Few large mammal research topics have received as much attention as elephant behaviour. The social organisation of elephants has been well covered by Douglas-Hamilton (1972), Moss (1976, 2000) and Poole (1996). Male territoriality and must have been researched in depth by Poole (1982, 1986, 1987, 1989). The aspects of behaviour included here are those which are directly relevant to elephant management in Namibia.

• Habitat modification
• Elephant movements dispersal and home ranges
• Human-elephant conflict

Habitat modification

The maintenance of a significant canopy cover of mature trees is not possible with high density elephant populations. Various authors have attempted to minimise the consequences of this, asserting that the impact on biological diversity is minimal or localised (Owen-Smith 2005, du Toit 2005). Pervading the publications on this topic is a tacit assumption that all management decisions relating to elephant are based simply on whether or not biological diversity is being conserved and consideration of the wider socio-political and aesthetic issues tends to be ignored. Perhaps conservation of biological diversity per se is not the fundamental issue. More important may be the type of landscapes we wish to create and the rôle we would like elephants to play in a multispecies land use system where sustainable development is the key issue.

Elephant movements, dispersal and home ranges

Owen-Smith (2005) notes that true dispersal involves the abandonment of previous home ranges

and the occupation of new areas. Much of what is taking place in the regional elephant range (Figure 10) may not fully satisfy this definition in the short term. Movements out of Botswana are still largely seasonal and it may take many years before new ranges in the Caprivi, Angola and Zambia are permanently colonised. Within Namibia, a similar situation pertains in the north-west and north-central areas.

In a radiotracking study carried out from October 1987 - May 1988, Lindeque & Lindeque (1991) found seasonal home ranges of 5,800-8,700km2 amongst the population of 1,000-2,000 elephants in Etosha. Elephants from Etosha moved almost to the Angola border and elephants in the Kaokoveld moved across previous bioclimatic demarcations of populations. This study raised questions about the validity of earlier classifications of elephant subpopulations in northwest Namibia. Viljoen (1987,1988,1989) classified the Kaokoveld elephants into 3 distinct populations, 2 of which were reportedly in contact with or were part of the Etosha population. Viljoen maintained that the western population ('desert elephant') was restricted to an area west of the 150mm isohyet and never left this bioclimatic zone. This paper disagreed. Some 30-50% of the Etosha elephant population were spending 3-5 months/year outside the park.

Rodwell (1991) used radio collars to establish that the majority of elephants in the Caprivi were seasonal inhabitants. The commonest movement pattern was residency during dry season and movement out of the Caprivi during wet season. One herd went from Mudumu through densely settled areas along Kwando to Sioma Ngwesi in Zambia in the wet season. Movements north into Angola were of very limited duration - probably due to the poor security situation. Whenever elephants were in the Caprivi, they used a small range and therefore placed enormous pressure on the vegetation. An elephant bull in Mahango used a dry season home range of 575km2 and a wet season range of 5,606km2. In Etosha home ranges varied from 2,851-18,681km2. Rodwell concluded that Namibian elephants display some of the largest home ranges in Africa.

The spatial distribution of Namibian elephant is very much determined by the availablity of suitable cover and surface water and the large home ranges are a result of the arid environment. This has major portents for management both within Namibia and across national boundaries. No single protected area is a self-contained range for elephants. Large areas co-managed by the relevant landholders and occupiers will be necessary to provide viable ranges, to distribute the pressure of elephants on habitats and to allow for population increase and expansion. The patterns of seasonal movement provide strong reasons for transboundary cooperation on elephant management especially with Botswana.

The vision statement developed under the ongoing UNDP/GEF assisted project in Namibia (UNDP 2005) provides for co-operation between stakeholders to achieve this far-sighted concept. Linkages would be established under a co-management system which would both secure and increase the available range for elephant.

Human-elephant conflict

Elephants destroy crops, damage water installations, compete with cattle at water points (and occasionally kill cattle) and are a physical threat to humans.

The Caprivi has the highest incidence of conflicts between humans and elephants in Namibia and the largest number of incidents occurs on the Kwando River frontage (O'Connell 1995a). Elephant damage to crops from 1995-2000 for the whole Kwando River region was estimated at about N$20,000 per year (O'Connell-Rodwell et al 2000). Although this may seem financially trivial, it has devastating effects on household livelihoods.

Recent data from the Annual Audit database (Event Book) monitoring system of the Caprivi Conservancies (NNF 2004) indicate an escalation in incidents involving elephants (Table 5).

The benefits which communities are receiving from elephants on their land are small. The number of elephants hunted as trophies and problem animals is very low (less than 10 over the time period of the table) and does not come close to compensating communities or individuals for their losses.

This situation is a potential 'time bomb'. With all conservancies being in a fledgling stage, there is a 'wait-and-see' attitude amongst their members. The recent rapid increase in the numbers of elephant in the Caprivi is probably due to a temporary tolerance of elephants while the conservancies are in their formative stage. But these communities will be evaluating whether a commitment to wildlife as a land use is worthwhile and, unless elephants contribute a great deal more to livelihoods, the present forbearance is likely to disappear.