Open Access

Ronald K. Mulwa

• The impacts of human activities, notably the conversion of tropical forests into farmland habitat, has profound impacts on biological diversity and ecosystem functions (Millennium Ecosystem Assessment 2005). It is widely debated to what extent human modified landscapes can maintain tropical biodiversity and their ecosystem functionality (e.g. Waltert et al. 2004, Sekercioglu et al. 2007). In this thesis, I have used a huge and temporarily replicated dataset to assess the value of different habitat types differing in land-use intensities for bird communities in tropical East Africa. I investigated bird abundance and species richness along a forest-farmland habitat gradient and assessed spatial and temporal fluctuations of bird assemblages and their food resources. I could show that forest and farmland habitats harbor distinct bird communities. Moreover, the protection of natural forests merits the highest priority for conserving the high diversity of forest-dependent bird species. My study, however, also shows that farmland habitats in the proximity of natural forest can support a high bird diversity. High bird diversity in tropical farmlands depends on a high structural complexity, such as in small-scale subsistence farmlands. From my findings, I conclude that the conversion of forest to farmland leads to substantial losses in bird diversity, in particular in specialized feeding guilds such as insectivores, while the conversion of structurally heterogeneous subsistence farmlands to sugarcane plantation causes erosion of bird diversity in agricultural ecosystems. Both findings are important for conservation planning in times when tropical forests and agroecosystems are under constantly high pressure due to increasing human population numbers and global demands for biofuel crops (Gibbs et al. 2008). From an ecosystem function perspective, my study demonstrates the potential of agroecosystems in supporting important ecosystem functions, such as seed dispersal by frugivorous birds and pest control by insectivorous birds. I could show that bird abundances in both frugivorous and insectivorous guilds were strongly predicted by their respective food resources, implying that seasonal shifts in fruit and invertebrate abundance at Kakamega forest and surrounding farmlands affect community dynamics and appear to influence local movement patterns of birds. The most interesting finding of this study was that feeding guilds responded idiosyncratically to resource fluctuations. Frugivore richness fluctuated asynchronously in forest and farmland habitats, suggesting foraging movements and fruit tracking across habitat borders. In contrast, I found that insectivores fluctuated synchronously in the two habitat types, suggesting a lack of inter-habitat movements. I therefore predict that insectivorous bird communities in this forest-farmland landscape may be more susceptible to the combined effects of land-use and climate change, due to their narrow habitat niche and limited capacity to track their resources. The fact that a number of bird species regularly moved across the landscape mosaic in my study system implies that birds are able to provide long-distance seed dispersal across habitat borders. Thus, birds may enhance forest regeneration in human-modified landscapes, such as those in most parts of tropical Africa, given that forest remnants are protected within an agricultural habitat matrix. In order to effectively conserve tropical biodiversity within forest-farmland mosaics, this study advocates for conservation strategies that go beyond forest protection and explicitly integrate farmlands into forest management plans and policies. This should emphasize the retention of keystone habitat elements within tropical farmland landscapes, such as indigenous trees, forest galleries and hedgerows, whose presence enhance species diversity. Such grassroot-level approaches can be operationalized for instance through providing incentives to farmers to maintain their traditional subsistence land-use practices and through community-based livelihood projects aiming at enhancing local habitat heterogeneity and inter-habitat connectivity.