Appraisal of two counting methods and the population dynamics of four harvested Antelope species in a Kenyan game ranch

Abstract

Wildlife Ranching and Research Ltd., Athi River (WRR)is a privately owned mixed game and cattle ranch, with the aim of demonstrating the environmental and economic viabi lity, as well as social acceptabi Iity of game ranching. The field work for this research was carried out at WRR between October f 986 and September 1987. The emphasis was on economically important species. Monthly total ground counts made it possible to evaluate levels of harvesting that could be accommodated by Grant's gazelle, kongoni and wi ldebeest. Wildebeest and kongoni could accommodate 32.6%and 14.0% average annual offtake rates, respectively without the population decl ining. Grant's gazelle, on the _~therhand, could only accommodate 0.05% annual offtake rate. The highly skewed sex ratio in favour of females was only partly responsible for the low recruitment in this species. An unknown mortality factor was considered responsible for the major part of the low recruitment value. Ground counts carried out in the wet season were more accurate than the ones carried out in the dry season for Thomson's gazelle, but not for Grant's gazelle and kongoni. Total ground counts were more accurate than aerial counts when applied to the small antelopes, i.e Thomson's and Grant's gazelles and impala combined. III No significant differences could be detected between resutts of the- two methods when applied to the WRRungulate community as a whole. Total ground counts were recommended as the counting method that best suits the WRRsituation. Results of road counts were disappointing. The results exhibited high variations, and two species, i.e. Thomson's gazelle and wildebeest were found to prefer the road side, resulting in overestimated population sizes. Ageing and sexing in the field indicated that Thomson's andGrant's gazelles, and kongoni breed throughout the year with possible birth peaks in the cases of Thomson's gazelle and kongoni. A correction factor was used when calculating the incidence of pregnancy in wildebeest for the bias resulting from WRRharvesting strategy of not harvesting pregnant and lactating females. The correction factor could not be extended to the other species. Equations describing the rate of foetal growth in wildebeest were given. Environmental factors affecting rate of foetal growth, variation in the rate of growth and the timing of the mating season were discussed. It was argued that a harvesting strategy of avoiding the killing of pregnant and lactating females may introduce artificial selective pressures that tend to increase the birth rate. Rates of pregnancy were low in young kongoni and wildebeest and high inold ones. Recommendations were suggested related to harvesting timing of the different age and sex classes of the different species. Subjective age classes based on eruption sequence and wear pattern of the maxillary molars provided an accurate ageing criterion up to the age of 28 months in the case of wildebeest. . Crown heights exhibited high variation in animals of the same age. The linear regression model provided a better fit than the exponential to the pattern of decrease of crown heights with age.