No abstract

Widespread bushmeat hunting represents one of the major threats to many mammals and birds in Africa. We studied the influence of illegal bushmeat hunting on large grassland birds in the Serengeti National Park (SNP) and adjoining protected areas, by using the ostrich (Struthio camelus) as a case study. First, we documented illegal hunting of both small and large birds by using a questionnaire in the villages on the western and eastern side of the SNP. Second, we studied the effect of illegal hunting on density by driving 4,659 km of transects inside SNP and on the adjacent protected areas, where the data were analysed by DISTANCE sampling. Last, we used flight initiation distance (FID, i.e. the distance between an approaching predator (human) and prey when flight is started), to assess possible impacts on behaviour from illegal hunting. We found that people from the western side of the SNP admitted to hunting both small and large grassland birds, and collect ostrich feathers and eggs. Although the Maasai also hunted small birds, only ostrich feathers and eggs of the large grassland birds were used. Surprisingly, we found no significant differences in densities between the SNP and adjoining partially protected areas, but ostriches had longer FID to an approaching human outside the SNP. Currently illegal hunting does not appear to affect the ostrich population, but given the extensive use of birds for consumption more awareness educational programs accompanied by provision of agricultural incentives within the protected areas are needed.

Most ground nesters lay pigmented eggs, and egg pigmentation generally matches the environment. Pigmentation of eggs has evolved as a protective device against predation, but dark-pigmented eggs can be susceptible to overheating when exposed to solar radiation. The Ostrich (Struthio camelus) lays white eggs that are unattended for the first few weeks before incubation, and are quite visible to predators. To evaluate the effect of colour on the surface and core temperatures, we painted some Ostrich eggs dark brown or white, and left some unpainted (control), and exposed all of them directly to the sun during the day. The surface and core temperatures of brown eggs were significantly higher than those of the white-painted and control eggs. In addition, the core temperature of brown eggs exceeded 37.5°C, which is the temperature at which embryo mortality starts to increase. In a second experiment, we placed eggs (brownpainted and control) in various types of vegetation to study their visibility to an observer walking towards them. The white eggs were discovered from a significantly longer distance than the brown eggs, indicating that the predation risk may be much higher for white eggs. The results thus suggest that white eggs minimise overheating and allow the Ostrich to leave its eggs unattended before incubation starts, but they are more susceptible to predation.

Ostrich breeding behaviour in the Serengeti ecosystem, Tanzania was investigated for differences in laying dates between low altitude western area (WA) and high altitude eastern area (EA) populations. Ostriches in WA laid eggs significantly earlier than in EA. The differences could be attributed to topography and rainfall pattern. Reliable rains in lower altitudes ensure availability of food that in turn influences the whole process of the reproductive cycle. Clutches were contributed by several females with a nest having up to 38 eggs. We also compared the frequency of observation of predators, ostriches, nests, 'singletons' (single eggs laid randomly) and broods between the two areas. There was no significant difference between WA and EA in 1) ostrich ⁄ nest ratio, indicating similar breeding densities; 2) ostrich ⁄ predator and predator ⁄ nest ratios, indicating that predation pressure was equally high; 3) nest ⁄ singleton and predator ⁄ singleton ratios, indicating that loss of nests did not vary between areas. However, there were significantly more predators, nests and ostriches compared to broods in EA than in WA, indicating a significantly lower reproductive success in EA. Using metapopulation terminology, ostriches in EA could be regarded as a 'sink' population and those in WA as a 'source' population, but investigations over longer time-periods are needed to further resolve if this is the case. RésuméLe comportement reproducteur des autruches dans l'écosystème du Serengeti, en Tanzanie, a été étudié pour voir les différences dans les dates de pontes entre les popula-tion similaires; 2) le ratio autruches ⁄ prédateurs et le ratio prédateurs ⁄ nids, ce qui indique que la pression de la prédation est aussi haute des deux côtés; et 3) le ratio nids ⁄ singletons et le ratio prédateurs ⁄ singletons, ce qui indique que la perte des nids ne variait pas entre les zones. Cependant, il y avait significativement plus de prédateurs, de nids et d'autruches par rapport aux nichées dans l'EA que dans la WA, ce qui indique une réussite de la reproduction significativement plus faible dans la EA. En utilisant la terminologie de la métapopulation, les autruches de la EA peuvent être considérées comme une population «puits» et celles de la WA comme une population «source», mais il faudrait faire des recherches de plus longue durée pour montrer si c'est bien le cas.

Rodent species abundance and diversity in Western Serengeti are evaluated and discussed in relation to different levels of conservation status [Unprotected Area (UA), Game Reserve (GR) and National Park (NP)] and broad site differences in human livelihood activities. A total of 2170 individuals, spread over 16 rodent species, were caught in a capture-mark-recapture study which covered both the dry and wet seasons. The more humid site (Tabora B) in the northern part of Serengeti had the highest diversity of rodents followed by the Mihale site at the western extension. The driest site at Robanda had the lowest overall species diversity. Diversity also varied between the three levels of conservation status whereby the UA had the least diversity while the NP, which enjoyed the highest level of conservation status, had the highest diversity of rodents. Unprotected Area and NP plots at Tabora B showed a rodent species similarity index of 40%; all the other paired plots scored over 50% similarity indices, suggesting that, within a site, species composition did not vary significantly between the three levels of conservation status. The Robanda site had the highest (56%) overall abundance of rodents; Mihale and Tabora B sites had about the same level of rodent abundance (20 and 24% respectively). For the Mihale site, Mastomys natalensis ranked first followed by Arvicanthis niloticus and Tatera robusta, each of which contained 40, 38 and 16%, respectively, of all individuals caught at the site. For the Robanda site, the figures were 66% A. niloticus, 22% M. natalensis and 9% T. robusta; while for the Tabora B site the scores were 37% M. natalensis, 18% T. robusta and 11% Lemniscomys barbarus. The differences in diversity, species composition and population abundance appear to result largely from physiognomic vegetation types, and habitat perturbations caused by livelihood activities in Western Serengeti.