Foothills Research Institute Grizzly Bear Program Deliverables

							  Foothills Research Institute Grizzly Bear Program Deliverables
                                      Available Spring 2009


The Foothills Research Institute Grizzly Bear Program* (GBP) is committed to providing
its program sponsors with research products to fulfill its primary objective:

       “To provide resource managers with the knowledge and planning
       tools necessary to ensure the long-term conservation of grizzly bears in Alberta.”

The GBP has completed habitat mapping for the entire range of grizzly bears in Alberta
(below).

                                               In January 2005 the Grizzly Bear Program
                                               delivered the first set of program deliverables
                                               based on research conducted from 1999 to 2004.
                                               These spatial models covered an area of over
                                               100,000 km2 in the central foothills.

                                               In January 2006 we provided our second set of
                                               program deliverables. These were updated and
                                               improved versions of the spatial models based
                                               on new work completed in 2005, and expanded
                                               south to the U.S. border, an area of 127,000 km2.
                                               Along with the models were new GIS
                                               applications to evaluate the impact of planned
                                               developments on grizzly bear habitat.

                                               In March 2007 we provided our third set of
                                               deliverables, expanding north to Peace River,
                                               covering 227,000 km2. Additions and updates
                                               were provided in July 2007.

In May 2008 we provided our fourth set of deliverables, expanding north and west to
cover the Clear Hills and the Rocky Mountains west of the Continental Divide - an area
of 301,000 km2, of which 252,000 km2 lies within Alberta.

In May 2009 we are providing our fifth set of deliverables, with improvements and
updates to our current maps and models .

These products are available to all GBP program partners once data sharing agreements
are in place. The GBP is currently supported by over 50 partners including forest
companies, oil and gas companies, universities, NGOs, and government agencies.

           * Formerly the Foothills Model Forest Grizzly Bear Research Program
                                                                                           May 31, 2009
                              Product Description
What’s New This Year:


      New RSF models for Northern Alberta
      (Chinchaga)
  Seasonal RSF models have been created for an
  area of 38,000 km2 northwest of Peace River to
  the BC border.


      Food-based habitat model
  This innovative method of classifying habitat
  quality uses key grizzly bear forage species to
  identify potential and realized habitat value.


      Pine presence layer
  This binary raster layer identifies pixels containing at least one pine tree more than 3
  metres high. This model has been developed for an area of 31,000 km2 in the Kakwa,
  north of Grande Cache.

      Provincial core and secondary Grizzly Bear Conservation zones – ASRD GIS
      layers
  As recommended by the Provincial Grizzly Bear Recovery Plan, ASRD has identified
  core and secondary conservation zones within which open motorized route density
  will be managed and maintained below recommended thresholds.

     ASRD watershed boundaries for provincial conservation areas

      New GIS tools to update landcover layers, and to automate calculation of safe
      harbour
  The landcover products (Landcover, crown closure, Percent Conifer) can be kept up-
  to-date using simple-to-use GIS tools that “burn” new disturbance features into
  existing layers.

      Annual landscape layers expanded to Kakwa area
  Anthropogenic disturbance features (roads, rights-of-way, cut blocks, and well sites)
  have been classified by year of construction over an area of 73,000 km2 in the
  foothills. This area of classification has now been expanded to the area from Grande
  Cache north to the agricultural zone south of Grande Prairie.

     GIS disturbance datasets have been updated throughout grizzly bear habitat to
     match the latest available high-resolution imagery, mostly to 2007.
Remote Sensing-based datasets:
   1. Landcover: this is a raster-based GIS
      layer, in ESRITM grid format,
      identifying 15 land cover classes
      derived from remote sensing imagery
      (Landsat TM7) at 30m resolution (pixel
      size). A sample is shown at right. This
      layer can be updated with new features
      such as roads, cutblocks, etc using the
      GIS tools provided.




   2. Percent conifer: this GIS layer
      represents the percent of conifer in a
      forested stand, also derived from
      remote sensing imagery (Landsat TM7)
      at 30m resolution (pixel size).




   3. Crown closure: this GIS layer
      represents the percent of canopy
      closure in a forested stand, derived
      from remote sensing imagery (Landsat
      TM7) at 30m resolution (pixel size).
4. Agricultural layer: This layer is
   comes in both raster and shapefile
   format and identifies five agricultural
   classes within the study area.




5. NDVI (Normalized Difference Vegetation
   Index) a time-series of 12 biweekly images (in
   georeferenced .tif format) tracking changes in
   plant phenology. It is derived from 250m
   MODIS imagery from summer 2005.




6. Forest Origin Year: this layer is derived
   from AVI, cut history, and fire history
   data, and comes in 30m grid format. A
   sample is shown at right.
7. Regeneration mask: this binary raster
   identifies areas of forest regeneration. An
   example is shown at right.




8. Orthoimage: a mosaic of cloud-free
   Landsat images, at 30m resolution, in
   georeferenced .tif format. (Note: this 2
   GB layer is not included on the
   Deliverables DVD but is available upon
   request).




9. Pine Presence Layer: this binary raster
   identifies the presence of pine in a 30m
   pixel (Kakwa area only)
Grizzly Bear Habitat Models:

   10. Resource Selection Function Maps
       (RSF): these are raster-based GIS layers, at
       30m resolution, showing the relative
       probability of grizzly bear occurrence on
       the landscape. They are derived from
       grizzly bear location data collected by GPS
       radio collars, combined with landcover and
       other GIS layers, and have been tested and
       validated with at least 2 years of GPS data.
       Models have been developed for three
       seasons (spring, summer and fall), and for
       all seven population units (Livingston-
       Waterton, Clearwater, Yellowhead, Grande
       Cache, Swan Hills, and Chinchaga).


   11. Mortality Risk Map: using spatial and
       temporal data on grizzly bear mortalities,
       we have produced a raster-based grizzly
       bear mortality risk map, which predicts the
       probability of human-caused grizzly bear
       mortality over the landscape. This map is
       based on the most current data for open,
       motorized linear access structures
       including roads and rights-of-way.




   12. Grizzly Bear Movement Corridors:
       Graph Theory has been applied to the
       Phase 6 RSF maps to indicate the location
       of grizzly bear movement. These layers
       also provide a ranking of the relative
       importance of movement corridors and are
       tested and validated using GPS grizzly bear
       location data.
13. Food-based Habitat Model:
    This model predicts the abundance of 12
    critical grizzly bear foods as a function of
    environmental variables including
    landcover, climate, terrain, and forest
    stand structure.




GIS applications:
   Python geoprocessing scripts, and associated GIS input layers, allow the user to
   predict changes to grizzly bear habitat caused by land use activities. Planned
   development features (roads, trails, cut
   blocks, well sites, and pipelines) can be
   incorporated into the landscape variables,
   and the RSF and mortality risk models
   regenerated. These scripts require ESRI
   ArcGIS 9.2 (ArcInfo functionality) with
   Spatial Analyst extension. Spatial
   datasets include:


14. Compound Topographic Index (CTI), a
    terrain-based index of soil wetness.




15. Terrain Ruggedness Index (TRI), a
    terrain complexity layer derived from a
    DEM.
Anthropogenic Disturbance:

   16. Linear Access layer (roads, pipelines,
       powerlines, railways, seismics): personal
       geodatabase feature class.


   17. Oil and Gas Well Sites: personal
       geodatabase feature class.

   18. Cutblocks: personal geodatabase feature
       class.




Within a 73,000 sq km area of the eastern slopes,
including the Clearwater, FMF Core, and Grande
Cache population units, these vector features and
associated landcover layers have been classified
by year of construction (2004-2008 north of
Grande Cache, 1998-2005 Grande Cache south
to Highway 11, 2002-2005 south of Highway 11
to Highway 1) to match landscape profiles with
grizzly bear GPS data and track habitat changes
over time.




For more information on how to become a partner of the GBP, please contact:
       Gordon Stenhouse, Program Coordinator
       Grizzly Bear Program
       c/o Foothills Research Institute
       Box 6330
       1176 Switzer Dr.
       Hinton, AB
       T7V 1X6
       Ph. (780) 865-8388
       Fax: (780) 865-8331
       email: Gordon.Stenhouse@gov.ab.ca
       website: http://foothillsresearchinstitute.ca/pages/Programs/Grizzly_Bear.aspx