Context and objectives
The population of the endangered African elephant has recovered over the last decades in Botswana. During the dry season, significant number of elephants migrate outside the protected areas, especially to water holding areas. When in ecological balance with its surroundings, this animal contributes substantially to the regeneration and conservation of the vegetation in its natural habitat. However, area reduction of its habitat and obstruction of its migration routes have led to overgrazing in and around the protected areas. Together with severe drought during the last years, this resulted in a change in vegetation cover and type and created deforestation problems in northern Botswana.
Purpose of the study : to evaluate how satellite images as LANDSAT TM and SPOT XS, used in combination with maps, aerial photographs and field observations, can within a limited period of time contribute (1) to identify relevant vegetation units, (2) to determine their structural characteristics and spatial organisation, and (3) to detect changes in vegetation type and structure, in the context of efficient planning for wildlife conservation.
Expected scientific results
- The use of satellite images is satisfactory for mapping vegetation in a limited period of time, and is especially recommended if the image processing is integrated in a GIS.
- For vegetation mapping, the use of satellite images taken at the beginning of the dry season is recommended. A bitemporal image set registered within one growing season increases the classification result.
- Field observations must be made along the tracks due to the poor accessibility of the area; it is recommended that the terrain observations be completed with aerial photographs taken at low altitude. The evaluation of the damages requires an extensive field campaign, as quantitative measurements are required.
- Important landscape elements such as tracks and waterholes can be identified visually on the SPOT scene.
- The cost of a vegetation inventory using the proposed methodology is estimated to amount to 10 USD/km2 (1993). The local forest services consider this to be cost effective.
The results of this study have served as a basis for a project proposal in relation to habitat mapping and wildlife and nature conservation in Botswana, which received financing from the European Community (DGXI).