Application of distance sampling to estimate population densities of large herbivores in Kruger National Park

Aerial surveys have been used in the Kruger National Park, South Africa, to count large ungulates since the late 1970s. After 1998, aerial line-transect sampling using fixed-wing aircraft and Distance analyses replaced the ' total' counting method. This paper investigates these methods and three sampling intensities for estimating the densities of nine large ungulate species in Kruger National Park. Estimates suitable for the detection of population trends and making management decisions were decided by examination of coefficients of variation (set < 20%, a priori). Despite the likely violation of some key assumptions of Distance sampling methods, analyses gave population estimates with adequate coefficients of variation for monitoring trends in impala, giraffe, zebra, kudu, white rhinoceros, and elephant bull populations. Significant improvements in precision were obtained at higher sampling intensities for kudu, giraffe, bull elephants and white rhinoceros, but these species already had sufficiently precise population estimates for the detection of trends at the lowest sampling intensity (15%). The estimates for warthog, wildebeest and waterbuck populations were, however, insufficiently precise for assessing population trends. Increasing sampling intensity to 22% and higher did not significantly increase the precision of the Distance estimates for these species. Shortcomings in interpretation of the data caused by violations of critical assumptions of analyses are identified and discussed.