Methods

Measuring cumulative disturbance in caribou ranges

The Cumulative Disturbance metric shows the total amount of disturbed area in each caribou range. Disturbed area was calculated using the most up-to-date publicly available disturbance datasets in each caribou range. Whenever possible, we used datasets that included the date disturbances occurred to enable annual tracking. Data included timber harvest, wildfire, roads, geophysical exploration lines (e.g., seismic lines), forests killed from mountain pine beetle, agricultures, and urban areas. Range boundary shapefiles were provided by provincial jurisdictions.

In Alberta, the Alberta Biodiversity Monitoring Institute’s Human Footprint Inventory, circa 2021, was used to quantify human land use. These data were combined with the Government of Alberta’s provincial wildfire boundaries, circa 2021. In British Columbia (BC), disturbance data varied between Boreal, Northern Mountain, and Southern Mountain Caribou ranges. Boreal and Northern Mountain ranges used publicly available data from provincial repositories depicting timber harvest, wildfires, oil and gas development, pest outbreaks (e.g., mountain pine beetle), linear features (e.g., roads and seismic lines, and other human-caused disturbances (e.g. urban areas, agriculture, and mines). Southern Mountain ranges used cumulative disturbance data provided by the Canadian Wildlife Service.  A full list of datasets is available here [link to table in Resources section once platform is live]. 

The Cumulative State of Disturbance is reported for all disturbances and also broken down by timber harvest, wildfire, pest outbreaks (where available), linear features, and all remaining disturbance types. Each of the six estimates are single values representing the cumulative area disturbed within a given range using the most up-to-date data.

For more information: Disturbances have known and unknown origin years, depending on the dataset used. Please see the Shiny app for uncertainty information. Data (area and % disturbed) are provided at the range level, though BC’s Southern Mountain Caribou herds are also summarized by core and matrix habitat in the app.

Calculating the net rate of disturbance after recovery

Not all disturbances are permanent, and some will recover over time. For a subset of disturbance types—timber harvest and wildfire—, we developed methods to assess vegetation recovery over time and when a disturbance would be considered recovered to “undisturbed” conditions.

After adding all disturbances in a range (see Calculating Cumulative Rate of Disturbance), we removed disturbances considered to be recovered. The net rate of disturbance was thus calculated as:

% undisturbed = (range area – area of unrecovered disturbances) / range area

Before disturbance, forests maintain a relatively stable level of “greenness,” representing the existing balance of coniferous and deciduous vegetation. When a disturbance such as wildfire or timber harvest occurs, there is an initial flush of young, leafy deciduous vegetation, which provides a temporary increase in forage for moose and deer. This surge in greenness, which we can measure using remotely sensed data, reaches a peak before gradually declining as coniferous vegetation re-establishes and deciduous plants are replaced. “Recovery” for harvested and burned areas is defined as the point at which greenness returns to pre-disturbance levels, signaling that the habitat is no longer providing elevated forage resources for moose and deer.

We modeled the amount of forage regrowing after timber harvest and wildfire using changes in the Enhanced Vegetation Index (EVI) derived from satellite data. The change in EVI is sensitive to changes in deciduous and herbaceous vegetation, and has been shown to be positively correlated with moose and deer densities (refs). Because the recovery process varies with climate and by site, we developed recovery thresholds by biogeoclimatic zone in British Columbia and natural subregion in Alberta. 

Thus, every year, new disturbances are added and existing disturbances subtracted once they are considered “recovered”. The net rate of disturbance can fluctuate up and down over time, reflecting both gain and loss of habitat, while the cumulative rate of disturbance is always increasing.

Other disturbance types were not considered in our recovery analysis at this point in time. Methods for assessing recovery on some disturbance types are actively being developed and will be added to future versions of the CHIP (e.g., seismic lines). Some disturbances are more permanent and unlikely to transition back to mature forest (e.g., agriculture), while some will eventually recover but had yet to do so within the time period of our analysis (e.g., borrow pits).

For more information: Disturbances have known and unknown origin years, depending on the dataset used. Please see the Shiny app for uncertainty information.