t h e s i s for s c h o l a r l y purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of Resource Management & Environmental S t u d i e s The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Date A p r i l 1 , 1999 ABSTRACT Maintenance of wide-ranging mammal species in protected areas often requires management beyond reserve administrative boundaries. This thesis examines the relationship between cross-boundary management of a grizzly bear population and the distribution of human-caused mortality in that population. Concurrent analyses of grizzly bear management practices and human-caused grizzly bear mortality distribution were conducted along the administrative boundaries of Waterton-Glacier International Peace Park (WGIPP) in Montana, Alberta, and British Columbia. From 1976-1997, the concentration of mortality was significantly greater outside the park than inside it. In addition, non-park mortality density was higher within 10 km of the park boundary than in areas 20-50 km from the boundary. Analysis of management policies and practices suggests that comprehensive preventative bear-human conflict management and more conservative legal harvest in the parks have resulted in lower mortality levels compared with outlying lands. Outside the park, dispersing bears and/or gradients in habitat quality may be responsible for higher mortality adjacent to the park boundary. Contrasts and interactions in management between the park and its surrounding jurisdictions may also render bears in the boundary region more susceptible to human-caused mortality by increasing gateway development and numbers of habituated bears. More detailed boundary analysis is needed in several locations to determine how management is influencing human-caused bear mortality patterns. Once this has been accomplished, a system of customised buffer zoning is recommended that incorporates a variety of prescribed management changes in each jurisdiction in effort to reduce mortality adjacent to Waterton -Glacier's boundary. ii TABLE OF CONTENTS Abstract ii List of Tables vi List of Figures vii List of Acronyms viii Acknowledgements x INTRODUCTION 1 CHAPTER 1: BOUNDARY-GENERATED GRADIENTS IN BEAR MORTALITY. 4 1.0 Boundaries and Bears 4 1.1 Generated Mortality Gradients 6 1.2 Grizzly Bear Management & Human-Caused Mortality 7 1.3 National Parks and Grizzly Bear Mortality Gradients 9 1.4 Identifying and Explaining Mortality Gradients 13 CHAPTER 2: CASE STUDY INTRODUCTION 15 2.0 The Crown of the Continent Ecosystem 15 2.1 The Human Environment 16 2.2 Grizzlies of the C C E 17 2.2.1 Population status 17 2.2.2 Habitat use 18 2.2.3 Seasonal movements 19 2.2.4 Population densities 20 2.3 Bear Management Jurisdictions 21 CHAPTER 3: ANALYSIS OF GRIZZLY BEAR MANAGEMENT IN THE CCE. . . 24 3.0 Methods: Management Analysis 24 3.1 Results: Management Analysis 25 3.1.1 Legal harvest management 25 Introduction 25 Waterton-Glacier 26 Surrounding jurisdictions 26 3.1.2 Illegal kill control management 27 Introduction 27 Waterton-Glacier 28 Surrounding jurisdictions 29 3.1.3 Human-bear conflict management 30 Preventative conflict management: Introduction. . . . . . . 31 Waterton-Glacier 32 Surrounding jurisdictions. . 33 iii Responsive conflict management: Introduction 37 Waterton-Glacier 37 Surrounding jurisdictions 39 3.1.4 Access Management 41 Introduction 41 Waterton-Glacier 41 Surrounding jurisdictions 42 3.1.5 Land-Use Management 45 Introduction 45 Waterton-Glacier 46 Surrounding jurisdictions 47 CHAPTER 4: ANALYSIS OF MORTALITY DISTRIBUTION IN THECCE 51 4.0 Methods: Mortality Analysis 51 4.0.1 Data sources and completeness 51 4.0.2 Review and organisation of mortality data 52 4.0.3 Data analysis 53 4.1 Results: Mortality Analysis 55 4.1.1 Summary of Data 55 4.1.2 Spatial Analysis 57 CHAPTER 5: DISCUSSION 61 5.0 Waterton-Glacier Park and Surrounding Jurisdictions 61 5.1 Mortality and Management in Relation to Distance from Park Boundary. . . . 63 5.1.1 Relation of access to observed mortality halo 63 5.1.2 Other possible contributing factors 66 Dispersal and distance-decay mortality patterns 66 Habitat quality and seasonal migrations 68 Concentrated hunter effort 70 5.2 Consideration of Individual Jurisdictions 70 5.2.1 British Columbia 71 5.2.2 Alberta 73 5.2.3 Montana 76 5.3 Summary of Discussion 78 CHAPTER 6: RECOMMENDATIONS 80 6.0 Reducing the Mortality Gradient 80 6.0.1 National park buffer zones 80 6.0.2 Customised buffer zones 82 iv CHAPTER 7: CONCLUSION 85 7.0 National Parks and Grizzly Bear Conservation 85 7.1 Final Notes 86 REFERENCES 87 APPENDED A: STATISTICAL TEST RESULTS 97 APPENDIX B: DETAILED MANAGEMENT CHARTS 106 v LIST OF T A B L E S Table 1. Categories of human-caused bear mortality Table 2.1. Grizzly bear density estimates for areas of the CCE Table 3.1. Persons interviewed for bear management information Table 3.2. Legal harvest management summary table Table 3.3. Penalties for illegally killing grizzly bears in the CCE Table 3.4. Citizen reporting options for wildlife violations Table 4.1. Units from which mortality data was requested Table 4.2. Dates and status of mortality and relocation data obtained Table 6.1. Age structure and sex of grizzly bear mortalities in the CCE, 1976-1997 vi LIST OF FIGURES 1.0 Current and historic distribution of the grizzly bear in North America 1.1 Anatomy of a generated gradient 1.2 Predicted bear mortality gradient surrounding a park 1.3 Categories of bear management and influence on mortality 1.4 Predicted bear mortality gradient 2: popular national park 1.5 Diagrammatic representation of mortality "halo" surrounding a national park 2.1 Waterton-Glacier Park's location in the CCE 2.2 Map of landmarks and areas of ownership/jurisdiction referenced in the thesis 2.3 Wildlife management units in the CCE 3.1 Categories of bear management and influence on mortality 4.1 Mortality analysis zones 4.2 Map of grizzly bear mortality locations 4.3 Annual grizzly bear mortalities in the CCE, 1976-1997 4.4 Types of grizzly bear mortality in the CCE, 1976-1997 4.5 Types of grizzly mortality stratified by jurisdiction 4.6 Mean annual grizzly bear mortality density by jurisdiction 4.7 Mean annual grizzly bear mortality density by zone 4.8 Mean annual grizzly bear mortality density by jurisdiction and zone 5.1 Types of grizzly bear mortality, park and non-park 5.2 Comparison of predicted gradient with observed mortality gradient 5.3 Seasonal distribution of grizzly bear mortality in the CCE 5.4 Access roads in the Kootenay Region of BC, 1952-1986 5.5 Annual grizzly deaths due to management action in Alberta's CCE vii LIST OF ACRONYMS AEP Alberta Environmental Protection AEUB Alberta Energy and Utilities Board AFLW Alberta Forestry, Lands, and Wildlife AFS Alberta Forest Service AKPP Akamina-Kishenena Provincial Park (BC) AMA Access Management Area (BC) ANRS Alberta Natural Resources Service BCEAA British Columbia Environmental Assessment Act BCMoE British Columbia Ministry of Environment, Lands and Parks BCMoF British Columbia Ministry of Forests BFWD Blackfeet Fish and Wildlife Department (MT) BIR Blackfeet Indian Reservation (MT) BMTJ Bear Management Unit (MT) BNESA Burlington Northern Environmental Stewardship Area (MT) BNRR Burlington Northern Railroad CAMP Co-ordinated Access Management Planning (BC) CCE Crown of the Continent Ecosystem CCEDA Crown of the Continent Ecosystem Data Atlas CEAA Canadian Environmental Assessment Act CORE Committee on Resources and Environment (BC) ESA Endangered Species Act (US) FDP Forestry Development Plan (BC) FNF Flathead National Forest (MT) FPC Forest Practices Code (BC) GYE Greater Yellowstone Ecosystem (MT) IGBC Interagency Grizzly Bear Committee KBLTJP Kootenay Boundary Land-Use Plan (BC) LCNF Lewis and Clark National Forest (MT) MDFWP Montana Department of Fish, Wildlife, and Parks M U Management Unit (BC) NCDE North Continental Divide Ecosystem NEPA National Environmental Protection Act (US) NFMA National Forest Management Act (US) NPS National Park Service (US) NRCB Natural Resources Conservation Board (AB) viii RBHCM Responsive bear-human conflict management RMZ Resource Management Zone (BC) UNESCO United Nations Educational, Scientific, and Cultural Organisation USDA United States Department of Agriculture USDI United States Department of the Interior USFS United States Forest Service USFWS United States Fish and Wildlife Department VAHC Vehicle Access Hunting Closure (BC) WGIPP Waterton-Glacier International Peace Park WHA Wildlife Habitat Area (BC) WMU Wildlife Management Unit (AB) ix ACKNOWLEDGEMENTS Throughout the preparation of this thesis, I was fortunate to have the support and assistance of many individuals and organisations. First and foremost, my graduate studies would not have been possible without the generous financial and logistical support of the Canada-US Fulbright Program and the University of British Columbia I am grateful to both institutions for allowing me the opportunity to complete my degree in Canada. I am also indebted to the members of my thesis committee; Dr. Les Lavkulich, Dr. Paul Wood, Dr. David Shackleton, and Dr. Susan Glenn, for their advice and guidance in shaping my ideas into a workable thesis project. I would like to thank all of those individuals who provided me with bear management and/or mortality information: Bob Forbes, Tony Hamilton, Rob Neil, Matt Besko, and Richard Daloise at BCMoE; Tim Manley and Helga Pac at MDFWP, Richard Quinlan and Sylvia Birkholz at AEP; Steve Gniadek, Kate Kendall, and Dave Shirokauer in Glacier Park, and Kevin VanTighem and Rob Watt in Waterton Park. Your data and comments form the core of this thesis project. In addition, I owe a great deal to Craig Stewart and Andrew Harries at the Crown of the Continent Ecosystem Data Atlas for their technical support in mapping grizzly mortality data over so many jurisdictions. To my fellow graduate students in RMES and my wonderful friends in Vancouver, thank you for listening, advising, and keeping me entertained through even the most unappealing stages of thesis preparation. And lastly, a million thanks to my mom, dad, and brother for bearing with me (and without me) for the past two and a half years. I love you very much. Introduction The distribution of the grizzly bear (Ursus arctos) in North America once extended from Alaska south to Mexico, and from the West Coast east to the prairies. Today this range is much reduced (see Figure 1), as a result of centuries of direct elimination by humans (Storer & Tevis 1955). In the southernmost part of their range, where grizzlies now inhabit only western British Columbia and a strip of land along the Rocky Mountain chain, direct human-caused mortality remains the greatest threat to population persistence (Knick & Kasworm 1989, Mace & Waller 1998, Wielgus & Bunnell 1994). Along the southern Canadian and northern U.S. Rocky Mountains, almost all deaths of grizzly bears older than one year are caused by humans (roughly 85-94% of 174 deaths in 11 telemetry studies) (Mattson et al. 1996a). In addition, rates of human-caused mortality have been unsustainable for several areas of the Rockies in the past few decades (AFLW 1990; Horejsi 1989; Knight et al. 1988, Mace & Waller 1998). For grizzly bears in this region, human-caused mortality is a deterministic factor that often affects population growth rate and stability more than any stochastic variable (Mattson et al. 1996b, Primm 1996). Figure 1. Current (1995) and historic (circa 1492) ranges of the grizzly bear in North America (graphic from BCMoE 1995a). 1 Several categories of human-caused grizzly bear mortality can be described, including legal and illegal harvest, control action mortality, self-defence and accidental kills (see Table 1). The amount of each type of human-caused mortality in a given area will depend upon a complex system of ecological, physiographic, bureaucratic, economic, and cultural variables (Primm 1996). By influencing the number, distribution, and behaviour of both humans and bears, these variables affect the rate of encounter between humans and grizzlies, and the probability that such an encounter will result in the bear's death (Mattson et al. 1996b). Table 1. Categories of human-caused grizzly bear mortality. LEGAL HARVEST MORTALITY: deaths from legal sport hunting of grizzlies ILLEGAL KILL MORTALITY: poaching deaths and deliberate vandal killing MANAGEMENT CONTROL MORTALITY: agency elimination of nuisance or threatening bears LEGAL SELF-DEFENCE MORTALITY: kills by citizens in defence of human life ACCIDENTAL MORTALITY: unintentional kills (research related losses, mistaken identity kills, vehicle collisions, etc.) As humans cannot easily control the distribution or behaviour of grizzly bears, management systems that regulate the number, distribution, and activities of humans in bear habitat are the most effective means of reducing human-caused bear mortality (Peek et al. 1987). Management within protected areas may represent the greatest opportunity for success in this endeavour (Hummel 1990, Mattson et al. 1996a). Protected areas, where bears may be shielded from human development, intrusive land uses, and/or sport harvesting, should become sanctuaries of lower human-caused mortality for grizzlies (Noss et al. 1996, Scheer 1991). Following this, lower mortality rates within protected areas would allow bears to increase in number and replenish depleted populations in outlying regions (Mattson et al. 1996a). Although there are many different types of protected areas, those occupying the most land area in Rocky Mountain grizzly territory are national parks of Canada and the United States. Within the ten Rocky Mountain national parks inhabited by grizzlies, we might expect human-caused bear mortality to be lower than it is in surrounding areas. Accordingly, lower in-park mortality should help reduce mortality rates of entire ecosystems. However, due to specific management interactions between Rocky Mountain national parks and their surroundings, I predict that areas of increased human-caused mortality may exist in association with the boundaries of some parks. This hypothesis is explored in the following chapters, and is tested via an analysis of the human-caused grizzly bear mortality distribution within and surrounding Waterton-Glacier International Peace Park. An analysis of management within the jurisdictions of this area is conducted in order to aid in interpretation of the observed mortality pattern(s). 3 V Chapter 1 Boundary Generated Gradients in Bear Mortality 1.0 Boundaries and Bears The grizzly bear's current distribution in North America spans many political and administrative jurisdictions. As individual grizzlies often range widely, they are more likely to encounter boundaries between jurisdictions than are individuals of most other species (Beuchner 1987, Woodroffe & Ginsberg 1998). Under most circumstances, bears do not recognise administrative boundaries as barriers to movement; these lines are usually intangible and are seldom drawn according to natural features of the landscape (Janzen 1986, Newmark 1985). However, the fact that bears do not recognise human administrative boundaries does not mean they are unaffected by them. Administrative boundaries separate jurisdictions and represent the limits within which a certain management system applies. Via the policies adopted by the management framework within a bounded area, the actions of human beings are regulated. For example, after crossing an administrative boundary, humans are required to act according to a set of laws, regulations, and management policies that may be different from those of the jurisdiction left behind. When policies and procedures for the management of grizzly bears on one side of a boundary meet and interact with different policies and procedures on the opposite side, bear populations spanning the border may be altered considerably. In short, administrative boundaries, by affecting human behaviour, can affect bear populations. Two ecologists, Schonewald-Cox and Bayless (1986) developed a conceptual "boundary model" to explain the mechanisms by which ecological changes can occur in association with protected area boundaries. The boundary model states that due to the establishment and management of an administrative boundary, and the interaction of forces that approach or cross it, a "boundary-generated gradient" in some component may develop (Schonewald-Cox & Bayless 1986). The generated gradient, as defined by Schonewald-Cox and Bayless, is simply a change in some factor that is associated with a contrast in human behaviour across the boundary. With regard to wildlife populations, this might be a change in species abundance, density, distribution, behaviour or any other relevant factor. 4 The development of a boundary-generated gradient is always due to a contrast in human behaviour or management over an administrative boundary (Schonewald-Cox & Bayless 1986). As a result, gradients in wildlife species are often seen in association with protected area boundaries, across which the contrast in protection can be marked (Schonewald-Cox 1988). For example, imagine a protected area within which introduced species are eradicated up to the boundary line. If there is no equivalent eradication policy outside the park, it is very likely that a generated gradient in the abundance of introduced species will develop in association with the park boundary (see Fig. 1.1). Conversely, if management practices regarding introduced species are identical on either side of the administrative boundary, no generated gradient should develop. generated gradient Figure 1.1 Illustration of a possible generated gradient (in the abundance of introduced species) which has developed as a result of the contrast in management over a protected area boundary. The protected area boundary is represented by the dotted line. The shaded areas indicate location of the generated gradient. The gradient will not always line up with the boundary, but its existence is always the result of a dichotomy in human management on either side of a boundary line. Figure adapted from Schonewald-Cox & Bayless (1986). A boundary-generated gradient is not always spatially coincident with the boundary line (Fig. 1.1) (Schonewald-Cox & Bayless 1986). In fact, a gradient may exist in complex spatial patterns across and along the boundary's length. In location, it may be entirely on one side of a boundary, or entirely on the other. Or, the gradient may overlap the boundary to some degree. Generated gradients may be abrupt or gradual, and they may not extend along the 5 boundary's entire length (Ambrose & Bratton 1990, Schonewald-Cox & Bayless 1986). Finally, the location and characteristics of a gradient a will likely change with time. In summary, although the existence of a generated gradient is always related to contrast in management over a boundary, the gradient's actual configuration on the landscape may be affected by many other factors, including natural features of the environment, climatic patterns, plant and animal distributions, or the consistency with which human management is applied. 1.1 Generated Mortality Gradients One type of gradient that has been identified in association with numerous protected area boundaries is a gradient in the amount and distribution of mortality for a given wildlife species (Forbes & Theberge 1996, Knight, et al. 1988). Most often, management policies offer greater wildlife protection within a reserve than outside it, and the density of mortality incidents increases to some degree beyond the protected area's boundary (Coggins 1987, Freemuth 1991). Thus, a generated gradient in mortality develops in association with the park boundary (Fig. 1.2). If the mortality gradient is severe enough, it can become a barrier to dispersal and colonisation because animals that move beyond the boundary are killed before they can return or disperse further (Janzen 1986, Primm 1996). Sections of the mortality gradient may also become population sinks (area for which mortality exceeds fecundity) which can drain internal park (source) abundance (Schonewald-Cox & Bayless 1986). If the rate of movement between source and sink habitats is very high, maintenance of a stable population could require extensive non-sink habitat area (Doak 1995). Species most vulnerable to boundary generated gradients in mortality are those with expansive home ranges, since their territories are more likely to encompass multiple administrative jurisdictions (Woodroffe & Ginsberg 1998). Home ranges of grizzly bears in the Rocky Mountains frequently exceed 200 km2 for females and 900 km 2 for males (Aune & Kasworm 1989, Hamer, et al. 1985, Raine 1991, Servheen 1983). In addition, low reproductive rates (females reach sexual maturity at approximately 5.7 years, and generally have litters of two cubs every three years) render grizzlies susceptible to overkilling (Weaver 6 et al. 1996). Therefore, the grizzly bear is an appropriate species on which to focus an examination of boundary-related mortality distribution. co t f o CO CD m T3 <D CO CO O I c CO E park admin, boundary Distance from Park Centre Figure 1.2. Density of human-caused mortality perpendicular to the boundary from the centre of a hypothetical park. Management within this park is more protective than that outside; therefore, the amount of human-caused bear mortality is greater beyond the park's administrative boundary. A generated gradient in human-caused bear mortality has developed as a result of the contrast in management between a protected area and its surroundings. 1.2 Grizzly Bear Management and HCM Any boundary-generated gradient in human-caused grizzly bear mortality will be related to a contrast in the types of management which affect grizzly mortality. Several types of human management may affect bear mortality; the most influential are: 1. Legal Harvest Management: Regulation of recreational bear hunting will affect the amount and distribution of legally harvested grizzly bears. 2. Illegal Kill Control Management: Prohibitive regulations against the killing of grizzly bears, enforcement of these regulations, and punishment of offenders will influence the amount and distribution of illegally killed bears. 7 3. Human-Bear Conflict Management: This type of management involves the reduction and handling of human-bear conflicts. It can be divided into the following: Preventative Human-Bear Conflict Management: These are measures taken in advance of human-bear conflict, in attempt to reduce the frequency and lethality of future conflicts. They include public education and management of bear attractants, which will influence the amount and distribution of all conflict-related bear mortality. Responsive Human-Bear Conflict Management: These are measures taken after a conflict between human(s) and bear(s) has arisen. They include policies for bear relocation, destruction, conditioning, etc., and will affect the amount and distribution of control action mortality. 4. Access Management: Regulation of the number and location of humans in bear habitat will affect all types of human-caused bear mortality. 5. Land Use Management: Regulation of human land development and land-use activities will influence all types of H C M . For convenience, I will refer to these groupings as categories of "bear management," as do most wildlife agencies, although they are primarily concerned with the management of human behaviour with respect to bears. Figure 1.3 summarises the management categories' relation to specific types of human-caused bear mortality. Across an administrative boundary, management in any of the above categories may differ; and could result in a boundary-generated gradient in bear mortality. As mentioned previously, protected areas often have a significant contrast in management over their administrative boundaries. For example, if management in all listed categories of bear management is more protective (greater management effort is made to reduce human-caused mortality of grizzly bears) inside a protected area than outside it, a gradient in mortality incidence will likely develop in association with the boundary. Such a gradient would resemble that in Figure 1.2. However, in reality, the situation is usually more complicated. There are different kinds of protected areas and different kinds of surrounding areas. Management in some categories of bear management may be more protective outside a protected area than inside it, and vice versa. Each type of protected area (and ultimately each protected area) must be examined individually in order to determine how management of bears differs in relation to the surrounding area, and what effect this difference may have upon mortality distribution. 8 Figure 1-3 Categories of bear management (outer boxes) and the types of human-caused grizzly mortality (central circle and contents) they primarily influence. Access management and land-use management influence the amount and distribution of all types of HCM (bold arrows), while the other management categories influence mortality of a specific type (narrow arrows). 1.3 National Parks and Grizzly Mortality Gradients In this thesis, I will focus on national parks of the Canadian and American Rocky Mountains. For a hypothetical national park in Rocky Mountain grizzly bear territory that is surrounded by multiple-use lands of various ownership, I would expect the density of mortality incidences to be greater outside the park than inside it, as in Figure 1.2. However, unlike in Figure 1.2,1 suggest that an area of even higher mortality density may exist just beyond the park boundary (see Figure 1.4). The prediction of this "halo" of greater human-caused mortality surrounding a national park is based upon expected contrasts and interactions in bear management over the park boundary line, as will be explained in the following sections. For legal harvest management, illegal kill control, and human-bear conflict management, I would expect management within the national park to be as or more protective than that of surrounding multiple use lands. National parks in the Rocky Mountains do not allow legal harvesting of grizzly bears, unlike many surrounding areas, and firearms are prohibited within the parks. More relaxed firearm policies and multiple ownership on outlying lands, on the other hand, probably make persecution of individuals illegally killing grizzly bears more 9 difficult outside the national parks. Attractants are likely more thoroughly controlled inside national parks because many attractants (apiaries, livestock, and most residential garbage) are restricted or prohibited in parks. Finally, education programs about bears are usually very extensive in national parks (Albert & Bowyer 1991, Gniadek & Kendall 1997), and they may reduce the frequency of human-bear conflict. A l l of these factors would suggest lower levels of human-caused grizzly bear mortality within national parks than outside them. ro o ro m T J <1> CO ro O ro E Dark admin. boundary Distance from Park Centre Figure 1.4. Density of human-caused bear mortality perpendicular to the boundary from the centre of a hypothetical Rocky Mountain national park that is surrounded by multiple use lands. A generated gradient in HCM has developed as a result of the contrast in management between the park and its surroundings, as in Figure 1.2. However, specific contrasts and interactions of management have made the area just beyond the park boundary particularly lethal for bears (see text). The final category of bear management, access management, presents a different scenario. While one goal of U.S. and Canadian national parks is protection of natural resources, including wildlife, the parks are also intended for the enjoyment of visitors (Dearden & Rollins 1993; Foresta 1984). Because "visitor use and enjoyment" is part of the mandates of both park systems, total visitor numbers are seldom limited, and the annual budgets of Parks Canada and the U.S. National Park Service are heavily weighted towards the provision and 10 maintenance of visitor services, programs, and facilities (Dearden & Rollins 1993, Foresta 1984) . As a result, popular national parks in the Rocky Mountains often receive millions of visitors each year, with most visitation concentrated in a single season. Although the framework of roads, trails and other travel corridors may be less extensive within a national park than outside it, the concentration of use within this network is usually greater (Martinka 1982a). Concentrated visitor use within a national park in grizzly territory has several consequences. First, it brings bears into more frequent contact with humans, leading to increased human-bear conflict incidences within national parks (Herrero 1985). The increased conflict, in turn, could translate into greater in-park control action mortality. Second, the concentrated visitation may increase bears' habituation to humans. Habituated bears exhibit diminished fear responses die to repeated encounters with humans that have no negative consequences (McCullough 1982). They are differentiated from "food-conditioned" bears, those that have grown accustomed to consuming human food and/or waste products. Bears spending most of their time within national parks are often less wary of humans than are those in non-park areas (Hamer, et al. 1985, MacArthur-Jope 1982, McLellan & Shackleton 1989) and both habituated and food-conditioned bears are more often the victims of human-caused mortality than are other bears (Mattson, et al. 1992, Mattson, et al. 1996a). While grizzlies that have habituated to humans within national parks may be tolerated within the park (to a certain extent), where visitors expect to view wildlife, these bears may not be afforded the same leniency on the other side of the boundary. On surrounding lands, where there are more developments, residents, legal hunters, and poachers, habituated bears may become easy victims of human-caused mortality. The third consequence of increased access within national parks is that it encourages development beyond the park boundary, in what is often called "gateway development" (Sax 1985) . Land-use activity and development (the fifth bear management category) is quite restricted in Rocky Mountain national parks, in comparison to surrounding lands. There are no industrial, agricultural, or extractive land uses allowed in the national parks, and residential and tourist developments are also limited to a certain extent. The demand for visitor services created by the large number of park visitors cannot be met within the park, and is satisfied just beyond the park's borders where land development is less restrictive (Sax 11 1985). Thus, interaction between access management inside the national park and land-use management outside it can result in more congested human developments just over the border. In addition, residential development often increases surrounding a national park simply because people find it attractive to live near an enormous "wilderness playground" (Ambrose & Bratton 1990, Rasker & Hackman 1996). Since human developments have been associated with increased human-caused grizzly mortality (usually self-defence, accidental, and control action kills) (Mattson, et al. 1996), gateway development may contribute to an area of greater mortality density just beyond the park boundary. r i established admin. boundary « , V visitors — — — - theoretical boundary .—. . , e ,u low human caused bear mortality I I human developments, facilities or communities -==• , , H H high human caused bear mortality Legal and illegal bear harvesters ^ ^ ^ B a- J Figure 1.5. Diagrammatic representation of the 'halo' of increased human-caused grizzly bear mortality surrounding a hypothetical national park. Visitors (V) are concentrated within the park, creating a demand for development of visitor goods and services that cannot be fully satisfied within the park. Developments (boxes) are therefore concentrated in a zone surrounding the park, which may increase human-caused mortality here. Increased visitation may also habituate bears to people, which will make them especially vulnerable to human-caused mortality if they range outside the park and into human developments, legal hunters, and poachers. 12 1.4 Identifying and Explaining Mortality Gradients Gradients in human-caused grizzly bear mortality may be identified by assessing the location of mortality incidents over a given time period and area. However, explaining the manner by which a gradient has arisen is much more difficult. It is not always possible to tell whether or not an observed gradient is in fact a "generated" gradient, one that has developed in response to differences in human management over a boundary (Schonewald-Cox & Bayless 1986). As discussed previously, gradients will not always line up with the boundary line, and some may have developed in response to natural features and patterns of the environment (natural gradients), or to differences in human systems unrelated to any boundary. The difference between a natural gradient and a generated gradient, however, is in how each has developed, and in what each responds to (Schonewald-Cox & Bayless 1986). In some cases, it may be possible to perceive a gradient develop after a specific management change on one side of a boundary. Under most circumstances, however, this is not possible, and determining the origins of an identified gradient involves much guesswork. If an observed gradient is believed to be associated with a certain boundary line, an analysis of the management on either side of that boundary can be carried out to identify contrasts in management across the border. Detected management dichotomies may or may not be responsible for the gradient; at this point, only an explanation most consistent with observed patterns and information can be derived. Large scale conservation problems that involve both human socio-cultural factors and complex, non-linear ecological systems (such as human-caused mortality of grizzly bears) often involve so many variables that they are difficult to explain using traditional positivistic science (Mattson, et al. 1996b, Primm 1996). Techniques for measuring or controlling for all variables are either unknown, unavailable, or prohibitively expensive (both economically and environmentally). As Mattson, et al. (1996b: 1020) explain: The amalgam of biophysical and social factors controlling grizzly bear mortality has defied scientifically reliable predictions and will continue to do so in the future. This is not to say that scientists will be unable to inform our insight or judgement by analysing broad historic patterns...or that we will be unable to characterise a system in ways that are directly helpful to management. Rather , this information will likely not allow us to predict changes in H C M to a standard that most scientists and managers would consider reliable science (i.e. 90-95% chance of not being wrong) 13 Although it may not be possible to determine precise mechanisms behind a mortality gradient's development, the information gained in the gradient's identification and in the analysis of surrounding management can be quite useful. When management most probably connected to the gradient is modified, the gradient may respond in kind. If it does not respond, alternative management strategies can then be adopted, and responses monitored. This process, often called "adaptive management," is one of the most effective ways to translate knowledge of complex human-environment systems into practical conservation measures (Shafer 1999). It is in this context that gradient analysis is most useful. The following chapters of this thesis are a case study of human-caused grizzly bear mortality and management within and surrounding two contiguous national parks: Waterton Lakes National Park, Alberta, and Glacier National Park, Montana. The parks are located within the Crown of the Continent Ecosystem, which includes portions of Montana, Alberta, and British Columbia, and is a critical area for the conservation of grizzly bears in the southern part of their range. M y goal in this case study is to examine, in two separate analyses, the patterns of grizzly bear mortality and management across park boundaries. Relationships which best explain the results of these analyses will be identified, and suggestions for direction of future management will be developed. 14 Chapter 2 Case Study Introduction 2.0 The Crown of the Continent Ecosystem The Crown of the Continent Ecosystem (CCE) is centred around the Lewis, Clark, and Livingston ranges of the Rocky Mountains and includes roughly 15,000 square kilometres of land in southeastern British Columbia and southwestern Alberta, Canada, and northwestern Montana in the United States. It is a rugged, post-glacial, mountainous area that rises abruptly but of the interior plains of Alberta and Montana to the east, and the Flathead River basin to the west. Elevation ranges from 1100 m in the valley bottoms to 3100 m atop the highest peaks (Gadd 1995). Both the Arctic Continental and Pacific Maritime weather systems influence the area's climate, contributing to a distinct east-west gradient in precipitation over the continental divide. Pacific storm tracks associated with the Pacific Maritime system send warm, westerly chinook winds and rain over the Rockies, bringing up to 2500 mm of precipitation to the west side of the mountain range each year (Parks Canada 1995). Much of the eastern Rocky Mountain front lies in a rain shadow where annual precipitation is usually less than 750 mm (Parks Canada 1995). Summers in the CCE are mild (20°-25°C daily maximum), and winters are cold (-15°C average overnight low) but moderated by occasional chinook winds (Gadd 1995). The climate and topography of the CCE have resulted in high floral and faunal diversity, producing an area where biota common in the northern Rocky Mountains and Arctic mtenningle with those found in the southern and coastal ranges (Environment Canada 1989). Generally, climax communities on xeric eastern slopes consist of Englemann spruce (Picea engelmannii) and subalpine fir (Abies lasiocarpa) at higher elevations, and shortgrass prairie at lower elevations (CORE 1994; Dood et al. 1986; Schmidt 1995). Serai communities in these areas usually consist of lodgepole pine (Pinus contortd) and/or aspen (Populus tremuloides) stands (CORE 1994; Dood et al. 1986 Schmidt 1995). Li western slope climax communities, Englemann spruce/subalpine fir forests mix with western red cedar (Thuja plicata), hemlock (Tsuga spp.) and Douglas-fir (Pseudotsuga menziesii) (CORE 1994; Dood et al. 1986 Schmidt 1995). Serai communities on the wetter west side include lodgepole pine and western 15 larch (Larix occidentalis) (CORE 1994, Dood et al. 1986, Schmidt 1995). In the alpine areas, above 2100 m, barren rock fields with small alpine meadows are common. Over 970 species of vascular plants have been identified in the region (Achuff et al. 1997), and faunal diversity is among the most diverse in North America, with over 300 documented terrestrial wildlife species (Environment Canada 1989). 2.1 The Human Environment Due in part to the region's diversity and ecological significance, a 4,627 km2 core of the CCE is protected wthin two national parks, Waterton Lakes National Park (Waterton), Alberta and Glacier National Park (Glacier), Montana (see Figure 2.1). Since 1932, the contiguous parks have been united as Waterton-Glacier International Peace Park (WGIPP), which was designated a World Heritage Site in 1995. Much of the surrounding land area in British Columbia, Alberta, and Montana is located within provincial, state, and national forests or recreation areas, tribal reservations, and private landholdings. The estimated 60,000 year-round human residents of the CCE are situated outside of the national parks (excluding a settlement of about 150 people in Waterton Townsite), primarily in montane or subalpine valleys (Dood et al. 1986; Noss et al. 1996, Servheen 1989). Most are employed in one of the region's main economies: agriculture (cattle ranching and farming), timber harvesting, mining, oil and gas development, and tourist services (Dood et al. 1986; Trant et al. 1995). At present, the CCE has a relatively low human population; however, parts of the area are quickly growing in terms of numbers of humans and intrusive land uses. The population of the area of BC surrounding Waterton doubled from 1971-1991, and the ranch/farmland area tripled for the Alberta portion vvithin that time period (Trant et al. 1995). In northwestern Montana, human population growth and development has been similarly accelerated (Dood et al. 1986, Martinka 1982a). As a result, there is some concern that the effects of coterminous land use and development may insularise the parks (Kellert et al. 1996, Martinka 1982a). Therefore, management strategies are needed for the grizzly bear that will account for the changes and complexities associated with a variety of human influences. 16 BfWISH MONTANA Figure 2.1. Location of the two national parks witrdn me Crown of me Continent T^system. Waterton Lakes National Park, Alberta (wlp) and Glacier National Park, Montana are officially united as Waterton-Glacier International Peace Park. 2.2 Grizzlies of the CCE 2.2.1 Population Status The grizzly bears of the Crown of the Continent Ecosystem are part of a greater population that extends far north and west into British Columbia and Alberta, and a short distance south into the Swan and Mission Mountains of Montana. Overall, the CCE is near the southern and eastern edge of the grizzly's current range in North America, as the bears have been extirpated from all but 5% of their historical range in the lower 48 states, as well as from the prairies of Alberta (Hummel 1990, Schmidt 1995). Due to jurisdictional constraints, no estimates have been made of the total number of grizzly bears in the CCE, although estimates of population size and density have been generated for smaller regions within and overlapping parts of the CCE (AFLW 1990, Dood et al. 1986, Dood & Pac 1993, Mace & Waller 1998, Martinka 1974b, McLellan 1989a). Reports of population trend are also varied and difficult to make accurately (McLellan 1991). In some portions of the CCE, bears appear to be increasing in 17 number (McLellan 1989c), in others decreasing (ALFW 1990, Horejsi 1989, Mace & Waller 1998), and in still others stable (Dood & Pac 1993; Martinka 1982a). There is, however, widespread concern over the health and status of grizzly bear populations in this region, as they are heavily impacted by humans and at the edge of a range which has been slirinking northward for over a century (Hummel 1990, Mattson et al. 1996b, Paquet et al. 1996, Wielgus & Bunnell 1994.) The following sections offer a condensed description of what is known about these grizzly populations and their use of the Crown of the Continent Ecosystem. 2.2.2 Habitat Use by Grizzly Bears The manner in which a population of grizzly bears utilises and travels throughout a given region depends largely on the spatial and temporal distribution of food items (Dood et al. 1986, Gibbard & Sheppard 1992). The omnivorous grizzly exploits whatever food sources are most abundant or most easily obtained in an area. Thus, habitat use often varies greatly between regions (Hamer et al. 1991). Based upon several investigations conducted vvithin the study area, the following generalisations can be made about the diet and food habits of the CCE's grizzly bears. As grizzly bears usually emerge from their dens before green-up, the digging of roots (especially Hedysarum spp., found on well-drained, sparse slopes and meadows) is an important food activity for several weeks in early spring (Hamer et al. 1991; McLellan & Hovey 1995). Depending upon the severity of the preceding months, winter-killed ungulates may also be a significant source of food protein at this time (McLellan & Hovey 1995). Later in spring, emerging green vegetation becomes the dominant component of the grizzlies' diet, and remains so until late summer (McLellan & Hovey 1995). Horsetails (Equisetum spp), Gramminoids, and umbellifers like cow parsnip (Heracleum lanatum) are especially important and are located predominately on moist avalanche chutes and in riparian valley bottoms (Hamer et al. 1991). Bulbs of the glacier lily (Eryihronium grandiflorum) are also an important midsummer food item (Hanna 1978; McLellan & Hovey 1995). In late summer (end of July to September), fruits of the huckleberry (Vaccinium spp.) and serviceberry (Amelanchier alnifolia) begin to ripen, and become a major component of the 18 bears' diet (McLellan & Hovey 1995). The production of these high-energy berries is closely tied to wildfire-successional patches in the CCE, and yields vary from year to year according to climactic conditions (Martin 1983; McLellan & Hovey 1995). However, fluctuations in the availability of berries within the CCE are not as severe as those in other parts of the Rockies, since two species of preferred fruits, serviceberries and huckleberries, are found here (McLellan & Hovey 1995). As the berry season concludes, bears often return to digging Hedysarum roots to supplement their diets (Hamer et al. 1991). Other food sources, such as insects, small mammals, whitebark pine (Pinus albicaulis) seeds, and fruits of the mountain ash (Sorbus spp.) or hawthorn (Crataegus spp.) are also eaten throughout the active season; however, these items are either less significant or less predictable than those described above. The foods eaten by grizzly bears in the CCE are located primarily in areas of high habitat diversity. Many are found in early successional patches or in relatively open areas near forest edges (i.e. avalanche chutes, burn areas, alpine meadows, riparian areas), and bears naturally favour a mosaic of forest with small openings that produce a diversity of vegetation types (Martinka 1974a, VanEgmond 1997). Such areas are found throughout the CCE. Alpine areas are found at higher elevations along the continental divide, and avalanche chutes occur on steep sided slopes on both sides of the divide. Berry-producing potential varies geographically; however, areas of Waterton Park, Glacier Park, and the Flathead valley have been burned by wildfire or cleared vrithin last century, resulting in some excellent berry habitat (McLellan 1989a; Hamer et al. 1991). Very generally, a west to east gradient in the quality of bear food habitat exists within the CCE. The richest habitats are found in the Flathead River valley (McLellan & Hovey 1995), and extend over the continental divide, while the poorest are located east of the divide, where the shortgrass prairie begins. 2.2.3. Seasonal Movements Seasonal movements of bears in this region have been observed to follow a general pattern, modified by considerable individuality in foraging behaviour (Martinka 1982). Grizzlies usually move from lowland riparian and grassland habitats in early spring to increasingly higher elevation shrubfields as the snow melts (Martinka 1982b). hi late summer and fall, bears will actively seek out the most productive berry areas of the year. Finally, a general dispersion from 19 concentrated berry sites occurs later in the fall. McLellan (1989d) observed two patterns of seasonal movement among grizzly bears of the Flathead River valley of British Columbia. Bears of one group were elevational migrants, as described above. Those of a second group were primarily mountain residents, rarely descending to valley floors. In any case, grizzlies are known to inhabit all jurisdictions of the CCE, whether their annual home ranges are centred on one area or extend over vast areas with long movements between seasonal habitats (Gibbard & Sheppard 1992). Telemetry studies have indicated that movements of bears across jurisdictional boundaries (park, provincial, and international), as well as over the continental divide, are significant (Aune & Kasworm 1987; Carr 1989, Dood et al. 1986, Dood & Pac 1993, Martinka 1982a) 2.2.4 Population Densities Densities of bear populations, like seasonal movements, are probably linked to food abundance and distribution (Martinka 1974). As with estimates of population size and trend, density estimates have not been generated for the entire CCE. However, several intensive studies have encompassed parts of the ecosystem, and density estimates have been calculated for these sections (see Table 2.1). It should be emphasised that methods involved in estimation of grizzly population densities are varied according to the study area and the researcher; therefore, separate estimates are not always comparable (Dood & Pac 1993). That said, it appears that densities in the Flathead River valley, British Columbia, are among the highest in the CCE, probably higher than contiguous areas east of the continental divide (McLellan 1989a). Glacier National Park's density estimate is also relatively high; densities in surrounding areas of Montana are thought to be lower (Dood et al. 1986; Martinka 1974b). Table 2.1. Grizzly bear density estimates and methods of estimation for study areas that include portions of the CCE. Researcher(s) Study Area Density Estimate Method of Estimation McLellan Flathead River Basin, 12.5-17.5 Derived using the % of aerial telemetry (1989a) BC km2/bear locations of radioed bears within a study area Martinka (1974) Glacier National Park,MT 15.4-20.5 km2/bear Derived from unduplicated sightings of individual bears and family groups Aune & Kasworm (1989) Rocky Mountain East Front, MT (includes two regions in the CCE) Badger-2Medicirie: 26-31 km2/bear St Mary: 69-97 km2/bear Derived using a modified minimum composite home range of marked bears. 20 2.3 Bear Management Jurisdiction Responsibility for bear management in the CCE is spread among many agencies and jurisdictions. Although Waterton Lakes and Glacier National Parks are officially united as an International Peace Park, the parks (and the bears within them) are managed separately by Parks Canada and the US National Park Service (NPS), respectively. South and west of Glacier in Montana, the bulk of the land is owned by the U.S. Department of Agriculture's Forest Service (USFS) in the Flathead and Lewis and Clark National Forests. On these lands, bears are primarily the responsibility of the Montana Department of Fish, Wildlife, and Parks (MDFWP), although the land itself is the responsibility of the USFS. Since the 1975 listing of the grizzly bear as threatened under the US Endangered Species Act (ESA), the US Fish and Wildlife Service (USFWS) has overseen grizzly management in the lower 48 states, although the bulk of practical grizzly management in Montana remains with the MDFWP. East of Glacier, on the Blackfeet Indian Reservation (BTR), the Blackfeet Fish and Wildlife Department (BFWD) manages grizzly bears. In Alberta, north and east of Waterton Park, bear management is the responsibility of the Fish & Wildlife Division of Alberta Environmental Protection (AEP). Lands in this part of Alberta are managed primarily by the Alberta Forest Service (AFS) in the Bow-Crow Forest Reserve and Poll Haven Community Pasture, and by municipalities. West of Waterton Park, in the province of British Columbia, bears are under the jurisdiction of the British Columbia Ministry of Environment, Lands and Parks (BCMoE). Most land area is under the management of the BC Ministry of Forests (Flathead Provincial Forest), and the BCMoE Parks Division (Akarnina-Kishenena Provincial Park). Several of the wildlife management agencies in the CCE divide their jurisdictions into management units. For convenience, the case study area of this thesis reflects those units that best fit the boundaries of the CCE (Fig 2.3). The same case study area was used for the management and the mortality analyses. 21 Figure 2.2. Map of study region detailing relative locations of national and provincial forests, native reservations, wilderness areas, protected areas, communities, major transportation corridors, and other landmarks mentioned in the thesis text. 22 Figure 23. Units included in thesis case study area. These include Waterton and Glacier Parks, WMUs 400, 300, and 302 in Alberta, MUs 401 and 402 in British Columbia, and a polygon of 25,760 km2 in Montana. 23 Chapter 3 Analysis of Grizzly Bear Management in the CCE 3.0 Methods: Management Analysis Analysis of management is an efficient way of assessing the relative strength of protection for the grizzly bear for the variety of different mortality types to which it is susceptible. Collecting and analysing data for all the factors that may directly affect grizzly bear mortality over five jurisdictions (i.e. road density, human population distribution, habitat quality, hunter effort and distribution) would have greatly exceeded the time constraints and limited resources of this thesis project. Because management policies and practices will change in some definable manner over an administrative boundary, their analysis is a useful tool for explaining patterns of grizzly bear mortality across multiple jurisdictions. Categories of management to be analysed were chosen based on a review of relevant literature. Several bear management plans were reviewed (AFLW 1990, BCMoE 1995a, Dood et al. 1986, Dood & Pac 1993, Tilson & Dodd 1989, USDI 1988, USDI 1998) as were other articles regarding bear management practices, and reduction of bear-human conflict (Brannon et al. 1988, Gniadek & Kendall 1997, Martinka 1974, Martinka 1982b, Mattson et al. 1996b, Peek et al. 1987). The resulting categories of management are those that I feel most significantly influence the amount and distribution of human-caused grizzly bear mortality. These are legal harvest management, illegal kill control, human-bear conflict management (preventative and responsive), access management, and land-use management (Fig. 1.3). Management information for each jurisdiction was obtained via review of relevant literature (government documents; federal, state, or provincial legislation; operational plans and policies) and interviews with pertinent individuals in the field of bear management (Table 3.1). Similar information was requested in each jurisdiction for every management category, and was organised in chart form (see Appendix B for completed management charts). In all cases, management information from 1970 to 1997 was sought; however, earlier information was not always available because of staff turnover and incomplete record keeping in some jurisdictions. Key bear management practices in the parks (Waterton and Glacier) and the surrounding jurisdictions (Alberta, Montana, and British Columbia portions of the CCE) are 24 described in this chapter, and their major differences highlighted. A brief introduction precedes each management category. TABLE 3.1. Persons interviewed in order to obtain grizzly bear management information. INTERVIEWEE TITLE AND ORGANIZATION Mr. Bob Forbes Regional Wildlife Section Head, BC Ministry of Environment, Cranbrook, BC Mr. MattBesko Habitat Biologist, BC Ministry of Environment, Nelson, BC Mr. Rob Neil BC Ministry of Environment, Nelson, BC Mr. Richard Quinlan Alberta Natural Resources, Claresholm, AB Mr. Tim Manley Montana Department of Fish, Wildlife, and Parks, Kalispell, MT Mr. Dan Carney Blackfeet Fish and Game Department, Browning, MT Mr. Steve Gniadek Glacier National Park, West Glacier, MT Mr. Rob Watt Waterton Lakes National Park, Waterton Park, AB 3.1 Results: Management Analysis 3.1.1 Legal Harvest Management Introduction Where legal harvest of grizzly bears is allowed, it is often a significant source of mortality and may constitute the majority of human-caused bear deaths (Brannon et al. 1988, Gunson 1996). Mortality from legal harvest is unique in that it is more controllable than other types of HCM, since management of the harvest largely determines the number and distribution of lawfully killed bears within a particular jurisdiction (McLellan 1991). Usually, the number of bears to be harvested is based upon a set quota, or a target percentage of the estimated population. Maximum sustainable rates of total mortality have been estimated at 10.7% (Bunnell & Tait 1980) and 13.2% (McCullough 1981) of a grizzly population, while recommended legal harvest rates range from 2-3% (Sidorowicz & Gilbert 1981) to 5-6% (Cowan 1972). A target harvest represents a different number of bears when based upon different estimates of population size, however, and methods and accuracy of population estimation vary substantially (Dood et al. 1986). Whether the actual harvest falls short, meets, or exceeds the target quota depends upon a variety of factors, including the number and distribution of hunters, hunter success rates, 25 season timing, and harvest monitoring. Limiting the number of hunters in a particular area, as through a limited entry lottery, can help prevent overharvesting and/or concentration of kills in a single location (BCMoE 1995b). Timing of the hunt can also influence the number and distribution of legal harvest kills. Spring harvests may reduce hunter success as ice and snow covered roads make access more difficult (AFLW 1990). In contrast, during fall harvests, more bears may be taken incidentally by those hunting other big game animals (Dood et al. 1986, Gunson 1996). Legal Harvest Management: Waterton-Glacier There was no legal harvest of grizzly bears in Waterton Lakes National Park or Glacier National Park during the period examined (1970-1997). Legal Harvest Management: Surrounding Jurisdictions British Columbia, Alberta, and Montana (with the exception of the Blackfeet Indian Reservation lands east of Glacier National Park) have all had legal grizzly bear harvests at some time during 1970-1997. BC maintained its harvest throughout this period, while grizzly hunting in northwestern Montana was discontinued in 1991. Alberta's southern management units were closed to grizzly hunting from 1970-1981, after which they were reopened. See Table 3.2 and the appended management charts for more detailed harvest management information in each jurisdiction. Management of the legal harvest in Montana over the period 1970-1997 was quite different from harvest management in the provinces of BC and Alberta. In Montana, the harvest was not limited until 1975, when listing of the grizzly bear on the Endangered Species Act (1973) prompted the establishment of an annual total mortality quota of 25 bears for the entire North Continental Divide Ecosystem (NCDE) (a 24,800 km2 region of the United States that includes the Montana portion of the CCE). This quota was reduced to 15 bears in 1985, and in 1991 the entire harvest was discontinued (Dood & Pac 1993). From 1970-1991, the harvest was conducted in the fall, with no limits to the number of licenses issued. In the years following the quota establishment, the target harvest was never exceeded (Brannon, et al. 1988, Dood & Pac 1993). 26 Table 3.2. Summary of grizzly bear harvest management in BC, AB, and MT during 1970-1997 MONTANA BRITISH COLUMBIA ALBERTA Harvest * , , , , „ , , , , „ 1 1 > H I A l l Montana HDs within the CCE: 1970-1991 MUs 401 and 402: 1970-1997 WMUs 300, 302, and 400: 1982-1997 Harvest • No stated harvest target 1970-1974 • 25 bear quota (all H C M within the NCDE) 1975-1984 • 14-15 bear quota 1985-1991 (Dood,etal 1986,Dood&Pac 1993) • 5% harvest 1970-1988 • ~3.8%harvest 1989-1997 (BCMoE 1979, B.Forbes, pers. comm.) • No published target 1982-1987 • 2% harvest 1988-1997 (AFLW1990) Uc^nsiiig General (no limit to # of licenses issued) • General 1970-1982 • Limited Entry 1983-1997 Limited Entry Timing FALL SPRING SPRING Instead of an annual quota of bears, British Columbia and Alberta have relied upon harvesting a certain percentage of their estimated grizzly populations. This target percentage has, at least recently, been slightly higher in BC (3.8%) than in Alberta (2%) (AFLW 1990, BCMoE 1995b). Additionally, British Columbia's official provincial grizzly bear population was changed from 6,000-7,000 in 1970-88 (BCMoE 1979) to 10,000-13,000 in 1990, after a different method of estimation was employed (McLellan 1991, BCMoE 1995b). Unlike in Montana, harvests in both the BC and Alberta portions of the CCE have been conducted during the spring, and licensing has been on a limited entry basis since the early 1980s. 3.1.2 Illegal Kill Control Management Introduction The extent of illegal killing of grizzly bears in the CCE (poaching and unauthorised killing of nuisance bears) is uncertain, as an unknown number of mortalities go unreported (Brannon et 27 al. 1988). Studies of radio-collared bears produce some of the better approximations of total illegal mortality, although there may be associated biases when one sex or age class is heavily represented in the collared population (Dood et al. 1986). McLellan (1989b) reported a significant amount of illegal mortality for radio-collared grizzlies in the Flathead Valley of BC, where 5 out of 9 known deaths were illegal. Overall, the province of BC estimates unreported illegal mortality at between 25% and 100% of the reported kill (BCMoE 1995b). Montana estimated an annual average unreported mortality rate at 4% of the NCDE population, based upon several radio-collaring studies (Dood et al. 1986). The Alberta Fish and Wildlife Division states only that it believes most grizzly bear mortality in the province of Alberta is reported, due to the bears' high profile status (Gunson 1996). Management may influence the number and distribution of illegal killing to some degree, via prohibitive legislation, enforcement, and punishment of offenders. Laws prohibiting the killing of grizzly bears, the sale and export of grizzly parts, or the possession of firearms may reduce the amount of illegal killing. However, what is considered an "illegal kill" may vary between jurisdictions, as will the strength and means of regulation enforcement. Finally, the severity of punishments for offenders may influence the amount of illegal killing by acting, or failing to act, as a deterrent. Illegal Kill Control Management: Waterton-Glacier It is illegal to hunt, kill, harass, wound, or capture wild animals in the National Parks of both Canada and the United States under the Canadian National Parks Act (1930) and the US Organic Act (1916). Firearms are not permitted in either Waterton or Glacier by anyone other than authorised personnel. In Waterton Park, the fine for killing a grizzly bear was a maximum CDN $10,000 until 1988, when it was raised to CDN $150,000 by an amendment to the Canadian National Parks Act (Keiter & Locke 1996). In Glacier, the same violation is punishable under the ESA's section 9 "no take" regulation, which sets out maximum criminal fines of US $50,000 and maximum civil fines of US $25,000 per violation (Table 3.3). Enforcement of illegal kill regulations is accomplished through warden/ranger patrols of backcountry and frontcountry trails (during the tourist season- roughly June to September) and park border areas (during brown and black bear hunting seasons in neighbouring jurisdictions) (R.Watt, pers. comm., S. Gniadek, pers. comm.). Toll-free numbers are 28 available in both Waterton and Glacier for visitors to report poachers or other suspicious activity, although no reward is offered by the parks for information (Table 3.4). Table 3.3. Current maximum penalties for illegally killing a grizzly bear in each jurisdiction of the CCE. Waterton Park Glacier Park Montana British Columbia Alberta • Fine of up to CDN $150,000 or < 6 mos. in prison (CariadianA/hft'ona/ Parks Amendment ^cfl988) • Criminal penalties up to US $50,000 and/or <1 yr. in prison • Civil penalties up to US $25,000 for each violation (US Endangered SpeciesAct1973) • Criminal penalties up to US $50,000 and/or <1 yr. in prison • Civil penalties up to US $25,000 for each violation (US Endangered SpeciesAct1973) • First offence: CDN Sl,000-S25,000 or < 6 mos. in jail. • Second and subsequent: CDN $6,<HMV$50JX>0or < 6mos but >30 days in jail (Wildlife Act1982) • Fine of up to CDN $100,000 and/or < 2 mos. in prison (Wildlife Act1987) Illegal Kill Control Management: Surrounding Jurisdictions British Columbia, Alberta, and Montana all prohibit the hunting or trapping of grizzly bears without a license and/or outside of a designated hunting season {Wildlife Act (BC) 1997, Wildlife Act (AB) 1987, Montana Code Annotated 1992). However, while it is legal in BC to kill a grizzly in defence of life or property (property includes pets, domestic livestock, etc.) as long as it is reported to a Conservation Officer (Wildlife Act 1997), it is illegal to kill a grizzly bear for any purpose other than defence of human life in Alberta and Montana (US Endangered Species Act 1975, Alberta Wildlife Act 1997). Possession of firearms on federal (excluding national parks), state, and private lands in Montana is legal, as long as the firearm is licensed to the carrier. The same applies for private lands and most crown (provincial) lands in the Alberta and BC portions of the CCE. All three jurisdictions restrict or prohibit the sale and export of grizzly bear parts (BCMoE 1995b). Fines and punishments for illegally taking grizzly bears are imposed in all three jurisdictions, and are comparable to those of the national parks (Table 3.3.). In the states and provinces, enforcement of wildlife violations is more difficult than it is in the national parks. The areas to be covered are larger and jurisdiction is more complicated. While the wildlife itself may be under the jurisdiction of one agency, land ownership is divided amongst federal, state, and provincial agencies, as well as private entities, making 29 consistent monitoring very difficult. However, wildlife agencies in BC, AB, and MT do employ enforcement staff in making year-round patrols of bear territory within their respective jurisdictions. In Montana, game wardens with the MDFWP make patrols of wilderness and non-wilderness federal and state lands, while US Forest Service backcountry wardens make additional patrols of National Forest lands (Dood et al. 1986). In Alberta, Fish and Wildlife Officers in the Enforcement Division of Alberta Natural Resources are responsible for patrols of WMUs 400, 302, and 300 (R Quinlan, pers. comm.), and in British Columbia, Conservation Officers with the BCMoE make patrols of the access points within MUs 401 and 402 (B. Forbes, pers. comm.). Each of the three jurisdictions has a toll-free phone number for citizen-reporting of wildlife violations; rewards of varying denominations are offered (Table 3.4). Table 3.4 Current avenues for citizen-reporting of grizzly violations or other suspicious activity in each jurisdiction of the CCE. Waterton Park Glacier Park Montana British Columbia Alberta • Wildlife • Waterton- • Turn In Poachers • Observe, • Report-A-Watch: toll free Glacier Guide: toll Program: toll free Record, Report: Poachen toll free number. No free number for park number. No reward. mail-informs and a number. Rewards rewards. No rewards. • USFWSGrizdv toll free number. up to CDN $5,000 • TJSFWS Bear Recovery Plan: Offers rewards up to for information Grw.lv Bear Persons supplying CDN $2,000 for leading to Recovery Plan: information leading to information leading conviction. Persons supplying conviction under ES A to conviction information leading may receive up to half to conviction under of the fine paid by the ES A may receive up offender. to half of the fine paid by the offender. 3.1.3 Human-Bear Conflict Management Human-bear conflicts are often responsible for a considerable portion of the total human-caused mortality in regions heavily influenced by people (Aune & Kasworm 1989, Martinka 1974b). Kills in defence of human life or property, or those due to management control action, are usually the result of unwanted conflict between the needs of humans and those of bears. Management of human-bear conflicts can be divided into two general categories: preventative and responsive measures. Preventative management measures are those taken in 30 advance of problem incidents in order to reduce the future frequency and lethality of conflicts. These include public education and attractant management. Responsive management measures are those taken once a problem has already arisen, and include policies and procedures for dealing with conflict incidents. Preventative human-bear conflict management: Introduction Public Education Surveys have indicated that knowledge of grizzly bears and their management varies widely among residents of the southern Canadian Rockies and contiguous United States (Dood et al. 1986, Kellert et al. 1996). Kellert et al. (1996) found that attitudes towards grizzlies in this region were often either very positive or very negative, and seemed to illicit four basic perceptions of the bear: aggressor, pest, obstacle, or symbol of wilderness. Conceivably, unfavourable attitudes towards grizzlies and lack of understanding about what encourages human-bear conflict could contribute to increased human-caused mortality in the CCE. Public education is an important factor in preventing situations that could lead to unwanted human-bear conflict (Keay & Webb 1989, Mattson et al. 1996a), and may also be instrumental in changing negative attitudes towards large predators (Dood et al. 1986, Kellert et al. 1996). Educational programs can be targeted at different groups of people (i.e. young students, adults, ranching/agricultural communities, hikers, community residents, and extractive industry workers), and may have differing foci (i.e. natural history, bear appreciation, prevention of conflict situations, or safety in the event of a conflict). They also may utilise a variety of different media (i.e. brochures, posters, videos, newspaper, radio, Internet, meetings). It has been suggested that targeting the negative attitudes of specific groups may be more effective at reducing human-caused mortality than providing general information aimed at a broader public (Kellert et al. 1996). Although the degree to which education programs influence HCM is neither easily measured nor well understood (Kellert et al. 1996, Mattson et al. 1996a), their effect, if any, will depend upon target audiences, program content, and choice of media outlets. Attractant Management Many human-bear conflicts are related to a bear's attraction to unnatural, human-related, food sources (AFLW 1990, Herrero 1974, Madel 1996, Martinka 1974b). Prevention of 31 situations that might attract a bear and result in a conflict is the business of attractant management. In short this involves the management of humans and human landscapes so that they are not an attractive or an available food source for grizzlies. Strategies for managing attractants usually depend upon the type of human activities within a region. For example, attractant management in recreational areas often involves food and garbage handling, while in ranching communities, it might involve carrion removal and safeguarding of livestock. Managing attractants can be very effective in reducing conflict-related bear mortality (Gniadek & Kendall 1997, Madel 1996) depending upon how well management actions fit actual conflict problems, and how consistently they are applied. Preventative Human-Bear Conflict Management: Waterton-Glacier Public Education Visitor education is one of the primary goals of both Parks Canada and the US National Park Service. Information about bears is distributed in parks not only to increase appreciation and awareness for the animals, but also for reasons of visitor safety (Maw 1989, Parks Canada 1997a). Written bear information in both Waterton and Glacier parks dates back to the early 1970s (Ralf 1995; USDI 1976), and currently includes several pamphlets concerning bear ecology, safety and management, an annual newsletter (Waterton-Glacier Guide) containing bear information, and warning signs at trailheads, visitor centres, and park entrances (USDI 1998; Parks Canada 1997a). Some of these materials are given to visitors on entering the parks; others are available at any information centre or park office. Verbal visitor information includes a continuous pre-recorded radio message with safety information about bears, as well as several interpretative programs, films, and hikes in both parks which emphasise bear safety and management (Tilson & Dodd 1989, USDI 1998). Glacier National Park also has an Information Distribution Plan, initiated in 1980, which spells out responsibility for dissemination of bear safety and management information to visitors, park employees, residents, and concessionaires (USDI 1988). New staff in both parks are required to attend training programs which stress bear management issues. Attractant Management Due to intense annual visitation, attractant management has become an integral part of the bear management strategies of Waterton and Glacier National Parks. Feeding of wildlife is prohibited in both parks, as is leaving food items unattended or improperly stored. 32 Backcountry areas of Waterton and Glacier maintain a "pack-in, pack-out" policy with respect to food and garbage, and storage devices (food hanging poles, storage lockers, or other bear-resistant containerisation) are available at most campsites (Parks Canada 1997a, USDI1998). After the first two grizzly-related human deaths in Glacier National Park (on the same night in 1968), the US National Park Service became very serious about bear-proofing garbage storage in Glacier's frontcountry and developed areas (Gniadek & Kendall 1997). By the mid-1970s, conversion to bear-resistant garbage containers was completed, with daily trash removal during the summer season (Gniadek & Kendall 1997). Dumps in Glacier were eliminated before 1970, and incinerators used to burn garbage at backcountry chalets were removed in the late 1970s. Waterton Park was slightly behind Glacier in instituting bear-resistant garbage containerisation; it was phased in during the early to mid- 1980s, with daily pick-up and removal to Lethbridge (R. Watt, pers. comm.). In both Waterton and Glacier Parks, information is provided to residents and commercial operators regarding appropriate garbage handling and storage (Tilson & Dodd 1989, USDI 1988). (There is a significantly larger population of residents in Waterton than in Glacier, as Waterton Townsite is located within the park. Glacier's residents are primarily inholders whose land ownership pre-dates park establishment.) If garbage regulations are not complied with in either park, enforcement action is taken against the offending party. Preventative Human-Bear Conflict Management Surrounding Jurisdictions Public Education While visitors are the target of most bear information programs within Waterton-Glacier, the surrounding jurisdictions must consider educating residents, hunters, and businesses about bear safety and management, in addition to tourists. The two provinces and the State of Montana have adopted distinct programs and techniques for public education. Montana has focused most of its efforts at bear education on big game hunters, launching an extensive "mistaken identity" campaign in 1984, in attempt to reduce the number of grizzly bears shot after misidentification as black bears (Dood et al. 1986). Misidentification kills comprised 5.4% of the NCDE's total human-caused mortality from 1975 to 1985 (Dood et al. 1986). The campaign continues to spread bear identification information via an impressive 33 array of media outlets (see Appendix A). The MDFWP, in co-operation with other federal and state agencies, also produces several visitor- and resident-targeted pamphlets concerned with bear management and the prevention of bear-human conflicts. These are available at agency offices through out the bears' range. In addition, department employees give bear education talks at schools during the winter months, and occasionally initiate one-on-one discussions with landowners regarding bear management practices (T. Manley, pers. comm.). In contrast, Alberta and British Columbia have not focused much attention on the education of big game hunters, aside from requesting that first time and/or juvenile license purchasers undergo training in animal identification and hunting regulations (BCMoE 1997; AEP 1997). Both provinces have produced pamphlets for residents and visitors that are available at agency offices and kiosks. These publications generally describe bear ecology and management or explain how to best avoid conflicts with bears at residences and in the wild. British Columbia has formulated electronic versions of its pamphlets and placed them on the Internet (www.orcabay.com/Bears/grizzliesvshumans). In the East Kootenays, where information programs primarily target residents and visitors, BCMoE makes available to the public several audio-visual programs that discuss bear-human conflict (BCMoE 1995 c). In southwestern Alberta, considerable energy has been directed at informing ranching communities about provincial bear management practices and strategies for reducing conflict with bears. Several brochures have been produced explaining wildlife damage prevention and compensation, and several well publicised workshops on bear management have been organised for landowners and livestock producers within MUs 400, 300, and 302 (R. Quinlan, pers. comm.). In the past year or so, with the introduction of a new Southwestern Alberta Grizzly Strategy, AEP biologists and wildlife managers have initiated many "kitchen table meetings" to provide information and receive feedback from residents, ranchers, and other parties (ANRS 1997, R. Quinlan, pers. comm.). Attractant Management Attractant management is another of those areas of management that is greatly simplified within the national parks, where a single agency has control over the land, the zoning laws, the regulation of human activities, and the wildlife within the park's bounds. With the many 34 types of land ownership and jurisdiction in Montana, Alberta, and British Columbia, attractant management is understandably harder to co-ordinate. In Montana's portion of the CCE, attractant management is most developed on National Forest lands. In these areas, campgrounds are "bear-proofed" according to their locations in a given habitat, and their history of use. Campsites removed from highways are generally "pack-in, pack-out." Each of the national forests had its own regulations regarding food and garbage storage until 1993, when a Special Order was developed regarding attractant storage in the Flathead, Lewis and Clark, Lolo, and Helena National Forests. The order dictates that from April 1 to December 1 each year, human foods, pet foods, livestock feed, garbage, wildlife carcasses, and other attractants must be attended or stored in a bear resistant manner during daylight hours. At night, attractants must be within 50 feet of an attendee or stored in a bear resistant manner. Burning of attractants is prohibited, and dead livestock must be reported to the USFS within 24 hours. The Special Order is a definite step towards establishing regional attractant management standards; however, the regulation is less stringent than those of Waterton and Glacier Parks, where merely keeping items within 50 feet of an attendee overnight is not considered "bear-proof." On tribal (Bureau of Indian Affairs) lands on Montana's Blackfeet Indian Reservation, garbage and food storage regulations similar to those of the USFS Special Order are in effect (BFWD 1996). Open dumps on the reservation were closed in the early 1980s, and non-bear-proof dumpsters are still in the process of being replaced by bear-resistant models (D. Carney, pers. comm.). On private lands in Montana, management agencies have no authority to require the elimination of attractive bear food sources. With the exception of a very few local ordinances, there are no food and garbage storage regulations on private lands (Servheen 1989). In the Alberta portion of the CCE, many bear attractants are related to the ranching/agricultural industry (AFLW 1990). Attractant management on rangelands is usually regulated by municipal bylaws; however, removal of livestock carrion has been obligatory throughout the province since 1971 under the Livestock Diseases Act. A free service is available that takes livestock carrion away to a rendering plant, although some ranchers are reluctant to use the service for disease or pride-related reasons (R. Quinlan, pers. 35 comm.). Beekeepers in the area have benefited from a government cost-share program for design and construction of electric fences around bee yards (a frequent bear attractant). A recent and interesting attractant management technique used in Alberta is that of carcass redistribution. In 1997, a program was initiated to distribute road-killed carcasses of deer, elk, moose, bighorn sheep, and beaver to selected high-elevation sites (ANRS 1997). This is an effort to reduce livestock depredation by keeping bears at high elevations until spring green-up. All provincial campsites in southwestern Alberta are required to have bear-resistant garbage storage. These camping areas are in the process of becoming privatised; however, new owners will be required to maintain previous standards (R. Quinlan, pers. comm). There are also strict regulations for garbage storage and disposal at industrial camps under the Public Lands Act (1980). All dumps in WMUs 300, 400, and 402, with the exception of one in Pincher Creek, have been completely removed (R. Quinlan, pers. comm.). Garbage is stored for frequent pick-up at fenced transfer sites or in bear-proof dumpsters. On private lands, it is left to municipal districts and towns to enact land-use bylaws that regulate food and garbage storage (Keiter & Locke 1996). In the East Kootenay region of British Columbia, sanitation on provincial forest lands was regulated by the Forest Act (1982) until the recent Forest Practices Code (1994) came into effect in 1995. At the present, most, but not all, Ministry of Forests and Ministry of Environment campgrounds and recreation areas have bear-proof garbage storage and food storage devices of some type (B. Forbes, pers. comm.). Provincial parks in British Columbia are either strictly "pack-in, pack-out" (as is Akamina-Kishenena Provincial Park in MU 401), or outfitted with bear-proof garbage containerisation. BC Parks also provides food storage devices in provincial park campsites. Private campgrounds and residences in BC are not required to have bear-resistant food and garbage storage. However, in some areas, municipal bylaws may regulate the storage of garbage and food attractants on private lands. In 1995, landfill operators began to be required to bear-proof facilities (usually via electric fencing) before they could obtain a "Waste Management Permit." Landfill sites have not been removed in the BC portion of the CCE, as they have in parts of Montana and Alberta, probably because BC does not have the liberty of transferring dumps "out of bear territory." 36 In 1996, after a two-year co-operative process by the Committee on Resources and Environment (CORE), a Kootenay Boundary Land Use Plan was drawn up which includes recommendations for attractant management in the region. The recommendations, which have no force of law, include bear-proof containerisation for industrial camps, conversion of all BCMoF, BCMoE, BC Parks, and private campgrounds to "pack-in, pack-out" only, and adoption of attractant minimisation plans by all landfill operators (Kootenay Inter-Agency Management Committee 1997). Responsive Human-Bear Conflict Management: Introduction The manner in which an agency deals with human-bear conflicts once they have arisen will influence the magnitude and distribution of human-caused bear mortality in that area (Brannon et al. 1988. Martinka 1982b, Peek et al. 1987). Bears engaged in activities that are seen as threatening or undesirable to humans may be dealt with in a number of ways. The most extreme management "control action" is the destruction of a bear or its removal from a wild population. Bears may also be relocated to areas where there are less opportunities to re-offend; translocation areas may be within a bear's original home range or several hundred kilometres away. As an alternative to relocation or destruction, offending bears may be "reprogrammed" to avoid conflict situations through aversive conditioning, in which behavioural modification is achieved by repeatedly pairing undesirable behaviour with a painful experience (i.e. rubber bullet, startling noise, encounter with bear dogs). Although it often requires a considerable amount of time and effort to be successful, aversive conditioning has become an accepted management tool in recent years (Clarkson 1989, Dalle-Molle & Van Horn 1989, Strickland 1989). Finally, human-bear conflicts can sometimes be dealt with by modifying the human element, instead of the bear. The point at which each of these management actions are employed differs according to the policies of each jurisdiction. Responsive Human-Bear Conflict Management: Waterton-Glacier Because protected areas theoretically have more control over preventative human-bear conflict management than do surrounding areas, one would imagine that bears would get in to fewer compromising situations in parks. However, this is not always the case. Many protected areas have so many visitors that the sheer concentration of humans in the summer months results in more bear-human conflict (Herrero 1985). Popular national parks like 37 Waterton and Glacier, especially, have to balance visitor safety and enjoyment with bear conservation, a difficult task which has been interpreted differently in each park. Throughout most of the 1970s in Glacier National Park, management action towards an offending bear was dependent upon where in the park the bear was found. In general, bears frequenting developed or frontcountry areas were relocated as soon as possible, while those frequenting backcountry trails or campsites were relocated only after visitor closures (human management) failed to deter the bear (USDI 1971, USDI 1975, USDI 1976). In this era, bears were destroyed only on their second or third "molesting" incident, or when relocation was not possible (USDI 1971). By the early 1980s, the management protocol changed to deal with the bear according to type of offence. "Habituated" bears, or those frequenting roadsides, developed areas, or backcountry trails and campsites (after closures in effect), but not interacting directly with humans or human products, are generally relocated to remote areas within the park (USDI 1988). "Conditioned" bears, on the other hand, those that display aggressive behaviour, damage property, obtain human foods, cause human injury, or become overly familiar with humans, are relocated out of the park or destroyed (USDI 1988). In the past few years (since 1990), management direction has changed again, allowing for a chance to re-condition offending bears before they are relocated or destroyed (USDI 1998). In Waterton Lakes National Park, actions towards offending bears are slightly more lenient than in Glacier. Grizzlies are only destroyed in WLNP when there is an immediate threat to human life or visitor safety, when a bear is very unafraid or very destructive, or when a bear has wounded or killed a person in a non-defensive manner (Tilson & Dodd 1989). Bears frequenting campgrounds, roadsides, or developed areas may be relocated, but are usually monitored or aversively conditioned. Those observed at immovable artificial food sources, those causing unwarranted property damage, and those displaying unnatural aggressive or fearless behaviour are relocated. Both Waterton and Glacier, when relocating a bear out of their bounds, attempt to place it somewhere within the greater ecosystem before initiating a long distance translocation. Li summary, bears displaying aggressive behaviour, obtaining human foods, or damaging property in Glacier Park stand a good chance of being destroyed, while bears in Waterton would probably be relocated for the same offences. This discrepancy in management 38 between the adjoining parks may be the result of differing bear-human conflict histories. Glacier has had far more bear-related human injuries and fatalities than has Waterton. Although the statistics are comparable once corrected for the sizes of the parks (Glacier is roughly ten times the size of Waterton), the fact that a single park has had ten grizzly-related human fatalities has likely increased public pressure to heighten visitor safety within Glacier. Responsive Human^Bear Conflict Management Surrounding Jurisdictions Li Alberta, bears involved in conflicts are categorised according to type of offence. Those that have attacked or injured humans, acted aggressively without provocation, or become a threat to domestic animals (livestock, pets) are considered "problem bears" (AFLW 1990). These bears may be destroyed on their first or second offences, or they may be relocated (AFLW 1990). Bears that are habituated to unnatural food items, frequent developed areas, or feed on crops are considered "nuisance bears," and are given two chances at relocation before being destroyed (AFLW 1990). However, until recently, most bears relocated from southwestern Alberta have been moved so far north (over 200 km) that they rarely return to offend again. Although it may be convenient that Alberta has the space to make such long-distance translocations, bears relocated to these northerly areas of the province are as far removed from the CCE population as are those that have been destroyed. Recently, a Southwestern Alberta Grizzly Strategy (ANRS 1997) was approved, under which offending sow, cub, and sub-adult grizzlies will be relocated within the local ecosystem on first offences. Second offences for these bears, and first offences for adult males will still be dealt with via long-distance translocations (ANRS 1997). Although it has been used intermittently in Alberta for several years, aversive conditioning of bears has only been employed as a structured management tool since 1997. Primarily, this conditioning is targeted at bears frequenting roadsides, private residences, and campgrounds, but not at livestock offenders (R. Quinlan, pers. comm.). An additional strategy adopted by Alberta for dealing with bear-human conflict in ranching and farming communities is monetary compensation for losses due to predators. Since 1974, some form of government or privately funded compensation has been awarded to livestock owners for confirmed grizzly-related losses (R Quinlan, pers. comm.). Similar compensation programs exist for crop and beeyard damage. 39 Guidelines for dealing with nuisance grizzly bears in Montana were adopted by the Interagency Grizzly Bear Committee (IGBC) in 1986 (1GBC 1986; BFWD 1996). As in Alberta, offending grizzlies are categorised according to type of offence. In the first category, bears involved in livestock depredation or property damage, those obtaining unnatural food items, and those habituated towards humans are relocated once or twice (depending on age and sex) before being destroyed. Second, females and cubs displaying aggressive (non-defensive) behaviour toward people that results in a potential threat to human safety or minor injury are relocated once before being destroyed. Male grizzlies are destroyed on first offences of this type. Finally, all bears causing substantial human injury or loss of life are destroyed. Grizzly translocations in northwestern Montana are generally over 65 miles, although bears are rarely moved out of the NCDE (Riley et al. 1994, T. Manley, pers. comm.). Damage compensation is not awarded by federal or state agencies in Montana; however, private organisations do offer money for confirmed stock losses. Like Glacier National Park, the province of BC tolerates very little bear-human conflict. Grizzly bears in the Kootenay region that exhibit unusually aggressive behaviour towards humans, have been habituated to humans or conditioned towards unnatural food sources for longer than one week, or are under two years of age at the time of offence, are destroyed (BCMoE 1996a). Relocation is attempted only for non-aggressive, healthy, mature bears with no past history of obtaining human food or garbage, provided suitable release sites are available (BCMoE 1996b). A "suitable release site" in the Kootenays must be at least 80 km from the bear's point of capture, and beyond a physiographic barrier from the nearest human habitation (BCMoE 1996b). Bears that re-offend after one relocation are destroyed. No aversive conditioning is attempted, and no wildlife damage compensation is awarded (B.Forbes, pers. comm.). In summary, protocol for dealing with human-bear conflict is comparable in Alberta and Montana, where bears are relocated once or twice for most offences before being destroyed (depending upon age and sex). However, translocation distances have consistently been much longer in Alberta than in Montana. In British Columbia, reactive human-bear conflict management is closer to that of the parks, where very little interaction with humans and human products is tolerated of grizzlies before they are destroyed. As with the higher legal harvest targets in British Columbia, this management policy may reflect the fact that with 40 thousands of grizzlies across the province, BC does not need to be as conservative in its management. Also, the BC portion of the CCE is less developed than similar areas in Alberta or Montana, and opportunities for bear-human conflict are fewer. 3.1.4 Access Management Introduction In the context of bear conservation, access management consists of regulating the number and location of humans and human activities in bear habitat. Management of human travel corridors (roads, trails, railways, etc.) is the primary means of controlling access, since the location, development and management of these corridors largely determine where humans will be found. Aside from their probable role in the fragmentation of habitat and displacement of bear activities, human access corridors have been found to increase bear mortality rates by bringing more humans (legal and illegal hunters, new residents, etc.) into contact with bears (Dood et al. 1986, Mace et al. 1997, McLellan and Shackleton 1988, Titus 1991). Li the Flathead Valley of BC, for example, most legally and illegally killed bears were shot from roads (McLellan and Shackleton 1988). Although there may be some reporting bias in this data, it is generally agreed that most forms of human-caused grizzly mortality are closely tied to ease of human access into bear habitat (McLellan 1991). Management of access corridors may reduce grizzly mortality where new road, trail, and railway development is limited and/or wisely located (see also section 3.1.5 Land-Use Management). On existing travel corridors, numbers of users can be restricted, as can the allowable types of transportation (motorised vehicle, horse, bicycle, foot, eta). Roads and trails may also be temporarily or permanently closed where potential for human-bear contact is high. Finally, comprehensive access management plans can be developed through which managers can consider the impacts of human access on a broader scale. Access Management: Waterton-Glacier The road and trail systems of Glacier National Park and Waterton Lakes National Park have changed little in the past several decades (Martinka 1982a), and major new corridor developments would be difficult to pass through the stringent development constraints of 41 either park (R. Watt, pers. comm) (see section Therefore, it is not the density or construction rate of roads and trails in the parks that necessitates access management, but the number and distribution of people using the existing infrastructure (Martinka 1982b). Neither Waterton Lakes National Park nor Glacier National Park places any limitations on total numbers of park visitors. Both parks are very popular tourist destinations, and annual visitation is high (In 1990, 360,000 people visited Waterton, and almost 2 million visited Glacier (Dood & Pac 1993, Parks Canada 1990)). Visitation peaked for the parks in the late 1970s and early 1980s, and has decreased only slightly since then (Dood & Pac 1993, Parks Canada 1997b). Backcountry use, which has also decreased since the mid-1970s (Dood & Pac 1993, KWatt, pers. comm.), is somewhat restricted in both parks via distribution of a limited number of campsite permits per night. Trail use by hikers is generally not restricted, although horses and mountain bikes are prohibited on many trails. Occasionally, both Waterton and Glacier temporarily close roads, trails, and campsites in the interests of grizzly conservation and visitor safety (i.e. when there has been a bear-human conflict encounter, if a sow and cubs are frequenting the area, if there is carrion in the area, if traps or snares are set, or if a bear has repeatedly displayed aggressive behaviour) (Tilson & Dodd 1989, USDI 1988). In Glacier, trails are also closed when bears are frequenting an area and feeding naturally or displaying neutral behaviour towards humans (USDI 1998). For a period in the late 1980s to early 1990s, an entire area of Glacier Park (the Apgar Mountains, where grizzlies congregate in late summer to feed on berries) was closed to humans during the weeks of highest grizzly use (S. Gniadek, pers. comm., USDI 1988). Access Management: Surrounding Jurisdictions Visitor use in areas surrounding Waterton-Glacier has been less concentrated than within the parks, although trends in recreational use are similar (peaking during the late 1970s and early 1980s) (Dood et al 1986). However, because road construction has not been capped to the degree it has within Waterton-Glacier, road densities are generally much higher on adjacent lands. The density of roads has increased significantly in parts of the CCE within the past few decades, especially along the northwestern boundary of Glacier and the western boundary of Waterton, where extensive mountain pine beetle (Dendroctonus ponderosae) salvage logging occurred in the mid 1980s (Dood et al. 1986, McLellan & Shackleton 1988). Oil and gas exploration has resulted in additional road construction along the eastern 42 boundary of Glacier Park (Dood et al. 1986), as well as north of Waterton Park, where development of the Shell-Waterton gas field led to high quality gravel roads in most valleys (Horejsi 1989, Parks Canada 1997b). In the late 1980s, the Alberta Forest Service began developing a pilot access management plan for the Castle River region (WMU 400) with input from government agencies, public interest groups, industry, and local authorities. The Castle Access Management Plan, which provides operational level direction for the recreational use of on and off highway vehicles, was adopted on a voluntary basis in 1996. However, because the plan has no force of law, some believe it has proven ineffective at managing access (Parks Canada 1997b). A similar plan for the Crowsnest Pass Area (WMU 400) has been proposed but not yet developed. The Alberta government has the capacity to close roads or restrict vehicular traffic under several statutes (Wildlife Act (1987), Public Lands Act (1980), Forest Act (1971)); however, closures and restrictions of forest roads, primary highways, and secondary highways (i.e. with physical barriers or signs) are not common (R. Quinlan, pers. comm.). Under British Columbia's Wildlife Act (1982), individual roads can be closed to vehicular access during hunting seasons to protect grizzly bears. Vehicle Access Hunting Closures, or VAHCs, have been instituted on several occasions in southeastern BC since the early 1970s (R Neil, pers. comm.). Entire regions may also be closed to vehicle traffic in Access Management Areas (AMAs), although these are less frequently established (R. Neil, pers. comm.). In the 10,921 ha Akamina-Kishenena Provincial Park (MU 401) adjacent to Waterton-Glacier, no motorised vehicle access or road construction is allowed. Here, BCMoE policies also allow for temporary closures of trails and campsites to accommodate seasonal cycles in bear habitat use or to respond to human-bear conflict situations (BCMoE 1995c). In the mid 1980s, a co-operative effort involving BCMoE, BCMoF, and the public was initiated in order to map and classify roads, and develop a planning process for creating access management plans. The Co-ordinated Access Management Planning effort, or CAMP, as it was called, never successfully implemented any plans (R. Neil, pers. comm.). When the Kootenay Boundary Land Use Plan (KBLUP) was released in 1996, following the CORE process, it noted several recommendations for access management, including a 43 reduction in open forest road density, more effective seasonal road closures, and avoidance of road development in important bear habitats (Kootenay Interagency Management Committee 1997). The KBLTJP recommendations are not legally binding; however, a recent Memorandum of Understanding between BCMoE and BCMoF may help implement parts of the KBLUP, as it includes an agreement to begin access management planning in priority landscape units in order to eliminate excess roads and uncontrolled access (BCMoE 1998). The Ministry of Forests plans to incorporate this access management planning into Forestry Development Plans (FDPs), which are prepared every five years by leaseholders, and require approval under the Forest Practices Code (1995). FDPs must include plans for the location, construction, maintenance, and temporary or permanent deactivation of roads. Most CCE lands in Montana, excluding Glacier Park, are within the Flathead National Forest (FNF) and the Lewis and Clark National Forest (LCNF). Under the National Forest Management Act (NFMA 1976), each national forest is required to produce a Forest Plan every 10-15 years that details how the forest is to be managed during that period (USDA 1994). Because the USFS has a responsibility to provide for the recovery and conservation of endangered and threatened species under the ESA, forest plans in northwestern MT are reviewed by the USFWS to ensure that they are not likely to jeopardise the continued existence of the grizzly bear. This approval process takes into account the direct and indirect effects of proposed road construction and management on bear populations. Forest plans were approved for the LCNF and FNF in 1985 and 1986, respectively. Both plans include the capacity to close roads or areas, seasonally or yearlong, in order to meet grizzly bear habitat management goals. In addition, human activity guidelines prepared by the Interagency Grizzly Bear Committee (IGBC) have been incorporated into both the FNF and LCNF Forest Plans. These guidelines include recommendations for closing roads in seasonal grizzly habitats, maintaining visual barriers along roads, bussing forest workers, scheduling road construction to avoid seasonal bear use periods, and deactivating roads no longer in use (USDA 1985). The 1985 LCNF Forest Plan concludes that, during the Plan's term, no permanent road construction will occur as a result of timber production in the forest's Rocky Mountain Division (all LCNF grizzly habitat). Timber harvesting and associated road construction on the Flathead National Forest, on the other hand, was such that a lawsuit challenging the FNF 44 Forest Plan was filed by several environmental groups in 1989 (USDA 1994). As a result of the environmental assessment that followed, an amendment was made to the Forest Plan that set long term objectives for total motorised access densities (open and restricted roads and motorised trails) at no more than 2 mi/mi2 (1.2 km/km2) for 18% of each Bear Management Sub-unit (there are 73 BMU sub-units in the FNF, each approximately 50 mi2 (130 km2)) (USDA 1994). Also, open motorised access (open roads and motorised trails) will be less than 1 mi/mi2 (0.63 km/km2) in 13% of each BMU sub-unit Security Core Areas (areas free from motorised use or high intensity non-motorised use of roads and trails during the non-denning period) will make up 55 to 100% of each BMU sub-unit (USDA 1994). Both the Flathead and Lewis and Clark National Forests include some designated wilderness areas (portions of the Bob Marshall (408,646 ha) and Great Bear (115,697 ha) Wilderness Areas are found within these two National Forests). In these areas, no motorised vehicle fraffic or road development is allowed. The Montana Department of Fish, Wildlife, and Parks does not have the authority to close or restrict road access on national forest or municipal lands; the department can only make recommendations for road closure (T. Manley, pers. comm.). On the Blackfeet Reservation tribal lands, the Blackfeet Fish and Wildlife Department does have the authority to close roads, although this is generally only done when there is a trap set in the area (D. Carney, pers. comm.). The Blackfeet Forest Management Plan also includes some access management provisions for grizzly bears on tribal lands (Sawyer & Talnagi 1994). Road traffic is closed in several areas on the BIR from April through June (see Appendix B-5). 3.1.5 Land-Use Management Introduction Land-use activities can affect the amount and distribution of human-caused bear mortality in many capacities. They can influence the number of humans and human developments in an area, as well as the degree to which an area is accessible. Thus, land-use activities are often tied to access issues, and will affect the amount and distribution of all types of human-caused mortality. In addition, certain land uses are inherently less compatible with grizzly 45 bear conservation than others; residential development and livestock grazing are two of the least compatible because they greatly increase the likelihood of bear-human conflict. Finally, some human land-uses may decrease (or, very infrequently, increase) the amount, quality, and connectivity of available grizzly habitat, which may secondarily affect levels of human-caused bear mortality. Land-use management is a broad and very complex topic. It includes policies and procedures for the authorisation, placement, and regulation of all types of human activities on the land, including residential development, agriculture and ranching, commercial development, mining, forestry, oil and gas drilling, energy production, recreational development, and innumerable other projects and activities. An exhaustive discussion of land-use practices and policies in each of the CCE's jurisdictions is not possible here; however, major regulations and restrictions placed on land-use activities are discussed below. Land-Use Management: Waterton-Glacier National parks of the U.S. and Canada substantially restrict the amount and types of development permitted within their bounds. Under the US Organic Act (1916), and the Canadian National Parks Act (1930), the parks are to be preserved for future generations, which precludes industrial activities (mining, logging, oil and gas drilling, etc.), agricultural operations, or ranching activities (although exceptions have occasionally been made). Residential development is prohibited in Glacier Park, but not within the Townsite of Waterton Park. Recreational and tourist-related development is permitted to a certain extent in both parks; however, recreational projects and activities are subject to environmental review under the National Environmental Protection Act (NEPA 1970) in the United States, and the Canadian Environmental Assessment Act (CEAA 1995) in Canada. Both Waterton Lakes National Park and Glacier National Park have developed use zones in order to ensure that activities and developments are not located in areas where they may adversely affect natural resources. However, as with access corridor construction, development of many facilities in Waterton and Glacier was completed prior to the institution of current land-use zoning. As a result, some developments are located in prime grizzly habitat or along travel corridors for bears (Noss et al. 1996, Parks Canada 1997b). 46 In Glacier Park, grizzlies have been afforded additional protection under the ESA since 1975. According to section 7 of the ESA, the NPS must consult with the USFWS whenever a proposed development or activity may affect the continued existence of the grizzly bear or adversely modify its habitat. Actions or developments which might place the species in jeopardy must be modified until they receive a "no jeopardy" opinion from the USFWS. However, designation of "critical habitat" for listed species became mandatory after the listing of the grizzly bear; therefore, critical habitat was never determined, making it difficult for the USFWS to determine whether actions will or will not adversely modify the bears' habitat (Rohlf 1991). As part of the USFWS Grizzly Bear Recovery Plan (USFWS 1993), however, five Management Situations for grizzly bear habitat (MS1-MS5) were developed by the Interagency Grizzly Bear Committee (IGBC). Each M S fits a type of land area where unique management direction applies (see Appendix B), and all federal lands in the N C D E are zoned according to these situations. In Glacier Park, most areas are MSI , in which management decisions favour the needs of the grizzly bear when grizzly habitat and other land-use values conflict. Land-Use Management: Surrounding Jurisdictions In the jurisdictions surrounding Waterton-Glacier, land-use restrictions are generally less prohibitive. Industrial and agricultural activities like mining, ranching, timber harvesting, and oil and gas drilling occur on surrounding lands, as do residential and commercial development. There are some restrictions on land-use activity in each jurisdiction; however, few are as conservative as in the national parks. The variety of land ownership (federal, provincial, state, private) makes consistent land management nearly impossible. Within Montana's L C N F and FNF (aside from wilderness areas), most of the industrial and agricultural activities prohibited within Glacier National Park are allowed. However, NEPA and the ESA apply to actions and developments on Forest Service land as they do in Glacier, and the two national forests have been zoned according to the IGBC's Management Situations. Management Situation zoning of federal lands, although required by the USFWS, is carried out by the land management agency (in this case the USFS); consequently, the resultant zones have been criticised for simply conforming to pre-existing human land uses (Primm 1996). The IGBC has also developed guidelines for the execution of human activities, which have been incorporated into the Forest Plans of the FNF and 47 LCNF. The guidelines include land-use recommendations for road building, buffering, closure and restriction; allowable noise levels; helicopter flight altitudes, frequencies, and landing areas; and the timing and placement of activities such as seismic and exploratory drilling, pipeline construction, and livestock grazing (USDA 1985, USFWS 1993). The National Forests are also regulated by the National Forest Management Act (1976), which imposes planning and management restrictions on the forest lands, including some restraints on clearcutting and timber harvesting practices. Besides requiring the preparation of Forest Plans (see section the NFMA provides for the maintenance of biological diversity through ensuring minimum viable populations of certain "management indicator species." These indicator species are selected by the individual National Forest, and are included in the Forest Plan. The Lewis and Clark National Forest has selected the grizzly bear as a management indicator species (USDA 1985), although the Flathead has not (Keiter & Locke 1996). Forest plans are subject to assessment under NEPA and the ESA; an Environmental Impact Assessment (EIA) was prepared for the Flathead Forest Plan in 1993, which resulted in an amendment to the Forest Plan that reduced the Allowable Sale Quantity (ASQ) of timber from the forest (USDA 1994). Timber harvesting is prohibited within all designated National Forest wilderness areas, as is residential and commercial development. Federal acts such as NEPA and the ESA also apply to developments and actions within tribal lands on Montana's Blackfeet Indian Reservation. In addition, the reservation has implemented a Forest Management Plan (Sawyer & Talnagi 1994), which includes a restriction on timber harvesting in the vicinity of active bear dens. State agency activities in Montana are regulated by the Montana Environmental Protection Act (MEPA 1971). This Act is based on the federal NEPA, and it provides for the preparation of an EIS whenever a state agency undertakes an action that may significantly affect environmental quality. The state, of Montana also has a Board of Environmental Review, which issues certificates of environmental compatibility to large projects like landfills, electricity generation plants, pipelines, etc. On private lands in Montana, land use activities and developments are regulated through County Planning Office, although few restrictive regulations are passed. Local resistance to development constraints on private lands is strong (Keiter & Locke 1996). Under the ESA's section 9, citizens are not allowed 48 to destroy or modify critical grizzly habitat on private land; however, as "critical habitat" has not been designated, this provision has done little to protect the grizzly from harmful land uses (Keiter & Locke 1996). The Alberta portion of the CCE is composed of mostly provincial forest (crown) lands (Bow-Crow Forest Reserve and Poll Haven Community Pasture) and privately owned lands. Crown land in Alberta is of two types, White Zone and Green Zone. White Zone public lands may be made available for settlement or agricultural development, while the Green Zone lands are not (except for grazing leases). Green Zone lands are managed primarily for forest production, watershed protection, and recreation. The Bow-Crow Provincial Forest Reserve is primarily Green Zone, while Poll Haven is primarily White Zone. Developments on Crown Land are regulated by two independent tribunals: the Alberta Energy & Utilities Board (AUEB), which regulates energy related developments, and the Natural Resources Conservation Board (NRCB), which regulates non-energy projects such as paper or pulp mills, water diversion projects, mining activities, and large tourist developments. Both boards must consider the social, economic, and environmental effects of proposed actions, before allowing them to proceed. Despite these controls, quite a bit of development, especially oil and gas exploration, has occurred in southwestern Alberta, with little opportunity for public comment or regard for wildlife impacts (Horejsi 1989). As in Montana, municipal bylaws regulate land use activities and developments on Alberta's private lands. The most significant of these is a bylaw enacted by the Municipal District of Pincher Creek, which established a Waterton Lakes National Park Buffer Area (Keiter & Locke 1996). The zoning buffer prohibits land subdivision into parcels smaller than 160 acres, preventing high-density development and additional habitat fragmentation. In the southeastern corner of British Columbia, most land is under the jurisdiction of the BC Ministry of Forests (in the Flathead Provincial Forest) or BCMoE's Parks Division (in Akamina-Kishenena Provincial Park (AKPP)). In both jurisdictions, major land-use developments (mining projects, energy projects, larger transportation projects, water containment projects, and recreational/tourist developments) are subject to assessment under the BC Environmental Assessment Act (BCEAA 1995). Land within AKPP, as it is a Class 49 A provincial park, is not open to timber harvesting or residential or commercial development under any circumstances. Forestry practices are not considered reviewable under BCEAA. The forest industry was regulated by the Forest Act (1979) (with little consideration for wildlife resources) until the Forest Practice Code came into effect in 1995. On Crown lands such as the Flathead Provincial Forest, the FPC regulates forestry, ranching, and recreational uses via a series of guidebooks, regulations, and standards for sustainable use of provincial forests. It also provides for legal enforcement of these standards, and requires leaseholders to prepare a comprehensive Forestry Development Plan (FDP) or Range Use Plan (RUP) every five years. The FPC mandates the establishment of Resource Management Zones (RMZs), which specify the degree of care to be taken in ensuring resource development is compatible with environmental variables. It also addresses the requirements of wildlife species and plant communities at several scales via Biodiversity Guidelines and an Identified Wildlife Strategy. As the grizzly bear is an "identified wildlife species," industry development and land-use activities will be modified to complement grizzly bear management within established Wildlife Habitat Area (WHAs). However, the FPC allows the Identified Wildlife Strategy to have no more than a 1% impact on the Annual Allowable Cut (Province of BC 1997). In addition, the Biodiversity Guidelines, Resource Management Zones, and the Identified Wildlife Strategy have not been implemented in many areas because landscape units have not yet been designated (M. Besko, pers. comm.). In 1992, the government of British Columbia created the Commission on Resources and Environment (CORE) to prepare comprehensive land-use plans for much of the province. Developed out of the CORE process, the Kootenay Boundary Land-Use Plan provides additional direction for the timing, placement and execution of forestry practices in bear habitat, and recommends the creation of several protected areas, including a 31,108 ha "special management area" along the Flathead River. The KBLUP has no force of law, however, and only a few of its recommendations have been fully implemented (BCMoE 1998, Keiter & Locke 1996). 50 Chapter 4 Analysis of Mortality Distribution in the CCE 4.0 Methods: Mortality Analysis 4.0.1 Data Sources and Completeness Grizzly bear mortality data were collected by five wildlife agencies in the CCE. Generally, for each morality incident, the date, location, type of mortality, and age and sex of the bear were recorded. I requested grizzly bear mortality and translocation data for the period 1970-1997 from all relevant agencies in the CCE (refer to Table 4.1 for exact units from which mortality data were requested). Translocation data were requested in addition to mortalities, because relocations over 200 km were included in my analysis as a form of "management action" mortality. (Grizzly bears relocated distances greater than 200 km were unlikely to return to their area of capture (AFLW 1990); therefore, such bears were effectively lost from the CCE population.) Table 4.1 Units within each jurisdiction for which mortality and relocation data were requested. See Fig. 2.2 for map of area. Jurisdiction Units Management Units (MUs) 4-01 and 4-02, Kootenay Region AJfcert* Wildlife Management Units (WMUs) 300, 302, and 400 i§§§| North Continental Divide Ecosystem Entire park area Entire park area Requested records were complete and available in some jurisdictions, although in others, record keeping during the early 1970s was less thorough; thus, dates of data received varied (see Table 4.2). Mortality records are thought to be fairly complete in southwestern Alberta from 1972 on (Gunson 1995), and in British Columbia (at least for hunter harvest data) for 1976-1996 (B. Forbes, pers. comm.). In Montana, the most complete records are probably those from 1980 to the present (H. Pac, pers. comm.), although mortalities on Montana's Blackfeet Indian Reservation occasionally went unreported (or were reported with no locational data) up to the late 1980s (D. Carney, pers. comm.). Translocation data was not obtained for portions of the study area in Montana; however, relocation policies in Montana 51 were such that very few relocations are likely to have been over 200 km (Riley, et al. 1994, T. Manley, pers. comm.). In the national parks, grizzly mortality and translocation data has been collected longer than in most surrounding areas, and is probably more complete. 4.0.2 Review and Organisation of Mortality Data Data from 1976 to 1997 were selected for analysis, as these data were available and reasonably complete in all jurisdictions. Mortality incidents were sorted into the following seven categories of human-caused mortality: legal harvest, illegal kill (includes poaching mortalities, malicious kills, and mistaken identity kills), legal self-defence, treaty Indian (First Nations use), management action (includes agency destructions and relocations > 200 km), accidental (includes research related losses and vehicle collisions, among others), and unknown non-hunting mortalities. Mortalities from confirmed or suspected "natural causes" were not included. Table 4.2. Dates of mortality and relocation data obtained for each jurisdiction. The method by which location was recorded is indicated in parentheses beside each set of dates. U T M = Universal Transverse Mercator Co-ordinates; T,R,'/4S = Township, Range, and quarter Section; T,R,S = Township, Range, and Section; T,R + L = Township, Range, and a landmark. Ju risdiction Agency Mortality Data Relocation Data British Columbia BC Ministry of Environment 1976-1997 (UTM) 1970-1997 (UTM) Alberta Alberta Environmental Protection 1972-1997 (T,R + L) 1978-1997 (T.R.ViS) Montana MT Dept Fish Wildlife & Parks 1970-1997 (UTM; TRS) 1989-1997 (UTM) Glacier Park US National Park Service 1970-1997 (UTM) 1970-1997 (UTM) Waterton Park Parks Canada 1970-1997 (UTM) 1970-1997 (UTM) Methods of recording locational data (mortality, capture, and release points) varied by agency (see Table 4.2). All locational data were converted to Universal Transverse Mercator (UTM) co-ordinates (recorded to the nearest 100 m) for the purposes of this analysis. The centre of each Section or quarter-Section was used as a conversion point where locations were given in Township, Range and Section (or quarter Section). In some cases only Township and Range were recorded, along with some known feature of the landscape (usually a stream, ridge, mountain, or community). For these data, a point on the landmark closest to the centre of the named Township was used to obtain UTM coordinates. Locations that were not recorded with UTM coordinates or with some unit convertible to UTMs (by the above methods) were not included in the database. 52 4.0.3 Data Analysis Converted data from all five agencies (approximately 330 mortality points) were entered into a singular database. Using ArcView software, jurisdictional boundaries within the C C E were digitized onto a digital base map provided by the Crown of the Continent Ecosystem Data Atlas (CCEDA). Outward and inward from the Waterton-Glacier boundary, six concentric zones of width 10 km were then digitized onto the base map and labeled Z.2o through Z40 (Fig. 4.1). Although data were requested from a much larger area, the mortality analysis was restricted to the collective areas of these six zones, roughly 20,080 km 2 . This was done for ease of analysis, and because the smaller area better coincides with the "boundaries" of the C C E as well as with occupied bear habitat in the region. The database fields containing U T M information (zone, northing, and easting) were imported directly onto the basemap, with each mortality or relocation incident appearing as a single point (Fig. 4.2). Average annual density of mortality incidents (kills/100 km2) was determined for each zone, as well as for the segments of zone within particular jurisdictions. Zone mortality densities were additionally categorised according to type of mortality. Figure 4.1. Diagrammatic representation of zones created inward and outward from the Waterton-Glacier boundary (red). Zones 10 kilometers in width (aside from Z.2o) were created and labeled Z.2o, Z . 1 0 (inside the park) and Z 1 0 , Z2o, Z 3 0 , and Zw (outside the park). Zones cut off by the frame of this illustration (Z 3 0 and Z40) were complete during analysis. 53 Figure 4.2. Locations of grizzly bear mortalities in the CCE. A single blue circle represents each mortality incident. Jurisdictional boundaries are drawn in red. In keeping with BCMoE policy, I was not able to print the BC portion of this map for reasons of hunter confidentiality. Spatial analysis considered data for all years, and tested the null hypothesis that there were no differences in mortality density outside the park with distance from the park boundary. One-way analysis of variance assumes that each data group is an independent random sample from a normal population, and that data groups come from populations with equal variances. Although data from all groups (zones) were temporally independent (see Appendix A , Runs Test; the sequence of observations can be considered random), some groups were not normally distributed (see Appendix A , Test of Normality), and variances were not equal (see Appendix A , Test for Homogeneity of Variance). Therefore, the non-parametric Kruskal-Wallis Rank Test was applied, which ranks sample values from smallest to largest with no regard to which group (zone) each observation has come from, and then tests for significance based upon the sum of ranks assigned to values in each group (Conover 1980). Similar analyses considered equivalent hypotheses for single jurisdictions (BC, A B , and MT), and individual mortality types (legal harvest mortality and management action mortality). Where differences were determined to be statistically significant, multiple comparisons testing identified the zone(s) or area(s) where this difference occurred. Finally, differences in 5 4 mortality density between total park area and total non-park area were examined using the non-parametric Mann-Whitney U Test for two independent variables. 4.1 Results: Mortality Analysis 4.1.1 Summary of Data The mean annual number of mortalities in the study area (all zones and jurisdictions pooled) was 14.8 for the period 1976-1997 (n = 22, S.D. = 5.48). The highest mortality numbers were recorded in the early 1980s; the year of lowest mortality was 1994 (Fig. 4.3). Legal hunter harvest was the leading type of mortality in the study area (44.3 %) when data for all years, zones, and jurisdictions were pooled (Fig. 4.4). Management destruction and long-distance relocation were also quite prevalent (26.6 % of the area's mortality). 301 * • • • • • • • • • •— 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 Year Figure 4.3. Annual grizzly bear mortalities over the period 1976-1997 in all six zones (Z. 2 0 to Z 4 0). 55 Figure 4.4. Distribution of mortality type using data from all zones, jurisdictions, and years. Legal hunter harvest represented the largest percentage of the total, followed by management action and illegal kills, respectively. Among individual jurisdictions, stratification of mortality type was quite varied (see Fig. 4.5). Total grizzly bear mortality was spread over a greater variety of mortality types in Montana and Alberta than in B C or either of the two parks. Recorded mortality in the B C portion of the study area was almost entirely legal harvest (99%), and mortality in the two national parks was primarily from management actions (88% in Waterton, 68% in Glacier) and accidental kills (13% in Waterton, 28% in Glacier). In Montana and Alberta, mortality was spread more evenly among legal harvest, illegal kills, management actions, native harvest and accidental kills. 5 6 c 4) « a. Mortality Type I lUnknown Non-Hunt I iTreaty Indian I Iself-Defence I [Management Action • i l l e g a l Kill fJ ^ L e g a l Hunter Harvest I [Accidental "<? % % /!>. % % 'A Jurisdiction Figure 4.5. Percentage of bear mortality types stratified by jurisdiction. Mortality in the two parks consisted mostly of management actions and accidental kills. British Columbia's recorded mortality was primarily legal harvest, while grizzly mortality in Alberta and Montana was spread among a wider variety of mortality types 4.1.2 Spatial Analysis British Columbia had the highest mean annual mortality density of any jurisdiction in the study area (0.197 kills/ 100 km 2), followed by Alberta (0.108) and Montana (0.056) (Fig. 4.6). The lowest mortality densities were found within the two national parks. By zone (all jurisdictions pooled), average annual mortality density followed the predicted trend, with mortality densities significantly higher for zones outside the park (Z ) 0 , Z 2o, Z 3 0 , and Z40) than those inside the park (Z. 2 0 and Z., 0) (U = 100, P = 0.001) (Fig. 4.7). In addition, mortality density in Zw, the zone immediately beyond the park boundary, was significantly greater than densities within zones further outside the park (H = 20.96, df = 3, P = 0.000). The concentration of mortality in zones Z20, Z 3 0 , and Z40 was statistically similar. 57 'in C Q j? I i _ n • 3 ST 0.400 4 0.300 0.200 5 : o.ioo-i o.ooo Jurisdiction Figure 4.6. Mean annual grizzly bear mortality density (kills/100 km2) by jurisdiction, 1976-1997. The two national parks (GP & WP) have lower mortality densities than do surrounding areas. The highest mortality density for the period occurred in British Columbia. When the density within zones was assessed by individual jurisdiction, the above trend did not persist in every circumstance (Figs. 4.8 a-d). In Montana, density in Z i 0 was significantly higher than Z20-4o at the a = 0.05 level (H = 26.348, df = 3, P = 0.000). In Alberta, where Z 4 0 was eliminated from analysis because the data contained too few runs to be considered independent (Appendix A, Runs Test, Alberta), mortality density in Zw was significantly higher than in Z 2 0 and Z 3 0 only at the a = 0.10 level (H = 9.348, df = 2, P = 0.009). In British Columbia, the Runs Test eliminated Z i 0 from analysis (see Runs Test, BC), and Kruskal-Wallis testing determined no significant difference in mortality density among remaining zones. When densities of the two most common mortality types were analysed., legal harvest mortality density was not significantly different among Z20-40; however, density of management action mortality was significantly greater in Z10 than it was in Z20 and Z30 (H = 16.963, df = 2, P = 0.000) ( Z 4 0 was eliminated due to temporal dependence of data). 58 -20 -10 10 20 30 40 Zone Figure 4.7. Mean annual grizzly bear mortality density (kills/100 km2) by zone for the period 1976-1997. All jurisdictions were pooled in this diagram. Zones within the national parks (Z.i 0 and Z.2o) had lower densities than did those outside (Zi 0 - Z 4 0). Density was highest in the zone immediately outside the park boundary. M o n t a n a 10 20 30 40 Zone (a) 59 Alber ta •31 10 20 30 Zone GO Brit ish Co lumb ia •3 1 10 20 30 (c) Figure 4.8 a-c. Mean annual grizzly bear mortality density (kills/100 km2) by zone in the jurisdictions of British Columbia, Alberta, and Montana. Montana and Alberta (a and b) followed the pattern observed when entire zones were considered (Fig. 4.7): mortality density in Z 1 0 was higher than in Z20-40. In British Columbia (c), the reverse was true, and mortality density was lowest in Zw. 60 Chapter 5 Discussion of Grizzly Management and Mortality Distribution in the C C E 5.0 Waterton-Glacier Park and Surrounding Jurisdictions In Chapter 1, I predicted the existence of a gradient in grizzly bear mortality distribution associated with the boundary of Waterton-Glacier Park, based upon presumed differences in management between the park and its surroundings. I suggested that comparatively protective park management policies and practices for legal harvest, illegal kill control, preventative and responsive human-bear conflict management, and land use management would result in a lower mortality density inside the park than that found in the surrounding area. In fact, mortality data for the region did identify a significantly greater density of mortality incidents outside WGIPP than inside the park. Management, however, was not clearly more protective within the park for all of the above management categories. Mortality Type [~J Accidental J~H Management Action Q Unknown Non-Hunt U Legal Hunter Harvest | Self-Defence • Illegal Kill J j Treaty Indian Figure 5.1. Distribution of mortality types for park (right) and non-park (left) lands in the CCE. Roughly half of all non-park mortality was legal harvest; management actions and unknown non-hunt mortalities were also prominent. Within the parks, management action was the most common means of grizzly bear death. 61 Legal harvest management and land use management were decidedly more protective within Waterton and Glacier than outside the parks, primarily because hunting activities and many land uses were strictly prohibited in the parks. Also, preventative human-bear conflict measures (attractant management and education) were initiated earlier in the two national parks than in most surrounding areas, and were developed more thoroughly. On the other hand, illegal kill control and responsive human-bear conflict management did not provide markedly greater protection for grizzly bears within WGIPP. Regulations against killing grizzly bears were comparable across all five jurisdiction of the CCE, as were fines and punishments for offenders. The ESA provided additional protection against illegal killing in the US, but was not specific to US national parks. It is possible that enforcement of illegal kill regulations was more thorough in the two national parks than outside them; however, relative stringency of enforcement was difficult to assess from the information acquired during my interviews. Lack of reported illegal kills from WGIPP (Fig. 5.1) could suggest either successful enforcement of regulations or a deficiency in reporting. Responsive bear-human conflict management (RBHCM) policies were also not distinctly more protective within Waterton-Glacier than outside the parks. Due to general uncertainty regarding the effectiveness of different problem bear management techniques, it was difficult to determine any particular RBHCM policy's level of protection for the grizzly bear. However, there was no clear dichotomy between "park" and "non-park" RBHCM policies from 1970 to the present- each of the CCE's five jurisdictions had unique strategies for dealing with problem bears. Glacier Park had one of the lowest tolerances for human-bear interaction in the region, along with the province of BC, where most bears that sought and obtained unnatural foods were destroyed (or aversively conditioned in Glacier, as of 1998). Waterton Park allowed for greater intransigence before bear destruction was considered, as did the jurisdictions of Montana and Alberta. Interestingly, with the range of RBHCM policies found in the CCE, the parks had proportionately more management action mortality than surrounding lands; 28% of the CCE's management action mortality occurred in Waterton and Glacier, although the parks made up only 21% of the total study area. Efforts at mortality reduction in Waterton-Glacier should be focused on reducing management control kills, as they comprise the largest percentage of park mortalities (Fig. 5.1). Since further reduction of attractants may not be possible, parks should consider tightening access 62 controls in areas where concentrations of both grizzlies and humans have previously resulted in conflict (i.e. Many Glacier area in Glacier Park, and Cameron Lake in Waterton Park). Despite the fact that illegal kill control and R B H C M were not decidedly more protective within WGIPP, relatively conservative land use management, legal harvest management and attractant management policies presumably increased protection enough to produce significantly lower mortality levels in the park compared with outlying areas. The absence of legal harvest kills alone may have been sufficient to reduce park mortality to a level of significance. 5.1 Mortality and Management in Relation to Distance from the Park Boundary Within the region of higher mortality outside WGIPP, I predicted the greatest density of mortality would be found just beyond the park boundary, creating, in effect, a "halo" of higher mortality incidence around the protected area. A gradient configuration of this type was indeed observed, as the density of mortality was significantly higher within 10 km of the park boundary (Zio) than further outside (Z20-Z40). Overall, the observed pattern of mortality was very close to that predicted in Chapter 1 (compare Figs 1.4 and 4.7). The rationale behind the predicted pattern was that uncontrolled tourist access within the two national parks could result in an increased number of habituated bears in Z10, as well as increased goods and services development just beyond the boundary, both of which should increase local human-caused bear mortality. In the following sections, the strength of this explanation will be discussed based upon my analysis of management and other relevant research. 5.1.1 Relation of Access to Observed Mortality Halo Results of the management analysis indicated that access restrictions were not numerous within Waterton and Glacier parks. However, i f access in the parks was uncontrolled, it was certainly no less controlled than that outside the parks. Although no area in the CCE placed limitations on total visitor numbers, in the national parks, visitor distribution was at least managed via limited backcountry permits. Types of transportation were also restricted to a greater degree within the parks. However, these restrictions did not have a considerable 63 effect on the number of visitors to Waterton and Glacier each year, and the parks unquestionably received the heaviest tourist use in the region. During the summer season (June to September), concentration of visitors was high (over 2 million people often visited Waterton-Glacier annually, the majority utilising roads, frontcountry facilities, and the most accessible trails). In many older national parks like Waterton and Glacier, major access corridors were constructed af a time when few impacts to wildlife were considered. As a result, some areas of heavy human were located in prime bear habitat and along important wildlife movement corridors (Noss et al. 1996, Parks Canada 1997b). Heavy concentration of park visitors in these areas may contribute to high incidences of bear-human conflict (Craighead & Craighead 1970, Herrero 1985) and bear habituation to humans (Craighead & Craighead 1970, Hamer et al. 1985, Herrero 1987, Maw 1989, Stokes 1970) in national parks. In fact, there is some evidence that bears spending the majority of their time in Waterton and Glacier parks are less wary of humans than are those in outlying areas. In a study of the reactions of grizzly bears to hikers in Glacier National Park, MacArthur-Jope (1982) found that bears in high human use areas had habituated to people, and that only 5% of grizzlies moved immediately away from hikers they encountered (33% eventually moved away). Conversely, McLellan and Shackleton (1989), in the adjacent (non-park) Flathead River drainage of BC and Montana, observed that bears fled from people on foot in every instance in areas of low human use, and in 63% of instances in areas of high human use. Although sampling biases in the Glacier study may be responsible for a portion of this discrepancy (observations were from trails only), it is likely that bears in the two areas did respond differently to people because of differing encounter frequencies (McLellan & Shackleton 1989). Habituated behaviours have also been observed in bears within Waterton Lakes Park (Hamer et al. 1985). As a consequence of its proximity to Waterton-Glacier, Zw probably contained more habituated bears than did Z20, Z30, or Z 4 0 . Increased unwariness may be partially responsible for the greater density of human-caused bear mortality observed in Z10, since in most cases, habituated bears are more likely to be involved in human-bear conflicts (Herrero 1985, Peek et al. 1987) and more likely to be killed (Mattson et al. 1992) than are their wary counterparts. One note of caution here is that very little is known about the spatial and 64 temporal specificity of bears' habituation to humans. For example, in MacAxthur-Jope's (1982) study, although a greater proportion of bears exhibited habituated behaviours in a high human use area than in a less heavily used area, no attempt was made to determine whether individual bears were consistently habituated in all locations. Although access to WGIPP was unlimited in terms of overall visitor numbers, development of new access corridors was quite restricted. A 1982 study of access in and surrounding Glacier Park found that road and trail infrastructure within the park had changed little since the 1930s (Martinka 1982a). Outside the parks, corridor development constraints were not as restrictive; access immediately adjacent to Glacier's boundary was unusually developed, as paved highways and secondary roads paralleled the boundary for 78% of its length, often within 2 km of the border (Martinka 1982a). Some of this corridor development was undeniably related to park access management, in that it was tied to WGIPP's popularity as a tourist destination. Road development which arose as a result of industrial activity (see sections and along the western boundary of Waterton and the northwestern and eastern boundaries of Glacier was less associated with park visitation, although it also increased access to the area Commercial and residential development surrounding Waterton-Glacier was similarly tied to park visitation. There were some development constraints on the mosaic of lands beyond the park boundary (see section, most notably in designated wilderness areas such as Montana's Great Bear and Bob Marshall Wildernesses, and in BC's Akamina-Kishenena Provincial Park. In Alberta, commercial and residential developments in the Bow-Crow Forest Reserve were also limited, as was land subdivision in the Municipality of Pincher Creek. However, considerable development has occurred in close proximity to the WGIPP boundary, particularly surrounding Glacier Park, where the communities of West Glacier, East Glacier Park, Coram, St. Mary, and Polebridge lie within ten kilometres of the border (Fig. 2.2). These peripheral developments were not allowed within the park, although they provide many amenities for visitors and park employees, such as food, fuel, lodging, campsites, and souvenirs. Along Waterton's boundary, tourist-influenced development was less pronounced; however, a few campgrounds and guest ranches did exist in close proximity to the park. 65 Roads and other permanent human developments have been associated with increased human-caused bear mortality in several studies (Aune & Kasworm 1989, Brannon et al. 1988, McLellan & Shackleton 1988). Most relevantly, a study in the Greater Yellowstone Ecosystem (GYE) determined that the majority of grizzly deaths were clustered within and on the periphery of Yellowstone Park, in "population sinks" that corresponded to gateway communities, recreational developments, and other areas of high human use (Knight et al. 1988). Therefore, access to park road and trail systems, by influencing the amount of peripheral development and the degree to which bears are habituated, may influence the amount and distribution of human-caused bear mortality adjacent to the park. To some degree, a phenomenon similar to that observed in the GYE probably occurred outside WGIPP during the study period, as suggested by the significantly higher level of mortality in Zio. 5.1.2 Other Possible Contributing Factors Dispersal and Distance-Decay Mortality Patterns At this point, it is necessary to consider some additional factors that may have contributed to the "halo" distribution of mortality observed in the CCE study area. To begin with, bear population density may have been greater in Zi 0 than in Z2o-4o, which could have translated into greater mortality density adjacent to the park boundary. As estimated population numbers were fairly stable in Glacier Park over the past few decades, the park's grizzly population may have been at carrying capacity (Martinka 1974b), and bears may have been dispersing from the park into the adjacent landscape. If we assume that Waterton-Glacier Park was a "source" area, and that all outlying lands were "sink" areas into which bears were consistently dispersing, the "halo" distribution of mortality might have developed even if human-caused mortality pressure were constant across Z10-Z40. This is best explained in terms of a distance-decay pattern, in which the number of dispersing bears available to be killed decreases with distance from the source. For example, if the mortality rate was 0.50 in Zio, Z20, Z30 and Z40, 50 out of 100 bears dispersing from the park might have been killed in Zio, then 25 out of the remaining 50 in Z20, and so on, assuming surviving animals continued dispersing in a direction perpendicular to the park boundary. Thus, bears dispersing from the park could account for the observed mortality halo without any additional anthropogenic pressure (via habituation or gateway development) in Zi 0 . 66 For a number of reasons, the distance-decay dispersal model inadequately explains the mortality pattern observed in the CCE. First, dispersing bears are most often sub-adults (Blanchard & Knight 1991); therefore, if dispersal from the parks onto outlying lands was substantial, sub-adults would represent a larger proportion of bears killed adjacent to the park boundaries. However, according to age structure data for the region (Table 5.1), the percentage of sub-adult mortality was no different in Zi0 than for the entire study area (31%). (Cubs and yearlings, on the other hand, represented a larger proportion of the total mortality in Zio (17%) than overall (9%), which may suggest that sows with cubs were displaced into more vulnerable habitat (developed land adjacent to park boundary) in effort to secure food resources while avoiding competition with adult males (Albert & Bowyer 1991, Mattson et al. 1987, McLellan & Shackleton 1988).) Table 5.1. Age structure and sex of grizzly bear mortalities within Z i 0 , and for all zones (park and non-park). Sex of bears killed in Z i 0 was not different from that observed overall. Age structure was also similar, with one exception: cubs and yearlings made up a greater proportion of deaths in Z i 0 (17%) than overall (9%). Cub/Yearling Sub-Adult Adult Unknown Z 1 0 Only 17% 31% 48% 4% All Zones 9% 31% 53% 7% Male Female Unknown Z 1 0 Only 62% 36% 2% All Zones 61% 36% 3% In addition, although immigration of bears has been reported in several surrounding areas, especially north and east of Waterton in Alberta (AFLW 1990; Wielgus & Bunnell 1994), and east of Glacier on the Blackfeet Reservation (Mattson et al. 1996a), evidence suggests the two parks were not neatly bounded source areas in a homogeneous sink matrix. For example, in the Flathead Valley of British Columbia, the survival-fecundity rate r8 (exponential rate of population increase) was positive (+0.081) (McLellan 1989c) and greater than that in Alberta's Bow-Crow Forest Reserve (-0.01 to +0.01) (Wielgus & Bunnell 1994), which suggests that bears may have been dispersing from southeastern BC into less productive areas, possibly even into Waterton or Glacier Parks (see section 5.2.1). Unfortunately, rates of population growth have been estimated for only a few areas in the 67 CCE, and parameters of dispersal in grizzlies have not been well documented (Weaver, et al. 1996). Habitat Quality and Seasonal Migrations There are other conditions under which grizzly population density might be greater in Zi 0 than in Z2o-4o- For instance, habitat quality could decrease naturally with distance from the park boundary, influencing the number of bears that can be sustained per unit area of land. This was certainly the case along portions of WGTPP's eastern boundary, where habitat shifted from montane forest to shortgrass prairie (see section 2.2.2). In these areas, an increased human-caused mortality rate may not have been responsible for the higher mortality density observed in Zio (see sections 5.2.2 and 5.2.3). At certain times of the year, seasonal migrations may also bring a greater number of bears to the WGIPP boundary than would be expected randomly. Early spring and late fall are characterised by a relative scarcity of food items for bears, which often prompts considerable movements in search of edible material (Martinka 1982b, Mattson et al. 1992, Servheen 1983). Elevational migrations (see section 2.2.3) bring some bears into the CCE's valley bottoms, where green vegetation and winter killed ungulates (and human developments with unnatural attractants) are more plentiful. Coincidentally, Waterton Park and Glacier Park (like many national parks) preserve the region's highest elevations, and portions of their boundaries lie along major valley bottoms. This is especially true for the western boundary of Glacier, which lies in the valley of the north and middle forks of the Flathead River. Therefore, it is also plausible that a greater concentration of bears existed along this boundary in the spring and/or fall due to seasonal migrations. If higher mortality densities in Zio were the result of seasonally higher bear concentrations along the boundaries of WGIPP, mortality would likely be greatest in early spring and late fall. Seasonal distribution of mortalities in Zi 0 indicate that mortality (all types pooled) was highest in the fall (September through November) but was quite low during the spring (March through May) (Fig 5.3a). Excluding harvest mortality (which was biased according to hunting season dates) did not substantially alter this pattern (Fig. 5.3b). Nor was the pattern markedly different from that observed when all zones (Z.20 though Z4o) were pooled (Fig 5.3c). Therefore, seasonal distribution of mortality does not appear to be different for 68 Zio than for the remainder of the study area. As a whole, bears died in the highest numbers during autumn. Migrations during this season may have brought them into contact with humans more frequently; but the phenomenon does not appear particular to Zi 0 . 70 "3 60 e S a v n f r o S E 3 50 40 30 20 a. spring autumn 50 spring summer autumn Season 69 spring summer autumn Season Figure 5.3, a-c Seasonal distribution of all grizzly bear mortality in Zio (a), and seasonal distribution of non-hunt grizzly mortality in Z 1 0 (b) and in all zones (c). The pattern was similar for all three diagrams; mortality was much higher in autumn than in spring. Concentrated Hunter Effort Finally, it is conceivable that legal bear harvesters deliberately sought out those areas closest to the park boundary, believing (whether correctly or incorrectly) that there was a greater likelihood of encountering a bear here. While hunter harvest did account for the largest percentage of mortality in the study area (Fig. 4.4), mortality density from this activity was not significantly greater in any of the non-park zones. Therefore, it is unlikely that hunter effort was concentrated in Zio. 5.2 Consideration of Individual Jurisdictions The complexity of boundary generated gradients was emphasised in the first chapter. Because gradient configurations are neither consistent nor predictable in time and space, it is difficult to assess their possible origins and suggest what, if anything, can be done about them. My system of mortality analysis by zones identified a gradient on a coarse scale, but a closer examination is needed to determine its continuity along the boundary. In the following sections, I will consider the mortality pattern and management of each of the non-70 park jurisdictions separately in attempt to better understand the origin(s) and distribution of the identified gradient, as well as to determine the influence of each jurisdiction on its neighbours. 5.2.1 British Columbia British Columbia had the highest average annual mortality density of any jurisdiction in the study area (Fig. 4.6). However, the southwestern corner of BC contains extremely productive bear habitat, and density estimates from this area are high even on a continental scale (McLellan 1989a). The likelihood of unusually concentrated bear populations in BC's Flathead Valley must therefore be taken into account when drawing comparisons of mortality density between jurisdictions of the CCE. Ninety-nine percent of recorded human-caused mortality in the BC portion of the study area was legal harvest. The virtual absence of non-hunting mortality probably does not reflect reality; the province concedes that unreported mortality may be as high as 100% of the reported mortality (BCMoE 1995b). Data from McLellan (1989b) also suggest considerable illegal mortality in this area, as 5 out of 9 radio-collared bear deaths were presumed illegal during a 1979-1987 study. Increased access in recent decades (Fig. 5.4) has probably elevated levels of illegal killing although the Flathead Valley remains relatively sparsely populated by humans. British Columbia did exercise some access controls (see section, and restrictions will probably increase once the Forest Practices Code gets further underway. The province constrained access more than did neighbouring Alberta, where the open and more heavily settled landscape probably made road closure more difficult from a political standpoint. However, although access is easier to restrict in BC's sparsely settled, forested, Flathead Valley, there is a much greater chance that illegal activities will go unnoticed in the event that roads are not closed. British Columbia is the only non-park jurisdiction in which a significantly higher mortality concentration was not found adjacent to the park boundary in Zio; the opposite trend was, in fact, observed (Fig 4.8c). This is largely a reflection of the wilderness character of Akamina-Kishenena Provincial Park, which buffers the Waterton-Glacier boundary along much of its length in BC (Fig. 2.2). Few mortalities were recorded within AKPP, where no settlement, forestry activity, road development, or motor vehicle traffic were permitted. It is interesting that although legal harvesting of grizzlies was allowed in AKPP, vehicle 71 restrictions either turned away most hunters or lowered the hunter success rate to the degree that no legal harvest was recorded here. In addition, no "halo" of mortality existed in association with the AKPP boundary, as Z2o, Z 3 0 , and Z40 had statistically similar concentrations of bear mortality. The provincial park effectively buffered Waterton and Glacier from gateway development in BC, and was saved from gateway development itself by declining to focus on tourism. "973 1986 Figure S.4. Two-wheel drive access roads in British Columbia's East Kootenay Region, 1952,1962, 1973, and 1986. The BC portion of my study area is at the southeastern tip of these diagrams (graphic from McLellan 1991: 118). Although the expected mortality gradient was not observed in BC, there is cause for examination of the area's relationship to the remainder of the ecosystem. The CCE grizzly bear population, taken as a whole, probably depends upon BC's Flathead Valley as a source area. Provincial grizzly management in this region should proceed with that in mind. 72 Human settlement of the southeastern corner of BC is increasing quite rapidly (Trant et al. 1997), and certain aspects of the current management system may need to be re-evaluated once the human population reaches some critical threshold. To counter the probable increase in non-hunt mortality that will arrive with human residents, southeastern BC's public education program should be developed more fully, and responsive bear-human conflict management expanded to include aversive conditioning techniques. At the very least, development standards and attractant management standards for private and public lands could be instituted while the human population is still modest. A relatively simple means of decreasing mortality in this area of BC, should the need arise, is reduction of the legal harvest (McLellan 1991). British Columbia is fortunate that in it has so many avenues for reduction of mortality; the options of other jurisdictions are limited by comparison. If the province repeats the mistakes of other areas, however, grizzly conservation effort in all five of the CCE's jurisdictions may be wasted. 5.2.2 Alberta Concentration of grizzly mortality in Alberta was significantly greater closest to the Waterton-Glacier boundary (H = 9.384, df = 2, P = .009). Management action mortality was the leading cause of grizzly death in Zi 0 (25%), and fluctuated considerably during 1975-1997 (see Fig. 5.5). Fluctuations were probably tied to annual differences in food availability, as bears may have sought alternate (often unnatural) food sources in lean years, increasing the likelihood of bear-human conflict. Despite annual variability, management action mortality was extraordinarily high in Alberta, and was significantly more concentrated in Zio for the entire study area (H = 16.963, df = 3, P = 0.000). Although promising if an indicator of increasing bear populations, the abnormally high number of management control actions in 1997 (Fig. 5.5) suggests that the potential for human-bear conflict has not been eliminated in Alberta. Many management action mortalities adjacent to Waterton Park were tied to the ranching industry (AFLW 1990, Gunson 1995), an expanding sector in southwestern Alberta (Trant et al. 1997). Horejsi (1989) recognised two areas, both within 10 kilometres of Waterton, where domestic stock operations were particularly threatening to grizzly bears: Poll Haven Community Pasture, and the Blood Indian Timber Reserve (Fig. 2.2). Mortality distribution 73 does appear to be clustered at these locations, as well as within the ranching community of Twin Butte, north of the park. Unnatural attractants (livestock foods, boneyards, calving grounds) usually play a key role in ranching-related grizzly mortalities, and although Alberta was rigorous about bear-proofing public waste receptacles and campgrounds, its private lands had few attractant regulations. Improvements in ranching attractant management are needed, whether they are achieved through legislation, incentives, or co-operative agreement. Southwestern Alberta's recent effort to focus education and information on ranching interests is a good one, and may reduce intolerance for bears in addition to minimising attractants. Year Figure 5.5. Annual grizzly deaths due to management action in the Alberta portion of the study area. Management action mortalities varied somewhat between years, and were very high in 1997. The program of responsive bear-human conflict management in Alberta also contributed to increased management action mortality by encouraging long distance translocations (usually over 200 kilometres). Long distance bear relocations (76% of all management action "deaths" in the Alberta study area), while socially more acceptable than destruction, were equally as damaging to the local population. In addition, high removal rates north of Waterton Park may appear sustainable only because of immigration from BC and/or the national parks (AFLW 1990, Horejsi 1989). To the degree that long distance tranlocations 74 can be reduced through attractant management or local relocations, they should be. The aversive conditioning program currently being developed in southwestern Alberta may also help reduce management action mortality and make possible local tranlocations. Illegal kills were the second most common type of mortality in Alberta's Z10 (14%). Grizzlies killed illegally by ranchers defending their livestock in Alberta were occasionally reported; however, most often, unlawfully killed bears were shot and left by persons and under circumstances unknown (Gunson 1995). Monetary compensation for livestock and crop losses, while it did not eliminate illegal control kills, may have kept mortality below what it would have been in the absence of damage compensation. Little is known about the degree to which access contributes to illegal mortality rates in this area, although most low elevation land north of Waterton has been roaded (Horejsi 1989). As road closures and/or restrictions were rarely instituted, and no overall access management plan existed for this portion of the province, unfettered access may have elevated mortality levels in southwestern Alberta. Legal harvest in Alberta's portion of the CCE was not permitted from 1970 through 1983, nor has the amount of legal harvest been substantial in the last decade, probably as a result of limited entry licensing and conservative harvest target percentages instituted in 1990. However, from 1975 to 1997, legal harvest comprised 11% of the total mortality in Alberta's Zio, and elimination of the legal harvest should be considered in this area if ever mortality needs to be reduced and alternate methods are not practical. Finally, while there was little park influenced gateway development in Alberta's Zio, there as an abrupt change in land use over the Waterton Park boundary, from protected parkland to ranching community. Origins of a mortality "halo" in Alberta were likely related to the proliferation of ranching in this area coupled with Zio's location in a transition area between forested mountain habitat (productive for bears) and grassland habitat (less productive for bears). Although large, contiguous blocks of ranch land adjacent to a park may create a shallower environmental gradient (allowing greater bear movement) than would commercial and residential subdivision (Knight et al. 1994), it is clear that some unique bear-human conflicts are associated with ranching in grizzly bear habitat. Conflict related non-hunting mortality was the greatest threat to bears in this area of Alberta, and grizzlies crossing over 75 the park boundary in search of food (natural or attractant related) may have been eliminated within one of the first ranching communities they encountered. As Waterton Park has one of the highest visitor concentrations of any Canadian national park (Environment Canada 1991), habituation of bears to humans may also have contributed to the observed mortality pattern in Alberta 5.2.3 Montana Grizzly bear mortality and management are more difficult to summarise for the Montana portion of the study area than for Alberta or British Columbia, as Montana had considerably more diversity in land ownership, habitat quality, and observed mortality types. However, mortality locations clearly lined the entire western boundary of Glacier Park (Fig. 4.2), and it is likely that kills abutting this western boundary were responsible for the significantly higher mortality density observed in Montana's Z i 0 (H = 26.348, df = 3, P = 0.000). Of the grizzly deaths along Glacier's western boundary in Montana, most fell into four categories: legal hunter harvest, illegal kill, management action, and accidental kill. Hunter kills were concentrated along the southern half of the western boundary (upper middle fork of the Flathead River), in an area encompassing a portion of the Great Bear Wilderness (see map in Fig 2.2). Before Montana's grizzly hunt was discontinued in 1991, there were no restrictions on the number or distribution of participating hunters (seasons were terminated once the quota was reached), and wilderness areas were open for hunting earlier than non-wilderness hunting districts. Beyond Zio, hunter harvest was also clustered within the Great Bear and Bob Marshall Wildernesses; Dood et al (1986) cited these areas as possessing the greatest portion of the NCDE's legal harvest from 1973-1985. Illegal kills and management action mortalities were fairly evenly distributed along the western park boundary in Montana Although there was no obvious clustering of kills in known community developments (West Glacier, Coram, Polebridge), Montana's portion of the Flathead Valley is generally quite roaded and contains considerable private landholdings. Coupled with the deficiency of attractant management regulations on private lands and the lack of restrictions on road development or use, it is not surprising that kills were concentrated along the unusually accessible and developed western park boundary. 76 Mortality here was clearly tied to gateway development arising from park popularity, and possibly also to increased habituation of bears within the park. As the (human) population growth rate in Montana's Flathead Valley is projected to be 17% by the year 2000 (Mace & Waller 1998), private land attractant minimisation and development restriction are needed. Accidental kills were comparatively high in Montana's portion of the CCE, where they were concentrated along the southern half of the western park boundary. Many of these deaths resulted from collisions along two major travel corridors which parallel this stretch of boundary, U.S. Hwy 2 and the Burlington Northern Railroad (BNRR) (see Fig. 2.2). Habituation to slow-moving cars within Glacier Park may have increased bears' vulnerability outside the park, where traffic moves more swiftly. Li 1991, the Burlington Northern Environmental Stewardship Area (BNES A) was created along the corridor from East Glacier Park to West Glacier (Dood & Pac 1993). The BNESA seeks to identify frequent track crossing areas for bears and make them safer, as well as to reduce the number and attractiveness of grain spills resulting from train derailment (Dood & Pac 1993). Since the stewardship area's establishment, average annual accidental kills have decreased. The eastern portion of the Montana study area was located on the Blackfeet Indian Reservation (BTR). Here the amount of mortality adjacent to the boundary was not as marked as it was outside Glacier's western boundary. However, the number of reservation mortalities used in my analysis was lower than the actual, as locational information was not recorded for a considerable number of mortalities on the BTR, and these data could not be included in the analysis. On the Blackfeet Reservation, most of the recorded mortality was either management action or illegal killing. No legal harvest was permitted. As in Alberta, ranching is prevalent, and unregulated attractants often prove deadly for grizzlies. Attractant management on the BIR is improving, although improvements are slightly behind those of neighbouring jurisdictions (see section Older dumpsters on tribal land are still in the process of being replaced by bear-resistant ones, while private lands and residences, as in most of the non-park CCE, remain largely unregulated and a significant source of attractants. Some gateway development does exist on the BTR, most notably in the areas of East Glacier Park and St. Mary. A clustering of illegal kills was evident near East Glacier Park, in an area identified as requiring special enforcement measures during the mid 1980s (Dood et al. 77 1986). Although the concentration of grizzly mortality was significantly higher in Zio than Z20-40 for all of Montana, this pattern was more obvious for the western half of the study area than for the eastern half. A more complete data set might have indicated a prominent mortality gradient along the park boundary; however, it is important to note that habitat becomes marginal for grizzlies the further east one travels on the BIR, and clustering of kills near the park border may have been a reflection of bear distribution. 5.3 Summary of Discussion In this boundary analysis, I have identified a coarse-scale gradient in grizzly bear mortality distribution over the Waterton-Glacier park boundary, with mortality density lowest in the parks and highest in the first 10 km beyond the park boundary. And I have illuminated major differences in management among jurisdictions of the CCE. The discrepancy in mortality density between the park and surrounding jurisdictions is consistent with differences in management, especially differences in legal harvest and preventative human-bear conflict management. The variance in mortality between Zi 0 and Z20-40 is more difficult to explain, but appears at least partially related to boundary-generated interactions between visitor management in parks and land-use management on surrounding lands. Other possible factors include dispersal of bears from the parks and distribution of habitat quality, which may have produced higher grizzly bear population densities adjacent to the park border. Currently, the "halo" distribution of mortality is not continuous around the entire international park (see Fig. 4.8), and might also be described in terms of a source-sink structure in which most of the mortality sinks are located within Zio. Low removal rates inside Waterton and Glacier Park may be mediating losses outside the park (Horejsi 1989, Martinka 1982b), although it has not been demonstrated that losses are unsustainable on an ecosystem-wide basis. Even with sustainable total losses, if this pattern of mortality continues, or becomes more prominent, it may effectively isolate bears with the park from those on surrounding lands, preventing genetic interchange between the two areas. This would not bode well for either group of bears, as neither may be numerous enough to persist indefinitely in isolation (due to environmental, genetic, and demographic factors such as reduced heterozygosity, inbreeding depression, or environmental catastrophes) (Franklin 78 1980, Noss et al. 1996). Therefore, it is important that solutions are found to reduce mortality levels adjacent to the Waterton-Glacier administrative boundary. 79 Chapter 6 Recommendations 6.0 Reducing the Mortality Gradient The preceding chapter established a need for the reduction of human-caused grizzly bear mortality adjacent to the adrninistrative boundary of Waterton-Glacier Park. Based upon my discussion, Table 6.1 suggests means by which each jurisdiction could reduce human-caused bear mortality within its bounds. Many of these management changes could be instituted on a temporary basis and monitored to determine their effectiveness before being continued or modified. In a more integrated approach to the reduction of mortality, several wildlife-boundary interaction studies (Forbes & Theberge 1996, Samson & Huot 1998, Woodroffe & Ginsberg 1998), have recommended the establishment of buffer zones as a way of increasing effective reserve size and filtering out inappropriate human influences on adjacent lands. In the following sections, I examine the feasibility and potential benefits of implementing buffer zones for Waterton-Glacier Park. 6.0.1 National Park Buffer Zones Both Waterton Lakes National Park and Glacier National Park are designated Biosphere Reserves, which, according to the United Nations Educational, Scientific, and Cultural Organisation (UNESCO), are supposed to consist of a core reserve surrounded by a buffer zone in which people are integrated with the landscape in a sustainable manner (Tangley 1988, UNESCO 1984). However, biosphere reserves in both Canada and the US (Waterton and Glacier included) have had difficulty implementing reserve systems that resemble UNESCO's model (Shafer 1999, Solecki 1994). Political backlash, primarily from private landowners' fear of government regulations, has transformed the "buffer zone" into a controversial concept, and North American biosphere reserves tend to foster weak partnerships in "zones of co-operation" instead of implementing regulations within a specified buffer. Resistance to buffer zones may be less pronounced in Canada than in the US (Shafer 1999), and Waterton Park has actually established a Management Committee to co-ordinate efforts in an area of unspecified dimension (Lieff 1985). However, this committee understands that the biosphere designation will not place restrictions on landowners, and it has primarily focused on less controversial issues than grizzly bear mortality management (Lieff 1985). 80 Table 6.1. Management recommendations for reduction of grizzly bear mortality adjacent to the Waterton-Glacier Park boundary and within the entire Crown of the Continent Ecosystem. Waterton Park mm ACCESS: a Consider,seasonal/temporary •ciosure;ofrradsandstrals;(orresta frequented by grizzlies before an incident arises " ][ t ' , > ATTRACTANTS Implement stiffer fines for improperly secured attractants , EDUCATION:*'Gonsideroffering a;'Hx>undary?^  interpretative hikerswalkcertam lengths oftboijndary.a^  boundary-related issues Try outreaching to the surrounding communities in addition to tourists •> > ' LAND^USEMANAGEMENT: .Build no new-facihtie^  that experience frequent bear-human conflict Glacier Park ACCESS: Consider seasonal/temporary closure of roads and trails (or restriction of visitor numbers) in areas known to be frequented by grizzlies before an incident arises. ATTRACTANTS: Implement stiffer fines for improperly secured attractants. EDUCATION: Consider offering a "boundary" interpretative hike: walk certain lengths of boundary and discuss park boundary-related issues. Try outreaching to the surrounding communities in addition to tourists. RESPONSIVE CONFLICT MANAGEMENT: Continue aversive conditioning for "conditioned" and "nuisance" bears. LAND-USE MANAGEMENT: Build no new facilities in high bear use areas. Consider relocating any structures/roads that experience frequent bear-human conflict. British Columbia ACCESS Monitor Forest Practices Code's success at restricting access and planning ecologically-conscious road ' networks. sMake landscape im VAHCs and AMAs while the FPC gets underway ILLEGAL KILL CONTROL Increase efforts to record and monitor all non-hunting kills -ATTRACTANTS: -Institute private land^ attractant regulahms before residential populations waste storage and disposal at all public and private campgrounds Bear-proof all landfills EDUCATION: Start greater outreach to residents and pnvate campground owners, of^ ^^  management and the province's grizzly bear management strategy RESPONSIVE CONFLICT MANAGEMENT Institute an aversive conditioning program within and surrounding developed areas » » ' LEGAL HARVEST Consider reducing the target harvest if non-hunting kills increase LAND-USE MANAGEMENT limit and plan carefully any resource use adjacent to Waterton-Glacier or AKPP s 1 Alberta ACCESS: Institute more co-ordinated access management in SW Alberta, especially in forested areas. EDUCATION: Continue with rancher-focused information campaign. ATTRACTANTS: Reduce attractants on private lands via legislation, education, or co-operative agreements. Continue government cost-sharing for electric fences. Monitor effectiveness of carcass redistribution. RESPONSIVE CONFLICT MANAGEMENT: Continue local translocation of female bears, and if successful, consider local translocation of males. Develop a structured aversive conditioning program. LAND-USE MANAGEMENT: Consider purchasing conservation easements on lands with frequent human-bear conflict. ACCESS No new road development adjacent to Glacier Park Consider independent monitoring and enforcement of road closure policies wilhm National Torests ATTRACTAOTS:;Reduceattractants on private lands via-legislation; educate NCDE Special Order, and modify it to require bear-proof containerisation for attractants during night hours EDUCATION: svWimm me GGE;;sw landowners and ranchers (especially in communities close to Glacier Park) ' v '"J. LEG AL'HARVEST: «Gonsiderelim matiott f i r j f j ^ S f ^ RESPONSIVE CONFLICT MANAGEMENT Place special emphasis on aversive conditioning adjacent to Glacier Park, especially along major highways/railways and in developments along the park's boundary * - >, * •> ( L A N T M J S E ^ I A N A G J ^ 81 Although Parks Canada and the US National Park Service favour "co-operative planning" in lieu of more direct approaches to boundary management, a recent article by Shafer (1999) suggests the implementation of enforceable buffer zones is far preferable to the mosaic of co-operative and uncooperative areas usually achieved using benign land use planing methods. In addition to increasing effective reserve size for wide-ranging species by moving harmful impacts away from the park's core, buffer zones can "turn hard edges into soft ones" (Shafer 1999: 54) by allowing only a subset of those human activities regularly permitted on outlying lands. Essentially, buffer zones can reduce the abruptness of generated gradients by lessening the contrast between human management systems of adjoining jurisdictions. 6.0.2 Customised Buffer Zones As generated gradients are very complex, buffer zones designed to reduce the sharpness of gradients should mirror their spatial and temporal complexity (Schonewald-Cox & Bayless 1986). Such customised buffer zones could have a variety of different (possibly overlapping) prescriptions of various dimensions and durations, according to the generated gradient's intensity, configuration, and presumed origin(s). For example, in the CCE, a buffer zone in which private land attractant regulations are established and enforced would be most beneficial along the western boundary of Glacier Park or the northeastern boundary of Waterton. Similarly, a buffer zone in which all long distance translocations are eliminated is needed most along the Waterton-Alberta boundary. However, buffer zones do not always have to be restrictive, nor do they have to be outside the park. Non-restrictive conservation methods can be stepped up within buffer segments; for example, rancher education (outreach to landowners) could be increased within 10 km of the park boundary to the north and east of Waterton-Glacier Park, or aversive conditioning efforts could be made a higher priority within a certain distance from the park boundary. Development of Customised Buffer Zones It is possible that some of the recommendations in Table 6.1 can be translated directly into appropriate buffer segments for minimisation of grizzly bear mortality adjacent to the Waterton-Glacier boundary. However, more detailed study will be necessary in many cases before efficient buffers can be designed (Ambrose & Bratton 1990). Priority could be given to sites where generated gradients are most pronounced, and localised collapse of protection is more likely (i.e. 82 north and east of Waterton; west and southwest of Glacier) (Schonewald-Cox & Bayless 1986). At these sites, more focused and comprehensive boundary analysis could be performed to determine gradient configuration and origination. Employing GIS technology can greatly facilitate such analyses once appropriate themes have been developed. For example, coverages of road networks, vegetation types, human population/presence, and grizzly bear habitat indices might be overlaid with bear mortality distribution at priority sites in order to interpret patterns and possible causal relationships. Establishment and Implementation Once specific and scientifically sound buffer zones are proposed, there is no guarantee that they will be established and implemented, and once implemented, there is no guarantee that they will be effective. Li the current political climate, it is doubtful that any buffer zones incorporating private lands will be established by legal force. Buffers affecting only a change in federal, provincial, or state lands are more likely to be implemented, although probably not as a result of actions taken by the park service(s). National parks have historically shown reluctance in using the law against their neighbours in transboundary issues (Sax & Keiter 1987). Aside from outright purchase of land by groups or agencies interested in developing park buffers, co-operative planning among government agencies and between governments and private interests may still be the most realistic way to implement customised buffer zoning. Participation in large-scale initiatives like the NCDE subcommittee of the IGBC, as well as smaller ones like the Waterton Biosphere Reserve Management Committee, can help build the working relationships and trust that are necessary to develop much-needed buffers in the CCE. Enforceable buffers are far more likely to be implemented if all relevant stakeholders (including and especially the local public) have participated in their design. Predictably, however, development of buffers through co-operative planning will require years of effort and considerable forethought on the part of planners. Adaptive Management As to whether customised buffers will work once they have been established, adaptive management is critical. If monitoring (preferably before and after buffer establishment) indicates that a buffer is not significantly reducing mortality adjacent to the Waterton-Glacier boundary, management should be altered and monitoring continued. Long-term monitoring at different 83 scales is especially important given the spatial and temporal plasticity of generated gradients (Ambrose & Bratton 1990, Schonewald-Cox & Bayless 1986), since what is an effective buffer zone today may not be appropriate in another decade. Understanding boundary effects will involve some trial and error, but should allow for the development of efficient buffer zones well-tailored to local conditions and trends. 84 Chapter 7 Conclusion 7.0 National Parks and Grizzly Bear Conservation For several decades, it has been recognised that national parks cannot function as ecological isolates (Coggins 1987, Freemuth 1991, Janzen 1986, Sax 1985). Surrounding lands must be managed in collaboration with park goals in order to ensure that regional resources are protected. Especially with regard to wide-ranging mammal species, parks often cannot provide sufficient area or solitude, and populations will depend upon the land and resources of multiple jurisdictions. These transboundary populations are affected by each human management system in each jurisdiction they encounter, as well as by the interaction between these systems. My intention in completing this study, therefore, was not find the national parks of the United States and Canada, or any other individual jurisdiction, accountable for grizzly bear mortality. My results strengthen the contention that bear management in all of the CCE's jurisdictions has contributed to the observed pattern of grizzly mortality in the region. However, very often in national park systems today, "external threats" are blamed for impending or actual environmental damage without due recognition of the park's role in the condition and management of outlying lands. For the grizzly bear, an animal whose chances of survival generally decrease with increasing human encounter frequency, wilderness areas, or even carefully managed multiple-use lands, may be more effective "critical cores" than are popular national parks. I suggest this not because bear mortality in the parks is excessive (although this can be the case), but because the dynamic between "crown jewel" national parks and their surroundings can produce a greater ecosystem with far more human activity than that associated with some wilderness areas or multiple use areas. On the other hand, the Rocky Mountain national parks of the US and Canada are extremely important areas for grizzly bear conservation, as are all areas which support grizzlies in the southern part of the bears' range. Additionally, national parks are an essential element of the North American landscape. Canadians' and Americans' national identities are based in part upon a connection with the unique environments of their home nations, and in this process, the national parks systems have had an instrumental role. Therefore, parks like Waterton and 85 Glacier are living histories, and popular use is fundamentally tied to their well-being (Foresta 1984, Sax 1985). Since reversion to a wilderness protected area is not desirable for Waterton-Glacier, even if it were possible, it makes sense to acknowledge and adjust to the fact that the parks cannot be purely preservational in their direction, just as we recognise these limitations of surrounding areas. Greater recognition of the interaction between parks and their surroundings, and of the shared responsibility for bear mortality across all jurisdictions, is a first step to designing more effective solutions to the problem. 7.1 Final Notes Protected area boundaries have been likened to the skins of organisms (Schonewald-Cox 1988) in that the condition of the boundary can indicate the health of a park ecosystem just as the condition of the dermis can attest to the health of an animal. If, as Schonewald-Cox and Bayless suggest, boundary processes are more important than are internal jurisdictional processes in many aspects of conservation, the Crown of the Continent Ecosystem, with its abundance of management jurisdictions, is an area to be kept under close surveillance. This study and others have demonstrated the need to emphasise boundary processes in planning park protection and evaluating protection effectiveness. In fact, boundary processes should be emphasised more thoroughly in all conservation of wildlife. Very rarely are administrative boundaries other than park borders examined or even recognised for their possible effects upon wildlife, which is disquieting given that there are 61 states, provinces, and territories in the continental US and Canada (representing thousands of kilometres of administrative boundary), each managing wildlife according to different policies. Although transboundary issues often seem difficult to address from a jurisdictional perspective, it is essential that we make the effort. If we cannot protect the majestic grizzly bear within and surrounding Waterton-Glacier International Peace Park, a protected area managed as an example of peace, friendship, and co-operation beyond boundaries, there is little hope for the bear elsewhere. 86 References Achuff, P.L., R.L. McNeil, and M.L. Coleman. 1997. Ecological land classification of Waterton Lakes National Park, Alberta: soil and vegetation resources. Paper presented at the Waterton Lakes State of the Ecosystem Symposium, Waterton Park, AB. Albert, D. M . and R.T. Bowyer. 1991. Factors related to grizzly bear-human interactions in Denali National Park. Wildlife Society Bulletin 19(3): 339-349. Alberta Environmental Protection (AEP). 1997. Alberta Guide to Hunting Regulations: 1997. Natural Resources Services, Edmonton, AB. Alberta Forestry Lands, and Wildlife (AFLW). 1988. Alberta Public Lands. Edmonton, AB. . 1990. Management Plan for Grizzly Bears in Alberta. Wildlife Management Planning Series No. 2. Edmonton, AB. Alberta Natural Resources Service (ANRS). 1997. Southwestern Alberta Grizzly Strategy. Ambrose, J.P. and S.P. Bratton. 1990. Trends in landscape heterogeneity along the borders of Great Smoky Mountains National Park. Conservation Biology 4(2): 135-143. Aune, K. and W. Kasworm. 1989. Final Report- East Front Grizzly Studies. Montana Department of Fish, Wildlife, and Parks: Helena. Blackfeet Fish and Wildlife Department 1996. Draft Bear Management Plan and Guidelines for Bear Management on the Blackfeet Indian Reservation. Browning, MT. Blanchard, B.M. and R.R. Knight. 1991. Movements of Yellowstone grizzly bears. Biological Conservation 58: 41-67. Brannon, R.D., R.D. Mace, and A.R. Dood. 1988. Grizzly bear mortality in the Northern Continental Divide Ecosystem, Montana. Wildlife Society Bulletin 16(3): 262-269. British Columbia Ministry of Environment, Lands, and Parks (BCMoE). 1979. Preliminary Grizzly Bear Management Plan for British Columbia. Victoria, BC. . 1995a. A future for the grizzly: British Columbia Grizzly Bear Conservation Strategy. Victoria, BC. . 1995b. Conservation of Grizzly Bears in British Columbia: BackgroundReport. Victoria, BC. 87 . 1995c. Bear-People Conflict Prevention Plan for BC Parks. Conservation Services, Victoria, BC. . 1996a. Human-Bear Conflict in British Columbia: Draft Discussion Paper. Victoria, BC. . 1996b. Nuisance Bear Management Policy. Kootenay Region. . 1997. Hunting and Trapping Regulations Synopsis. Victoria, BC. . 1998. Memorandum ofUnderstanding between BCMinistry of Forests andBC Ministry of Environment, Lands, and Parks regarding Instruction for the Preparation of the 1998 Forest Development Plans. Kootenay Region. British Columbia Ministry of Forests (BCMoF). 1989. Forest Planning: A Guide to Co-ordinated Access Management Planning. Integrated Resources Branch. Bruns, E. Grizzly bear management in Alberta, in R.K McCann (ed.). Grizzly Bear Management Workshop. Proceedings of March 20-21,1991, Revelstoke, BC. Buechner, M . 1987. Conservation in insular parks: simulation models of factors affecting the movement of animals across park boundaries. Biological Conservation 41: 57-76. Carr, H.D. 1989. Distribution, Numbers, and Mortality of Grizzly Bears in and around Kananaskis Country, Alberta. Fish and Wildlife Division Management Series 3. Clarkson, P.L. 1989. The twelve gauge shotgun: a bear deterrent and protection weapon. Bear- People Conflicts. Proceedings of a Symposium on Management Strategies, Northwest Territories Department of Renewable Resources: 55-60. Coggins, G.C. 1987. Protecting the wildlife resources of national parks from external threats. Land and Water Law Review 22(1): 1-27. Commission on Resources and Environment (CORE). 1994. East Kootenay Land Use Plan. Victoria, BC. Conover, W.J. 1980. Practical Nonparametric Statistics. Second Edition. John Wiley & Sons: Toronto, ON. Cowan, I.M. 1972. The status and conservation of bears (Ursidae) of the world. International Conference of Bear Research and Management 2: 343-367. Craighead, J.J. and F.C. Craighead. 1972. Grizzly bear-man relationships in Yellowstone National Park. International Conference of Bear Research and Management 2: 304-332. 88 Dalle-Molle, J.L. and J.C. VanHorn. 1989. Bear-people conflict management in Denali National Park. Bear- People Conflicts. Proceedings of a Symposium on Management Strategies, Northwest Territories Department of Renewable Resources: 121-127. Dearden, P. and R. Rollins. 1993. Parks and Protected Areas in Canada: Planning and Management. Oxford University Press: Toronto. Doak, D.F. 1995. Source-sink models and the problem of habitat degradation: general models and applications to the Yellowstone grizzly. Conservation Biology 9: 1370-1379. Dood, A.R., R. Brannon, andR. Mace. 1986. Final Programmatic Environmental Impact Statement: the Grizzly Bear in Northwestern Montana Montana Department of Fish, Wildlife and Parks, Helena, MT. Dood, A.R. and H.I. Pac. 1993. Five-Year Update of the Programmatic Environmental Impact Statement: the Grizzly Bear in Northwestern Montana. Montana Department of Fish, Wildlife and Parks, Helena, MT. Environment Canada. 1989. Waterton Lakes National Park Background Information: Management Plan Review. , 1991. State of the Parks: 1990 Profiles. Parks Service. Minister of Supply and Services Canada, Ottawa. Flath, R. and J. Gunson. 1995. Discussion Paper on the Handling of Bears in Alberta. Occasional Paper No. 15. Wildlife Management Division, Alberta Environmental Protection, Edmonton, AB. Forbes, G.J. and J.B. Theberge. 1996. Cross-boundary management of Algonquin Park wolves. Conservation Biology 10(4): 1091-1097. Foresta, R. 1984. America's National Parks and their Keepers. Resources for the Future, Inc.: Washington DC. Franklin, I.R. 1980. Evolutionary change in small populations. In M E . Soule and B. A. Wilcox (eds.). Conservation Biology: an Evolutionary-Ecological Perspective. Sinauer Associates, Inc.: Sunderland, MA. Freemuth, J.C. 1991. Islands Under Siege: National Parks and the Politics of External Threats. University Press of Kansas: Lawrence, KS. Gadd, B. 1995. Handbook of the Canadian Rockies. Corax Press: Jasper, AB. 89 Gibbard, M.J., and D.H. Sheppard. 1992. Castle Wilderness Environmental Inventory. Special Publication No. 1 of the Castle-Crown Wilderness Coalition. Gniadek, S.J. and K.C. Kendall. 1997. A summary of bear management in Glacier National Park, Montana, 1960-1994. International Conference on Bear Research andManagement 10: 1-5. Gunson,J.R. 1995. Analysis of Grizzly Bear Mortalities in Alberta during 1972-1994. Alberta Environmental Protection, Natural Resources Service, Edmonton, AB. . 1996. Analysis of the Limited-Entry Grizzly Bear Hunts in Alberta during 1988-1996. Alberta Environmental Protection, Wildlife Management Division, Edmonton, AB. Hamer, D., S. Herrero, and K. Brady. 1985. Studies of the Grizzly Bear in Waterton Lakes National Park. Prepared under Contract WR-149-83 for Parks Canada, Calgary. . 1991. Food and habitat used by grizzly bears. Ursus arctos. along the continental divide in Waterton Lakes National Park,, Alberta. Canadian Field-Naturalist 105(3): 325-329. Hanna, W.L. 1978. The Grizzlies of Glacier. Mountain Press Publishing Co.: Missoula, MT. Herrero, S. 1974. Conflicts between man and grizzly bears in the national parks of North America Bears, their Biology and Management Third International Conference on Bear Research and Management 3: 121-145. . 1985. Bear Attacks: Their Causes and Avoidances. Nick Lyons Books: New York. Horejsi, B.L. 1989. International land-use threatens an international grizzly bear population. Conservation Biology 4(4): 220-223. Hummel, M. 1990. A Conservation Strategy for Large Carnivores in Canada. World Wildlife Fund Canada, Toronto, ON. Interagency Grizzly Bear Committee (IGBC). 1986. Plan for Determining Grizzly Bear Nuisance Status and for Controlling Nuisance Grizzly Bears. Janzen, D.H. 1986. The eternal external threat In M Soule (ed). Conservation Biology: The Science of Scarcity and Diversity. Sinauer Associates, Inc.: Sunderland, MA. Keay, J. A. and M G . Webb. 1989. Effectiveness of human-bear management at protecting visitors and property in Yosemite National Park. Bear- People Conflicts. Proceedings of a Symposium on Management Strategies, Northwest Territories Department of Renewable Resources: 145-154. 90 Keiter, R.B. and H. Locke. 1996. Law and large carnivore conservation in the Rocky Mountains of the U.S. and Canada. Conservation Biology 10(4)1003-1012. Kellert, S.R., M . Black, CR. Rush, and A. J. Bath. 1996. Human culture and large carnivore conservation in North America Conservation Biology 10(4): 977-988. Knick, S.T. and W. Kasworm. 1989. Shooting mortality in small populations of grizzly bears. Wildlife Society Bulletin 17:11-15. Knight, R.L., G.N. Wallace, and W.E. Riebsame. 1994. Ranching the view: subdivisions vs. agriculture. Conservation Biology 9(2): 459-461. Knight, R.R., B.M. Blanchard and L.L. Eberhardt. 1988. Mortality patterns and population sinks for Yellowstone grizzly bears, 1973-1985. Wildlife Society Bulletin 16: 121-125. Kootenay Inter-Agency Management Cornrnittee (KJMC). 1997. Kootenay/Boundary Land Use Plan: Implementation Strategy. Lieff, B. 1985. Waterton Lakes Biosphere Reserve: developing a harmonious relationship. Parks 10(3): 9-11. MacArthur-Jope, K. 1982. Interactions between grizzly bears and hikers in Glacier National Park Montana: Final Report. Contract no. PX 1430-1-0623. Oregon State University: Corvallis, OR. Madel, M . 1996. Rocky Mountain Front Grizzly Bear Management Program: Four-Year Progress Report 1991-1994. Montana Department of Fish, Wildlife, and Parks: Helena, MT Mace, R.D., J.S. Waller, T.L. Manley, L J . Lyon, and H. Zuuring. 1997. Relationships among grizzly bears, roads, and habitat in the Swan Mountains, Montana. In R.D. Mace and J.S. Waller (eds.). Final Report: Grizzly Bear Ecology in the Swan Mountains, Montana. Montana Department of Fish, Wildlife, and Parks: Helena, MT. Mace, R.D. and J.S. Waller. 1998. Demography and population trend of grizzly bears in the Swan Mountains, Montana. Conservation Biology 12(5): 1005-1016. Martin, P. 1983. Factors influencing globe huckleberry fruit production in northwestern Montana International Conference on Bear Research and Management 5: 159-165. Martinka, C.J. 1974a. Ecological role and management of grizzly bears in Glacier National Park, Montana. Third International Conference on Bears, Their Biology, and Management. Paper 13: 147-156. 91 . 1974b. Population characteristics of grizzly bears in Glacier National Park, Montana. Journal of Mammalogy 55(1): 21-29. . 1982a Effects of co-terrninous land use on grizzly bears in Glacier National Park. Paper presented at the American Association for Advancement of Science Symposium on External Threats to Ecosystems of National Parks. Washington, DC. . 1982b. Rationale and objectives for management in grizzly bear sanctuaries. Forty-seventh North American Wildlife and Natural Resources Conference. Wildlife Management Institute, Washington DC. Mattson, D. J., R.R. Knight and B.M. Blanchard. 1987. The effects of developments and primary roads on grizzly bear habitat use in Yellowstone National Park, Wyoming. International Conference of Bear Research and Management 8: 33-56. . 1992. Yellowstone grizzly bear mortality, human habituation, and whitebark pine seed crops. Journal of Wildlife Management 56(3): 432-442. Mattson, D. J., B.M. Blanchard, R.G. Wright, and C M . Pease. 1996a Designing and managing protected areas for grizzly bears: how much is enough? In Wright, R.G. National Parks and Protected Areas- Their Role in Environmental Protection. Blackwell Science; Cambridge, MA. Mattson, D.J., S. Herrero, G. Wright, and C M . Pease. 1996b. Science and the management of Rocky Mountain grizzly bears. Conservation Biology 10(4): 1013-1025. Maw, R.R. 1989. Visitor Attitudes, Perceptions and Knowledge concerning Bears and Bear Management Practices in Waterton Lakes National Park Canada. Ph.D. Thesis, University of Alberta. McCullough, D.R. 1982. Behavior, bears, and humans. Wildlife Society Bulletin 10:27-33. McLellan, B.N. 1989a. Dynamics of a grizzly bear population during a period of industrial resource extraction. I. Density and age-sex composition. Canadian Journal of Zoology 67:1856-1860. . 1989b. Dynamics of a grizzly bear population during a period of industrial resource extraction. II. Mortality rates and causes of death. Canadian Journal of Zoology 67: 1861-1864. . 1989c. Dynamics of a grizzly bear population during a period of industrial resource extraction. III. Natality and rate of increase. Canadian Journal of Zoology 67: 1865-1868. . 1989d. Effects of Resource Extraction Industries on Behaviour and Population Dynamics of Grizzly Bears in the Flathead Drainage, British Columbia and Montana. Ph.D. Thesis, University of British Columbia Vancouver BC. 92 _. 1991. Current status and long term threats to grizzly bears m British Columbia ;>? S. Rautio (ed.). Community Action for Endangered Species. Vancouver, BC. McLellan, B.N., and F.W. Hovey. 1995. The diet of grizzly bears in the Fkthead Wver drainage of southeastern British Columbia. Canadian Journal of Zoology 73(4): 704-712. McLellan, B.N. and D.M. Shackleton 1988. Grizzly bears and resource-extraction industries: effects of roads on behaviour, habitat use and demography. Journal of Applied Ecology 25: 451-460. 1989. Immediate reactions of grizzly bears to human activities. Wildlife Society Bulletin 17: 269-274. Newmark, W.D. 1985. Legal and biotic boundaries of western North American parks: a problem of congruence. Biological Conservation 33: 197-208. Noss, R., M. Quigley, T. Hornocker, and P.C. Paquet. 1996. Conservation biology and carnivore conservation in the Rocky Mountains. Conservation Biology 10(4): 949-955. Paquet, P.C, T. Clark, and A. Peyton-Curlee. 1996. Large carnivore conservation in the Rocky Mountains of the United States and Canada. Conservation Biology 10(4): 936-939. Parks Canada 1995. Waterton Resource Guide: Waterton-Glacier International Peace Park World Heritage Site. Waterton Park, AB. . 1997a Rocky Mountain District: Draft Bear/Human Conflicts Management Plan. Banff National Park, AB. . 1997b. An Integrated Core Monitoring Program for Waterton Lakes National Park: Discussion Draft. Waterton Park, AB. Peek, J. M , M.R. Pelton, H.D. Picton, J.W. Schoen, and P. Zager. 1987. Grizzly bear conservation and management: a review. Wildlife Society Bulletin 15:160-169. Primm, S.A. 1996. A pragmatic approach to grizzly bear conservation. Conservation Biology 10(4): 1026-1035. Province of British Columbia 1997. Forest Practices Code: Procedures for Establishing Wildlife Habitat Areas: Volume 1. Queen's Printer for British Columbia, Victoria, BC. Raine, R. M 1991. Grizzly bear mortality in Yoho and Kootenay National Parks, British Columbia, in R.K McCann (ed.). Grizzly Bear Management Workshop. Proceedings of March 20-21, 1991, Revelstoke, BC. 93 Ralf, R. 1995. History of Bear/Human ConflictManagementin Jasper National Park: 1907-1995. Jasper National Park, AB. Rasker, R. and A. Hackman. 1996. Economic development and the conservation of large carnivores. Conservation Biology 10(4): 991-1002. Riley, S.J., K.A. Aune, R.D. Mace, and M J . Madel. 1994. Translocation of nuisance grizzly bears in northwestern Montana International Conference of Bear Research andManagement 9(1): 567' 573. Rohlf, D.J. 1991. Six biological reasons why the Endangered Species Act doesn't work- and what to do about it. Conservation Biology 5(3): 273-282. Samson, C. and H. Huot. 1998. Movements of female black bears in relation to landscape vegetation type in southern Quebec. Journal of Wildlife Management 62(2): 718-727. Sawyer, P. and K. Talnagi. 1994. Forest Management Plan for the Blackfeet Indian Reservation. Bureau of Indian Affairs, Blackfeet Agency, Browning, MT. Sax, J. 1985. A rain of troubles: the need for a new perspective on park protection. In P. Dooling (ed.). Parks in British Columbia: Emerging Realities. Proceedings of the Symposium on Parks, February 17-19, 1984. D.W. Friesen & Sons, Vancouver. Scheer.B. 1991. The role ofBC Parks in the preservation of the grizzly bear, in R K McCann (ed.). Grizzly Bear Management Workshop. Proceedings of March 20-21,1991, Revelstoke, BC Schmidt, K.J. 1995. The Crown of the Continent Ecosystem: Infonnauon Paper No. 1. Glacier National Park Homepage (www.nps.gov/glacMol.html). Schonewald-Cox, C M 1988. Boundaries in the protection of nature reserves: translating multidisciplinary knowledge into practical conservation. Bioscience 38(7): 480-486. Schonewald-Cox, C M and J.W. Bayless. 1986. The boundary model: a geographical analysis of design and conservation of nature reserves. Biological Conservation 38:305-322. Servheen, C. 1983. Grizzly bear food habits, movements, and habitat selection in the Mission Mountains, Montana. Journal of Wildlife Management 47(4): 1026-1035. . 1989. The management of the grizzly bear on private lands: some problems and possible solutions. Bear- People Conflicts. Proceedings of a Symposium on Management Strategies, Northwest Territories Department of Renewable Resources: 195-200. 94 Shafer, C L . 1999. US National Park buffer zones: historical, scientific, and legal aspects. Environmental Management 23(1): 49-73. Sidorowicz, G. A. and F.F. Gilbert. 1981. The management of grizzly bears in the Yukon, Canada. Wildlife Society Bulletin 9(2): 125-135. Solecki, W.D. 1994. Putting the biosphere reserve concept into practice: some evidence of impacts in rural communities in the United States. Environmental Conservation 21(3): 635-641. Stokes, A. W. 1970. An ethologist's view on the management of grizzly bears. Bioscience 10:1154-1157. Storer, T.I. and L.D. Tevis. 1955. California Grizzly. University of Nebraska Press: Lincoln, NB. Strickland, D. 1989. Teach your grizzlies well. Wyoming Wildlife, October 1989. Wyoming Game and Fish Commission. Tangley, L. 1988. A new era for biosphere reserves. Bioscience 38(3): 148-155. Taylor, J.S. 1984. Bear Management in Canadian National Parks: Fifteen Essential Elements. M.Sc. Thesis, Faculty of Environmental Design, University of Calgary, Calgary, AB. Tilson, D. and O. Dodd. 1989. Bear Management Plan for Waterton Lakes National Park. Parks Canada Warden Service, Waterton Park, AB. Titus, K. 1991. Forest management and brown bear ecology in southeast Alaska in R.K McCann (ed.). Grizzly Bear Management Workshop. Proceedings of March 20-21,1991, Revelstoke, BC. Trant, D., C. deBoer, and K. VanTighem. 1995. Environmental Change in the Crown of the Continent Ecosystem: Socio-economic Trends around Waterton Lakes National Park. Statistics Canada, National Accounts and Environment Division. UNESCO 1984. Conservation, Science, and Society: Contributions to the First International Biosphere Reserve Congress. 26 September-2 October 1993, Minsk, USSR. United States Department of Agriculture (USDA). 1985. Lewis and Clark National Forest Plan. USDA Forest Service, Lewis and Clark National Forest, Great Falls, MT. _. 1994. Forest Plan Amendment #19: Objectives and Standards for Grizzly Bear Management and Timber Production: Environmental Assessment. USDA Forest Service, Flathead National Forest. 95 United States Department of the Interior (USDI). 1971. Bear Management Plan. National Park Service, Glacier National Park, West Glacier, MT. . 1975. Bear Management Plan. National Park Service, Glacier National Park, West Glacier, MT. . 1976. Bear Management Plan. National Park Service, Glacier National Park, West Glacier, MT. . 1988. Bear Management Guidelines: Glacier National Park. National Park Service, Glacier National Park, West Glacier, MT. . 1998. Bear Management Guidelines: Glacier National Park. National Park Service, Glacier National Park, West Glacier, MT. United States Fish and Wildlife Service (USFWS). 1993. Grizzly Bear Recovery Plan. Missoula, MT. VanEgmond, T. 1997. A Literature Review of Selected Management Indicator Species for Waterton Lakes National Park. Prepared by Axys Environmental Consulting, Calgary, AB. Weaver, J.L., P.C. Paquet and L.F. Ruggiero. 1996. Resilience and conservation of large carnivores in the Rocky Mountains. Conservation Biology 10(4): 965-975. Wielgus, R. and F. Bunnell. 1994. Dynamics of a small, hunted brown bear population in southwestern Alberta, Canada. Biological Conservation 67: 161-166. Woodroffe, R. and J.R. Ginsberg. 1998. Edge effects and the extinction of populations inside protected areas. Science 280:2126-2128. Woods, G. 1991. Grizzly bear management in the Kootenay region, in R K McCann (ed.). Grizzly Bear Management Workshop. Proceedings of March 20-21,1991, Revelstoke, BC. 96 Appendix A Statistical Test Results 97 1. Significance Testing for Mortality Density among Zio. Z ? Q. Z30. and Z A P (all jurisdictions) Runs Test for Data Independence Z10 Z20 Z30 Z40 Test Value .16300 4.9000E-02 4.4000E-02 3.9000E-02 Cases < Test Value 10 7 9 4 Cases >= Test Value 12 15 13 18 Total Cases 22 22 22 22 Number of Runs 13 11 10 7 Z .260 .000 -.514 -.035 Asymp. Sig. en-tailed) .795 1.000 .607 .972 la Median Runs Test H 0 : data are independent Hv data are not independent a = 0.05 p > 0.05 for all zones therefore, H 0 is accepted. Data are independent. Histogram: Zona Ton 0.00 .06 .13 .19 36 3\ J 8 Danaky of MortaUtias (kms/ log aq. km.) In Z10 Histogram: Zona Twonty 0.000 .026 .050 .075 .100 .125 Dandy of Mortalities (kffla/ioa aa, km.) m Z10 Histogram: Zona Thirty I... 16 20 DanaRy of Mortalttlaa (KU1U100 sq. km.) bl ZJ0 Histogram: Zona Forty 0.000 .026 .060 .076 .100 .125 DansKy of MortaJWaa (WUs/100 aa. km.) In Z40 98 Test of Normality Density Kolmogorov-Smirnov Zone Statistic df Sig. 10 .160 22 .148 20 .193 22 .033 30 .154 22 .191 40 .225 22 .005 a Lilliefors Significance Correction Test of Normality Ho: Within groups (zones), data are normall distributed. Hi: Data are not normally distributed, a = 0.05 Pzio = • 148 HO is accepted. PZ20 = .033 HO is rejected. Pz3o=191 HO is accepted. Pz4o = 005 Ho is rejected. Not all data are normally distributed. c. Test of Homogeneity of Variance Levene df1 df2 Sig. Statistic Density Based on 7.711 3 84 .000 Mean Based on 6.632 3 84 .000 Median Test of Homogeneity of Variance Ho: Variances are equal Hi: Variances are not equal a = 0.05 p = 0.00 therefore Ho is rejected Kruskal-Wallis Ranks Zone N Mean Rank density 10 22 65.70 20 22 40.61 30 22 37.41 40 22 34.27 Total 88 Kruskal-WallaceTest Statistics density Chi- 20.963 Square df 3 Asymp. .000 Sig. a Kruskal Wallis Test b Grouping Variable: zone Kruskal-Wallis Rank Test H0: Mortality density is the same among zones. Hi: Mortality density is significantly different among zones. a = 0.05 p = 0.00 Therefore, H 0 is rejected. 99 2. Significance Testing Among Zones in Alberta Only a. Runs Test Zone Ten Zone Twenty Zone Thirty Zone Forty Test Value3 .13700 5.3500E-02 4.3500E-02 .00000° Cases < Test Value 6 11 11 0 Cases >= Test Value 16 11 11 22 Total Cases 22 22 22 22 Number of Runs 9 13 9 1c Z -.127 .218 -1.092 Asymp. Sig. (2-tailed) .899 .827 .275 a - Median b All values are greater than or less than the cutoff. Runs Test cannot be performed. c - Only one run occurs. Runs Test cannot be performed. Runs Test (Alberta) H0: Data are independent. Hi: Data are not independent, a = 0.05 p > 0.05 for zones 10, 20, and 30. H 0 is accepted for these zones. Too few runs to perform test in zone 40. Therefore, zone 40 is removed from analysis. b. Ranks Zone N Mean Rank kills/100 sq. km. 10 22 43.32 20 22 28.84 30 22 28.34 Total 66 Test Statistics3''' kills/100 sq. km. Chi-Square 9.384 df 2 Asymp. Sig. .009 a. Kruskal Wallis Test b. Grouping Variable: Zone Kruskal-Wallis Rank Test (Alberta) H0: Mortality density is the same among zones Zio, Z2o, and Z30 in Alberta. Hi: Mortality density is significantly different among zones Z 1 0 , Z2o and Z30 in Alberta. a = 0.05 p = 0.009 Therefore, H 0 is rejected. Mortality density is not the same among zones. Multiple Comparisons Dependent Variable: kills/100 sq. km. Tamhane Mean 90% Confidence Interval Difference Lower Upper 0) Zone (J) Zone 0-J) Std. Error Sig. Bound Bound 10 20 .19645* .065 .052 2.3610E-02 .36930 30 .19495* .065 .055 2.1512E-02 .36840 20 10 -.19645* .065 .052 -.36930 -2.361 E-02 30 -1.500E-03 .065 1.000 -6.581 E-02 6.2808E-02 30 10 -.19495* .065 .055 -.36840 -2.151 E-02 20 1.5000E-03 .065 1.000 -6.281 E-02 6.5808E-02 Tamahane Multiple Comparisons Test (Alberta) Zone 10 is significantly different from Zone 20 and Zone 30 at the a = 0.10 level. Zone 20 and Zone 30 are mutually similar. *• The mean difference is significant at the .10 level. 100 3. Significance Testing Among Zones in Montana Only a. Runs Test MT Zone MT Zone MT Zone MT Zone Ten Twenty Thirty Forty Tost Value3 .12000 3.7000E-02 1.6500E-02 3.0O00E-02 Cases < Test Value 10 9 11 10 Cases >= Test Value 12 13 11 12 Total Cases 22 22 22 22 Number of Runs 11 10 12 15 Z -.180 -.514 .000 1.142 Asymp. Sig. (2-tailed) .857 .607 1.000 .253 a. Median b. Ranks Zone N Mean Rank kills/100 sq. km. 10 22 68.05 20 22 37.70 30 22 36.80 40 22 35.45 Total 88 Test Statistics3'1' kills/100 sq. km. Chi-Square df Asymp. Sig. 26.348 3 .000 a- Kruskal Wallis Test b. Grouping Variable: Zone Runs Test (Montana) H0: Data are independent. Hi: Data are not independent, a = 0.05 p > 0.05 for all zones; therefore, H 0 is accepted. Kruskal-Wallace Rank Test (Montana) H0: Mortality density is the same among zones Z 1 0, Z 2 0, Z30, and Z40 in Montana. Hi: Mortality density is significantly different among Z 1 0 l Z2o, Z30, and Z40 in Montana. a = 0.05 p = 0.00 Therefore, HO is rejected. Mortality density is not the same among zones. c. Multiple Comparisons Dependent Variable: kills/100 sq. km. Tamhane Mean 95% Confidence Interval Difference Lower Upper (I) Zone (J) Zone (l-J) Std. Error Sifl. Bound Bound 10 20 .10445* .019 .001 3.9608E-02 .16930 30 9.8000E-02* .019 .002 3.0686E-02 .16531 40 9.9636E-02* .019 .001 3.2596E-02 .16668 20 10 -.10445* .019 .001 -.16930 -3.961E-02 30 -6.455E-03 .019 .994 -3.798E-02 2.5071E-02 40 -4.818E-03 .019 .999 -3.556E-02 2.5924E-02 30 10 -9.800E-02* .019 .002 -.16531 -3.069E-02 20 6.4545E-03 .019 .994 -2.507E-02 3.7980E-O2 40 1.6364E-03 .019 1.000 -3.514E-02 3.8416E-02 40 10 -9.964E-02* .019 .001 -.16668 -3.260E-O2 20 4.8182E-03 .019 .999 -2.592E-02 3.5560E-02 30 -1.636E-03 .019 1.000 -3.842E-02 3.5143E-02 Tamahane Multiple Comparisons Test (Montana) Z 1 0 is significantly different from Z2o, Z30, and Z40 at the a = 0.05 level. Z2o, Z3o, and Z40 are mutually similar. *• The mean difference is significant at the .05 level. 101 4. Significance Testing Among Zones in British Columbia Only Runs Test BCZone BCZone BCZone BCZone Ten Twenty Thirty Forty Test Value9 .00000° .12850 .11150 .27900 Cases < Test Value 0 11 11 11 Cases >= Test Value 22 11 11 11 Total Cases 22 22 22 22 Number of Runs 1c 11 9 13 Z -.218 -1.092 .218 Asymp. Sig. (2-tailed) .827 .275 .827 a. Median b. All values are greater than or less than the cutoff. Runs Test cannot be performed. c Only one run occurs. Runs Test cannot be performed. Runs Test (BC) H0: Data are independent Hi: Data are not independent, a = 0.05 p > 0.05 for Z 2 0 , Z3o, and Z40. H 0 is accepted for these zones. Too few runs to perform test in Zi 0; therefore, Z10 is removed from analysis. Ranks Kruskal-Wallis Rank Test (BC) Zone N Mean Rank kills/100 sq. km. 20 22 33.43 30 22 30.16 40 22 36.91 Total 66 TestStaasacs3" kills/100 sq. km. Chi-Square 1.479 df 2 Asymp. Sig. .477 a. Kruskal Wallis Test b. Grouping Variable: Zone H0: Mortality density is the same among Z2o, Z30, and Z40 in BC H1: Mortality density is significantly different among Z 2 0 , Z30, and Z40 in BC. a = 0.05 p = 0.477 Therefore, H 0 is accepted. Mortality density is the same among Z2o, Z30, and Z40. 102 5. Significance Testing for Hunter Kill Density Among Z10. Z20. Z30, and Z40 (all non-park jurisdictions) a. Runs Test Zone Ten Zone Twenty Zone Thirty Zone Forty Test Value 3 2.70E-02 2.5000E-02 2.2000E-02 3.900E-02 Cases < Test Value 7 6 9 10 Cases >= Test Value 15 16 13 12 Total Cases 22 22 22 22 Number of Runs 8 9 7 13 Z -1.038 -.127 -1.872 .260 Asymp. Sig. (2-tailed) .299 .899 .061 .795 a - Median Runs Test (Hunter Harvest Data) H0: Within groups, data are independent Hi : Within groups, data are not independent. a = 0.05 p > 0.05 for all zones. Therefore, H 0 is accepted. Data are independent. Ranks Zone N Mean Rank kills/100 sq. km. 10 22 46.23 20 22 43.77 30 22 42.82 40 22 45.18 Total 88 Kruskal-Wallis Test (Hunter Harvest) H0: Legal harvest mortality density is the same among Z 1 0 l Z2o, Z30 and Z40. H1: Legal harvest mortality density is significantly different among Z-|0, Z2o, Z30 and Z40. Test Statistics3" kills/100 sq. km. Chi-Square .235 df 3 Asymp. Sig. .972 a - Kruskal Wallis Test b - Grouping Variable: Zone a = 0.05 p = .972 Therefore, H 0 is accepted. Legal harvest density is the same among all tested zones. 103 6. Significance Testing for Management Action Mortality Density Among Zm. Z ? n . Z30, and ZAP (all non-park jurisdictions) Runs Test Zone Ten Zone Twenty Zone Thirty Zone Forty Test Value? 2.58E-02 1.1364E-03 3.0000E-03 4.364E-03 Cases < Test Value 11 21 19 18 Cases >= Test Value 11 1 3 4 Total Cases 22 22 22 22 Number of Runs 13 3 6 3 Z .218 .000 .000 -3.077 Asymp. Sig. (2-tailed) .827 1.000 1.000 .002 a- Mean Runs Test (Management Action Data) Ho: Within groups, data are independent. Hi: Within groups, data are not independent. a = 0.05 p > 0.05 for Zio, Z 2 o , and Z M . Therefore, HO is accepted for these zones. P < 0.05 for Z40. Therefore, H 0 is rejected for this zone, as data in this group are not independent. b. Ranks Zone N Mean Rank kills/100 sq. km. 10 22 43.50 20 22 27.32 30 22 29.68 Total 66 Test Statistics 3 ' ' kills/100 sq. km. Chi-Square df Asymp. Sig. 16.963 2 .000 a. Kruskal Wallis Test b. Grouping Variable: Zone Kruskal-Wallis Rank Test (Management Action Mortality) H 0 : Management action mortality density is the same among Z 1 0 , Z 2o, and Z30. Hi: Management action mortality density is significantly different among Z 1 0 , Z 2o, and Z30. a = 0.05 p = 0.00 Therefore, H 0 is rejected. Management action mortality density is not the same among Z 1 0 , Z 2o, and Z 3 0 . Multiple Comparisons Dependent Variable: kills/100 sq. km. Tamhane Mean 95% Confidence Interval Difference Lower Upper (I) Zone (J) Zone d-J) Std. Error Sig. Bound Bound 10 20 2.4636E-02* .006 .008 5.7794E-03 4.3493E-02 30 2.2773E-02* .006 .016 3.7421 E-03 4.1803E-02 20 10 -2.464E-02* .006 .008 -4.349E-02 -5.779E-03 30 -1.864E-03 .006 .735 -6.867E-03 3.1394E-03 30 10 -2.277E-02* .006 .016 -4.180E-02 -3.742E-03 20 1.8636E-03 .006 .735 -3.139E-03 6.8666E-03 Tamahane Multiple Comparisons Test (Management Action Mortality) Management action mortality density in Z 1 0 is significantly different from that in Z 2 0 and Z 3 0 . Management action mortality density is mutually similar in Z 2 0 and Z30. *- The mean difference is significant at the .05 level. 104 7. Significance Testing for Differences in Mortality Density; Park vs. Non-Park Lands a. Runs Test park density non-park Test Value3 2.2000E-02 7.70E-02 Cases < Test Value 8 11 Cases >= Test Value 14 11 Total Cases 22 22 Number of Runs 8 12 Z -1.271 .000 Asymp. Sig. (2-tailed) .204 1.000 a- Median Runs Test (Park and Non-Park Data) H0: Within groups, data are independent. Hi: Within groups, data are not independent. a = 0.05 p > 0.05 for park and non-park data. Therefore, H 0 is accepted. Data are independent. b. Test of Normality Kolmogorov-Smirnov0 park/non-park df Sig. DENSITY 1 .117 22 .200* 2 .260 22 .000 *. This is a lower bound of the true significance, a. Lilliefors Significance Correction Test of Normality (Park/Non-Park) Ho: Within groups, data are distributed normally. Hi: Within groups, data are not normally distributed. a = 0.05 p =.200 for non-park data. Therefore, Ho is accepted. These data are likely from a normally distributed population. P = .000 for park data. Therefore, Ho is rejected. These data are likely from a non-normally distributed population. 105 Appendix B Comprehensive Management Charts a O -a 3 CO ed o &td ID T< c! g • a O . C cd 3 3 T H + • » •S 8 CO a o & I f a .S © . * t 9 8 .a o 3 co o a V a 2 -r e !"2 =s g £ 0 0 1_ ON <L> ~ PL| •° £ ^ a N -a N -5 '§>'£ -3 -3 • - & £ 8 +JS< : [1 g 2 ::+.JMC-:*> 00 00 ON .2 « P M "3 e o s .2 •3 « — ' co° O a V a a a o CO •c O H S o o o o o" o 0 9 a, O c E 1 J •SI 1 00 e cd CO e o co "5 CO £ SP o c a cd 1 -8 a0 a0 T3 cd a o - a 2 U O H C —< <D 5 6 cd co U H <U a O a a o O H 1 * . cd PL, O I O H - S PL, j g a a <L> SP a cd cd P H "cd a o CO CU cd a o ts 107 cu o o CO c -r) co cO b „ o is ~ .2 S3 » 60 X> ^ rt O -3 « h co co rv I cO <0 > cu 5 to C O _ o 3 x> cu C o o § O o 5* «r cu Q cu 5 » ••§ « S 5 0 * 0 . 2 CM .p " i f i -a * IS | § s ^  s 1 OH 00 c o s cO -*-» cn "> cS C U I-I p " o u **** s ca 00° ON ON cu cO o OH CO CO CD •a C O >, O "3 x> cu X ! co . C O S to c  O cu +-' .59 d 3 O •s s xi & B • ~ CU H cu > C U O c0 •«-> •a cu & cu I ca i—I " " ^ J CO t - . C O <u X> CQ k or S"1 uj « CU ca ^ 3 < s a C O C O CD CQ ca cu •S M g & ^ S CU c ^ C U CO J 3 CO C U X> I c CU E S' o o o C U T 3 co C JS § C O C M O C U co C O O ca > O • a -o 5 o ca - M -Si ^ I § CN 5 CO C C U C U CO co "cO CO co CO CU CO C U X ) C U > '5b . /—^ o s •a 0 0 2 ON co cO O o o co cu cu >. T 3 O P c o o \f2 6 w C U N^ CO ^ OH C O 3 OH O § a S I to a cu o s ^ C«Q co 00 .5 ' 1 ja cu 00 C U c^ O •K - o o eg UH CU OH O Ui OH CH-I O T 3 C U c2 .S cu CO u. CU CO c 3 O cu cO X ) • • ca 108 o H «< U Q •> O -J O N Q 1 » CO DO cu I > cO 2 » 6 -° cu M -o 5 4 - 1 e o OH 2 3 S OH C >•> o-r3 eu "° I > co l i X . cu OH •3 CO u, cO cu i-H .s § §•9 CO r-00 +3 CU a «2 OH cO ffl 00 c s i s ON cu S H h l ft 1 1 a . 5 as CN '3<* O OH ro CO cu + J 'co t cO CJ C 3 O o o CO ON 00 O N co S op*o CO l-l § o Q c2 o CO cu n3 2 H . CO *o OH CU SP i *o o <2 CU I g -i C cu CU X I CU > CO CU M CO 2 xi &1 co * b -g O 3 00 O >> cO — X ) 2 £ 00 £ I CU co O •B LH c2 CU •8 CU T 3 O c2 xi cu % X ! co CU co I fro •s e 3 s 2 8 cfci cu OH co j_r O cO x> S L l cO co cu cu 8 | .S-S V3 > WO C 3 O o o CO X> cu a cO CJ CO O <4-( "TO CU > o & CU 1-1 cu X ) o _ CO ON CU 00 a 3 o CJ "E o cb T 3 O Q o CO M g § .2 * o 2 CU C M r3 o £ co _ co CO o c3 'cO O is N O 3 o CJ CJ cO X ) s 'cO •g o o C+H o o CO* O N cu oo X> O N T3 CU T3 - a - a ' o a > o O H C ^ . a -c CO c cu x! u CO O . cu ^3 C 3 CO CO "> CU •5 ^ CO C eu cc u. O cO •=! -3 eu co II x i CO I 109 Cd cd o -ti 3 D H ' O 5 ,cd 6 ° « 3 4) O £ « *J t i tj .5; 03 -4_> ^ 3 <D ° < X X O ^ »1 >> CO CU -l-J S o I CO g O H -3 2 °o SP to o - ° - O \ 0 T3 . . cd C "3 , ao o o T 3 O H CO 3 3 « . « C O 1^  g O H 1) 60 3 cd j * i . ID -  | x ca ^ 3 s cd j 3 cd o 4) T 3 3 3 o 3 S cd w -a ^ & O H o PQ § P H cd r-o\ a \ cd -o cd 3 cd o CO t"» 4) ON co H cd cu 60 cd 6 cd •a CO cd 0) <L> CO 3 <2 X CD r 2 'co CO O O H J+H -a CU CU 4> X 3 co O -a e cd CO 60 ON .3 oo 3 O N -T3 •3 O Q s o 3 H CU w I t s I 53 B * - WH 3 t-cd O B "3 3 a> ffi £ "> ,2 ON oj ON t i * H co ON 'Z 00 cd ON 4) * H J3 s o ~3 cd o s 4> E? e a o O 3 cd B cd cu PQ cu H-» cd w !5 "s I o B cu -a VH UH 2 8 cd <U 60 -ir; \ 3 cu 8<§ ^ 2 CD "3 +-» £H cd O co cU 4) . y > cd h .60 3 .s cd r-ON ON cd cd 3 cd O CO cd ON ON 3 3 O o 3 cu CU UH X cd .SP %—» CO CU > 3 cu cd c 2 P. £ H 3 -3 3 O 4) 3 ti cu cd <D <D 60 c 5 § 3 O o cu £ PQ cd cu cu >> CO X> O t5 " cd cd 3 ii O cd .2 '« CO CO >. O X O H Q H - X . ' CO 60 c3 .S 3 cd 8 ^ g & c S O cd U B > > 1 H O 3 is g 5 ° 3 cu ° ^ -3 cu H~» O cd cd B cu c -3 .5 -a . cu I I > cu o > O H3 4) Tj ^ S t i o 3 O '§ ^3 4) - O 4) _> 'co 3 Cii 4) -o 3 O 3 cd „ 4) CO cu o o 3 . CU o > 4) O t i 3 -a ' f f 'cd cd "wS 3 3 3 •g S u CO > CO CO 3 3 O 4) O H CU UH X ) O 4) H-> > CU j -.s « 6 Id § o 3 O co UH 4> O H cd -a J D 3 <! g "3 73 .2 ON <D ON H 00 ON co .3 cd cd B B 3 cd cd cu cu H H 3 9> cd co O _ CO CD X > S (D ^ +2 T3 4) 3 ON O CO CO cd u cd 4> X I 4) H •3 T3 O Q 3 O CO H i N M W « CN 60 cd X 4) -3 r. O H J ? CU 2 ^ M co" NO ffO t< ,4) H-; cd ^! S <° * «2 £ . 2 * a o M (1 X 3 g P 5 . 2 O H O H3 «I Ci_H cd O ^ 3 Id g4) O O fd t i J3 4) | i •o 5 d X cdx i C+1 O 4> 4) ob 4) "3 C O ^ 4) • O t i O H cd 4) X <+-c O 4> _ 3 co "t3 cd 4) X O « i S g O H >. 4) <§ s to <U 4) 60 "3 S „ 3 O cd 6 ii o 1 2 cd •3 cd 3 cd CO CO cd P H 3 O O co 4) - g 3 •2 cdi cd P H o O > O i - cd a . cu x i 3 cd cd o g E C CO CU 4) 3 4) 4) > H3 3 O "g X 4> X 4) 3 4) 60 >, 3 4) 60 O " m M u ids & 1 € 1 •em-. 110 x> O d 3 o '§ X! cu x> CD > '-8 CU o CU > co c C M cu 1 3 "cd cd c C M o "3 rt cu ON u, ON CCJ r-1 CO cd cd 1 3 CO 3 s ° o «s .5 o . 3 , cu 'co 1 3 cd >^  co d cd cu m s |m Si cd r-ON ON cd cd d cd o CO 44 cd \ M CLi CO O cu - c o OH co d o u, cd cu X> cd O cu 1 3 CU -O 'o cu 1 3 cd C cu M cd CJ d cu 2? CU s cu <? s CU td cu CtH CU CJ _cd "a, CU M cd 1 d o 00 Xi 60 3 O VH O X! H-> cd 00 .S <=! E: cu ON ' ON cd  C M cd 1 3 cd d cd o CU co C cd 2 >-d cd CO g . M , 0 5 O O '-3 cd co 3 cu .tJ 1 3 M O cu +J S i 3 « 5 .2 CJ CO OCU d 1 3 CU M t-i cu ,o 00 c o Xi CO 1 3 a cd cd cu i— cd d o cd 3 • cu Xi 00 M CU X! O CU 'co O '3 . . o 5 1 3 2 CU r-td 8 -CU "O i- d d o S oo 1 3 OH cu c CU "~l > g cu X) OH O o 5 *-> d M cd cu CJ o !§ >> a | e > o OH o £ -d « ? ; — . cd « O CO M CU OH 3 X H e o CU co X) T 3 c 2 CU .O X! <*->^ .2 co cd cj cu • ° cd cu cu O Xi d " t l UH O S >> CU 3 . . cr r- cu ON cfc| ON H H oo ON cd cd cu 00 cd s cd 1 3 S3 £-8 ^ CO 1 3 cu CU ' H_> CU H_> -H CO r-.s >^ § CU j3 S H co co «a 3 C co ?5 M H_> H_» Ol H H 3 O '§ XI cu x> cu .> 'co CO CU u. oo . 00 co cd C _ cd 2 S 3 2 •a § o 3 co CO co n) •-3 .52 ON 00 ON 1 3 1 3 O Q c o i CN cn i n cd u-0) c5 CO c O 1 3 CU • a o o cu X) >. cd C cd d o CU OH 'o CU. OH CO 1 3 CU 1 3 Ui & cd cu X> ^ = 3 oT •> oo ? ON -° 2 3 ^ O O xs Q • I * s i t a .5 'cj >-, g OH CU co rv CU ^ CO co" C _a> o "cd '2 cd O o o CU T 3 O Q o cd cu X) UH cu O H 1 3 CU cd CO C3 O o o C M O S I 'E s cd cu CO cu X! - M o CU td o _c co CU CJ § C/H) s § 1-^  CO I CO • — N ON CU O N oo -d oo O N £ O N xi o o _o M cu CJ a -o CU 1 3 '3 % cu X) CS c o o o cd ON ON ed 1 3 cd O cd O to cd o cd d o CO M CU OH CJ CU OH CO CU •d -S J 3 O co & M o CU -+-1 d cd co T3 cu M CU X> "ed 3 "2 'S 1 3 .5 X! d o CJ M -2 cd >* cd <U C M CU • M CU . - ^ O x> d o '•§ o o r~~ d ON cd o rd O H CO M —H cd g X> CO Q cu cd x j pq 111 - a CD O N O J £ O N ^ -S ^ 3 "3 3 o _ o O ^ — a co -rj O ' 3 jo cd r3 CU Ul i w cd i u m 3 3 u, Q w 0 ° P H ."3 4> to _ Q ' > 1 * „ , cu -3 o CU .3 "3 ed lid CU or, <- 4 ) CU +-> g 4 ) > o a co O H-> CU & 3 <+-< 4) • — cd O H OS cd . cu £>JO cd 4 ) co _ 5 co o « * I ON oo ON 4 ) 'co 3 O cd C O C O 3 O 'id o _o 13 3 O ' 3 o _o Is 4 ) 3 cd O H-» CU > o S cu JO 4 ) a •3 3 O O CU co cd •3 4> cd 4> PQ cd -3 O O 3 cu a cu 3° cd b cd "ed 'o CU O H 3 O J 3 3 cd u 4 ) a cd rH cu cd <i) -3 3 cd cu C O g 3 • a WH.'*3 O co H 4 ) C O •> a O <H-< 4 ) 3 O ii '53 J 3 H-> o .a a -5 4> • ;i ^ i i 4 ) *T3 co 3 g Ml C O 4> . t i > , co cd cu a co cd O H "ed 3 O 3 CU JG -4-» 4> 3 O •3 CU co C O ^ cd co Ov —i cd O N cu 4 ) S PQ cd , cu cd "53 r-I i O N CU O N JO i- i + 1 cd o •3 3 na '% Ca C/J C O Beai ["Par ai cd JO "cd 4 ) a cu o 4 ) ' O H T=f 4 ) Ii 3 H-* o a +-» CO 3 3 3 cd 3 O 4 ) 3 O •3 co cd C O l_ cd 4 ) <41 O 60 3 a ' "3 3 cd SP " 3 4 ) 3 O , o H-r 3 £ 4> ^ > co 4 ) ^ "2 S g .5 cu a a o o 4> +-» "3 3 cd "3~ 4> 3 O a cu Ii cd ed cu PQ -3 CU ti 3 a 3 CU 4 ) J O C O cd J 3 ed O H cu -—^  *•* a .a a g»8 "3 co 5 >i O CU \3 O H '•a H-T § 1 o ^ cu 5-a £? ° ON 4 ) O N ti rH C O 8 o .a-fj CZl ed <2 -3 3 cd 3 3 O o a o <a -3 3 cd C O S *3 o 8 ^ O 4 ) T 3 Ii — cd CU C O 3 3 o o cu ti 3 * ed co a 3 -3 J O 3 o o J 3 SP' '-3 ^ 3 "O .3 O 2 c2 cd CU JO 8 ,o o C J I ^ - H T 3 O 4 ) co « (u 3 co co O J3 .a co O CD 1-2 - 0 "3 4 ) 3 -3 4 ) cd •E. a 4> 4 ) a o co 4J « O &o;f>j3 ^ J 3 .2 « 1 5 rr'/rrt* 'tr/r t i ft u x-:-x->y.'X-:v»:v: c w.: •.ys.v.v.-.-S.<<<<<fSSM<<'. WSSSS/.*. 'SSSSA ... J 112 6 m co o 8 "9 8 8 S T3 I 2 £ " S i b i ! PQ cu 3 CN ! CD 1 cu g cu 00 cO c cO £ -t-» O • • — » <3 c o U t3 cu PQ i C cO B £ cu .> C o cx co CU p4 CO P H "cO C o co CU cO c o t : cu -*-» cO is « 113 CD o ^ g <U ON 60 ^ B i s « 60 52 • i ^ . s « 5 c <*> OH 3 co S3 CD CO P O a .h p .2 > cO td « cO S •S a -S a 2 <o «> m S CQ CD O .2 8 a U -ti p > UH ° CD CD CD —H a ™ '3 3 « J 3 £ g "S! » S ° •> B co i-c .a P «» 5, cr CD ^ l - i e s ^ :> O " " - » - > . _ CD 3 cO co -C iT9 I I CD a w CO u . H CD 3 OH CD a to 8 £ P CD •ft! OT ^ >> cO = 5 co 3 co 3 CD CD CO a 60 CD co 3 O CD - 0 ^ CO c p co CD _ a cO o 2 *§ ~ a '-S -o p CD 1 - 1 CO H-> CJ C CD .23 • O CO S X ! 60 co 3 CO <r—t 3 3 P 5 co 3 j>> *N N •c 60 T3 CD 3 g CD o C P C J co CD u . cO .2 T 3 CD co O a CD CD 0) p 6 CD u, CO CO '3 l - l H "g CD 3 * l o ^ CD a .2 Si. t-c CO co "a, O co CD X > CD CO CD >~> CD CD « 1?-C «> £ to J2 v-. JU *T CD *—' CO "5 .2 3 a CJ .2 a co CD co tj 2 CO ^ 5 GO a co ° a c0 CO c .1=1 CD CD X ) CO CD co u , cO cO a <o 3 CD 60 j ! a g . CD — 2-• C & 2cS a £ co CD X > <-> 3 CD J -co "J3 ° '§ CO OH l _ cO 73 CD CD CD ,—v X > O N cO OO ON OH .-I -o p p Z Q cO r-ON ON cO cO ^ c S H CO T 3 • — g CO co "Q CD cO >-i a co 3 TO »«* t.\j S.]3 ^6 CO rt 60 a a co U CD OH X > P T3 CD 3 a .SP ' c o CD T 3 T 3 CD -*-» C J •c +-» CO CD u> CD 3 CO CD M a '3 g 3 P CO CD .-a OH CO CO p 3^ co CD — i5 cO _ P "3 H-> P a CD cO c o CD CD CD .2 2 c0 «J H co 53 S =3 P a CD s CD a c0 CO CO CD CJ C J <5 cO P H "3 a p • 3 % CO CD cO a o tJ CD 3 114 6 0 CD ed O '5b o o° • a £ a .a Cd _, ^ °* 2 -« J2 cd cn « H cj JS «rs +-> • -3 S e » C 4) (B <U O T 3 CL, « o a a 8 8 * 3 ccj 3 X P H co . 3 co 4) 3 c2 td •3 3 cd cd CO •3 00 OO ON T 3 CD • 3 3 cu 1 o" cn o\ 3 cj _ I ap o C+H o 3 O cd a ed CD 3 -w CU < OH j 4> £ d 3 & a c3 .52 cu o "3 OH -3 /3 T ? CU F_< X 4> —1 9. . r3 cu § I - S B ' S . ed co .3 3 3 co S ed OH CU 3 > £ CU « 53 -3 £ ^ 3 g cd HI O H S 2 I co cu ™ 3 T3 •O ,0) ojj "O o 5 * — > u, cd > cu 1 s ' § > S * - a £ - N * § - a 2 <u CJ cu H-> OH VH CU O 43 ."3 03 *^ « C M e o cd r r — _2 a > 6 0 - 3 g > 60 O 53 ~ ^ CJ o ed 8 5 5 O G Sf? -5 o 9 > '+± o • * IK e ea ' 1 e • * 3 « « cd g XI o g £ X rN - K Cd OH H-> IH £ cd cd >% 03 • a - d 3 3 O co cd 3 cu X 3 -T= 3 -?5 > "tj - 3 fa, 3 VH . 3 3 <H 6 0 ; OH 6 0 cd .3 cd I 09 I i l l OS i m I j cu _ X . a H X 3 ed 3 cu s P4 I -1 c 3 CU * a CO CO 3 CU CU co 3 co -9 -3 > « <o CU •3 a >, g -3 0 - 3 in "3 -> ,cd 3 ^ I /—v cd £ g C£N cd S o w cu 4 - X <f I i •«-» T 3 C 3 CU X *-< X o 1 O <u o u 3 o co •S i2 T3 3 ^ cd 3 OH & .2 2 S 3 CU Cd OH M CO VI cu 5 c2 co CU VH 3 • 3 CU o o <u a •=r> - 3 2 rt OH - 3 —, 3 13 CO . f CU • 3 3 2 & ^ 3 0 OH CU rt o ^ O H J 3 3 -3 co CU 3 CU 3 O OH O co b X ^ H 3 o „ CO >; tu cu cu o 1 6 0 t-1 53 H ^ OH •s ^ -s ed O 3 Id cu a CO CO CD CO cu C £ X O 6 0 6 0 _ 3 Icj a cu co O OH o ,4) CH ^ 53 JKi CO M tfl M Cd CU OH O^ « cZT . CO * * J 13 < 0 3 co 2 13 t3 5 -a .cu cd 3 cu •3 C+H CU X co -*-» O CU 83 41 e cu 1 ¥ c M O CD w a c ^ a § C S O ed 3 g c t t a 3 cd o a 1— 3 CD a 3 o _ co . a ~o o co H-» CD co i w o co O tu ^ c O il) OH a lM 3 o 53 a > 2 60 2 W OH (-=3 .2 "> cd - 3 3 - cd U 5 / 5 3 3 2 U X r^ N - G H 0 ct; X IH +> « O - 3 OH CJ J3 3 o 60 co cd CO VH CD CD co "3 cd d—1 W*J wX CO cu 6 0 4) VH J—» CO ^ O « O <D cd 6 0 3 cu 3 CJ O •-H OH 3 - 3 5T id O " 3 - 3 " .9 5 CO ^ CJ cu CO CO X S CO HH « < 6 0 cd S (J co co ^ cZ <D o "3 r- cd CJ 1 . 2 - a 2 « CD cd E ' a o s f § a « 2 .§ I 2 S a w _ CU CD VH CJ . 3 53 ? cu X -3 CU OH • i . s > -o CJ 4) T3 H-J 3 S X +-> 1 g cd cu ed CO 4) co 3 W 0 a N 3 U "3 co C 3 cd £ i a 5 T3 cd IU co § 3 N 2 CU OH 4) Q a a o 0 co IH CU OH s: -3 CU H-j > ed cd ^ X cu cd 0^ cj CO cd cu 3 RT 5 1 / 3 OH 3 4) « .2 X QH-a X T co 3 2 0 . 3 X 4 ) X +J J H co ed > X « * o •2 rt 5 ed OH 3 f « 2 W VH .3 ed -rt > PQ CJ 4) 3 4) a 4) 60 cd 3 cd - 3 3 cd cd PH "3 3 .2 'HH cd CO CU M ed 3 O t : 4) td 115 116 o cO co § 9< CO 3 CO t i O OH CU o o o oo oo 00 NO o .00 a 32 cO C cu X> cu CO « > T3 CU eO X5 eu *3 •5 1° C cO .2 cu0 5 «2 £ oo E to O •fl ^ 6 ,3 o a o OH > a — i X3 a co C CU 5 co ea £ g 2 ' 1 OS o s C73 co I B % u cu ^ CO c3 «a cu o PQ a N cO • g o © j> c/3 ""rj c0 a cu OH ' > "3 O CU a C+1 O Ct-l *e3 X o -4-> OH 3 cu T3 li cO E UH cu a cu S CU 8P a cO 3 "cO 00 cu •a cO AH *e3 a o o CO 5 a o CO 3 OH | T 3 eu CO *c i j 3 a 3 u i«8 *e0 00 cu ro 117 •3 >rj s g 3 co CD "2 •C O H c<-i CD ° 13 60 |C3 O O H -c § <3 >^  « (H-» *T3 r3 'co CO C co O <U o. CD co >-. CD O O CD = 3 * CD CO OH OH to „ - C £ co O (=1 13 .2 •8 a CO c e o . ^ H •»-» 3 CD a CD 3> C «0 •c a .22 T3 C oo O w oo "5 «8 2 l i s s ° co O O \ H 00 CO Os C a "J3 co CO CD UH 'cO C _o 'co CO CD CJ C O CD co 3 O B o o eO T3 CD H-» CO cO a -o cO o oo 00 Os a °^  a ; : C on Q H £ o ^ CO eO M CD oo I—I 3 w x> a <o CD C s -2 c c .2 eg to e c cO t — I CO - M X J i-» .tr, JS £ 13 <D cO 00 * Q 00 « o >> o PQ 2 CO a * j g os eO co f""1. ^ ID & 3 a -CD O 8 O x> -° += ^ ?3 a CD 3 "3 CD f—i CD i—) _ CO oo ao O S UH <-I O CN uo so 118 . . c3 £ 2 cd ed 60 00 53 O N Q 00 fl.S .a 55 .23 CD >•£ P H & a w 60.5 9 -a D cd HS — " & I S § C VH CU CU co -rt CU g s 2 e : « cd > <u Z. J 3 OH C cu 2 i » " cu 0) CD C H - S 2 | CO cd CD cd oo a 2 60 ^ 00 CO 5 2 » ^ O N J D "co "O a a O cd & >P VH C+H _ cd cu co O 60 co o VH _ H CD O ex 3 . C M m o ON CO »-H cd • S -cl "O dco cu >. o t CD VH 2 1 £ g « T 60 3 s ^ P H Q 60 u] ,s p» -a ^ s 2 ^ °^ 2 S •a °^  ^ IH CD co CO 53 P 00 oo ON CD •5 ° c3 p a* „ ON 9 a CD a CD _ 60 LQ g «-J cu 0- PQ i* a w a oo 8 p 60 w cd 8 cd 6 IH cd cu JO a CD a CD 60 cd a ed a P H »2 C N Q 00 o t3 i — i cd • »H C N ^ ON & 8 cd o e3 a * "S s H3 Q oo CD CD VH cS a cu I> <a *-> " f i O " C o . & H cu CD C+H a o -a ftp's CD C+H •S3 co o .2 CO ed -3 a o H O la CD JS 60 CD CD PH' *0 U J co SP * J cd CD 5 H CO ja H HHC ^ J 3 S 2 O H a O N H2 p o rVn IH P H OL, J R . oo O N J> H H I S i s " *3 O N O —I I a o cu • f i O H CD -a 2 O N •S °^  •&r 2 a 1 2 cu 5 *a S " o O N ( M • — i CO CD O N VH c2 « <U TO H O CD -O td CD 5^ 5? CO J2 ed O T3 U cu ™ CO 53 o^ *c31 ed o £ S C co CO eg 6 2 a -S Q P Q C N 119 SO ON Os " - 1 O N .S ^ I I TO" V £ is •« "S • 8 CO Q co co *T3 cO 9 8' S 8 I I CO O i a 3 .a 8 2; I* eg .2 • a * co -a a 1 3 3 Q « co 00 -, CO O x> r-3 2 00 ^ |c+-, -O f l 3 1 a « e a CD o U CO Onto CD 2 d o \ o os 8 o c2 52 CD CD _ Q CO a £ ^ 8 a, co - x> 3 O i M o cO OH 3 O a CD .5 oo "cj cO J=L< CO -a CD e •3 CD Ii CD B CD Xi a o T3 CD in Os CO e CD X! co CO * £z +-> —1 zn o > O !> ^ co SO * co Os W rH OH c n a^ > | 2 -a o <2 -§ cO CO CD O 00 OH OS oo ~ c !=i I S T3 "—' o ^ c2 a - 2 C N O 00 OS 00 m OH wo s o 00 00 a I o CD CO O XI c2 CD | '3 S CD •a cO a CD UH co 8 8 ^ H_> H H Is Q .2 ID co *ZZs . a a CO r> "g 13 O CD *1 - 00 Os -o 1 — 1 a Os Si; « 120 0) a co T3 O o VH CD 4) 8 a OH 4> co 3 o cu u 3 o ' VH « o •o cu cu o cu CJ C O Ul ccj CJ X> C O co CJ O cd / - N CU VO co «-• Cd 3 ,«J c .to £ O 5H -o ° c 0 g CJ O cu cj co CU ° s "3 ° 4) T3 4) VH O H 3 Cd OH CJ cd C O " V - ! C O C O ed cu C O C O *•> C O CD cu VH 60 O CO "1 b 1 * I E co C o co sj c§ ca <S g co a ' a «a 5 60c+H 6 * ~ O 4) +-1 trl co ^ CU £ O tj C O VH ccj CJ J D wT e3 CJ CJ 8 -° a g VH J) a 9 c -S 1? 0 3 4) CJ cd co O X "J3 OH VH co CJ "~ i f C O cd 4) C O cd 4) co 3 X O VH OH CJ C O O 0 o •ti 3 co 1 « 8 jg cd o X) co o <5 . s CJ "O 5 \ 3 CJ .S co ^ td a ts e CJ co O PQ £ cfa 7 3 3 0) o X cj Cd VH CN CO "S Sri J*f t/tf *Sss t , w **f>m /4 t. tWt JV J ty S/s/'M f j. j. * i $4's f ft * *t f f * frits t , , f. 121 122 CO >>. "3 1 "3 8 |x> eO c o CO . 0 3 X> 3 "3 1-5 I "2 -o C cu CU . y 1-2 § £ O O " O co c o cj Q T3 CU a 3 CO eu •g 60 eu Id cu o B f CO I^ O 3 00 ON ON p 00 ON ON co CO X j CO eu o 9 co • « S CO ~ 3 S S 2 • 3. o ^ " co CO o -o eu o eu O s eu > o I o cj eu X ) 1 eu a * 00 I cO *•» fa X 3 OX) o 3 $ e co co Jjj co eu « X ) eu 00 > S3 -55 '5 c ° « s 2 " ? ^ S3 eu o C+H o c o CO eu •o J 3 CU _c X . cj 1 3 O 'I 43 eu X> CO 3 3 o 8 g o - ° CO T3 X> U co" 2 <*> a eu .-a &•§ eu -a cu OH "eu -a •S 3 eu X ) co eu X^ CJ CO eu Q ~a eu 3 2 '•3 K 3 O 1 00 43 s - ° c" SP 3 'OO C T cO CU Ii to cO S3 S J 3 >> J5 x> C41 T3 O CU S3 ca eu eu > CO CO CU I i 00 3° -a cu >> cO ~ CO c0 2 "3 . ii T3 3 cu -3 6 g o +3 CO co 3 T3 cj cu CO ss cu '3 6 £ o ^ CJ co eu cO X> X ! OH O • 3 I X J .S JS x 3 X ) cu =3 x> 5 co « s s O O O co «U « lo O 3 ^ - O S3 «r .2 S • -co - a s s 3 ° X ! O SH § 2 - 3 *° -"5 eu > >, > O fa to" eu cO 2 *a c0 co **-« I I O I o T CJ eu co" e c • 3 c0 X i 3 © X ! o -t3 CO eu 3 3 g. o £ o Ii M s i r-i CO eu co 1 OH C+H O eu JT3 'co O W Q O x>X ! eu 'co S3 eu CJ S3 cO ti B S O cj cu 0 c0 cu 4H c £ eu eu <u .trj X> co CO cu 5 f co" "T3 6 § 5 o CO co^'-O 1 ^ CU CJ Ii 3 2 CO o -a CU « : o 2 ii •*-> a >> 9 0 3 cO £ _u R 13 cu u •fa CO o . ^ ^ X '3 CO co f CO "co CO u -o cu co «3 & '5 X! 3 "5 13 a I I co co m 2 c0 T3 eu « -3 cO CU CJ j5 CO -3 13 " O S eu cu 2 x -2 ^ OH CO eu B -° co o cO eu X> S © 1 eu T3 eu >> O ha HM CO CU na d o co" O O . C cj *T3 CJ u et -2 5. OH H^ - o a cO co O H ' -• -o CU CU -a 'S «S 3 O T J • O « « a. * S eu • O cO eu x ! 3 * 5 co X> eu cO X ) cu • eu co -2 3 ?3 CO CO s £ « cu C fa w cO eu co X i «3 2 cu C4 X ) pfi T3 B a -2 •ti  u T3 C I I l o CS a OS e cu co XI C5S CO « cuCO S C3 e _© es u C O E 2 s . © 73 1 % «3 g S o ON . S c ts. C M eu 08 T3 .5 cu cu .5 o co O § c SN ON W> ^ © eu S CU x> -0 cu 3 cj o cu x> 2 cu •3. co cO cu UH CO CU co 3 If a s f | •S3 * * * § ° 3 CO 5 - co cS a cu s3 . . CJ f - cO ON X ) f- 1 cu euO s i s i •8 c2 OH eu CJ O cfcl co I T ) CO CO CU CO T3 eu > o CO CU X> -a cu HI co cu CJ CO O 3 "to fa 9.S ll CO H -5: cu c/3 co NO I IT) r~ ON CN I CU o 5 -<-> « - a o c o S I 2 2^ 3 -o CU OH c *H o 3 CO OH S3 o CO cu co o eu o CJ CO „ S3 -a «> g eu O fa 2 £ o O H C « CU CU eu eu xi fa cfa *3 OH ^ b 2 x> a * * eu CO CO co C O O m U .1 u C CO CJ co O cO S3 cu 00 » "3 S3 o X> CO C<1 o S3 o 'a* CJ 13 CO CU 00 eu T 3 ON CU cT & ^ 3 00 ex eu CO CO 13 t-< o. o CJ O c+-< *-> "S. 0 3" cu W ' S 3 00 ON ON CU 1—1 C N 00 ON CO B 5 CU ^ 00 % S c 2 S o cu co 2 M CU ^ rS i t . • a £ | CO • 13 >, > CO ^ ^ S 32 123 CD I i T D cd CD t—\ *l z SP ..s ^  60 Is CD 4 * CD 4 3 3 CD 4 3 C*-c • ^H c 60 O . 4*! CD rr, CD C M CJ 1*3 «| 3 „ 4 3 4 ^ "t2 CD •r CD .« § CO l o CD OH co 3 cd « <D O T ! 4 5 ~u 0 0 CD oo r ^ , — 1 JH Q-g . CO 8 CD I H P H o oo -2 ffi 0 0 6 0 - 0 ON CO CD 2 w P H CD 'co ~ * § co O "i-a CO c d g 4 3 .ii ° g 2 2 «2 HI* 11 i i cd OH oo ON ON Q co g co CD O ON ON cd i i NO CD a.g .3 60 CD I r i cd 3 -S CD CD co C O cd CD £ i f id < 4 D ed 4 3 O •3 CD Sf •a 3 CD s CD cd° 3 O O i i cd CD PQ i 3 ed E a CD . g ; ' co 3 O OH CO & i P H Id 3 O '3 W CD '3 ed 5 124 ON Q oo C O •a . o "3 . € if o o 8 ccj s '! . CD 'o % ~3 '£* g-CD C O '"8 s a, o . co 3 O U i 60 9* § co CD O r-ON 60 CD CD CD C O C CD S C O CD U l 3 co O cd f-is ON u, ~ O i—i T3 Q 3 oo 3 ^ 60 ^ 1 ^ « 81 cd .3 I | £ 5 CD OH > U , o S 3 O 3 O CD SP 1 t s - s CD CD * 2 ON H-J i - i 3 I CD o -a ON ca co SP'S .3 CD CD CD CD cd -3 CD co t - CD cd X C O < ^ •3 ° CD 60 -3 .S ci § > OH 1 8 C O U i cd „ S co I 5 C O • 2 .<§ 3 co C O Q H -2 i f <D ° m -a u, r-3 O ON CD is Q cd 00 "S cd i i 9> 5 -a cd 3 O cd O cd — i CD 8 O Q 5 CD U i 1 •8 o 3 1 . "3 cd cd 8 O -O CD 2 2 .5 5 3 2 ccf b CD .2 cd -•3 _ e 3 ° 8 X) CD .CD cd CD co CD CD -3 60 cd S CD X cd 3 CD •g "3 CD co O 00 ON ON Q 00 p cxf oo 2 Q O OO 8 o o co CD OH cd id 1 £ C O -3 cd O U i •a 3 cd "3 i a1 CD 1 co J D CD > -3 CD "3 ^ 2 § -3 2 •3 CD "S-CD U i CD X CO -3 8 3 §> I cd CD 2 . 3 T3 O CD , _ OH 3 O O CD "2 « X M X 00 CD ON > 1 8P J" 60 ^ cc? 00 ^ ON .23 _ n "3 i—i co SH R — V 8 oo 3 00 8 2 8 h-H 2 Q oo B ON co ON CD I H CD co 3 Q ON i — l CN « oo £ON CD 3 O O •3 E O O j 3 ON H C O i i O .8 U U S CO *f co <D < Cd Q H ±j 3 CD CD H "3 M h cd 5 J = 8 I H <N cn NO O O M«c«««/«WW/i ,i'i'^AViy/ ****** * / .* • w * * * * * * * yV*£*£*&***-*.******.V***************************** 3 «JSi< s** * ^ f,*f> **S "* * % n * "5"" ' "*%' ^ ' ^ / /// *********** ****** **/** ** * * * * * * * * * * * * *** **•* ***** */ * / * * **• * *. * * ** * . * * * * V.**.*Sf.*S*,*^ '*l*l*&ffi£*l*£K*^^^ 125 of) g '5 ISCO _ eu •.e co •c T 3 OO •s >> . u 2 oj |c2 CD oj T3 3 ' 3 § c 3 CD c o 3 CD CO CD OH t o CD 3 CO CO cd co . OH ^ , 3 2P .a CJ 2 £ 5 \< a w .9 s 3 O1 " CD 00 42 o II B CD O . - O o a II CO CD-\3 * C O CD '+- ' IS tj CD .2, WH ^ u- 3 O co ~ CD co C CD 3 tj S c0 oo 00 £ C CD xj g § 42 IH CD »-. > O CD 2 "2 3 CD M 3 J 2 00 to co -c f 1 42 g OH O X \3 4> 3 co CD CO h 60 g ^ -d • O 3 £ CD 42 C M O CD § CD CO aj *a co 3 > > CD g X! O X co cO > 13 QJ 5 "2 cO co C to - 3 3 CD C - * t r ! *0 _ OO cO ON *i cu co . 5 CD 32 3 0 O g> £" 3 / — • v J3 X) O e0 oj CQ CD J = I O .-2 "2 1° •3 kH o o CJ eO cO P M ON • 2 £ * % 3 I CD O * N - O co V cO X! ...,.v-y.., yy " A'r >' *' ' ' * ' r fr' 'V > / a M , ; , 126 ctt >>3 1* CD CD CD ccj ^ 3 -1-31.1 " g CD O 3 a a CD a >~> "co ~1 a C+H o a 2 ._, CD 53 CO 6 -2 o *± a a : CD CO S CO CD O w CO TD OH cD CD ™ s I CD -3 00 a a £ 1 3 •a ^ 2 6 0 ^ 3 Cj - 8 CO .52 8 CD CO a. -a o ss a <-i a CD sts 211 TD 00 g CO § S •a £- & co *a CD .2 a g a CD S OH » o o a OH OH TD T3 1| CD 1 8 « 3 « 127 I cd co > CU S « n co |J8. U hp o cu 5 o ^ 8 S 3 co 3 3 . » § CO CO 3 _o 'co '3 cu -° -d O - M fU o eu cu X ! - O .s £ to o £ -3 .-3 eu »£2 cu a g cu 3 \>? s cu eu o - ~ O O , 3 £ 0 3 O 3 ••§•3 co eO cu > s CU S 3 5 3 3 ,_. O •a c £ a * >. 11 eu - r x> Sb >? cd S)*8 V H > £ O o -5 ' • eu J O 'co _CU x> ~* >> „ N O N a >» cu <* S eu - a OJD cu 2 tS -tat :: & i •§ I L 55 ys CO eu & eu cu eu to oo cu 1-4 3 ° O o co & CU X) 3 3 O co ro CU co eu X) s° 3 *o CU co 3 3 cO £ -3 eu CO O "o eu X) >> cO 3 oo 5 oo e0 0\ 3 3 O 2 ro 00*3 OH CU < c2 C N 3 cu e eu SP 3 cO cu co CO CO PM "3 3 O eu '3 eO 5 128 0> 1) •e o a, CD CN 53 ^ s o o 3 0 »—; co O OH ~ ~3 CD 2 o co CD «4H o o N NO « 3 C O 60 J D C O C O o o C O o 3 *3 C O 2 co 3 VH —1 "3 00 o H CD"cd s cn J" i ? i ** > C cd V CD CO - H CD a 1 OS 00 T3 CD _ V H rt 'co I ^ •5 T1. l o w-i CD NO 0 C H • * H H rrTf 5 CD 3 .3 cn o CD Q o •c 8 « V H CD X &0 "3 PH t -1 ON 3 60 •c OH &0 a CD co CD tm a cs c?\ f.y/syk**ysssssssss$-« ' ' ; 60 _c "to 3 CD CD CD 8 "3 T3 00 CD OS J-I I O CO 1— 00 C N C N f Sf* S"ff M M m •3 CD > I tH o <L> 8 CD ^3 CJ >^> X> m CJ 3" I VH C O CD >> ~(H ' CO CJ H O cO CJ 00 t-ON 1-H ' ' o do' ON ON B B o C J co' CO 2< CD ° -VH <1 CO • u c i -2 s £ 00 M H . 1 2 d 3 CD B CD co CD t C O Id 60 CD HH I cd 1 O O X CO PQ 129 £ g O ~ oo U C ^ 3 -»-» C d .2 3 3 « CD t0 u a , a> CO cd 60 2 u a s CO > N cd cn a 2 A a -o e c O cd CO •C 2 a ^ • s 2 o c o o V O c«_, u-T o CN V5> v» .a c CU a c o co Icu o eu co ca i . . o c3 o I s ~ cu CO JO 3 cd T3 C o o .a eu o \ H H CW CO X) 0\ cu cu cu o a CU o o o eu CO C o U eu fits* S88S V 4£ 5 130 |vfi» "it 1 ca CO i-l O s Ui co" Ui O +J 'co •> co" C 4s N-V55 1 lo 'co' 00 he w ° £ a I f 5 -3 ca a1 ."3 'co 1) cd cu 'I 5 co V s • * I 5 0 Q ca I t ,3? '« 31 1 ca x a -a § CU cj I* - O ca P "5 u •»« °< L; 3 "J5S •a G cd Q ui ca T3 ~ CU ^ JO w ' - a - T ? i § w i a o ^ * & co' a o 3 "g g ° .2 DC O «D S n o * ! m 2 ^ \ j ^ x td g -a x * * § co cu - r & « ca •>—' S Qs v, t PH R III S U < PQ CN 3 « C+H cd CO Ui cd CU X "c3 u CJ c$ co "3 OJ CJ a "3 O co CJ '3 o GO s o '{§ CJ CO CJ Ui PH C+H co I CJ ON PH s s .2 O CO ^ " ^ 2? a O CO > cd a c 3 o S P « » § I 9 SP CO O « X a cu CJ C*H ed CO cd CJ CJ § ? a o o CJ 2 CO § s § ) a la Is 1 t •g CJ X -4-» CO^ • iH X -u» 00 O OJ PH a" CJ e CJ SP 1 CJ H-» CO •S c£ 5 o X ) ° § O <u cd 3 > cy 2 CJ a u o co 1= 3 cd Q g ON I .a cu va x> 3 ° <r ft ** S f t t ft t £ f '4 ryffyjffffffffys.'yy s ~yt' X* *$ t*4*t * * , , , ^ , f, tff * t * * f * * , f tt y tt t ft tStfttStftf ffff f tf tt ty/A/f t tt *S ttttt tt/ ty t/t tyfjytSjySSs ftyy, * t/t Sstt. S S* S ± i */,*/'$' 't ' t ' S tt' ttttU^t/Xt f t t t/sfs St t ft yt t f f S ty / **<r t t ttttty y ttt. tfffff tft f f „ tt*t t y j * S, *• •* rytfyftssryftsttStttttttttttttttttttttttttttytttsrtytf* ttyty * *t t t t tt tt/tytt tt 131 2 .a .§ e & <-> £ & « 2 3 8 1 5 8 s-o a 2 an - ° -55 . . cu — s T3 CO 05 o PH 00 O 2 S « -fa cO CU O cO CU ON 'S .8 5 o ale 3 ^ 60 cu 5 cO ^ co i_ co cu O ' 60 to c^ —Hj co " ° 3 o3 o •s 00« 3 „ <<-c CO v> o ^ 3 O •< ° fa € 8 & cu CO 1 X ! 3Q H X > * OH . 1 ^ £ o CD g\ cO > ON OH cO CN CO CU o e 13 cO I • 3 O c2 -o 3 cO CU SP « 60 c+-c o cu Sf o to-CU X S cu 1) cu UH co X ) "eO OH '3 '5 I co" - 3 3 JO cu S cO £ •n 8 OH oj O O J S U M ! 5 * * y* CB £ cd ca 'S O 60 S3 . £ § cu 45 H-» <B O O , eu , ccj i_ | P H O S3 S3 O M •a o 3 § co ° 15 g |U ° In "S S3 t J •-H *<-H H-> C J o cu 1 c .2 1 , 9 eo S3 | eu co "St I * S S £ o 2> ^ ON eu, - s o c2 •a i S3 S3 o S3 - o •a H-» X I o co M -a "a, 8 OH O + H "O cu 3 X ! ro cu x> CU X , H UH ••3 O •S X ! cu x> co CU t? X C J CU 4* cl 3 CO •a 3 ccj •a cu a C J O *«3 M 3 cu cu x> ""co 3 O 'S co ccj X k H ccj cu X) i i H 3 ccj ca eu X) >, X) -3 eu eu CO 3 C M cu T3 3 O 3 - 4 CN co O _ —, X ca C J kH • -< cu £ 60 O H . 2 ° c 2 g > -T co 8 cu O eu k-co 3 O O X ! * - T3 O ca £ 3 <i i co ca 2 3 l l > -3 I « •2-2 cu > 3 co O co o eu S fa 60 O 60 co ca c I It c2 ca CO O g -3 o cu 3 to 8 "fa H ca to g eu fa TJ <U — { - ° 3 33 O £ --fa ^ " 2 CN c2 «5 3 O OH CO JU w co x> £ CU CU X ! eu e3 X ! -2 5 If Q Q 3 "3 3 ° t 3 3 2 x> 3 l2 eu £ ca § 5 S .2 CU <H O 3 i - O P i cj 60 CO CU - o eu X) cu o £ g 2 ^ § co cu rr> I-  j?  ca o is C 4) ( H K ^ • 8 I 1 ? cu cu S OH O kH •a -o •3 CU 1 CU •3 eu X> O i CD 45 kn CU , kH 3 cu CU X) >> •3 ca eu •a eu ca X3 i i H-> co & CU PQ eu I "ed ro eu . " f a ' c O 3 O 3 o o o 3 9 eu g O - o eu a Jta 42 CU co >-> s 8 eu k H Ji - ca^  eu ON T 3 ON «> ^ x> pq O o -2 S 3^ • ?3 m CO < £ b ON ca <U r-H l | x fi o w k n Q Q 3 O '1 -o ca cu cu CO <H 3 8 g .2 a cu " X5 •3 ^ O cu .•fa >" s s % B S3 T 3 .2 S 3 eu •2 S • f a ca Q -to X3 cj > ca ^ eu cu - 2 1-8 kH H-> OH ca 3 cu eu 43 S2 *3 ON ON „ O co O 1> C J > 2 ^—' 8 CO CU a t : cu cu UH a eu s 12 'co 3 O cj O CO 13 ca cu a *c3 > I cu kH C+-. ° 3 3 o S | co 3 3 OH 2 2 II O C M P H O CO co* ca JB £ ca ca S 2 ,eu C+H CO 33 » 3 eu •3 43 ca CM 60 O 8 cu C J 3 ca £ 3 cu > 60 O co 3 ca 33 "3 •S ro 43 » O co >-< fa OH g 4S X> 60 cu *:B CU 2 > Z .2, kH n CU co •S fa 3 2 § -o X j "3 ^ £ >H -2, O g g a .2 "3 H-» e3 -3 C J 8 O ca -35 « 8 T3 ca o nb O M 60 ca +-> cu O X> 3 kH ^ , C M eu k H „ " to to eu cu £ spa "3 a ^ u o o OH eu o a 60 ^ -s 3 co ._ , 'S -o §^ fa co ca ^ - V fa CO r -H ^ 60 r , ca B 3 6 3 S ft '„ 133 T D cu .a & CD IH CO «D OH O o 1 cd C+H o a • PH § a -a o u-> ^ o ed o ed VH OH VH CD cS 8 •B CD t H *OH CU -a ed o o e o t> CD ^ a CD <u m -| | £ !3 cd 2 - 2 PH 4J 3 cd O O •5b ^ o O co O CU H O 'j? ON cd OH CO CU TD CU VH RCU C+H CU CO CU C+H H O 3 "O a cu y 2 -2 OH M a cu .2 4 3 o .3 S § a .2 HH a -a o o O co — i H g 3 •a ^ CN H3 ,H a^  g ° ? & . g .S o cu cu S3 S3 >> >> CU CU 43 43 a o o c ed 43 a o ed cu cu I 'ed 5 cu a *a ed cu — . VH CO cd OH a o & »cd C+H cd C+H O t : ed OH IH cd CU x> 60 C CU I 4 ^ > co cd 5 £ "S cd CU a cu O 43 a « cd cu a .co. cu a 3 O co CU VH _ Q OH a a ° a >> a C+H cu cd 1.3 W PH T D O eg a I Hi a T D co CU co <H cu • rt S 8 >> +^  £ s cd CO co .SH co co ed a ta a J H -a C+H 43 O TD cu « +s a cd o cu £ O cu o cd J H * *C3 o 0,73 3° ti cu s co cu H—» -H H—• .ti 2 6 2 S3 S3 cu cu m P Q c2 c ^ CO 10 . « 'H^ ed* 3 cu PQ 3 .2 s 1 S3 .a CU 3 P Q vi ^ c N m r j - i n N o r ~ - o o O N r - i .2 2 cd —. cu cd cu 3 <D co g M 5 cu o JJ a a 3 ? ti © p cd VH r" r* TD O -5 »/-) ON O co 2 ON cd e • a 6 C+H O b a o 5 CO >-H 60 cu cd CO ed OH co O » a cd £ a .a cu % PH a'«' cd CU 4 3 X> cd cu VH co a 4 * „ c3 ^ PH ^ VH CO Q H CO " CU „ 60 PQ 60 cu cd 3 a co O c a cu b „ 4 3 CU co cd T i CD ^ 13 -a ^ s o VH O I B cu cu a -° o ^ ^ C+H ^ 2 ON a ON O w S o § — CO cd C .a cd a j 3 CO TD o CU TD >> "3 cd 2 6 « 60 S3 >> *1 o eg TD o 5 <2 i ^ c/ 3 ^ 3 • 3 CU • leg CU 4 3 VH O OH ed 43 +J 4 3 O cd +H r3 cu cd fd cd a cu I cu 43 • ° 4 ? cu -u> 3 o .a 43 3 60 co 3 C+H O 3 TD CU B o ^ >, 3 a ^ I CO u cd ^ cu O 4 3 ^ 4 ^ 'CU 13 s ^ 181 a co 3 • *H o v -td S3 £ CU HI 4 3 ed O £ S ) 3 ° H ^ g c2 « « -a ON - -H ON TD 1—I CO S *<r s StS'* * **** <f " J- * ''sV , ** */ Vs*/*** * f t ft tftj. f t f t , <r* t f * s * * <f t * t f S* 134 CD t3 CD s CD IP S3 CD IPG-CD cd CO • 1—) § ^ s ON S3 ON CD i P-l M 1 o <u 10 -S CD ccj S3 ccj s 42 •M VH o 53 . t i 4 3 4 * 13 ed e +3 . CD a I ScT CD CD O * C+H +H O co CD S* a " S § i c § cfc T3 S3 .53 .2 CO LH 00 a •fet CO cd •5 3 n CN C o .13 I s a M *-> CO CD S3 CD 13 3 8 4 3 OH I f co T J OH SH O c^ s « C+H 4 3 T 3 CD ed Cd 83 2 3 S3 < j </3 &0t cd 6 6 o CD co 8 o. "•e o PH a o e 'S S o. CD VH 60 O VH o. (3 O '3 CO a CD g-VH • * o. cd (3 CD > CD S3 CO H-t o U PQ (3 CD B CD cd° S3 O O S3 CD 1 B & CD ' c o S3 O O , CO Cd 1 O U 43 CO PQ 135 is,; 'y—s'/ -136 • *-< rn 2 T 3 8 cn e § CU CJ •n 3 -a 138 C O ti TD o x J* H u C O TD cd 2 "<5 • t t CO Co cu . cu' -8 •43 co E M TD c CO co CU u cu OH CO o C O a" to B CU .a & «U VH CU •B C O C O CU VH TD TD cO ON TD ON -a 8 H 0 ft. CO I? 5 3 3 « •< . cO ffi 3 « 0) H-» C O co +•> o o o to ^ S 3 M cO t > C+H t-H O O O CO CU VH CU co cu -s-a S I S ° C O 3 1 co CU _e "8 TD TD I f "<3 TD 1 S i x CU H-> ao to co 1 o •8-S s i TD 3 c cu s . & jf a .s ^ to J3 3 CU a i •3 - ° J - C O OJ cO > CO VH TD cO .2 >, « g cu O 2 *° +3 a> M t3 . « a> H 3 OH M • a c cu cu C + H cu g 3 2 S H X cu CO O OH PQ 2 <*« OH ° o a o a o '5b CD 52 E OD 88 W + H O °£ T D TD 3 + H co cO cO g W i—t to w O $3 ti «< 2 - M OH C CD *> X £ f . . e td •o 5 g % 3 _2 SP a g § O 3 eg co CD TD OH 6 a co § to a ' ? > T D S --2 CD co cu _a .2 -a 73 E £ . S it 2 o £ I a to S to a o cO X 'lo "S cO cd c 3 > S P « CO § ' t i to to to to a xi to C+H O HT D v © I NO ON ON "fr a ON a ON to B to a CO B cO -+H X a O o 13 VH to C O T D •a * .3 .-a C+H o x o to CO O . X g M - H S _ , CO o O C+H co eO >  cCD cO ^ co cd cO X "ft ' fffftt * V mt y Art t <rt •* VtytX tttttt tt ttttttt t tt tt t t mVt\ j. t t ttj, t* <ft t tt '** y 'tsy t ttttt ty * t t r * St 4-tyt t t a to B to a ed to C O u 1 I i o U -a C O . :£ PQ 139 CU 3 S •»» o 'S 3 cu CA CO b o a J2 Sf? cu • a cu cu p5J,,'; 13* , H, ~/ • <t i s l l l l l l l l l l '4, , ' C+-1 O •a § CU . 3 & tS on eu i h -3 fx 3 ^ w c2 S co 3 s eu 'cd S E eu O eu T. 2 ^ O , o O co " <0 - H co cu s 1 o CO CO CO eu t-l eu .3* 3* M £ c§ co 9> 60 -S 5 t t 6 o cu CJ ,1 « g <+H eu eu co J"""* eu ON kn ON 13 60 3 o . 3 cu > g ^ cu CJ 13 eu -a * H H ' ' CO ed • S 3 3 * CO CU co a a kH JO c2 £ 2 8 ? CU o s * 5 eu X i co co cu o S X X) «J 3 co ' """ CU T3 O cd •g eu CJ =3 3 cd 13 £3* CO ed S '-3 as -1 cj X ! _ « < CO" S -o .2 cu 5 I a5 p 2 " H—i CO c3 r p kn </S/M * f f A m*A "tA* *r$ 4f ft " *• S<r s •"> ' • ,,-v, ; r'trf'tiy;,",, 1 ft s /" / A"SSJAA ft ' ' 's S f f ",V AV*f.f 3 eu s cu 3 cd eu co •s cd I a 1 O O •a. PQ 140 CN ON ON 4} 5 O C ON 0. ON < 5 J C co o _ ^ u 00 o. I I O CN O 00 ON ON « © a e o u CO s OH cT y—s j d o *c ON ON '3 a LW cn <J ON ON CN *"H 00 CD ON CD _ C CD 'co o _g £ 'to ul de en CD .IS -o 6 lite a CN O ro © O o O cn CD e CD C O CD cd w •a 00 CD Cd t : CD wO 3, O ON ON CD CD • v 141 >>.<a cd - C M II CO > O H X> Cd g CO UH a o cu 1-2 g |1 • S i g e o CO f O i UH 8 cu OH OH 8 o co cd cu CO NO cn i> CN S3 NO 3 CN S3 «i .2 td .. "° . -»™s cd 00 o cu 12 "co o oi) S3 Ja 2 ^ — 3 o •o x i ca ^ g £ cd 3 a cd I X ) o • M CO C U .|> .i '•o , C U 3 fa . s i r N C M O © U- C U O H CO C M 3 « ,eu C M C U -3 cu C O 3 c+-< cu T3 C U 3 33 cu x> cu X ) 3 Ji> cd .a co C O g •a e' = 8 .a 2 H-> W HH O H cd X> OH eu X> C M O tJ O OH >T eu UH O C U cd co eu •a 3 O 3 cd X ) ed co cu UH T T 7 cd ? co' O 2 •3 ^ •5*v .3 »- i ~ O 2 o 8 cd vo o o ° o O O OH —< 3 55 eu £ s Q « O co 3 OH CU 3 .2 C M — 1 ° CTCN eu ^. O CN ~" ON 3 .-fa cd S i M x l 00 8 CN ON CO 3 O cu co 00 ON - o op <: C M 13 I 5 H 2 *3 O X ! ^ cs £ ' « * n » ' T ' " T co o *§ T3 •3 u X3 co S-f ,o ' c O •£ Q cl eu e eu O UH eS 3 w 3y -/ ft 3 3 00 g J4 eu eu cd co eu "13 2 12 oo S 3 jd - +-> * S oi O . 3 - U 3 OH O CO w . 2 "3 cd «> o 2 8 3^ co . ed 3 ai PH W O •••• •• HV ' ' O o 00 cn • CN NO I o o 00 3 O .2 S & 8 3 Q CO H-> oo 5P 3 .2 i : .2 c2 a C U UH •2 c 2 2 £ s % 'ob o «- S cu O O H-* O H ^ fa a £ eu cu 00 8 cd > ed co i i cd cu eu X3 eu X3 -s X ! ^ C M O co CO _ C U "3 oo 8 cd O cj 3 O -3 eu 1 cj cu X> B CO cd eu -3 eu co eu 8 1 / 3 § 3 3 cu 1 3 : 3 3 cu C U 00 cd 3 cd "cd oo cu cd •e cu X 142 143 o . 3 co to Ul CO JJ . OJ cj |<g § . 6 >> o S 3 •co I ! i f O . „ cd , . t r 60*3 2 c -S 1 S ' C 3 -3 ^ to l i s to ed o CJ a CJ SP 3 cn 3 « C/3 a .55 3 cj , ° cd E «J e3 .2 CJ T 3 f i o CJ co 3 CJ CJ CO o SP « . g I J 3 * I*f5 ed cj O —; co g'3 3 IP4 cd 2 C7\ W •< CO CU CO cd CU co 4* CJ o CO CU > CJ 4 3 to •a 3 3 O cj C+H O cd o 3 to co to r-ON cj I S 4= S 60 . C to cj to t> «- i ON O ON to CO O I I « to 3 -3 e3 co to .co CO s*"1 CO - ed o CO to J = 3 a . § 5 ed — P4 _s 3 3 to co •3 ± to 43 +i 60 cd *3 to "(3 .2 ' if * IS to +J co o — n f r- co 4 3 1 2 S In I S ' C 60 2 -ft! 3 I'-5 & £ IH U 15 5> 4^ .2 s >> -3 to 43 ^ 3 - 2 co 5 *2 ed cd cd CJ ~ "3 u OH to cd cj u CJ cj OH «rf< £ ^ H co *3 to OH eT _ed 3 a CO -3 to t d > •c OH 3 O to c d ° CO to SP •I 60 T 3 O c2 Ui <2 CO 3 O '•i 1 to Ui *cd CJ to OH CO O 3 to 3 to Ui to 43 H a a o v C+H o 3 O ti 1 3 O cj 3 .SP 'co to -3 . 3 CO to co to PQ NO CN cn NO 00 4 ' -V . /' *S <r Cd CO . 2 cd ~ . X! ^ ••a . 2 *> 2 co cd s ^ til 8 - o 8 2 3 .2 .-r x> » 2 S T J -3 ^ cd o co *.J"3 • ^H cd kH O cd O CD > X> O o CO co CU l-S cn" s _~ 14= u T J 8 ] T J 3 ca o IS £ I « U H 1 ~ CU . . X> g CN l« . S CO s 11 ^ cu ^ 2 cu cd o ^ r cu O JX 2 ON ^ OH H-» 3 O X -*-> > _>> .55 CO CO CU UH oO 00 cd -3 CU o p « w cu "3 > cu 13 -3 ^ § s O £ f-j 5> X > O co •*-> .2 3 3 xi co 73 X i 00 " O co UH CU a ^? 3 « co cd cu cu CO OH 3 a7 -3 I CU CO " C - O u cd cu -^ j CO OH O 5 cu cu a 3 -3 >, .a § 3 -3 CU Id 3 H-» •I X s cu X> co i i X! cd cu X I U es cu PQ cu u c « co IS 3 O CO cd cu kH 3 cu cu x> CU I X! cu 1 -3 O c2 St /»—-. cu" cu cd co X) « a kH Cd OH ^ O co OS C 2 s 3P 3 O CO •3 CU CH—( co •a 3 2 00 I o C+H O co cu 3 cu 1 2 '53 cu kH CO cd X! cu CO cd cu ro cu CO 3 9 s o ON ON s s o ct5 CO 3 .o 'o ca cu •3 CU > O •s I 3 oo - 2 ON cd cd cu 3 cu cu 00 3 1 2 . cu 2 x ! O cu H-> to cd OH . 2 CO cu 8 3 ° J£ a § 8 3 3 k o"cS^  § o is 3 2 o « +-> 3 co CU •a *a cd -*-» cu _. £ 1 *-> cd kH O <u a & 'Xi cu 2 -cd co S CO cu o co CU 33 cd .2 6 "3 OH .2 a' ° a §1 •HS 2 8-° cu cu cu xi •a a <=> " 2 S > a? T J 8 C*H O 3 O § I § a « CO S I ca o 2 •§ | 8 & 8 -2 § I Si S 3 UH 3 OH g ON > 2 5 p ^ ^ H 8 ~ 2 CO £P 2 •§ a § S 3 CU >> ."3 X> xi oo 6 2 (S r 2 > .a x i H r w 3 cu ct! p o ~ &.9 cB a p o Cd CU X ) CO^ """3 -3 § ^ . 2 "cd fa U cd cu 3 X ) CU H-> cu p 3 3 cu ~ kH 3 P "i | U -3 oo 9 a ~ oo 3 "8 I . 2 § £ V 3 eu co •<-> •=?a cd 5 0 n cu £ cu 3 & o s *7" 2 cu 3 p eu 3 cd CO 1 cd CO cd ^—/ eu a p eu X i eu cd •3 cu >> <§ CO CU -3 cu 13 co X I 3 O XI ON +-} £ % _1 P s i 2 -2 cc 73 ^ 3 "1 ON « ON Xi , ^ 3 T J cu cu a p H-> T J I co 5 ctt cd T J eu co eu cu o a _ kH CO eu 3 aid *H CU ca cd 8 P eu X> 2 CO eu 1 "C 00 00 8 T J 8 ,cu CO 3 ed T J XJ ft a eS ^ T J . 2 ^ >•> * J ro - 2 CU «C> OH 2 5 f*> CO cd ed* eu p p o 'C cu CO kH p CO 00 3 CO 1 s cu 2 2 CO "K £ P s S CO .2 » X CU co h 33 cd S a o cu aj > H C Z I CN ^- J< 0 0 < < - 3 ON w—i CN cn Si a ' •5?"' ' ' ^ *f/* s is s s jy/fSf- i s s ff,wfsssssffSfssssttsfffssssssssMnfAssfs.'fs\r.-.< '//'i,,/?,,;^/,/ "/,z,//'H':A s WA f t t *t trtw ' y * A s s s s f ft* J f t * s * 5 > 145 to •S S IT) rfc ON ^ co to o c O •a CN n o CD Id o to U. cd /cos J O 3 I cn S3 s < X - w S3 x, E 3 to ON 6 0 O . £ ca "3 S ON « cn a* § w << cd to a 'HP I T 3 to > I to X> ed to "a to CO C u , O c o in ON ON c o cn c OH to j -* d •a t: to p cn •>-* cd - 3 -o -a « co o •c * o a ao o §1 » S3 ed co to O X O O r ^ j v CD a S i ON -n ^ x 60 •c • - p co -o co dX S3 to O o to I 'cd CO CO to O to cd to to § X to X -a 1 u CO S3 X a X o T3 to •5 a 22^-gj 9 | t t & ft 1 si 11 tt s ft* B IS *5 » % so "S? 2 146 cn w «—i CN cn < O CD £ « * ^ S3 3 CN I 2 c2 O CD O 6 0 VH ""j O H CD T3 T3 1 3 CD CO CD T 3 £S.§ CO cd # . 2 r « CO 3 CD f O O CD & CD CO VH o "3 CD cd a 1 6 0 3 'co CO c2 O m ^ C o 3 gn CD CD § S3 •a cd 3 cd C+H O c o •3 •3 p co CD cd S > CD c ~o . 3 C M o © « ft CD H O O I o '£ o OH s" cd CD t+3 c 'co & 9 8^  6 0 co a rt '-3 o CD O CD \ 3 CD CD o, O 7^  +H O CO C j _ , £ 2 d ^ Z. o §S +3 3 cd CO co S § c2 O i n CD a VH c2 o o o O N O N o 00 3 CD CD i 3 jo O MBE*" * * * * ' /*/*/*•/*, ******* "* */s ******* ** y*/ *• * * <y* / * We**"** f/s J* ** * ** **f W****y***************************-** mm^^m^Vm^^^§M*^^ * ***** * ***  J* ** */** * V * * * * * * ****/*w*fty** f ** * ******** * ,* * ***/*/***/> ** * * **.* * * * ** **f***** * **** j*** */%** f , / * " * /* n? * */ 1 "/* &*****%••*: 147 cu > 'a OH 5 6 £ <-> a O C O eu eu CU X a o ed Ol C O s M o CJ o * H i co > cu XI cu CO CU XI •<-* 3 © O N O N cn ON o co C O C O cu :s ccj eu X> ' C 0 0 T 3 CU I 3 * 8 1 - O a cd O ef o c2 3 o a C O a cu it o cj \=> o x O o I 0 0 s o s e o o cu d^ '3 o co o cd % X> C+H O CO O •a l 2 2 a g OH O UH u 1 "cd K cd s X) eu X ! H-" >% x> -a eu UH eu 8* "5 'i >% - a o eu I ed - O OH P c2 S 04 5 I 3 o "5 & o CJ eu cj 3 cd UH 3 C O a cd C O J H a> § eu > cj \ 3 OH eu l i s CT U H O CU co cd ® cu 0 0 — 2 * I cd 0 0 T 3 *s a . 3 eu cd x> •cd c° 6 g CO .o UH -S «J a OH CO CU 3 eu eu OH eu eu PQ e .21 « CO e cu Q E ©i U| co! CU! • i x: cu eu PQ cu PQ< s ocj UH c2 _eu 3 'ob CU 3 ° a UH eu OH o 3 cd >> X) " 3 cu C O O 1 CU 3 cu eu X) cu I "43 2 CJ T3 3 3 • 3 < M 0 0 O cd co g co 3 S - 3 £ O O .2 3 cu 0 0 cu UH 3 cu eu 0 0 cd 3 O O 3 eu PQ I 3 cd s a CU ' co 3 O OH C O cd X) 3 S X) 43 1 148 ,to H-! 35 to 5 to 60 Ui cd ed eg T 3 cd O to - o co 2 .9 i cd a o •o - to • u to a O N o co to ^ S to o Z B to u S a 3 .8 3 *-* O to to J> "3 ^ cd j o 8 1 u c2 CO ccj >> cd 'to c? > cd to cd OH CO r N 4 3 to ON co" 3 2 3 J2 >^  3" •3 U 8 Q 4 co" to 4 ? 7 1 -3 to > •3 to CO •c 2 I o 5 ^  S i " co -3 cd 2 CO _o "3 o H-» "3 to co 3 to 4 2 u CN co -3 cd O U Cue O CO to Ui 3 O a o CO cd to CO Ui cS "3 to co 3 to r-O N I CO 2 to CO o "to 2 -3 to co 3 O CO - a § Be C» 1 to I 3 cd '3 a 3 O -3 CO *c 3 '.=> to -a 3 to 4 5 to 1 co" J 60 cd •3 3 O to to co to CO O CO <H> 1 S 6 ^ B i ^ 8 5 S <1 cu 1 0 ',S/'SSS,/, 3 to B to SP 3 cd CO CO to to I cd 4 3 " W0>. 149 i I CO CU O H CU cu a 1 g: « M Q c cu 33 ca -a "c 3 > o, S o So O C cfl N S I cu 2 r H C . . j£ N e3 * / * * * * * * * * * * * * * * * ***************"*** * * ** ,* * ****** * **** ** */** * * "•$* * * ** **** ***$**'*'t * ************** ***/y-* */ * **% ** '"/* , ~ , ' * *** fef'"'*, fJ * ^ S * B 8 8 ^ ' * * * A V S / ^ ' * ' * *"***'%** ,*,& V**W********}*ffi*^ ^^**^A**v*^A^^^^^*i***^My^****^^^& '****S /* ma. * ************* * *** **** * v* % **//'* V ****** * ' * '* $*M **$**** ' * $** * /'* * Z' i * * ffiM **/*****.***,** ***, * ** * S*r * * * * * * * •************. 150 to O N <D O N 8 *i — ' O to a is to _ 'So S co CO o H H cO co •£) <D 3 £ to •3 I CU OH to O 0 <o a. 15 co TD o OH o U i OH to — C to -a < f2 •M CU a -ft 1 2 3 C O 03 CU e a w cu §3 c S •2 « " to -2 e « C O © co £ " to C M co "e5 <! + J a cu co B o L ^ § J*«s © I—1 to r 5 «s B ^ "g •311 •3 S a *• 8 P £ £ 'I cd to 03 O , ' t o 'a I C O to N •c O •S 3 H3 to i ON ON U « < B to E O to •a s S °* .5" 5 u <u II to -a r„ C O HO ccj TD ">> to X C O a TD a ed -a " 2 ^ to cd to o 23 cd a c TD « S ° cd to £ 3 -g > is ' C 23 AH TD co cd 3 PH UH c a ° « « _ a - H vo 03 3 o\ w S ON •a & ~ tS 3 O •a co <y £ 22 VH O "*—' co .a •=< CU co to a o t i to I cd — a 3 O a co cd •« a" 2 I S Cd TD ^ c C U IH ^ •3 fto t i cd ?*H cd fll C+H + J c 3 5 « - ° " i a 1 to o to • 4 H to 2 -o •S 5 g PH > 53 C+H ^ O " O cn T D •Si -a s; cd «-a cu C O to OH o a 8 ^ co a to .-3 O H .a .a ITS " / '" is/;,,//'* V///S/S//SS. 1//", '/,',?' /,"„/, *? lllllt 'ft "//, ' ; 7"f',' i i " ,/' 151 NO ON ON . '« CJ O , U H cu , 4J 8 3 33 [£] .3 1 O ^ ccj 5 ^ ^ -g IB ^ _ J S _ - O flj . o o <u ZfflPQ .if "3 -3 TTS 3 * J « J UH OJ > , CO CJ ~ 3 CU B 8 o "S3 3 cu £ cu s CCJ CU CO cd •c cu CCJ .S3 cn ON ir> P H '3 8 ' £ O P H CU •a CO "o3 > 3 cu PQ mm 1 ^ 152 5 •B P H 1 O • '3 1.55 VH to • »-H, • CJ ccj IS H-> OH C J I 8" O . to •5 C M O a o ' f i o OH ccj s ccj a" o to a o , co | T 3 ccj 4 H to to cj •SJ a -3 CCJ 3 I-a ~ ca 0 s-w Q <J 55 .S I 15 o CJ Cj_| co O 3 a H3 y O « - a-a * a ccj to H O NO C+H O cd O HO 3 rt C+H O §•§ * C J "3 *c3 o g .+-> »to C+H O co CO ccj 3 J- J cdS-3 2 ctf ^ § O 2 H 2 i i i o *r> m r- oo ON ON ON C+H 3 CO Co" « cd ™ I 00 co' - J S3 - ' 0) co . S3 3 to C+H •° o f—I +H C+H O ° 2H cd ^ +H • • O *™H 3 ON CT1 ON »/-> NO oo M 00 w A ON ON ON ON ON ON W , I—H »-H i -H I-H *-H i -H 3 -3 ^ O ON 3 ON *i to 3 cd OH . I T3 O O >; -3 ~M cd H g " ' o V to *H3 O 3 CO to td "3 3° '-3 3 to 3 O CO cd to CO 3 3 3 O co cd to CO C+H o to VH O § ON I o ON ON ON 3 O H3 00 to a ^ O cj>j to n S § -3 a 3 co XT ' ed i cj cd co 3 SP 5 O .3 S3 ^ "S 3 to <u ^ •£ £ 2 ON W • cd S i to 1 > o £ to t3 •J o H-> VH to to H-> OH to 1/3 a ^ c§ t 3 • 1 = 3 ^ * 60 3 o -a ON O H ON in © ~ H t~- ON ON ON 3 +2 CJ 3 cd 2 OH CN O ON O 2 a cd -+H o & -3 to ed H-» 2 cd -a 3 -i-> 3 O " O 3 3 to B *a . 3 « -a 3 ti ti -3 ON 3 3 ON « to j -: -3 -3 to to -a -a 3 3 3 3 1 1 O r- ON ON ON T - H »-H I I o in ON ON CJ cd P H 3 J to ed T3 3 O 3 ° 3 D a ^ S 00 .£ 2 "C OH CO . . H O ON O 2 a cd ti |»^ "* * lllll • *s * •* : , * 153 a I 2? 154 PQ 3 oo .S2 c T3 O .g to T3 , C U s cu" C D ' 1 O H C o co C U "H. ON o o t > 8 <« to cu • mm eu O t/3 T3 C U oi c CQ B H i ccj op cu 6 0 • ™ ' O o X3 co c3 cd - C If 53 CO cd M a c § C U a o p G O 1 cu cu co cu • M C U cu O H co C U co o >> co C cu cd > , 53 - M m 1-ed cu 93 +•> to C3 B « B O NO ON ON C U cu o S "S I cu a <3 cu -a a 2 ? 1 § ) to ° 1 SP I % s cu 6 a c2 cu ^ UH £ OH .2. 0 0 C U O u B C3 I cu 93 5 cu 1 C U T3 C M 13 CO CtH o co C U co cd cu * * s*f •* , rftj**^** ******* f ffsttf $ •*•* * * **&*/* f V / •* jut*/ JV •* * /•* S y t * * , v ff»y4'',/<S'v/t/s' sf, ,y 155 CO I V i i CO 1 O rn O to O cd O <u 8 * ' VH 0 eg co O T tJO 1 * I<5 CU bp CO O r-00 " O - s CN cd 00 w - < OH O tt o a s CU o C5 B « CO CD VH O O O CN" •a a l If OH a C+H O co <D SP T 3 o a o co •c OH a • i - H S3 CD >> OH ' " H V I H o T3 c cd o O O o" A T ) & 9 03 o +H OH a C+H o CO CD 'M S 3 ° a -a cu <d o< o -a •£ "S u lm CD S CJ cd CD VH c2 © © © ^ CN < y » 02 CO g 1 1 •S CD >CD $ cd 2 CO T3 B pfi • M u •<-> 8 o « 6 | co CO 12 d <a « ^  ti 3 VH" a> 3 «J 3 CJ > ' a s CD cd +•> cj co ed » ed d +3 CJ 60 Jd H 'S E £ ? J 3 & cd CD CN co CJ VH cS o 1 CD > ' t i CD OH CO C+H O CO 1 OH CJ 4KH O CO *a CO a CD T3 I a a o cj J*\ CJ cd CQ CD O CD CO tt o PH 03 ro •a 60 CD CD ti CD OH CO 3 CO C+H O a o 'a* 60 CD O 1 o cj P H on H J o T3 d cd 156 I x> -a fi o CO • 3 £ Q< ft ~& I O "3 O <> o & H 00 4) VH C O I - | 5o 3 c o m -a a 3 , o >» -3 3 CM O OJ 3 c g ^ . . t > S C O « CO M l o p 3 fid h * i c o CO H vf I - o 2 a, C J VH CS VH P H B 3 H CO tN CO C>H « -in C+H -V "o VH .irK. J$ s tt CO 1 CJ a , T3 C cd c o " "O c-"cd CO OH *3 £3 O 3 00 s CJ co 73 3 cd CJ T3 & ° ^  Id g CJ cj co rt •—> CO co § s VH VH C+H OJ O H3 .2 8 S3 00 CJ 3 £ ° O co P H cd 8 8P 13 00 CJ cd 3 cd "S O 157 3 O S 3 i s 3 3 3 CU «_ X i cd _ eu O o ^ 1 CO CO T J 2 °-ft CO 3 P H e 60 cd - M OH - s (11 ^ WH eu O H X eu H X i H-> _ CO O k H ccj X ! u 60 - ° 3 >> k H N -2 .2 • • 60 s 3 (WD M co ft-2 * 00 U ON • . ^ •5 3 = "kH , « OH TJ co M 8 S 1 W T J ^ cd •3 4 1 ccj c +- CCJ « SP cu 2 > £ • - H " 60 pL, CO T J O Q T J cu "—' N c*i co 55 5 2 o e 3 1 3 60 T J 3 ccj M o ccj X> 8 cu cu H-» , cu X ) X ! co . 3 " 5 . ! ! H-» O co Ct_ T J .2 co CO T J •5 in CO .H x +r: O cu co "3 c CO S £ OH CCJ co 60 3 .SP co X £ * 3 60 •5.5 2 s - r t kn « 2 CCJ OH CO CO kH WH CU CU CO 3 O 3 OH X CU H-> CCJ H-» CO CO kH CU H-> CO 0 0, T J 8 CCJ co C 60 CO T J ccj CU X I CCJ kH H-» CO CU O H-> kH ed kH CU OH cd OH CO CU CO O CU CO CO CCJ cd CU 0 * o O co" O cu 'H-> "C cd 0 cu H-> 3 CO T J co eu kH cu cu G H—* 3 3 3 _o XS 'co T J 3 >> CU cd CU CO H-> H-» T J X ) an xhi 0 CU T J kH cd .3 kH Ct-, 69 in. C3 ca ca IP i k T J 3 cd N l k Ik CU s „« cd « 60 o >< £ £ ^ O T J T J O a C HH > cd O > ~ -2 x £ S t ^ cu •» kH Cd & « 0 5 £ £ T J C O 3 "cd ra 3 C O „ C O H-> ^ 8 ^ 5 co PH CO B N = © 00 '2 S es ^  • "H N . CU « R &> a"^ .5 b co cu ^ PQ ^ ^ a > ca co ^ > , X 5 .-3 H-< H-» ? CU T J .2 J3 cd o k H * ^ cu OH T l O N • 2 O kH , O 60 8 co . 3 co T J UH eu 43 2. T1< o UH OH CO H-> co" 3 £ -cu cd O OH kn eu X ) T J C O C O cd eu X> N N 'G 60 CU X ! H-> CO CU CO CO 3 cu CO P H P H Q X> T J CU cu 3 T J O P H PL, Q cd cu ed 4) > « < O H-» c £ 6 cd • CO eu *OH O eu OH O O o" iri >> x» T J eu eu T J 3 cd C+H cd CO kH cd cu x> 6 0 , _3 . 'co co 3 O co ca C>H csT ca 2 ,cu X I Cd co co . CO U - i _ , cd o cu 5 X i c O T J 3 cd CO cd 5 UH cd T J OH 00 ON co eu -*-» 'co OH £ cd cu T J 3 cd CO T J cd CU X j cd ed OH cd Q 'hHt o o 04 > N kH cd H-* 3 cu £ 3 cu o T J cd kH eu •*-> co O OH O H CO •2 cu o cd kH ed X cu T J 3 cd co T J CU CU 3 ~*Hk R a o k ca QQ *1H* to cn o C+-C « o .2 3 c: O eu co 3 H-> CU co > eu eu 'z kH O OH tu, ed g X> 60 =o .S £ x! hq T J c is. cd Q 60 O 3 ^ . 'SH ?> £ S cd k. CU f~H cu o ? ca 3 ^ O ^ 3 O X> cd 3 cu kH 8 O U kH cj cd co cu cd OH eu Q .cu X I co . ^ H 60 3 H-> cd CJ 3 T J CU co CO eu .2x1 ^ S x>G kH OH H-> 2 OH OH 3 - ° CO CO 3 CJ CO cu X ! H-> T J 3 cd co" kH eu X J o cd O OH 60 3 8 kH 3 3 O "H-» cd cj 3 60 X O — O O X cj CO P H PH Q cu CM r-' cj 3 cd O 3 '3 cu > - 3 W H co cu CO H_I X cu H-> cu 8 cS O ^ e cu cd • 3  H ^ to ca R l k kH cu PQ H-> •S x! 60 3 k- M 3 ^ T J ed 3 O ?? eu T J g T J H ^ O O 43 ed -a « ft g '•UJ 3 cu cd CU -UH S £ ^ UH CU . - H .O ao 1 - 1 ^ cd ft .3 C j -C d Hj—> 'cu 6 ' % 60 kn ^ cd cd 3 P H « — _ X cd - ° C O + ^ cu 1 - cu o 3 ± ^ i Q ft S C3 eu c S T J o M*M / f t * st tt ttt ***** * *** * ***** ** & ** <r * * * ** * ****** ** a . 'S p ' //SO,//,"',',,, H P * * ** *. * * ** * * * **** * **** * ***** * **. v^y^y^wyyv***.'******************** * ;**%*: / * '***'*<*$' **'**$£ **, * * ** * y * ** * * *** <y * **** ** * * S*^^.*.*.***.^*^****,*/************^^ * * * * ** **** * S* ********** ******** * *** * s**ss****s*******s**s******S*s.*.-. * ***'"y^'* * /f* ? 158 13 GO "eS c o CS o CU Oil 3.5 bfl eg •O » s I 2 8 co s S S < +-> - " X c o * ^ eu JB ^ 2 co O 53 cu M - O >; « C O «» P C o u t a * s u u g o cu «J co ^ 1 5 pIN 9 I * ? St; i '/,'74", , s„r. 'v'-M"",/" Jf* ** *****, * ** t*** "***p*/* MMM*W*WM*% * ** * */** * * * * * * ***jf,***/* •Y*V*VS*VS*S*SSSSS***S. ^^mmmmW^^^^&W^^mW. w V A V V *** ' * * * * * * * * * * ********** ** J ****** ********************* MSSSSSSJ.*-****.*.*.***.**.*********************************************^^ * s**^**, *• ** * ^ * * * ***Sy**** * *^********* * y* *****,$* ****y* * "»4;y,v iu, m% .. ... v****,-A***********.-******* '**** ?/ *% / * ***** _ ' *'*% *%*kfi •*Stf***S**S****f**^S**S**$f**fssj*^ * f **/>****** / *,/** *** * f f * f * ** ** * * **********************************************. *************** *** ******* ***$***$** 4VZ. ' * ss ***** V* * * * ** ** * * 159 ed I to J$ «S ^  £ & © - S JS (50 CJ u 2 a CJ • to-T 3 H— co co C « T3 .2 § ^ 00 co ed « T 3 £ £ <» M S >-O k « | c O &Q .2 U to K ed 60 >^ • g & s o o ti c2 | to H c o to to to cd £ >> SfiffiFte tig* •* ** *fs t ty Syy* ' SsPj, ff/$*f $„ **y, s'/ii'Jti!, ty*ff y.*Miy£.t*ttf*********** "*"*/**/ !%'*** " W '**fif* * '**'/s, ****** jmmfflsm * * * ** * * * * f ^**^tyyi**\***iyy * * ********* * * w * W * ***'*$ %"%* * <*y * y *** * * ** try* **** ** * ***** ***** *****¥** **S*W *** * * *********** ***S* ** S***V. ** **&**** ******** ***** / ** * ** *yy ***, * *** * * * * * * *, * * *SS * * ********** * * **S ** ******* * **} *S**/* ***** ft ** ** **// ******** ** **SS* ******* * Sr S*SS*** S*:**?ts*y:*: \**V*yyS*ys/S*VS*v\ ' s *** * 160 t-c CU -3 H-» O T3 CU 3 S 3 PQ cu 3 o o .g CO 3 O to ^ co s> S cu u, i i cd O cu • a ccj co * ^ ccj CQ CU CU X> s-s eg CO 3 « - H cu l> co a M P CO CO O H * H CU g P -S ,cu 'S3 33 co 3 o cO 3 !±i .S3 '3 1 3 s X CO S3 PQ V 3 O • 3 161 163 eu o 53^ eu eu ed .5 .22 oO eu •3 « '•5 S i PQ § 5 cu cu T 3 CU P .2 CO SB is* * , , ~ ^ CO , . S X> x! *o a 5 T S b eu 1 T3 CU cu cd cu o g 3 S 3 164 165 PQ «J 1 .2 < « i x 3 CO J D "ca s OJ PH 3 • 3 < CJ *cd s CJ PH CJ cj I o PH1 « 1 -2 —< C N C O - 3 3 O O PH* 5 S PQ 12 o U C N - 3 3 O U d i s s §11 £ s P4 PH* C N C O - 3 3 O U PQ C+H o co CJ •g Jo 3 CJ o t-i X 3 CJ a.x CJ 3 oj 2 3 O ti 1 CJ -3 o cj "3 31 ccj CJ CO ti cj Q CO **S***f*S///s .J J- *** ** * */* ** * * w?S¥w ****S****s ft t ****y *** ** ** vy*Y**K** wy*SSS^*S*S*..\***S*****y*.********^ •*********yyyj:ty.*******y*j^**^ "ty*******" *t$* * //**** * S ** ^ '•>//1^>^^1w>»'AwW>X : .• .v X- v / / .v .v / * **'**. r • ************f************************AV****'**W /** S***y* * •* ***** *"*'* " &$-&ffi<>s, *&?**/**%* ds&sMSiK ft/*/**/*//**** *****/ * * * ** if ***** ,f * ******y** ft * * * *•********** ********** ** \A*s*y*t*tSSfssss^ ** . * * * * * * > '* ** * * ? f v q ^ ™ ^ * ? s j g » J m '*#*/*,* *$* * p p * *"*"*• T"' " * ' * " * ' * **"/**** * Mm f*******/*******^**********/***/*********^^^^^^^^^^^ *s"*'*.'**/**/*, **/$** ***• *»:< f co 3 * o 2 & CJ OJ ° 2 2 B W H - H ""3 3 3 2 P i S ty'*******/*A-***/*****lf***t *• * yt** * .:•*********.•****** %Mm$$**w% ****.•****************, 166 CO I0 CU > » o . to i CA CU '1 § . 8 ts PH cd cu td CU J2 "<3 >-. a cu • f cu OH cu to; CU 2 CO CU CN A CO I 23 'cO I i O "3 CO Cd* cu >> CN A If 3 O CU •a B cu PH t*^*******f**f*/*****f*A**A**^fyKf^Aj^y*A***/W*. wjA&&MJ!*M&MM*f^Mr.*jfy,******A*********A** •X •• "tttttttt/ ttt t * *** ****** ft t t t ffftt tttt ft ft ; "As,/;?/,;,",;,'/, "ft" . t ,t t *** tt t tttt w t t t tt t w A* t * t */**' ft t . ttt t * A* ** tt ft tfttf t t •* *f-*tf$ttf v.fS.f.*.f.fAt.fj fff1r ffftfffffAtM ' ****** ****A /At t tfftt t/ S<*t t ttt fftt /ft tfff't ttt tAt ttt ttt*, ^  tt ^ * ^* */* * t/t ts 1 m I P 1 '** *'/**• %* *S* "^'t ' *'%*''*"***'ft tAttt^tt'^tt t **tt ** * 167 o •c Q CU CO co B PQ 3 CU £ a . ^ co" .S3 CU ••s o C/3 r-< cu CO Q CU co CO CO 1 60 £ § S C CO • - H cu s • s . « wo cu cu "3. ^  CO CO CU CU -a .a . CO CU CU CO CO CU "33 UH co CU cu CO CO CU c •c OH CO 00 3 O «3 -3 3 cO c ^ Q T3 •c o 6 cu 3 cu cu £ cu X) CO co CU e UH cu cu -3 60 ' 8 -i 3 2 •c 60 cu cu •5 cu CO eu u o "oS c CQ ccj C eu "3 W OH eu cu cu cO UH c2 cO 'C cu U -g i" CO CU 3 If 3 * ; 3 C | o 3 co J£ UH S 2 ° x o co 4> ^ cO o CO w S £ ts f l ttty****Ay t */** /** y*t t , */» * */ t /1 t tt //ttt t t. t tttt * tt ttt t t . ' t * *f*V * * *S * ? t ttft * * * t tt tt/ttt t t t ttt f t tt ****// * t tttt t/ t t t t /tt/t t tttt ' ft t sssttt /// "t / /Art* ttttttt/sttttttttjttt */ tt'tttt t t* / t t t , * * * ttt t t * **4A*SSV " '//'""J * t fat t tt/t, t t tt/t / ' * -w t t ***** • V* t //* *//*./ ' - ^ W ^ « ^ X S « ? » \/*jVjr*.*y.v*.*yyyyyy/yyyy^ ' W , " / / , 4, ';;/,,,» 4';,s:7;w:',,„;,; '' ';:%:';;',„ %&tr<v$&8ra™M*J" V"S" ' ' " ' s ' " » V V S...SSS..SSSS.?...* /'»',£, $****%£#///*•! ttttttt**t****tttt *******t y* <t ttt sSf* " "\t*Vt / V *7'* t * /**^t,* t t't?j%& * /* * 168 > N XJ 3 o C O C O a> C A 60 -a oo § CD kH o •1 o 6 1 CA M "3 o o r=s H3 £ 2 3 OH 3 CD X) S o C M O B I T 3 CD OH .& & a> CD X> I C A CD CD r3 CO O 00 CD o • O CD H-> OH a> o o CO CD X) CO CD X) PH S < T 3 kH 1 3 -3 cO CD .a CD CA CA CD CD -3 8 O X ^ CD S CO CD " C A CN CA CD '3 CA CD CN f > U 93 P H "eS B O • PH • M CO l a CD es o CD 1 co CD CD co cO CD "3 kH CO T5 8 93 h J HM B V E 0 i 93 B 93 •a B 93 J o s 93 CD s CD "3 C O co CD CD CO CO J& "CD kH § CD X> V i i II.' 169 170 , f* w * s<fjy *** sfo Ss&ffijyfi yy f * * * * * * syw '* I o O 3 CJ PQ £ CJ . 'co C o OH CO • cd o 171 1 1 172 hH CO > J , kH C < ^ S I •a 1 IS o ^ O cu CD £ " & O S o U o cn 1 -8* o co CD •c CD 3 o m CD T3 Q CO CD I T ) 8 ccj CD 3 O CO C O ccj ii 173 CO c 2 S cct o S3 S O N S U J3 td in ed .to _ e to CO O "to Co" CD 00 a 2 CO 2 * l a cd cd to ed <2 6 C a to c f 00 C O § • H i t | - a a o $ I C M o I 7i cd to H 3 Q g 4) to •B o T3 C ed to to T3 s f to I to *•» •!-! op co t%/t S fftf/*tt 7 /Sf i 'jtfft/'/Sf// fft ttt •$MfW%%M$W$Mf mmmiWyWfWm ', '''' ,<•'' *' f' "tt'' Vffitttttttttftff*V?tftty> /f, ,,S,f. ',' t fft f&',V t / t t t (A C O CO to u, O P H • a C o a a o co T3 cd O to •c o to co _o "to to u to to CO +•> co to t H o P H CO to -3 H o a is g § 00 ed C cd C O C O to o to < I cd 3 cd a* o 174 CO CO co C e+i • * H O 4) 2 S p <-> V+H co 2 "H -I 2 ° ° ° 3 --3 # "4s * * **/,, ******** * * ** * ***** * *y* *. * ** ***** *'** ***** *s * s 4/ ****** ****** ** **y %* ''US'' WtfMi, ******* * **S* * ** * ********* * >* ** f**/*"**^ ******/* * ** **** ******** * ;;'/,;,"//; , < * **** * ****** ,**.***** * * * :*.***.**.*.**•******************************** f * * ** *,**jr***y**/s * */ *, **** •***************,*************f**** •**f*******S****^*****s****ss* * r** * * * **** * ** **** ************* f&k "S' ' '"SV * '* '4 ' 4 * ' ' , s;. **$*** "** %**$*** ***/ / ' , , / Sf, „s„ , ,4, 175 cu CO CD VH o • t o oo ON VH I<2 cd cu o l o cu op •c 1-3 c o 'ai '§ lb e • a a' 8 ir> oo ON CU c/5 i 4& to-rn W*W** •*sj£*S*j£jp *A *** 4?**, ** * * ******* / *••*.*>* **/ ** ********* ,*/•%• * **.jf A,, ** ** * * ******** * * * * * ******** * * / * * * * * * • > . **•*/*•***, *'>*** A * * ** ***"*** *** * ***? * * ** * * A****** * >*\ "**• '*A A * ***** * '** ************ **•*"*?* y*y******** **** ^* '*y***Sym*y*y'***&MrffifrfYy^ 3> CD CO CD t5 .2 co ed c i 1 5 « PQ 3 C/J-_ : O Cd CN CO I cd CD cd J* a <-> C+1 O O PH tJ O ' CD XS •<-> 3 o 8 O 8 o o o CD > o oo" oo -8 3 O PQ t co * * wV IP* %!{{{{{'/ * V  f f t* ttt t ttt * t  rV ****** /**s***v**3&** */t 'w/*** ^tPtttt^ttt^^t' '*Mi$$*wvfWi& *^4fttt*4f4*^t-t, ttt, ttt *f / * •'ttt S 's t* 178 cd -3 £ o 3 ~ £5 .a °o a co .a cd cd C cd t\ O CD 2 a ^ cd g co s a. CN J e s s o, o a g v co co a^ o, ed fd O o. a co a eg !T5 ^=2 1 CD CO cd CD O V H 60 <= c § * 2 1 ( 5 ca ^ jo a .-s CD cd 5« -e OH C r. Cd -r Q a S •McecVfUiWiWi'WMr^  .V*,«V/«yWWA,A,A 179 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items