Logo image
Back
Remote sensing derived edge location, magnitude, and class transitions for ecological studies
Journal article   Open access   Peer reviewed

Remote sensing derived edge location, magnitude, and class transitions for ecological studies

M.A Wulder, B.P Stewart, M.E. Andrew, M. Smulders, T. Nelson, N.C. Coops and G.B Stenhouse
Canadian Journal of Remote Sensing, Vol.35(6), pp.509-522
2009
DOI: https://doi.org/10.5589/m10-007
pdf
remote_sensing_for_ecological_studies.pdfDownloadView
Published (Version of Record) Open Access
url
Link to Published Version *Subscription may be requiredView

Abstract

Regionally intensive human activities related to resource extraction (i.e., harvesting, oil and gas extraction) are increasing the occurrence of edges found in some forested landscapes. Edges between different land cover types represent important transition zones for abiotic and biotic processes. However, boundary detection methods often identify edges solely in areas of high contrast, such as transitions between forest and non-forest areas, and are insensitive to the relative contrast and orientation of different transitions. Edge contrast and orientation can determine the magnitude and even the occurrence of ecological edge effects and should be measured to provide information on landscape condition and habitat potential. Wombling was applied to the wetness component of a tasselled cap transformation (TCT) of a Landsat scene acquired over a portion of the eastern slopes of the Rocky Mountains in Alberta, Canada. By incorporating wombled edge contrast and orientation, and edge class transition type obtained from a land cover dataset, the nature of all transitions between land cover classes within the image was characterized and quantified. The consistency between edges identified by wombling and other common methods of edge delineation (such as spatial clustering) and methods of edge quantification (such as landscape pattern indices, or LPIs) was also assessed. Land cover transitions showed a broad range of edge contrast. Comparisons of edge contrast and the LPI edge density showed a positive correlation (r2 = 0.33); however, the strength of this relationship varied with the dominant land cover type (e.g., r2 = 0.016 for broadleaf open forest to r2 = 0.48 for dense coniferous forest). Stratifying edge contrast to higher values (i.e., >1 standard deviation) increased agreement with edge density, indicating that the LPI is preferentially relating high contrast edges. This study demonstrates how unique edge characteristics may be generated from a remotely sensed continuous variable (TCT wetness). This knowledge of the location, magnitude, and class transitions found at edges provides insights into the nature of the edge effects and enables the development and testing of hypotheses informing wildlife habitat use and selection.

Details

UN Sustainable Development Goals (SDGs)

This output has contributed to the advancement of the following goals:

#13 Climate Action
#14 Life Below Water
#15 Life on Land

Source: InCites

Metrics

168 File views/ downloads
106 Record Views
4 Times Cited - Web of Science

InCites Highlights

These are selected metrics from InCites Benchmarking & Analytics tool, related to this output

Collaboration types
Domestic collaboration
Citation topics
3 Agriculture, Environment & Ecology
3.40 Forestry
3.40.195 Biodiversity Conservation
Web Of Science research areas
Remote Sensing
ESI research areas
Geosciences
Logo image