Several theories for he Decline of Roan, Sable and Tsessebe have been put forward. Dunham and Robertson (2001) demonstrate the correspondence of the observed numbers of tsessebe in Kruger National Park from 1977-2000 with the long term cumulative deficits and surpluses in annual rainfall and the late dry season rainfall.

In a study of tsessebe on a mixed cattle and game ranch in south-western Zimbabwe, the same variables accounted for an observed decline in tsessebe (Dunham et al 2003). An additional variable, that of competition with cattle, had to be incorporated into the equation in order to fit a population model closely to the observed data.

Wilson and Hirst (1977) saw the whole subject of factors limiting roan and sable as a complex one involving disease, malnutrition and habitat quality and thought that nutritional status might be affecting populations.

Grant et al (2002) review the various hypotheses which have been in place at various times to explain the decline in roan, sable and tsessebe in Kruger National Park and the effects of the management actions which have resulted from these hypotheses.

Dunham and Robertson (2001) and (Dunham et al 2003) reviewed hypotheses put forward to account for the tsessebe decline in both Kruger and Shangani and were able to dismiss all of the following - based either on the fact that they could not account for the situations over the full time period or that the data supporting them were inadequate -

• Competition with other wild herbivores;
• Predation;
• Fire;
• Effects of artificial waterpoints;Illegal hunting;
• Emigration;
• Excessive sport hunting and capture offtakes; and
• Direct food shortages caused by poor rainfall in any given year.

Animals were suffering from undernutrition - and might even have been killed by predators because of this - but these were proximate rather than ultimate causes.

The only satisfactory explanations for the overall population performance under a range of rainfall conditions lay in the relationship with the long term cumulative rainfall deficit/surplus and the late dry season rainfall.

Rainfall in late dry season appears to be critical, affecting animals' condition, survival rate, late stages of pregnancy and early stages of lactation. On Shangani Ranch, an increase in cattle stocking rates midway through the time span under consideration caused direct competition for food which resulted in a negative correlation of cattle and tsessebe biomass. The more serious indirect effects of competition with cattle in the longer term were the structural changes in habitats caused by bush encroachment following a high biomass of cattle.