DissertationRodents, one of the diverse groups of mammals, are important contributors to biodiversity and good ecological indicators of habitat disturbance and their assessment is an important component of the study of forest ecosystems. Forested mountains in East Africa, which are home to diverse flora and fauna with sizeable endemics, include the Eastern Arc mountains (EAM) region and collectively with its adjacent lowland coastal forests recognized as one of the top 25 biodiversity “hotspots” worldwide. As conservation of any landscape requires knowledge of the available resources in the area, it is critical to study the rodents of the EAM due to the high degree of endemism in these mountain ranges and the rapid degradation of the remaining montane forests. However, research on rodents in sub-Saharan Africa has focused on pest species, and data on non-pest species is rare, however, needed for the implementation of effective conservation measures. The Ukaguru Mountains within the EAM are of particular interests because the forests of this range have received little/no attention on rodents and are facing alarming rate of anthropogenic disturbances such as grazing, firewood collection, tree/pole cutting and clearance of forest for agriculture which may impact on the diversity, population structure and demography of rodents. This study, therefore, aimed at understanding the ecology of the rodents, population and demographic parameters of the dominant species in disturbed and undisturbed habitats of the Ukaguru Mountains. There are two seasons in the landscape: wet and dry seasons; the wet season is October to May while the dry season is between June and September. Three methods were employed. In the first, the habitat association of the diversity and abundance of small mammals in the Ukaguru Mountains was assessed, by estimating the vegetation parameters using a Modified-Whittaker method and live-trapping of small mammals by Capture-Mark-Recapture (CMR) technique for three consecutive nights per month in farmland, disturbed and intact forests from June 2018 to February 2020. Two 70 m x 70 m grids, 600 m apart, were set in each habitat; each grid consisted of seven parallel lines, 10 m apart, and seven trapping stations per line, also 10 m apart (a total of 49 trapping stations per grid). The two grids in the disturbed forest were in proximity (50 m) to human settlements and farmlands and were characterized by cattle grazing, illegal hunting, tree cutting and wood collection. The two grids in the intact forests were devoid of human activities and were farther than 3 km away from disturbed forests. With the second method, the data derived from the CMR trapping in the disturbed and intact forests, was used to investigate the impact of anthropogenic forest disturbance on the abundance, breeding patterns and sex ratios and two demographic parameters: survival and maturation rate of the dominant rodent species i.e. Praomys delectorum. With the third method, removal trapping technique was used in the survey for rodents trapped for 13 months to determine the dietary composition of rodents in the landscape. Fifty snap (kill) traps were set separately in farm/fallow lands and forest habitats and trapping of rodents was conducted for three consecutive nights per month. Micro-histological analysis of the stomach contents of rodents collected from various habitats was carried out to determine the food categories. Small mammals were identified to species levels using relevant keys and confirmed by sequencing the mitochondrial cytochrome b gene. A total of 1196 individuals of small mammals comprising 13 species were captured from the CMR trapping in farmland, disturbed and intact forests; in the order of relative abundance, these are: P. delectorum (52.26%), Mus triton (21.07%), Mastomys natalensis (17.14%), Mus minutoides (2.93%), Lophuromys kilonzoi (2.26%), Crocidura hirta (1.59%), Beamys hindei (1.59%), Grammomys surdaster (0.59%), Graphiurus cf. raptor (0.17%), Lemniscomys rosalia (0.17%), Hylomyscus arcimontensis (0.08%), Pelomys fallax (0.08%) and Xerus sp. (0.08%). Mastomys natalensis, Mus triton and Praomys delectorum accounted for 90% of total captures. Mus triton and Mastomys natalensis were the most abundant in farmland with 46.75% and 42.77% of total captures respectively while P. delectorum accounted for 90% and 80% total captures in disturbed and intact forests respectively but the mean abundances in both habitats were not significantly different (p = 0.72). Species diversity and evenness in intact forest were much higher compared to disturbed forest. Principal component analysis (PCA) explained 87.70% of the variance with two factors. PCA 1 (67.47%) consists majorly of canopy cover, ground cover, shrubs and litter depth, with P. delectorum, B. hindei and L. kilonzoi loading positively. In the same component, herbs, Mastomys natalensis and Mus triton loaded negatively. Distance to human settlement contributed greatly to PCA 2 (20.23%) with G. cf. raptor, L. kilonzoi and C. hirta having greater than 0.5% loading. Mastomys natalensis and Mus triton associated with the herbaceous vegetation while P. delectorum correlated with litter depth and trees. The results suggest that vegetation and habitat disturbances are factors responsible for the observed diversity of small mammals in the Ukaguru Mountains. Analysis on sex ratios, population structure and breeding parameters were carried out on the dominant species i.e. P. delectorum in the two forest types. The population density of P. delectorum varied temporally with peaks attained in the wet seasons in both disturbed and intact forests. Population structure showed few juveniles in most months and sub- adults accounted for the population increases in the wet seasons in all habitats, high number of adults from the late wet season to the end of the dry season in disturbed forest but a greater number of adults in the wet season in intact forest. The sex ratios showed highly significant number of males in intact forest compared to disturbed forest (χ 2 = 10.71, df = 1, p = 0.001). The proportion of reproductively active individuals variedv temporally in both disturbed and intact forests while habitat disturbance showed no effect on the breeding activity of this species. Overall, these results suggest that rainfall and habitat disturbances as factors responsible for the varying sex ratios and population structure of P. delectorum in the landscape. Furthermore, the effects of habitat disturbance on abundance as well as two demographic traits: survival and maturation of P. delectorum were investigated. Abundance was estimated using the Minimal Number of animals Alive and subjected to a generalized linear mixed model. Survival and maturation probabilities were estimated using multi- event capture-recapture models using the software E-SURGE. There was no variation in abundance or maturation between intact and disturbed forests, but habitat type did affect survival. However, this effect was sex-dependent since female survival was higher in disturbed forests while male survival remained similar across the two forest types potentially due to differences in predation pressure or food availability between the two habitats. These results highlight the need for continuous demographic monitoring of P. delectorum in the EAM given that the increasing human population surrounding the landscape is leading to higher deforestation rates, and expansion of the pine plantation in the forest reserve. In the food analysis, Mastomys natalensis, Lemniscomys rosalia and Mus triton, seed/grain showed greater importance (> 35, > 55 and 60% respectively) whereas for Lophuromys kilonzoi (farm/fallow land), plant materials were more important (> 55%) than other categories. Seed/grains were more important (> 20%) in the diet P. delectorum, while plant materials were more important (> 45 %) in the diet of forest- dwelling L. kilonzoi. Food diversity for rodents in farm/fallow land was relatively low compared to forest dwelling rodents. Niche overlap between L. kilonzoi and P. delectorum was found to be relatively high. Given the interspersion of the forests with farmlands and their proximity, this may favour the establishment of farm generalists in the forests. The findings of the present study are of concern in conservation of this landscape and necessitate urgent conservation actions in this EAM block to limit the effects of anthropogenic activities in the landscape. As the human population surrounding the Ukaguru Mountains and their encroaching activities increase, continuous exploitation of the forests of this landscape should be discouraged as this was shown to modify the habitats of the small mammals, reducing their diversity and evenness and affecting other demographic parameters.