THE ROLE OF REMOTE SENSING FOR SUSTAINABLE ELEPHANT MANAGEMENT IN SOUTH AFRICA. FOUR MEDIUM SIZED GAME RESERVES AS CASE STUDIES
Keywords: Ecosystem, Change Detection, Management, Monitoring, Multispectral, Classification
Abstract. Loxodonta africana (African Elephant) are running out of living space so the protection of what space they have is essential. Existing areas of suitable elephant habitat need to be protected not only from human development but from the elephants themselves.
As most elephant populations in South Africa are enclosed and multiplying, there is some increasing cause for concern as the damage caused will escalate and could reach unsustainable proportions. This study examined the utilization of satellite images for the detection of elephant induced ecosystem modification. A pilot study was conducted on four medium sized Game Reserves (each ±30 000 ha) in South Africa. The aim was to ascertain the feasibility of using image analysis as instrument by which Game Reserve managers could assess biodiversity richness, habitat loss, and population-habitat viability.
NDVI as indicator of primary production in vegetation is one of the instruments used to evaluate whether the carrying capacity for elephants of each Game Reserve has been reached and to compare the current biomass with those of previous years. The study also looked at the use of the woody canopy cover as target for change detection analysis.
Spectral characteristics of specific trees species which are known for being preferred by elephants were used to conduct a temporal analysis on satellite images starting from the period when the elephants were re-introduced into each Game Reserve, thus attempting to identify possible impact on the biodiversity of the respective Game Reserves. Images from satellites such as Landsat, SPOT, Quickbird and SumbandilaSAT provided the needed data and maps.