No abstract

This account considers progress in research on small rodent ecology since about 1970. While some ecological survey has been undertaken throughout much of the continent, the main research thrust has involved the detailed examination of population processes in a range of habitats from deserts to tropical rain forests. The deserts and semi‐arid regions support a range of species with some better adapted to these harsh conditions than others. The adaptations manifest themselves in different life history strategies e.g., opportunistic with high turnover species, or low turnover‐low reproduction species, in their physiology e.g., adaption to water retention, and in their behaviour e.g., avoiding heat of day. Long term studies show large numerical fluctuations with a quick response by some species to favourable conditions caused by relatively high precipitation. In the savanna and grasslands, studies have centred on population dynamics, the relationship between food and breeding, reproductive and life history strategies and the determination of niche. These have been approached through the study of groups of species in particular localities as well as by the examination of individual species. In these habitats populations are more stable and less erratic than in semi‐arid regions. Perturbations, particularly fire, can have various effects on the numbers and species composition of the disturbed area. In the tropical rain forest population fluctuations are modest and may take up to 2–3 years to pass from a peak to a trough and back again. Breeding typically extends over much of the year with, apparently, the greatest number of rodents occurring on the ground surface. Bush and temporary farms derived as a result of forest modification support more animals (and often more species) than forest. The niches occupied by forest rodents are poorly understood and very little is known of the ecology within the canopy. In most localities a small number of species are abundant and the remainder much less frequent.

The population dynamics of the Grass rat Arvicanthis niloticus were studied in relation to their diet and the climate in the Nakuru district of Kenya between September 1980 and May 1982. Numbers in the study area were low at the end of 1980 with groups of animals isolated in patches of grassy habitat. Local populations declined and some became extinct during the first half of 1981. Breeding began in April, one month after the start of the rains, but with a small initial population and high juvenile mortality, numbers did not increase significantly until September, when animals born earlier in the season were old enough to reproduce. This second generation reproduction was probably very important to the survival of the population over the following dry season when adult mortality was estimated at 15% per month. Reproduction was facilitated by a nutritious diet, consisting mainly of young grasses early in the season and seeds later on. Seed production was itself related to rainfall and as that declined towards the end of 1981, breeding ceased. Arvicanthis were considerably more numerous then than one year previously, with the difference being ascribed to the longer rainy season experienced in 1981 than 1980. The density of animals in favourable habitats increased with immigration from less suitable areas between January and March 1982, and then declined until the end of the study. Breeding resumed in April, one month after the start of the rains. Peak numbers of Arvicanthis were recorded in the middle of the dry season, when few crops were in the ground. They would only become a serious pest to agriculture if they were still abundant at the start of the breeding season some months later. Expansion then could result in very high numbers by harvest time, in August or September.

With 24 figures in the text) 1148 small rodents representing 16 species were caught in Mayanja Forest, Uganda from regular trappings between September, 1967 and August, 1968 inclusive. These were restricted to five compartments which had been subjected to extensive felling in 1962-65 leaving at the time of the survey a dense herbage and scrub. Rain fell in all months of the year with the wettest periods from September to November and February to May.The most frequently caught species on the ground (Praomys morio and Lophuromysflavopunctatus) showed small differences in their responses to three baits (peanut butter, banana/ wheat flour, powdered rat diet), the only exception beinghphuromys which wasnot attracted to the rat diet. O f f the ground, where peanut butter was the bait, there was little or no decline in the catch rate over the first three nights trapping in one situation for Hylomyscus stella, Grammmys dolichurus and Thamnomys rutilans and only. at most, a small drop after six nights. The catch of Praomys fell quite steeply after the first night and thereafter more slowly to the sixth.Trappings revealed differences between species occurring on and off the ground. Lophuromys flavopunctatus and L. sikapusi, Hybomys univittatus, Malacomys longipes, Aethomys kaiseri, Lemniscomys striatus, Mus minutoides and M. triton were almost entirely obtained on the ground; H. stella, T. rutilans. G. dolichurus and Graphiurus murinus off the ground and P. morio and Oenomys hypoxanthus spent an appreciable amount of time in both situations. P. morio, L. flavopunctatus and H . stella were apparently the most numerous species. Of the remaining 13 species the maximum number caught for any one was 41 and for each of seven of these species less than ten were caught. Detailed information is provided on the species of plants, heights and nature of habitat from which animals were trapped above ground. No large scale fluctuations in the numbers of rodents were recorded throughout the year.Observations on their ecology are generally consistent with those from elsewhere in Africa.The recent man-induced modifications of the habitat enriched the fauna without eliminating any of the more typical forest species that might be expected.The condition of reproductive tracts, body weights and extent of molar tooth wear have supplied information on the time of reproduction and population structure of the species obtained. The reproductive data include litter size. abundance of placentation scars, testes weights and occurrence of lactation. In Lophuromys flavopunctatus there are apparently two main breeding seasons coinciding with the rains whilst in Praomys breeding appears to be more protracted. The limited data on the remaining species indicates breeding to take place at various times of the year with indications of coincidence in some species with the wet seasons. L.flavopunctatus, H. stella and P. morio bred in captivity. In development, pigmentation of the skin was followed by appearance of dorsal hair, then belly hair and finally openi...