PJB-2008-37
LINKING REMOTE SENSING AND ECOLOGICAL VEGETATION COMMUNITIES: A MULTIVARIATE APPROACH
RIFFAT N. MALIK1 AND SYED Z. HUSAIN2
Abstract
In spite of few attempts for mapping land-cover types in Pakistan, remotely sensed data has not been used widely; and its potential is not being explored for providing information on mapping vegetation cover in general and ecological communities in particular. The present study was undertaken in the Lohibehr scrub forest in the Foothills of Himalaya, northeast of Pakistan. The objective of the study was ot find out the relationship between remote sensing data and vegetation communities of ecological importance using multivariate techniques such as TWO WAY INDICATOR SPECIES ANALYSIS (TWINSPAN), Principal Component Analysis (PCA) and Correspondence Canonical Analysis (CCA). Floristic data were compiled for vegetation types and Digital number (DN) values were extracted from SPOT XS image for visible and near infrared bands (NIR). Classification and ordination methods were used for the classification of floristic data and to describe the relationships between floristic species composition and DN values. Ordination analyses indicated positive correlation between floristic species composition and DN values along the first ordination axis, with the NIR. The ordination methods proved effective in summarizing basic, general structure of the plant community types and to some extent indicated correspondence with their spectral signatures. The results highlighted the potential of remote sensing data in providing information on different plant community types that could be used in planning, management and conservation of subtropical forest.
To Cite this article:
Download
