Methods: Natural Landscape Indicators

The backfill vegetation layer is a reference landscape representing a hypothetical modern landscape if no human footprint had occurred. It was created by the ABMI through the amalgamation of existing information on vegetation, habitat and soil throughout Alberta. To obtain a seamless reference vegetation layer, vegetation in human footprint was replaced by vegetation predicted to be present in the absence of human footprint (i.e., human footprint was "backfilled" to native vegetation). More information on this product can be found in ABMI's Publications^[1]. 

Forested vegetation is classified into seven categories: upland deciduous, mixedwood, Pine, or White Spruce (including true firs), and lowland treed bogs, treed fens, and treed swamps. For all forested vegetation classes, except for treed fens and treed swamps, we assigned age classes of 0–9 years, 10–19 years, 20–39 years, continuing in 20-year classes through to the oldest class of 160+ years. Non-forested vegetation is classified into seven categories: upland grass and shrubs, shrubby fens, bogs and swamps, graminoid fens, marshes, open wetlands, and bare areas.

The ABMI’s backfill vegetation layer combined with the ABMI Human Footprint Inventory is used to summarize the effects of land base change on different vegetation types, and as part of statistical modeling to determine cumulative effects of habitat alteraion due to human footprint (Biodiversity Intactness, Sector Effects) and land base change (Attribution) on species habitat. Note that this analysis does not account for vegetation recovery.

Results for nine vegetation types are summarized in this analysis: 

• Upland forest types: deciduous, mixedwood, Pine, White Spruce. 
• Lowland forest types: Black Spruce, treed fen
• Other vegetation types: swamp (treed and non-treed), open wetland (shrubby and graminoid), and upland grass and shrub.

In addition, we summarized information for old forests (both upland and lowland). For this report, we have used a generalized threshold of 100 years to define "old forest'.

To report on the status and trend of vegetation (Section 2.1), the ABMI presents the percentage of each vegetation type directly after accounting for human footprint on the land base, ranging from 100% coverage of a vegetation type (no visible human footprint) to 0% (completely modified by human footprint). 

Trend in vegetation types in Al-Pac’s FMA area was assessed by overlaying the 3×7-km detailed inventory of human footprint onto the backfilled vegetation layer and summarizing percent area of vegetation types left undisturbed.  Results for trend in overall cover for the nine vegetation types is available for: 1950, 1985, 2000, 2001, 2004–2023. Trend in old forest categories (>100 years) only covers 2000–2023 due to the high uncertainty associated with back-casting forest ages to historical times.

We calculated annual rate of change in vegetation types from 2000-2023. There are multiple approaches for defining the annual rate of change. For the information presented here, we used the methods defined by the Food and Agriculture Organization of the United Nations^[2].

Annual rate of change = (A₁ - A₂) ^{1 / (t2 - t1)} - 1

Where A₁ and A₂ represent the area at the time points at t₁ and t₂.

The Native Cover indicator provides a baseline measure of remaining habitat and is a core metric for assessing cumulative effects. This indicator was developed as part of a joint effort between ABMI and Alberta Environment and Protected Areas^[3] and it offers a transparent, repeatable way to track habitat loss and retention over time.

Native cover is expressed as the percent area of remaining habitat free from visible human footprint for two broad habitat classes:

• Aquatic & Wetland Native Cover: Wetlands, bogs, fens, swamps, marshes, lakes, rivers, and riparian areas.
• Terrestrial Native Cover: Upland habitats such as forests, grasslands, bare ground, and woody debris. 

We define our reference condition as the total area of Alberta that would be covered with terrestrial habitats (~66%) or aquatic and wetland habitats (~34%) if no human footprint had occurred. It is important to note that this reference condition is not a historic re-creation of the provincial landscape, as it does not capture historical Indigenous land use or historic wetland loss.

We applied an established method to model the recovery of coniferous and deciduous harvest areas using species‑specific recovery curves^[4]. In general, harvest areas recover over time, with harvested areas greater than 80 years old being considered as equivalent to undisturbed habitats. Other footprints are treated as permanent as we have not developed recovery trajectories for these feature types.

Through mapping all human footprint in the province, the HFI allows areas of native terrestrial habitat to be identified. There are areas in the province that have not been visibly disturbed by humans, although natural disturbances (e.g., wildfire, insect outbreaks) and indirect effects of humans (e.g., pollution) still occur. The interior habitat indicator tracks the amount of interior native habitat that is spatially separated from the influence of human footprint at varying distances^[5]. 

To report on the status of native habitat (circa 2023), we present:

• The percentage area of land cover that has no visible human footprint although land uses like grazing may still occur^[6]. 
• The percentage area of interior native vegetation using three edge distances (i.e., buffers) applied outwards from human footprint—50 m (the area closest to footprint, representing the strongest edge effects), 200 m (following recommendations from AEP 2022^[5]), and 500 m (to represent the longest reported or policy-based edge effects, such as those for Woodland Caribou) . 

The 50 m and 200 m base buffer distances are adjusted in certain cases to account for two factors following AEP 2022^[5]:

• Width of human footprint. For narrow linear features (less than 20 m wide), the edge influence that extends into native habitat is reduced because these openings have little surface wind, are shaded in most orientations, and are considered to be under “forest influence” for many species. As an exception, roads are assigned the full buffer distance regardless of road width, to conform to with AEP 2022^[5] methods for interior habitat.
• Recovery of forestry footprint. The edge distance is reduced as forestry footprint recovers (ages). 

No adjustments are made to the 500 m buffer distance.

We used the 2010 and 2023 HFI datasets to report total changes in native habitat and across the three interior habitat categories during this period. Changes in area for the four buffer zones (0–50 m, 50–200 m, 200–500 m, and >500 m) are expressed as a percentage of their 2010 values. We report on change in interior native habitat for the following categories: all native, lowland, upland, deciduous, mixedwood, pine and White Spruce. 

Note that calculations of interior native habitat exclude open water and bare ground, resulting in a smaller analyzed area than the total Al-Pac FMA area. 

See Section 2.3 for native habitat results.

Habitat quantity alone cannot capture the ecological effects of cumulative land-use impacts. Landscape connectivity measures how landscape composition and configuration affect the ability of organisms to move between habitat patches, which is key to wildlife movement and long-term biodiversity health in working landscapes. This indicator was developed as part of a joint effort between ABMI and Alberta Environment and Protected Areas^[7]. 

Landscape connectivity is expressed as Equivalent Connected Area (ECA), a graph theory–based measure that incorporates patch size, dispersal thresholds, and movement resistance between patches. We report landscape connectivity as the ratio of ECA calculated under both current (i.e., 2010 and 2022) and reference conditions (no anthropogenic disturbance) for two broad habitat types:

• Upland forests habitats: Upland forests such as Pine, White Spruce, deciduous, and mixedwood stands. 
• Lowland forest habitats: Wetlands, bogs, fens, swamps, and marshes.

We account for the recovery of coniferous and deciduous harvest areas using a two step process. First, harvest areas impact the ability of species to move between patches (resistance), which declines over the first 15 years^[8]. After this period, harvest areas are treated as undisturbed landcover that recovers over time (starting at age 15) based on the harvest area recovery curves^[9].

Connectivity can be estimated at multiple spatial scales. For this analysis, we assessed the connectivity of upland and lowland forests for each watershed (HUC-8) that falls within or touches the regional boundaries. This means that land-use changes outside these regions that impact connectivity may be reflected in this indicator. 

See Section 2.4 for landscape connectivity results.