SummaryThis paper presents a description of the temporal and spatial distribution of rainfall within the Serengeti Ecosystem, Tanzania. Ecosystem and Regional patterns are discussed with respect to synoptic processes, vegetation zonation, phases of primary production and the distribution and migrations of large mammals.

Interspecific competition and facilitation have both been proposed as processes promoting species separation and co-existence in African ungulates. In one group of grazers on the Serengeti plains, comprising wildebeest (Connochaetes taurinus), zebra (Equus burchelli), and Thomson's gazelle (Gazella thomsoni), these processes have also been suggested to regulate the populations. Censuses of these populations over 20 years have shown changes that allow a test of which, if either, process regulates population numbers. Wildebeest numbers have levelled off as a result of intraspecific competition for food following a five-fold increase due to release from disease and an increase in food supply. Zebra numbers have remained stationary throughout the same period. Gazelle numbers have declined in the last 10 years. These results are contrary to the facilitation hypothesis, which predicts that wildebeest numbers should not have increased if those of zebra did not, and that gazelle numbers should have increased since the wildebeest population increased. The gazelle results are consistent with the interspecific competition hypothesis, but the zebra results are contrary to it. We propose an alternative hypothesis that predation regulates the zebra population, and we suggest ways of testing this.

No abstract

No abstract

Summary A method for sample counting the Serengeti migratory wildebeest using vertical aerial photography to sub‐sample randomly located transects is described. The population estimate achieved was 754 028 animals, with 95% confidence limits of ± 8‐5% of the population estimate. The application of this method to other wildlife populations is considered.