Monitoring of a national-scale indirect indicator of biodiversity using a long time-series of remotely sensed imagery

Nicholas C. Coops, Fabio M.A. Fontana, Gillian K.A. Harvey, Trisalyn A. Nelson, Michael A. Wulder

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


Remote sensing provides continuous, large-area coverage that supports synoptic, consistent, and repeatable monitoring of vegetation and, therefore, can be used to derive indirect indicators of biodiversity. We used a 21-year archive of Advanced Very High Resolution Radiometer (AVHRR) (1987–2007) data to assess changes in an indirect indicator of biodiversity, the Dynamic Habitat Index (DHI), which has proven useful in predicting patterns of species richness and abundance at broad spatial scales (>1 km2). The index has 3 components reflecting integrated vegetation greenness, vegetation seasonality, and minimum vegetated cover. We used a Theil–Sen nonparametric rank-based test and computed the proportion of cells with positive or negative DHI trends within both the Canadian protected area network and the Canadian ecoprovinces. In general, the smaller protected and natural areas close to urban development had relatively larger proportions of trending DHI components over the 21-year period, than the larger, more remote protected regions. Most protected areas and ecoprovinces observed an overall increase in integrated greenness and a reduction in vegetation seasonality over the analysis period. We conclude that the ability to derive trends from long time-series of remote sensing data helps focus local biodiversity monitoring programs by guiding actions to areas of the greatest observed change.

Original languageEnglish (US)
Pages (from-to)179-191
Number of pages13
JournalCanadian Journal of Remote Sensing
Issue number3
StatePublished - May 4 2014
Externally publishedYes

ASJC Scopus subject areas

  • General Earth and Planetary Sciences


Dive into the research topics of 'Monitoring of a national-scale indirect indicator of biodiversity using a long time-series of remotely sensed imagery'. Together they form a unique fingerprint.

Cite this