Cumulative Effects Assessment Practitioners' Guide

Trans-Canada Highway Twinning Phase IIIA: Case Study Highlights

VECs: Elk, moose, wolf, coyote, grizzly bear and black bear

Issues: Loss/alteration of wildlife habitat, wildlife disturbance due to alienation, collision mortality of wildlife, and disruption of wildlife movements due to habitat fragmentation

Approaches: Spatial analysis using a GIS to determine loss or reduction of habitat value due to various types of disturbances to wildlife

Lessons learned: Qualitative conclusions and ranking systems are useful to communicate results if supported as much as possible by quantitative analysis that is defensible


In 1994 an EIA was submitted for the proposed twinning of a section (IIIA) of the Trans Canada Highway in Banff National Park, Alberta (Parks Canada 1994). The project consisted of upgrading 18 km of two-lane highway to four lanes and the construction of interchanges. The section to be upgraded was approximately half the distance between the townsites of Banff and Lake Louise, and closely followed the Bow River along its entire length. The Bow River valley forms a long and narrow mountain valley through the park.

The project proponent was Parks Canada, the federal department responsible for that portion of the highway. The project was assessed under the Environmental Assessment and Review Process Guidelines Order (the predecessor to the new federal Act). The highway, Canada's major east-west route, required upgrading due to increased traffic volumes (especially from trucking and tourist visitation). The Phase IIIA proposal was able to make extensive use of assessments previously done for earlier Phases I and II, also in the Park.

Assessment Approach

The assessment identified two study areas: the Middle Bow Valley within the immediate vicinity of the highway right-of-way, and the Central Rockies Ecosystem. The areas were 1,150 km 2 and 43,000 km 2, respectively. Most of the analysis was done in the Middle Bow Valley, with the results of that assessment extrapolated to the Central Rockies Ecosystem. A time limit (i.e., temporal bounds) was not defined as the affected environment was assumed to exist indefinitely in a national park.

Environmental components of most concern were wetlands, the river channel, fish-bearing streams, wildlife (particularly large carnivores and ungulates with regional movements), vegetation and fish. The residual effects on wildlife were summarized in the EIA, which preceded the CEA, for nine species and five wildlife categories. Based on these results, vegetation, fish and wildlife were identified as most likely to be affected cumulatively.

The assessment consisted of three parts: 1) screening of other projects and environmental effects (based on the results of the local effects assessment which preceded the cumulative effects assessment), 2) within the Middle Bow Valley, quantitative analysis of effects on six wildlife species and vegetation and a qualitative analysis of effects on other wildlife species and fish; and 3) within the Central Rockies Ecosystem, quantitative analysis of overall habitat loss due to human disturbances.

Cumulative Effects Screening

A project inclusion list of 28 projects and activities was first identified. Linear projects were separated from areal projects due to the different analytical approaches used for each. The projects were screened to determine "which projects are likely to make a measurable contribution to cumulative effects" in the Middle Bow Valley. A screening table rated the likelihood of effects on ten environmental components (i.e., terrain, hydrology, air quality, vegetation, fish, wildlife, recreation, history and archaeology, visual, socio-economic) of various projects, ranked on a 4 point scale (negligible, low, moderate and high). The rankings, qualitatively determined, indicated that wildlife was the most affected environmental component.

Effects on Indicators in Middle Bow Valley

Most of the quantitative analysis was performed on six wildlife indicators: elk, moose, wolf, coyote, grizzly bear and black bear. These were selected based on the availability of data (i.e., life history, movements, population dynamics) and their scientific and public profile. The assessment used available Ecological Land Classification and wildlife data to analyze effects.

Sixteen projects were identified in the screening as possibly contributing to cumulative effects. For each species, the contribution of these projects to cumulative effects were ranked in a matrix against habitat loss/alteration, disturbance due to alienation, collision mortality, and disruption of movements due to habitat fragmentation. The relative contribution of each project to overall cumulative effects in the Middle Bow Valley ecosystem was also ranked. The actual rankings were qualitatively determined, but based on the results of the GIS analysis.

The GIS calculated a series of indices that were based on the degree by which habitat suitability was reduced or lost due to various disturbances. A map of the 16 projects was overlaid on a habitat map for each species. The area of spatial overlap was multiplied by the habitat suitability rating for that ecosite and a modifier specific to the type of disturbance and species being modelled. Modifiers were estimated, based on professional judgement, in recognition of the animal's known behaviour, and in some cases on empirical data. Modifiers were used to represent alienation effects (i.e., "0" represented no disturbance, "5" represented a highly disturbed area), a barrier factor for blockage of movement effects, and a collision mortality factor for death due to collisions with vehicles. A specific distance around disturbances (i.e., "zone of disturbance") was also defined for areas of human use to define a spatial buffer in which alienation effects were expected to occur.

Finally, the total of each index for each species and type of effect (loss, alienation, fragmentation and mortality) was calculated and the incremental changes compared between three scenarios: existing, existing with proposed project, and all possible future projects. The final results were discussed and summarized as qualitative rankings. For example, the results for grizzly bear indicated that: the overall cumulative effect from all activities was "major" (i.e., long-term adverse effects on population in study area), the main contributors to the overall cumulative effect on bear were the existing Trans Canada Highway and the Lake Louise townsite, the incremental effect of the proposed project as a contribution to overall cumulative effects on grizzly bear was "moderate", and two other projects (controlled burns and existing powerlines) contributed to an equal level of incremental effect as the proposed highway twinning project.

Effects of Human Disturbance in the Central Rockies Ecosystem

Areas of human development were mapped at three levels: no and low development, moderate, and high. The density and distribution of three wildlife indicators (elk, moose and bear) were also mapped. Spatial overlap of these maps indicated how severely the cumulative effect of these developments may have already impaired habitat supporting these species. Areas of high habitat potential that were disturbed by areas of high or moderate development were considered of concern. For example, 30% of the Central Rockies Ecosystem had high or very high habitat capability for large carnivores, yet 87% of that area occurred in areas of moderate or high disturbance.


The greatest concern was blockage of movements of bear and ungulates through the park. As mitigation, wildlife underpasses at various points along the highway, and possibly one overpass (i.e., wildlife cross a bridge over the highway) were recommended. However, later evidence indicated that underpasses elsewhere along the existing sections of twinned highway were not used (especially by bears) to the extent originally predicted. This resulted in the recommendation for overpasses only. Furthermore, the effectiveness of the overpasses would be monitored, and fencing along the highway would again be used to reduce collisions and guide wildlife to these crossings. "Carnivore Conservation Areas" were also recommended to ensure that large areas remained undeveloped.

Lessons Learned

  • Despite the use of quantitative analysis that provides "precise" numbers, conclusions often had to be made qualitatively. This is generally more true the more complex the effects are, and is particularly true for conclusions about effects on animal species. The numbers provided only one (albeit important and sometimes pivotal) source of information - the balance of the decision based on professional judgement of the assessor and of experts solicited for advice. It is always important, to assist decision-makers in making an informed decision, that an explanation is always provided that describes how such tables of numbers are "translated" into rankings.
  • When performing a "quantitative" assessment, care must be taken to fully explain assumptions and uncertainties associated with the derivation of certain quantities. In this assessment, the various modifiers represented a significant source of qualitative input into a "quantitative" analysis. Given the complexities presented in a CEA, this is not in itself unacceptable if the assumptions and uncertainties are explained. At least, upon later examination, reviewers will be aware of the limitations of the method and data used, and weigh the information provided by its conclusions accordingly.
  • The presentation of results from the Middle Bow Valley analysis relied on matrices with qualitative rankings and associated qualitative discussion. The presentation of results from the Central Rockies Ecosystem analysis relied on qualitative discussion and GIS maps. This combination of information is useful as the matrices organized results and summarized discussion points. Although the derivation of the matrix rankings were not provided, the results were discussed in detail.
  • Performing a CEA in an area under only one principal jurisdiction, such as a National Park, has a significant advantage. This case study had the benefit of relatively well defined land use goals (e.g., from the Park Management Plan and other guidelines), data from extensive ecological research, and ready access to the descriptions of other projects and activities in a large regional study area. Most assessments do not benefit from these advantages.
  • The mitigation in response to cumulative effects represented two important approaches: 1) implementation of conventional or innovative on-site design modification (i.e., overpasses); and 2) long-term interjurisdictional planning and monitoring on a regional basis to create a forum in which various stakeholders communicate their concerns, support monitoring and reach consensus on land use goals.