Biodiversity is quickly diminishing across the planet, primarily owing to human pressures. Protected areas are an essential tool for conserving biodiversity in response to increasing human pressures. However, their ecological effectiveness is contested and their capacity to resist human pressures differ. This dissertation aimed to assess the ecological effectiveness of different protection levels (from strict to less strictly protected: national park, game reserve, forest reserve, game-controlled area, and unprotected areas) in biodiversity (both mega diverse butterflies and mammals), maintaining habitat connectivity, and reducing anthropogenic threats at the wider landscape in the Katavi-Rukwa Ecosystem of southwestern Tanzania. To achieve this overarching goal, I employed an interdisciplinary approach.
First, I analyzed butterfly diversity and community composition patterns across protection levels in the Katavi-Rukwa Ecosystem. I found that species richness and abundance were highest in the game reserves and game-controlled areas, intermediate in the forest reserves, national park and unprotected areas. Species composition differed significantly among protection levels. Landscape heterogeneity, forest cover, and primary productivity influenced species composition. Land-use, burned areas, forest cover, and primary productivity explained the richness of species and functional traits. Game reserves hosted most indicator species.
Second, I modelled the spatial distribution of six large mammal target species (buffalo Syncerus caffer, elephant Loxodonta africana, giraffe Giraffa camelopardalis, hartebeest Alcelaphus buselaphus, topi Damaliscus korrigum, and zebra Equus burchellii) across environmental and protection gradients in the Katavi-Rukwa Ecosystem. Based on species-specific density surface models, I found relatively consistent effects of protection level and land-use variables on the spatial distribution of the target mammal species: relative densities were highest in the national park and game reserves, intermediate in forest reserves and game-controlled areas and lowest in un-protected areas. Beyond species-specific environmental predictors for relative densities, our results highlight consistent negative associations between relative densities of the target species and distance to cropland and avoidance of areas in proximity to houses.
Third, I examined temporal changes in land-use, population densities and distribution of six large mammal target species across protection levels between 1991 and 2018. During the surveyed period, cropland increased from 3.4 % to 9.6 % on unprotected land and from ≤0.05 % to <1 % on protected land. Wildlife densities of most, but not all target species declined across the entire landscape, yet the onset of the observed wildlife declines occurred several years before the onset of cropland expansion. Across protection levels, wildlife densities occurred at much greater densities in the national park and game reserves and lowest in the forest reserves, game-controlled areas and unprotected areas. Based on logistic regression models, target species preferred the national park over less strictly protection levels and areas distant to cropland. Because these analyses do not support a direct relationship between the timing of land-use change and wildlife population dynamics, other factors may account for the apparent ecosystem-wide decline in wildlife.
Fourth, I quantified land-use changes, modelled habitat suitability and connectivity of elephant over time across a large protected area network in southwestern Tanzania. Based on analyses of remotely-sensed data, cropland increased from 7% in 2000 to 13% in 2019, with an average expansion of 634 km2 per year. Based on ensemble models, distance from cropland influenced survey-specific habitat suitability for elephant the most. Despite cropland expansion, the locations of the modelled elephant corridors (n=10) remained similar throughout the survey period. According to ecological knowledge, nine of the modelled corridors were active, whereas one modelled corridor had been inactive since the 1970s. Based on circuit theory, I prioritize three corridors for protected area connectivity. Key indicators of corridor quality varied over time, whereas elephant movement through some corridors appears to have increased over time.
Overall, this dissertation underpins differences in ecological effectiveness of protected areas within one ecosystem. It highlights the need to utilize a landscape conservation approach to guide effective conservation across the entire protection gradient. It also suggests the need to enforcing land use plans and having alternative and sustainable forms for generating income from the land without impairing wildlife habitat.
Tropical forests worldwide support high biodiversity and contribute to the sustenance of local people’s livelihoods. However, the conservation and sustainability of these forests are threatened by land-use changes and a rapidly increasing human population. In this dissertation, I focused on the effects of land-use change on forest biodiversity in the rural landscapes of southwestern Ethiopia, against a backdrop of human population growth. These landscapes are being progressively degraded, encroached and fragmented as a result of different pressures, including the intensification of coffee production, farmland expansion, urbanization and a growing rural population. Understanding the drivers of biodiversity loss and the responses of biodiversity to such pressures is fundamental to direct conservation efforts in these tropical forests.
This dissertation aimed to characterize biodiversity patterns in the moist Afromontane forests of southwestern Ethiopia and to examine how biodiversity patterns are affected by land-use and land-use changes (mediated by coffee management intensity, landscape attributes and housing development) in a context of a rapidly growing rural population. To achieve this goal, I take an interdisciplinary approach where, first, I examined the effects of coffee management intensity on diversity patterns of woody plants and birds, spanning a gradient of site-level disturbance from nearly undisturbed forest interior to highly managed shade coffee forests. Results showed that specialized species of woody plants (forest specialists) and birds (forest specialists, insectivores and frugivores) were affected by coffee management intensity. The richness of forest specialist trees and the richness and/or abundance of insectivores, frugivores and forest specialist birds decrease with increasing levels of disturbance. Second, I investigated the effects of landscape context on woody plants, birds and mammals. Community composition and specialist species of woody plants and birds were sensitive to landscape context, where woody plants responded positively to gradients of edge-interior and birds to gradients of edge-interior and forest cover. Further results showed that a diverse mammal community, with 26 species, occurs at the forest edge of shade coffee forests and that the leopard, an apex predator in the region depended on large areas of natural forest. A closer examination of leopard activity patterns revealed a shift in the diel activity as a response to human disturbance inside the forest, further highlighting the importance of natural undisturbed forests for leopards in the region. Together, these findings demonstrate the value of low managed shade coffee forests for biodiversity, and importantly, emphasize the irreplaceable value of undisturbed natural forests for biodiversity. Third, I investigated the effects of prospective rural population growth (mediated by housing development) on the forest mammal community. Here, population growth was projected to negatively influence several mammal species, including the leopard. Housing development that encroached the forest entailed worse outcomes for biodiversity than a combination of prioritized development in already developed areas and coffee forest protection. Fourth, to understand the motivations behind high human fertility rates in the region, I examined the determinants of women fertility preferences, including their perceptions on social and biophysical stressors affecting local livelihoods such as food insecurity and environmental degradation. Fertility preferences were influenced by underlying social norms and mindsets, a perceived utilitarian value of children and male dominance within the household, and were only marginally affected by perceptions of social and biophysical stressors. Results further indicated a mismatch between the global discourse on the population-environment-food nexus and local perceptions of this issue by women. My findings suggest the need for new deliberative and culturally sensitive approaches that engage with pervasive social norms to slow down population growth.
Overall, this dissertation demonstrates the key value of moist Afromontane forests in southwestern Ethiopia for biodiversity conservation. It indicates the need to promote coffee management practices that reduce forest degradation and highlights that high priority should be given to the conservation of undisturbed natural forests. It also suggests the need to integrate conservation goals with housing development in landscape planning. A promising approach to achieve the above conservation priorities would be the creation of a Biosphere Reserve and to promote the ecological connectivity between the larger forest remnants in the region. Finally, this dissertation demonstrates the importance of placed-based holistic approaches in conservation that consider both proximate and distal drivers of forest biodiversity decline.