DefinePK

Abstract

ABSTRACT The Salt Range area of Pakistan is a local biodiversity hotspot. Shrub forests of the area are faced with an increasing problem of forest fragmentation and degradation, which is eroding the original biodiversity. The main objective of this study was to identify priority remnant forest patches for biodiversity conservation planning. Extensive use was made of remote sensing and GIS techniques to achieve the objective. Forest vegetation was selected as a subset of total biodiversity. Major forest mosaics of the area were identified from satellite image and then visited on the ground to collect locational, environmental, and biological data. Field data collection was done through Line transect sampling design. Image classification was performed by using the Supervised and Knowledge-based classification methods. Patch size of the remaining semi-natural forests was selected as the most important criterion for the identification of priority areas. However, patch shape index was also measured. NDVI (Normalized difference vegetation index) was found to be a good predictor of forest vegetation of the area, particularly, when remote-sensing data is collected during summer rainy season as in this study. Results of image classification obtained by using NDVI were comparable to those obtained by using all spectral bands of the Landsat TM satellite image. However, for the drier western forests of the study area, classification accuracy was quite low A total of 17 priority forest fragments for biodiversity conservation have been identified which could form the core areas in any proposed reserve design. Remote sensing and GIS are powerful and useful tools for biodiversity assessment, mapping and conservation planning at the ecosystem or landscape scale.