Canadian Forest Service Publications
Biophysical indicators based on spatial hierarchy for informing land reclamation: the case of the Lower Athabasca River (Alberta, Canada). 2017. Thiffault, E.; Webster, K.; Lafleur, B.; Wilson,S.; Mansuy, N. Ecological Indicators 72:173-184.
Issued by: Great Lakes Forestry Centre
Catalog ID: 38849
Availability: PDF (request by e-mail)
Available from the Journal's Web site. †
† This site may require a fee
In the Lower Athabasca region of Alberta (Canada), surface mining for bitumen from oil sands creates highly disturbed environments, which need to be restored, after mine closing, to equivalent land capability in terms of biodiversity and ecosystem services. We demonstrate a method to characterize ecosystem diversity and conditions using biophysical indicators of the Lower Athabasca meant for informing land reclamation planning and monitoring by identifying and creating a typology of the main assemblages of topography, soil and forest vegetation at the watershed, landform and ecosite scales, and analysing the relationships among land units of various scales. Our results showed that watersheds could be classified into distinct groups with specific features, even for a region with a generally flat or gently rolling topography, with slope, surficial deposits and aspect as key drivers of differences. Despite the subtle topography, the moisture regime, which is linked to large-scale cycles that are dependent on the surrounding matrix, was of primary importance for driving vegetation assemblages. There was no unique and homogeneous association between topography and vegetation; the specific landforms each displayed a range of ecosites, and the same ecosites were found in different landforms. This suggests that landscapes cannot be defined in a qualitative manner but rather with quantitative indicators that express the proportion occupied by each class of ecological units within the coarser units, therefore requiring during land reclamation that sufficient care is given to create heterogeneity within a given landform in terms of soil texture and drainage so that a mosaic of ecosite conditions is created.
Plain Language Summary
They are suitable units for land use planning, thus methods that predict ecosite distribution on the landscape are useful tools for decision-making. We present an automated approach for ecosite classification based on readily available national datasets at 250 m resolution and a fuzzy k-means classification approach, using the Lower Athabasca River (LAR) region as a case study area. In this dry, gentle relief landscape, the moisture gradient along hillslopes was an important driver of ecosite types. The k-means classification produced 12 categories of ecosites consistent with Alberta provincial and national ecosite classifications. In the LAR region four ecosites were dominant and represented 67% of the landscape. Four ecosites, three of which were relatively rare on the landscape (i.e., < 5% of the landscape), had >40% of their area in existing or planned oil sands mining or in-situ oil sand extraction development. This automated method of mapping ecosites, in addition to delineating unit relevant for resource management improves fundamental understanding of the dominant drivers structuring the landscape. A coupled understanding between process and pattern in the LAR is an important first step in setting reclamation targets to restore function and services of disturbed areas.