You are here: Home » Project »
Experimental Design and Methods:
1. Tweedsmuir-Entiako Caribou Range Base Line
Spatial information will be organized in a GIS and used to meet the landscape modeling, data distribution and cartographic objectives of the project. Data from Cichowski and MacLean (2005) will be integrated, including caribou telemetry locations, habitat interpretations and lichen mapping. Additional data modeling of topography will be done to capture soil moisture (Fall and Morgan 2000) and to generate terrain curvature and ruggedness indices (Apps et al. 2001).
Satellite imagery will be georectified to the existing spatial data and land cover will be classified. A confusion matrix will be constructed by comparing the classified image with existing forest inventory data and linked project plot data (FIA-FSP M075047) to validate the classified image. Tweedsmuir Park spatial ecological data will be generated based on old (1950s) generalized forest cover data, topographic modeling and the classified imagery.
The historic reconstruction of the GTE will use existing 1973, 1993, 2001 and 2006 Landsat imagery. Additional images will be purchased to create a 5-7 year time sequence of landscape change. Classified imagery will be verified with current and historic ecosystem plot data.
2. Synthesize Existing Information
Team will review existing literature on caribou ecology and response of caribou to natural and human disturbance, with a focus on the Tweedsmuir-Entiako caribou. Based on this information, conceptual models of seasonal moose, deer, wolf, and caribou habitat, and models of threats to caribou populations will be constructed, including models of predator-prey dynamics. Conceptual models will be drafted into influence diagrams and describe the likely state of seasonal caribou habitat given observed environmental condition, the potential status of other ungulates and wolves. The diagrams will define the key environmental correlates, latent variables and response variables.
Empirical data and expert opinion will be gathered through formal workshops with caribou and forest ecologists to refine the influence diagrams. The team will formalize the diagrams into Bayesian Belief Networks (BBNs) to capture the environmental correlates, disturbance factors and response conditions following the methodology of McNay et al. (2006).
Results of caribou habitat and population ecology studies in adjacent areas, such as the Itcha-Ilgachuz and Rainbow mountains (Apps et al. 2001) and Chase, Scott, Takla and Wolverine herds (McNay et al. 2006) will be used to calibrate and validate components of the models developed for the GTE.
First Nations will be engaged through our traditional knowledge protocol (see First Nations section).
3. Scenario Analysis of Future Landscape Condition
There are two existing landscape projection models, one in the Nadina (Steventon 2006 M07-5006) and one in the Vanderhoof Forest District (DeLong 2006). These two models and the data for Tweedsmuir Park will be integrated into a Tweedsmuir-Entiako Landscape Model (TELM). The area will be represented at a resolution of 1 ha per grid cell, and will incorporate recent timber supply reviews (Morice, Lakes Timber Supply Areas and Vanderhoof Forest District). The landscape models will be implemented using the SELES (Spatially Explicit Landscape Event Simulator) modeling system (Fall and Fall 2001). This software is a flexible tool for building and processing grid-based, spatio-temporal models that has been used for a wide range of related projects, including the provincial MPB projection model (BCMPB, Eng et al. 2005), and other caribou projects in west-central BC (McNay et al. 2006). Using it will facilitate interoperability of our models with those from linked projects. The landscape model will include processes of forest management, stand aging and succession, MPB dynamics (downscaled from BCMPB), road development, and fire. Output from the landscape projection model will be input into the BBN models to conduct caribou analysis of the landscape projections.
A set of scenarios will be designed to evaluate a range of future conditions, such as 1) current management, and 2) management applies no salvage, moderate salvage or aggressive salvage. Models will be verified by aligning current timber management scenarios with timber supply review results, sensitivity analysis of key model parameters, and by having model output reviewed by timber management and forest ecology experts.
4. Enhancing the Resilience of Caribou Population to Landscape Change
Scenario analysis results will be used to gain insights into the factors that enhance or limit the capacity of the caribou herd to adapt to landscape level disturbance, such as, structure of post-disturbance stands, availability of alternative habitat, movement barriers, predator-prey dynamics, and the impact of forestry activities. Strategies will be identified that provide the most robust approach to maintaining caribou habitat when faced with landscape dynamics resulting from a range MPB and fire scenarios and the uncertainty of how caribou will respond to landscape change.