Damage to vegetation

Elephants are able to survive in a very wide range of habitats across the extremes of rainfall in Africa. Namibia provides the prime example of elephants adapted to living in desert conditions where annual rainfall is less than 150mm. Except for the most extreme desert areas, all of Namibia is suitable habitat for elephant. Even within the desert areas, elephants are able to make use of the watercourses almost as far as the coast and, following good rainfall, elephants may use areas below the 100mm and 50mm rainfall isohyets on an occasional basis (Figure 8).

Parker (1984) and Parker and Graham (1989) show that maximum elephant densities increase with rainfall across Africa and, invariably, this places elephants in competition with human beings for habitats. With few exceptions, elephants are displaced by human beings. Kingdon (1971) showed that the range available to elephant in East Africa decreased by about half over the 50 year period from 1920-1970. From a situation where human populations had been islands in a sea of elephants, the transformation had taken place to make elephant populations islands in a sea of humans. Martin (et al 1985) adapted Parker's relationship between elephant and human densities to estimate elephant numbers in areas of Africa where no elephant surveys had been conducted but where human population densities were known.

In northern Namibia elephants and humans are competing for the same resources in a zone extending from the Cunene River in the west as far as Impalila Island at the eastern extremity of the Caprivi. The experience throughout Africa is that no amount of protection by the State can save elephants in areas into which humans wish to expand. The final outcome of this competition will depend entirely on the value (both economic and intrinsic) which the people living in the north of Namibia place on elephants and the extent to which they are able to realise that value.

Damage to vegetation

In the case of elephant the question to be asked is not so much what elephants require from habitats as what elephants do to habitats.

The term "damage to vegetation" is resisted strongly by many elephant researchers who prefer to regard what elephants do to vegetation as "modification". However, Anderson (1973), Swanepoel & Swanepoel (1986), Lindeque (1988) and Murindagomo (1992) used the term elephant damage to describe the loss of canopy trees from woodlands.

Over the past 30 years the relationship between elephant densities and loss of trees has been quantified in many parts of southern Africa (Martin 1974, Thomson 1975, Barnes 1983 and 1985, Coulson 1992). Both Craig (1992) and Martin (1992a) independently arrived at the finding that, almost irrespective of the vegetation type, to retain more than 50% of canopy trees required elephant densities less than 0.5 per km2. Trees with rapid growth rates such as Acacia species appear to suffer higher rates of mortality from elephants than Brachystegia and Baikiaea woodlands: however the regeneration potential is far lower for the hardwood species. O'Connell (1995) reported significant damage to many species of trees in the Caprivi including Acacia erioloba, A. nigrescens, Terminalia sericea and Baphia massaiensis and notes that Baikiaea was not heavily used by elephants. She remarks - "Acacia nigresecens specimens do not have much chance along the Kwando".

Whether the changes in vegetation wrought by elephants are harmful will remain a hotly contested point. However, the work of Cumming (et al 1997) showed clearly that biodiversity in the mid-Zambezi Valley in Zimbabwe had been reduced by the action of elephants. Comparisons were made of the presence or absence and, where possible, densities of a range of plant and animal species (including large mammals, small mammals, birds, bats, and insects) inside and outside State protected wildlife areas and it was found almost universally that the biodiversity in the State protected areas was lower than immediately outside them. If the aim of a conservation area is to conserve biological diversity then, in this area, they were not meeting their objective. Attention has been drawn to the possible negative impact that elephants are exerting on the other rare and valuable species which Namibia would like to conserve (Martin 2002, 2003, 2004a).

Cumming & Cumming (2003) examined the trampling effects of ungulate communities (including both wild mammals and domestic livestock). The impact of elephants was 3 times higher than that of cattle (for the same biomass of animals) and that of cattle was double that of smaller domestic livestock (sheep and goats). Cattle and elephants both had a far greater impact than any multi-species wildlife communities or small domestic stock.

Elephants do not reach equilibrium with their habitats. Lindeque (1988) found no evidence of density-dependent regulation in the Etosha National Park. Whilst Laws (et al 1975) certainly found that the high elephant densities in Bunyoro National Park were having marked effects on age at first conception and reproductive performance, there was no indication that elephants would level off at some asymptotic density - indeed they appeared to be on the brink of a major collapse at the time a major culling operation was undertaken. In Tsavo National Park, Kenya, elephants destroyed their habitats and then crashed: although it is tempting to attribute this to the severe drought at the time, the result was probably inevitable (Parker & Amin 1983). In all the large culling operations carried out in Zimbabwe between 1972 and 1987 on populations which were considered to be above the 'carrying capacity' for the area, there was no post-mortem evidence in any of the animals that their body condition was poor or that their reproductive capacity was lowered.

Caughley (1976) hypothesized that the relationship between elephants and their habitats was cyclic. As elephant numbers built up, trees would decline. This would be followed by a decline in the elephant population. Once elephants had been reduced to low densities, trees would begin to increase. And so the cycle would repeat itself. The hypothesis is probably correct - but it lacks a spatial dimension. Given the finite areas available to elephant in the 21st century and the relatively high human populations on the continent, it is very likely that the troughs in the cycle proposed by Caughley might result in local extinctions wherever elephants are unable to move away from their own 'mass destruction' of trees or wherever there are no adjacent populations to repopulate the devastated area.

Even if the option of elephant population reduction were not a highly volatile political and emotional issue, the hard facts are that there is no simple recipe for elephant management which will simultaneously maintain biological diversity and substantial populations of elephant.