Wildlife Research Wildlife Research Society
Ecology, management and conservation in natural and modified habitats

Utilising Airborne Multispectral Videography to Predict Habitat Complexity in Eucalypt Forests for Wildlife Management *

Further information about this research can be found on the World Wide Web at http://www.ffp.csiro.au/nfm/mdq/

N. C. Coops and P. C. Catling

Wildlife Research 24(6) 691 - 702
Published: 1997


Airborne videographic remote sensing is a relatively recent technology thatcan provide inexpensive and high-spatial-resolution imagery for forestmanagement. This paper presents a methodology that allows videographic data tobe modelled to predict habitat complexity in eucalypt forests.

Within the eucalypt forests of south-eastern New South Wales, plots werelocated on the imagery, and the local variance of the videography within eachplot was computed on the assumption that changes in local variance provided anindication of forest structure, and thus the habitat complexity of the site.The near- infrared (NIR) channel demonstrated the most variation, as thatchannel provided an indication of photosynthetic activity and, as a result,the variation between canopy, understorey, ground cover, soil and shadowprovided a highly variable response in the video imagery. Habitat-complexityscores were used to record forest structure, and the relationship between theNIR variance and field habitat-complexity scores was highly significant(P < 0·001)(r2 = 0·75;n = 29). From this relationship, maps of thehabitat-complexity scores were predicted from the videography at 2-m spatialresolution. The model was extrapolated across a 1 1 km subset of the videodata and field verification showed that the predicted scores correspondedclosely with the field scores.

Studies have demonstrated the relationship between habitat-complexity scoresand the distribution and abundance of different mammalian fauna. This methodallows predictions of habitat-complexity scores to be spatially extrapolatedand used to stratify the landscape into regions for both the modelling offaunal habitat and to predict the composition, distribution and abundance ofsome faunal groups across the landscape. Ultimately, the management of foresthabitats for wildlife will depend on the availability of accurate maps of thediversity and extent of habitats over large areas and/or in difficult terrain.


© CSIRO 1997

Rent Article (via Deepdyve) Export Citation Cited By (34)