The Botswana population is nominally some 90,000 animals in 150,000km2 - an average density of 0.6 animals/km2. This is consistent with the carrying capacities based on rainfall (Figure 7). Over most of the Botswana range densities should lie between 0.5 and 1/km2, with possibly higher concentrations in the extreme north-east adjacent to the Caprivi (1-2/km2).

The distributional data of DWNP (2000b) show the marked concentrations of buffalo herds in the dry season in the Okavango Delta and along the Chobe and Linyanti Rivers (Figure 17). In the wet season, the herds are widely dispersed.

ULG (1995) recognise 4 subpopulations within the main Botswana buffalo population:

1. The Kwando subpopulation
2. The Delta subpopulation
3. The Chobe subpopulation
4. The 'Border' subpopulation (the Botswana-Zimbabwe border). It is not obvious from their distributional maps whether these subdivisions are based on home range characteristics or whether they are simply an artifice for partitioning hunting quotas over the full buffalo range.

Data Quality and Interpretation

• Botswana DWNP have systematically carried out annual surveys in both dry and wet seasons for most years since 1987. The estimates which have been used are those for the 'potential population' (DWNP 2000b) - obtained by extrapolating the estimate for the area surveyed in the given year up to the total area of 145,605km2 which is the wildlife range in northern Botswana. In most of the years concerned the actual area surveyed is more than 95% of the total area so that this correction is unlikely to inflate the results.
• The wet season estimates are significantly lower than those of the dry season because of the reduced visibility when woodland tree canopies are in full leaf.
• The distributional data associated with these estimates show that buffalo are more widespread in the wet season than during the dry season when water restricts the daily movement. ULG (1995) note that this effect is most pronounced along the Kwando and Linyanti Rivers but is also present within the Okavango delta where water is widely available throughout the year. This observation suggests that movement between Botswana and the Caprivi is likely to be at its lowest during the dry season and, therefore, the dry season estimates for the Caprivi are indicative of the size of the "permanently resident" buffalo population. To a limited extent, the lower estimates for the Botswana population during the wet season could be attributed to movements into the Caprivi. However, (1) there are no survey data for the Caprivi during the peak of the wet season to confirm this hypothesis and (2) the estimates of buffalo in the Caprivi are so small compared to the overall size of the Botswana buffalo population that it does not seem reasonable to conclude that emigration from Botswana could account for the lower Botswana wet season estimates.
• ULG (1995), in considering the results from 1988 to 1995, presented a regression to show a significant downward trend in the population. This deduction was supported by an apparent shrinkage of range. However, the high estimates from 1999 and 2001 suggest that no such conclusion can be reached yet.

Figure 17: 1999 Wet and Dry Season Buffalo Distribution in Botswana

Figure 7: Crude Carrying Capacities for Buffalo in Southern Africa