Analysis of Fragmentation and Anthropogenic Disturbances in the Himalayan Forests: Use of Remote Sensing and GIS
Subrat Sharma, L.M.S. Palni
G.B. Pant Institute of Himalayan Environment & Development
Kosi-Katarmal, Almora 263 643, India
Fax: + 91 -
5962 - 31360; Email: prasad@nde.vsnl.net.in
P. S. Roy
Indian Institute of Remote Sensing
4, Kalidas Road, Dehradun 248 001, India
Fax: +91 - 135 - 741987;
Email: psroy@del2.vsnl.net.in
Key Words
District Almora, Forest, Fragmentation, Himalaya
Abstract
The vastness and rugged "difficult to reach" nature of Himalayan terrain poses serious limitations on field observations; consequently field experimentation and information collection to develop regional planning is often difficult. Satellite remote sensing and GIS can be utilized to extract information through simple analyses and/or by the use of models to answer specific questions. Only limited attempts have been made to use these technologies in describing landscape dynamics for biodiversity management in the Himalaya. In the present approach forest vegetation has been explored using IRS 1C -LISS III FCC in an area of ~3167.5 km
2 (entire Almora district) and various landscape elements have been analysed. Forested area accounted for 42.2% of the total area. Most common vegetation type was pine forest (Pinus roxburghii), occupying 85.3% of the total forested area, but it was highly fragmentary in nature. Total 414 pine forest fragments were recorded over the entire landscape. Area of each patch varied from 1 km
2 to 179 km
2; however, majority of fragments (77% of the total) were below 1 km
2. These smaller units are possible sites of pine forest "extinction" due to anthropogenic pressures. A greater shape index (1.1 - 9.2) was observed for pine forest fragments. Edge effect analysis of the pine forest fragments indicated that the impact area ranged from 19.9 ha to 855.7 ha. These areas are subject to microhabitat alterations without canopy cover loss. Human dependency on natural vegetation appeared to be the main cause of fragmentation. Disturbance index model revealed that about 34% of the total landscape area is under severe anthropogenic pressure as apparent from high index. RS & GIS based landscape approach is an emerging tool for identification of hot spots for biodiversity conservation in the mountains, and especially to appropriately include human dimension in the management planning.
1. Introduction
Mountain environment is characterized by its complexity. Large variations in climate, water characteristics, soil, and geology have, over millenia, produced a wealth of biological diversity in flora, fauna and microbes. The Himalaya are massive mountains that occupy over 1 million km
2. They produce a distinctive climate of their own and also influence the climate of much of the Asia. The Himalayan range contains some of the most spectacular biodiversity on earth, but much of it is under-explored and lacking in effective protection. The vastness and rugged "difficult to reach" nature of Himalayan terrain poses serious limitations on field observations; consequently field experimentation and information collection to develop regional planning is often difficult. There is a wide spread agreement that global biodiversity is being reduced at an accelerated rate (Myers, 1980) and current approaches to understand the landscape ecology are highly diverse. Landscape harbors all grades of biological hierarchy, from ecosystem level to species and genes, that are targeted for biodiversity inventories and conservation (Noss & Harris, 1986). Landscape may also include agricultural, forested, protected and ecologically sensitive areas, which interact considerably (Forman & Godron, 1986), and upon which humans have a major influence (Naveh & Lieberman, 1990). Remote sensing can be used successfully to identify the frequency, boundaries, sizes and shapes of various landscape components (Scott et al., 1993). Use of satellite remote sensing in vegetation related studies was introduced around mid eighties in the Central Himalayan region (Singh et al., 1985, Tiwari et. al., 1985). The technology was used to identify landscape/landuse patterns (Singh et. al., 1984), and approaches were developed for mapping, monitoring & change detection (Rathore et al., 1997, Sahai & Kimothi, 1994). Only limited attempts have been made so far to use satellite remote sensing & GIS in describing landscape dynamics and for biodiversity management. Using such an approach Dinerstein (1998) has recently assessed the biological importance and status of habitats & ecosystems of the Himalaya to identify gaps in conservation and protection, and a total of 16 Himalayan ecoregions have been identified. Literature review indicates paucity on printed information on the loss of Himalayan biodiversity, related issues, and landscape studies on fragmentation. In this rational the present effort is to explore the vegetation status (major forest types) and the role of human habitation in a mountain landscape of the Indian Central Himalayan region with particular reference to landscape attributes and fragmentation in respect of a predominant pine (Pinus roxburghii) vegetation in the middle Himalaya.
Study Area
The study area, District Almora, lies between 29
o23'-30
o00' N and 79
o00'-80
o05' E in the Indian Central Himalayan region (Fig. 1) and is one of the 13 hill districts of Uttaranchal State. Geographical extents of the district are spread in an area of about 3167 km
2, and majority of landscape geologically falls in the Lesser Himalayan belt, however, administrative boundaries also extend to Siwaliks (0.2 %) in the southwestern part. The district is inhabited by 836 thousands persons (1991 Census) with a population density of 282.9 persons per km
2.
Figure 1. Location of Indian Central Himalayan region and district Almora.
Methods
Attributes of any broad vegetation type that is structurally homogenous and distinct such as a forest or grassland, are readily inventoried with remote sensing. Using maps and remotely sensed data, with verification from careful ground truthing, it is possible to draw limits around vegetation types (Roy et al., 1985). IRS-1C, LISS-III false colour composite (FCC) of bands 2, 3 & 4 of January 1998 was used to identify various forest types and landuse/landcover classes. Visual interpretation (interpretation key is given in Table 1) was done followed by ground truthing, and interpreted details were transferred to the base map. Forest, Landuse/ Landcover and other thematic maps were digitized and different vector data layers were created in PC
