Eastern Slopes Grizzly Bear Project

• A collection of experts (ecological and social) and interest groups spent four days reviewing data on the grizzly bear in the Central Rockies Ecosystem, and the most recent research and future modeling efforts for grizzly bears throughout North America.
   
• The local human population is expected to increase in the Central Rockies Ecosystem region at an annual rate of 4%, potentially resulting in severe stress to grizzly bears with major impacts on grizzly bear habitat and a reduction of the population from current numbers. Extensive and successful management will be necessary to maintain local populations and to reverse current negative trends.
   
• A recovery plan is necessary to restore the function of enough compromised habitat to counter the effects of continuing development and expanding human activities elsewhere.
   
• Joint action between Alberta, British Columbia and Parks Canada is required.

EXECUTIVE SUMMARY

Introduction
Stephen Herrero, ESGBP

Rapid and continuing human population growth and infrastructure development has occurred in the last 30 years in Calgary, Canmore, Banff, and surround. This has resulted in grizzly bears being challenged to live in one of the most developed landscapes in North America where they still survive.

The Eastern Slopes Grizzly Bear Project (ESGBP) began in 1994 in response to an urgent need for scientific understanding of the grizzly bear population and habitat, and relationships with people. Several major new pieces of environmental legislation, such as the Canadian Environmental Assessment Act (CEAA) of 1992, and the Alberta Environmental Protection and Enhancement Act (EPEA) of 1993, made cumulative effects assessment of major projects mandatory. Grizzly bears, because of their potential rapid decline with certain human influences, became a focal species for cumulative effects assessments where this species occurred (Herrero et al. 1998).

ESGBP research is based out of the University of Calgary and is conducted in cooperation with many agencies and stakeholders. Research is carried out by graduate students, professors and associates and is vetted through the thesis review and defense process, and peer-reviewed and other publications (Herrero et al. 1998). Strategic directions for research and funding are overseen by the ESGBP Steering Committee. The Project has no formal links to policy or management decisions although it has had significant influence in this regard (Herrero et al. In press).

Figure I-1. Map showing the Central Rockies Ecosystem study area of the Eastern Slopes Grizzly Bear Project. Map courtesy of Scott Jevons, Geoworks GIS.

To meet their needs, grizzly bears, especially males, move long distances, often using lands managed by several jurisdictions. Responding to the extensive use of land by the mobile grizzly bear, the ESGBP focuses research on a large landscape. This approximately 42,000 sq. km. area has been called the Central Rockies Ecosystem (CRE) (Komex International 1995) and includes lands in Alberta and British Columbia (Figure I-1). As an ecological unit it has significant but not complete closure. Indeed, one of the ESGBP objectives is to contribute to maintaining connectivity for grizzly bears between the CRE and other regions.

Coordinated land and wildlife management throughout the CRE is a daunting task because it requires interagency communication, coordination and cooperation, even though agencies may have very different management objectives regarding grizzly bears. For example grizzly bears are hunted in portions of the CRE in Alberta and British Columbia, while they are not on Parks Canada lands, or in certain provincial areas. Blending different management processes around a potentially common goal of grizzly bear population persistence has been recognized as important (Dueck 1990, Herrero 1994, Herrero et al. 1998, In Press). But, only recently, with the formation of the Rocky Mountain Grizzly Bear Planning Committee, has there been a structure to try to coordinate activities of different jurisdictions.

Such a simple sounding task as preparing a grizzly bear habitat quality map that spans jurisdictions has been mired in the different and incompatible map products used by each of the jurisdictions. Attempts are now being made to develop a Landsat based map that will cover the entire CRE (and other regions as well), but this task has not been completed, let alone validated with data from actual grizzly bear habitat use. The PHVA workshop was constrained by lack of a scientifically vetted grizzly bear habitat map for the CRE.

Three regional scale projects of the ESGBP have contributed to the goal of understanding the cumulative effects of human activities throughout the CRE. Grizzly bear mortality from 1971 through 1996 has been summarized and analyzed (Benn 1998). We also have five years of data collected from 52 radio-marked grizzly bears regarding relationships with development and demographic parameters (Gibeau and Herrero 1999b). Finally, we have completed a "security area" assessment for the CRE (Gibeau and Herrero, in press). This GIS-based analysis identified landscape units 9 sq. km or larger that were without major human activities, and that contained potential grizzly bear habitat. These are the size of units thought to be important to meet the daily foraging needs for adult female grizzly bears (Mattson 1993). This analysis contributed an important understanding of our CRE landscape---the usable habitat is naturally fragmented by the snow and rock of the Rocky Mountains. Human developments and activities further fragment grizzly bear habitat into smaller units. This raises concerns for the long-term persistence of reproductively successful adult female grizzly bears.

The PHVA workshop is the ESGBPs first attempt to integrate available population, habitat and human impacts data regarding grizzly bears, and to develop population and habitat modeling to forecast future trends based on what we now know and assume. We attempted to do this by combining outside expertise with our own. We invited the CBSG of IUCN/SSC, chaired by Ulysses Seal, to structure the workshop and to share their accumulated experience gained in managing similar workshops for other potentially threatened species. To try to insure that we had some of the best population modeling expertise in North America regarding grizzly bears we also invited four internationally recognized experts: Mark Boyce, Rick Mace, Dave Mattson, and John Weaver. Perhaps most importantly we asked about 50 regional stakeholders, whose activities in the CRE influence the fate of grizzly bears, to donate four days to the workshop process. We express our sincere thanks to all participants.

The specific scientific framework for the workshop was a Population and Habitat Viability Assessment (PHVA). The variables affecting the grizzly bear habitat or population are well known human activities such as transportation, residential development, resource extraction, intensive recreation, and hunting. Through the workshop process, and based on ESGBP data, we tried to understand and quantitatively predict the cumulative effects of development and human activities on the grizzly bear population and habitat.

To the extent possible model variables were derived from ESGBP project data. However, and inevitably, assumptions had to be made for some variables. We caution interested readers to carefully examine the strengths and limitations of our data. We strongly believe in the modeling process because it forces quantitative values to be entered for all relevant variables. Once done, others can dispute or change the variables based on more complete data or understanding. At least the process is explicit and subsequent iterations can be run as data change or improve.

The PHVA workshop process
Ulysses Seal, CBSG

This report is from a Population and Habitat Viability Assessment workshop (PHVA) conducted 28-31 January 1999 at Camp Chief Hector, near Seebe, Alberta. The workshop was organized by the Eastern Slopes Grizzly Bear Project (ESGBP) Steering Committee in collaboration with Conservation Breeding Specialist Group (CBSG/SSC/IUCN). CBSG has 800 volunteer member experts around world and 15 years of experience with workshop processes. Workshops are always conducted at the invitation of local wildlife agencies. The process and report are advisory not prescriptive and derive their strength because they are locally engendered with local ownership. CBSG has done more than 150 workshops in 50 countries with more than 4,500 experts and other local stakeholders participating.

The CBSG team for this workshop included 8 people who provided a wide range of expertise as a resource and assisted in facilitating the individual working groups. Steve Herrero and the ESGB steering committee assembled 87 people, a remarkable accomplishment attesting to the work done on this project and its implications for management of the entire ecosystem. This provided a high concentration of expertise on the ecosystem, bear habitat, bear biology and human impact activities in this system. Included were 4 top grizzly bear experts from the USA with extensive knowledge of the Yellowstone system and other bear populations, managers and field biologists from the range of agencies with responsibilities in this ecosystem, representation from commercial and industrial interests, and university based researchers.

The workshop process extended over 3 ½ days building on materials provided in fieldwork and studies done to date. The 60-70 participants, after initial plenary presentations of background material and guidelines for the process, were divided into working groups that then remained together over the duration of the workshop. This provided for a rapid building of mutual common ground to allow a focus on the tasks of the group. Working groups included a big picture group based on an ecosystem and landscape mapping approach, a habitat and bear distribution group, a population modeling group, a secure areas group, a human access impact group, and a physical impact group for six working groups. All individuals conducted a structured analysis of the goals and problems in small working groups to ensure intensive participation. The intense commitment of time and energy by all in this beautiful site was creative and productive.

A draft report was prepared by each working group with plenary reporting each day to ensure effective flow of information between groups. This was augmented by exchange of individuals between groups as needed to transmit or obtain information or guidance on particular questions. Strategies and specific actions with measurable outcomes were then developed in each group to respond to the carefully identified and defined problems. These actions constituted the recommendations of the workshop. The final report was prepared from a compilation of the working group reports and thus is a direct product of the participants in the workshop.

The first day's agenda began with an opening presentation by CBSG (Seal) on the workshop process, the use of thinking tools, the use of small working groups to do the analyses and prepare the report, and basic facilitation guidelines for conduct of the group sessions. This was followed by technical presentations summarizing grizzly bear status in Eastern Slopes.

The first plenary session was opened with each person introducing himself or herself and stating their primary goal for the PHVA workshop. This process and the expressed goals provided a unifying experience of shared values and interests in the outcome for the workshop and their shared vision to:

'Maintain the presence of grizzly bears in the regional landscape'.

A common characteristic of management decision-making processes throughout the world is to argue for delay of management decisions to wait for additional or definitive (often unattainable) information. This poses a real threat to effective management before a population reaches a severe decline or terminal crisis stage. Also it undervalues the accumulated wisdom of the managers and local field biologists and their knowledge for making management decisions. There is a substantial body of knowledge to inform management decisions for the grizzly bear in the CRE here and now. The results of application of these management decisions will need to be part of an adaptive management program and there is a strong need for continuing the long term studies through a longer fraction of the life cycle of the species since there clearly are ecosystem specific characteristics of its demography.

Working group summaries

Landscape, Mortality, and Risk Modeling

This working group focused on issues related to modeling the viability of grizzly bear populations and implementing the VORTEX population viability model using grizzly bear data from the ESGBP study area. It was recognized that a strong link to habitat conditions was necessary if viability projections were to have any meaning. It was also recognized that conditions were not uniform throughout the ESGBP study area and that the prospects of local grizzly bear populations varied accordingly. The boundaries of the ESGBP study area were furthermore recognized to be open to movement of bears into and out of the study population. These conclusions lead the group to adopt VORTEX models that were open to emigration and immigration at the margin of the modeled population and structured to allow for internal differences in birth and death rates. Kananaskis Country and Banff National Park were chosen to be the focus of finer-scale simulations. These conclusions also lead to the development of two related approaches to modeling the effects of habitat on death rates of grizzly bears. The output of these proposed habitat-based models would be used by VORTEX to project population viability in a way that was sensitive to temporal and spatial differences in habitat conditions.

The modeling working group provided estimates of demographic rates that were used in VORTEX model simulations presented at the workshop. Credible estimates of grizzly bear birth and death rates were available for the ESGBP study area. Comparable estimates of immigration and emigration rates were lacking and so a plausible range of exchange rates was used in model simulations. The working group recognized the importance of uncertainty in vital rates (e.g., birth and dearth rates) and the related importance of sensitivity of viability projections to variation in each of the rates. This lead to a description of uncertainty as well as an analysis of sensitivities for each rate.

The habitat-based models of grizzly bear death rate were based on the premises (1) that most adult grizzly bears die because humans kill them and (2) human-caused deaths will occur at a rate governed by the frequency of encounter between humans and bears and the likelihood that a human will kill a bear during a given encounter (i.e., lethality of contact). Frequency of contact between humans and bears is affected by, among other things, the numbers of humans in grizzly bear habitat, the amount and dispersion of road and trail access, and the quality of grizzly bear habitat near human facilities. Areas secure for grizzly bears (i.e., security areas) are those areas sufficiently productive to attract bears and sufficiently remote from humans to ensure survival of the bear while it is there. Lethality of contact will be affected by administrative jurisdiction and whether a legal hunt occurs. These concepts were the basis for the habitat-based models which correspondingly used maps of jurisdictional boundaries, human facilities, roads and trails, human populations, and grizzly bear habitat productivity to describe and predict variation in grizzly bear death rates.

Habitat and Distribution

The Habitat and Distribution Working Group served as a resource providing data regarding habitat quality, effectiveness, supply, and distribution from existing research results and habitat evaluation for input into the habitat-based population viability model. This group: 1) defined terms and scale of assessment; 2) analyzed existing information on the spatial distributions of animals over the landscape and broke it down into measures of population density, home range size and minimum daily movements relative to habitat, at the appropriate working scale; and 3) developed a rational basis for linking existing mortality data to habitat-based landscape evaluations.

The overall recommendation from this working group is to establish and fund a technical group to validate, and incorporate into a revised PHVA model the following parameters:

• habitat quality polygons (ground-truth LANDSAT greenness polygons)
  
• linkage zones
  
• population density, home range size and other values by habitat class
  

These tools should then be used to determine where conservation efforts should be focused when implementing the following specific strategies (listed in priority order based on the results of paired ranking analysis):

1. Implement seasonal recreation and/or road closures in areas that have high habitat quality.
   
2. Maintain an open population/landscape, including high quality dispersal linkages, to minimize extinction risk.
   
3. Locate/relocate roads and trails in lower quality habitat.
   
4. Optimize/restore amount of secure habitat. Implement management actions to increase habitat effectiveness, especially in higher quality habitats; at a minimum, maintain habitat effectiveness region-wide.
   
5. Based on current knowledge implement management actions to achieve a minimum of 60% habitat security in areas where we want to maintain females within their home ranges.
   
6. Prioritize conservation efforts (eg: restoration of security) on high quality habitats
   
7. Secure existing areas that currently have high habitat effectiveness by eliminating future development of roads or modification of habitat (=protected areas)
   
8. Document a minimum of five crossings of the fenced portion of the Trans-Canada Highway (TCH) within the next two years. Failing this, implement active measures to increase secure crossing areas and enhance habitat quality.
   
9. Create administrative flexibility to meet habitat goals by renegotiating existing leases and tenures.
   
10. Each jurisdiction should identify explicit measurable goals for habitat effectiveness and habitat quality in resource and land use plans.
    

Population Modeling

The need for and consequences of alternative management strategies can be modeled to suggest which practices may be the most effective in conserving the grizzly bear in Alberta and eastern British Columbia. VORTEX, a simulation software package written for population viability analysis, was used as a tool to study the interaction of a number of life history and population parameters treated stochastically, to explore which demographic parameters may be the most sensitive to alternative management practices, and to test the effects of a suite of possible management scenarios.

Working group participants used the best available information on life history and demography of the Eastern Slopes grizzly bear to develop a series of stochastic simulation models of grizzly population viability. Initially, demographic sensitivity analysis was employed in order to assess which population demographic parameters (such as birth rate, age-sex specific mortality, etc.) influence population growth rate most strongly. The group found that their risk assessment projections depend most heavily on two demographic parameters: the percentage of adult females breeding and the level of adult female mortality. Interestingly, female breeding success was one of the most important parameters we need to improve our understanding of, even though it is also one of the better-estimated parameters. This is most likely because small differences in breeding rate (3 to 5 years) can almost double the net lifetime reproduction for bears. If mortality is (or becomes) as high as in some other populations, bears could be in trouble. (Low fecundity in the area places extra emphasis on keeping mortality low.)

Workshop participants noted that the primary goal of grizzly bear management in Alberta and British Columbia is to prohibit population decline or to promote a modest increase in the provincial bear population. Consequently, the group defined "extinction" as the probability of population decline below current levels (this is technically referred to as a "quasi-extinction" probability). Under this definition, modeling efforts indicate that the population is not secure: the provincial goal of maintaining or increasing the population above today's numbers is not likely to be met under current conditions.

The modeling group was also able to use estimates of growth in the human population in and around the Central Rockies Ecosystem (CRE) to derive models assessing its impact on the local grizzly bear populations. These models show unequivocally that the impact of the expanding human population could be severe. The CRE grizzly population probably cannot sustain increases in adult female mortality (or decreases in fecundity), so it will be imperative that the impacts of humans be reduced even while the numbers of humans in the region increase.

Human Impacts - Physical Components

Physical developments on the landscape have the potential to significantly affect grizzly bear habitat. Historically, human activities such as fire management, oil and gas development, and logging have been the economic drivers of land uses that can strongly influence grizzly bear habitat. Using information on current and possible future land use activities, estimates were made of direct effects of human activities on grizzly bear habitat over the next 100 years assuming best management practices.

In general, only low to moderate changes were forecasted in most land uses for most areas. This prediction occurs because existing land use is already relatively intense throughout much of the CRE. However, there is potential for high increases in some areas when existing land use is compared to that approved in land-use plans.

This analysis of current and future land use conditions for CRE suggests that core protected areas alone will not sustain viable grizzly bear populations in the CRE. State-of-the-art management of development activities on intense recreation and multiple-use lands will be required. Five activities will most directly affect grizzly bear habitat: timber harvesting, fire management, oil and gas development, recreational developments, and residential developments. Recommendations for specific land-use strategies and management actions for all areas designated as occupied grizzly bear habitat were developed for each of these primary human impacts and these can be found in the working group report. In addition, the following overall land-use strategy was proposed:

• All land use will take place within the context of cumulative effects models that utilize grizzly bears as key indicators.
  

Three actions steps were identified to implement this strategy: 1) Prepare a grizzly bear habitat map for all lands within the CRE; 2) Prepare a map of existing land uses and 3) Develop a CEM to evaluate the effects of current and proposed land uses.

Human Access Impacts on Habitat

Data suggest that 80-90% of adult grizzly bear mortalities in the CRE result from human/grizzly encounters and that grizzly/human encounters result from human access to grizzly territory. In terms of access, the rule of thumb is that mortality correlates to:

1. The number of roads and trails in a given region;
   
2. The number of people using those roads and trails;
   
3. The activities pursued by the people using those roads and trails and, specifically;
   
4. The use of firearms by the people pursuing those activities.
   

The working group identified types of access and their attributes, identified and rated the impacts associated with each type of access in terms of: a. Mortality; b. Displacement (of bears from habitat, especially security areas); c. Reduced reproduction (through stress to the population); and d. Habitat reduction and degradation. The group then worked to create scenarios that capture the current state of affairs, a probable future, and a possible future, allowing for spatial and temporal variation (i.e., within different jurisdictions in the CRE over time) of impacts.

The group of experts concluded that the increase in access has a direct correlation with human/grizzly encounters in leading to both displacement and mortality. The projected increase is 4% annually. Through an aggressive management program that involves education, management of human food, aversive conditions, law enforcement, and cooperative strategies, this increase could be mitigated by only 50% at best. This leaves a residual impact of 2% annually. Even with our best efforts the model clearly demonstrates that a 2% decline would result in a population collapse within a few decades, especially for the east slopes subpopulation.

Restoration scenarios must be developed within a decade and should be implemented for at least a decade to reverse this trend. The level of restoration must approach 2% annually. Principally, restoration will involve closing and restoring access to particular areas, and relocating recreation activity to areas with low grizzly bear density. The following recommendations were made to realize a more intensive "restoration scenario" focusing on adult female home ranges:

1. Select restoration pilot project (s) for high profile communications vehicle (one which create success story for grizzly population recovery). Possible areas could include the Smith-Dorrien Road, the Elk Lake Grizzly Bear Conservation Area, and the Fairholme Benchlands.
   
2. Reexamine existing overflight and jetboat regulations and consider extending legislated restrictions.
   
3. Establish tighter quotas and controls on high-use areas (consider Lake O'Hara management schemes for such areas).
   
4. Open new camping/recreational areas in low-density grizzly areas in exchange for existing closing/restricting camping/recreational areas in high-density grizzly areas.
   
5. Select an area for "holistic" restoration - a multi-stakeholder approach with a social and economic component as well as ecological.
   
6. Hold a follow-up workshop based on the findings of this workshop, but dedicated to developing vision, strategies and actions for the long-term.
   

Human Impact - Secure Areas

The Secure Areas Group utilized a five-step process to structure their discussion. The first step described the current situation to establish a common understanding of the issues and to record the information as base line data. This scenario was projected to a probable future if trends continued and then a possible future created from which a vision statement was generated: To establish and maintain a viable population of grizzlies by accommodating individual security needs in high quality habitat with emphasis on the survivorship of adult females. A gap analysis between the probable and possible futures established the categories for future planning with feedback loops to the base line data.

A security area was defined to be 9 km2 of habitat used by a female grizzly bear every day. (ESGBP research has demonstrated this to be the average size of an area used by an adult female.) This 9-km2 area moves with the bear within her home range (Gibeau and Herrero in press). A home range that contains connected security areas can enhance female survivorship. Disturbances higher than 20 human parties per week were considered to cause significant and adverse behavioral changes in grizzly bears.

The group recognized that there needs to be a joint management response from the governments of Alberta, British Columbia and Canada to manage the grizzly bear as a unit. Without input from BC within the group, however, it was recommended that similar work be done in that portion of the Central Rockies Ecosystem. Socio-economic pressures require a focused management response to protect security areas for female grizzly bears and sufficient information currently exists to act. Due to scientific uncertainty, however, the precautionary principle needs to be applied and the burden of proof should shift to the developers. Some general and specific recommendations were made to maintain and increase the number of security areas in legally protected areas as well as in landscapes not protected. Using the security area concept combined with existing knowledge, specific and detailed recommendations were made within the public context in relation to science, management strategies, and legislation.

Literature cited

Benn, B. 1998. Grizzly bear mortality in the Central Rockies Ecosystem, Canada. Master Project, Faculty of EVDS, University of Calgary, AB. 147pp. plus appendices.

Dueck, H. 1990. Carnivore conservation and interagency cooperation: A proposal for the Canadian Rockies. Master's Project, Faculty of EVDS, University of Calgary, AB. 151pp.

Gibeau, M. and S. Herrero. In press. Managing for grizzly bear security areas in Banff National Park and the Central Rockies Ecosystem. Ursus 12:000-000.

Gibeau, M. and S. Herrero 1999b. Eastern Slopes Grizzly Bear Project: Progress report for 1998. Prepared for the Eastern Slopes Grizzly Bear Project Steering Committee, Calgary, AB.

Herrero, S. 1994. The Canadian National Parks and grizzly bear ecosystems: The need for interagency management. Int. Conf. Bear Res. and Manage. 9:7-21.

Herrero, S., D. Poll, M. Gibeau, J. Kansas, and B. Worbets. 1998. The Eastern Slopes Grizzly Bear Project: Origins, organization and direction. Proceedings Canadian Council on Ecological Areas Annual Meeting, Calgary, Alberta, November 1995.

Herrero, S., J. Roulet, and M. Gibeau. In press. Banff National Park: Science and policy in grizzly bear management. Ursus.

Komex International. 1995. Atlas of the Central Rockies Ecosystem. Report to the Central Rockies Ecosystem Interagency Liaison Group, Calgary, Alberta. 49 pp.

Mattson, D.J. 1993. Background and proposed standards for managing grizzly bear habitat security in the Yellowstone Ecosystem. Cooperative Park Studies Unit, University of Idaho, Moscow Idaho. 17pp.

Home | About the Project | Grizzly Bear Basics | Bear Safety | Research Publications
Media & Products | How to Help | Supporters | Contact | Links