Consequences of the worldwide biodiversity decline for ecosystem processes are rarely understood. Important ecosystem processes are mutualistic plant-animal-interactions such as seed dispersal. Especially in tropical forests more than 90% of all plants rely on animals for the dispersal of their seeds. A loss of animal dispersers might lead to a break-down of seed dispersal processes with long-lasting consequences for forest regeneration. The aim of BIOTA subproject E11 is to investigate the influence of forest fragmentation and disturbance on avian diversity and its consequences for seed dispersal and regeneration processes in the heavily fragmented and disturbed Kakamega Forest, western Kenya. During the first project phase, the research has been focusing on 1. the monitoring of animal dispersers, seedlings and fruiting trees and 2. seed dispersal and regeneration processes in selected tree species (Ficus thonningii, Prunus africana).

In the second project phase we will focus on seed dispersal processes in P. africana and F. thonningii along a disturbance gradient in different East African forests. This approach is crucial for the identification of an optimum cost-benefit relation between biodiversity conservation and its sustainable use. The contribution of dispersal and pollination to gene flow of P. africana will be studied as the genetic diversity is an important monitoring tool for assessing human disturbance. On a landcsape level we will assess the importance of isolated trees as regeneration centres in the farmland adjacent to Kakamega Forest. A study on the regeneration dynamics of secondary forests and plantations in Kakamega will be carried out in order to evaluate whether they facilitate forest restoration and can be used in reforestation programs. A community-based bird monitoring in Kakamega Forest will be carried out in a joint project with our counterparts NATURE Kenya within their IBA-program.

Our research activities in the second BIOTA phase are dealing with ecosystem processes and sustainability on the interface of seed dispersal, genetic diversity as an indicator system for disturbance, landscape approaches to seed dispersal and forest restoration. All our studies are oriented towards important questions in sustainable conservation management and aim to derive management recommendations. Although each study stands on its own, they are closely linked to each other. As seed dispersal is inextricably linked to many other research fields from zoology to botany, close collaboration between our subproject and other subprojects is essential. However, seed dispersal processes are also closely linked to socio-economic questions. In most of the present tropical forests, exclusive conservation is not feasible and desirable. Due to high population densities especially in African countries, forest ecosystems are extensively used by local communities for a variety of purposes such as fuelwood, building material, medicinal plants etc. Consequently, natural regeneration of these heavily used ecosystems is important for the regeneration of important tree species, i.e. species used by the local communities. Seed dispersal processes contributing to regeneration processes, genetic variability and plant diversity, are, therefore, crucial for sustainable use and for the countries economic stability. Consequently, our research will have a close link to the socio-economic subprojects. Together we will be able to identify a point of maximum maintenance of biological diversity and ecosystem function and of profit for the local communities as sustainability is important for the economy and stability of the local people.

Monitoring biodiversity
The monitoring activities which have been intitiated in Kakamega Forest during the first project phase will be continued in the second phase in combination with extensive capacity building and in a joint project with NATURE Kenya in the framework of the IBA-program of BIRDLIFE International.

Understanding dispersal and regeneration processes: The main focus of the second phase will be on the process level and its implications for sustainable use of forest ecosystems. The study of processes of seed dispersal and regeneration will be extended from a species and community level to a genetic and landscape level.

1. Ecosystem function and sustainability: dispersal processes as model approach
   The consequences of forest disturbance for seed dispersal and regeneration processes will be studied along a disturbance gradient in different East African forests. These studies will be carried out both, on a species and a community level (WP1).
   
2. Genetic diversity as indicator for forest disturbance
   Forest fragmentation and disturbance do not only affect biodiversity on a species level but have been shown to influence gene flow and the genetic diversity of animal and plant populations as well with serious consequences for the long-term maintenance plant communities. We will be studying gene flow through seed dispersal and pollination of an insect-pollinated and bird-dispersed tree species by measuring and interpreting population genetic variation in small and fragmented populations using microsatellite markers (WP2).
3. Seed dispersal and regeneration on the landscape scale
   Many tropical tree species are particularly vulnerable to habitat alteration due to their low densities. If dispersal of pollen and seeds is disrupted by fragmentation, the remnant forest plant populations will become genetically and demographically independent. Habitat connectivity might be enhanced through tree patches in the surrounding matrix which may provide avenues for exchange of pollen, seeds and consequently genes between patches and may also contain tree individuals that can contribute to the local dynamics of adjoining patches. Furthermore, free-standing remnant trees might be important foci of seed deposition. We will investigate the importance of tree patches in pastures as possible stepping stones for frugivores and habitat connectivity (WP3).
4. Forest Restoration
   Kakamega Forest has a long history of forest disturbance and use of forest products by the local communities. Consequently, it does not include primary undisturbed forest, but mainly large tracts of disturbed, recovering secondary forest consisting mainly of exotic Guava (Psidium guajava) thickets and also partly of plantations, both indigenous and exotic ones. Guava or plantation trees with fruits attractive to frugivorous animals might act as foci of seed deposition and might therefore facilitate forest restoration. We will be studying frugivore visitation, dispersal and regeneration in differently degraded forests and plantations (WP4).

BIOTA research and results of the first phase
The consequences of forest fragmentation and disturbance for the diversity of dispersing animals, for fruit diversity and for dispersal and regeneration processes has been investigated by BIOTA subproject E11 in a comparative approach in Kakamega Forest, western Kenya, and its five peripheral fragments. Our baseline biodiversity monitoring data over one year indicate significant lower species numbers of both, forest specialist and frugivorous bird species, in forest fragments and more disturbed areas under weak protection through the FD as compared to continuous forest and less disturbed areas managed by KWS. Our baseline monitoring data on fruit diversity indicate general low fruit abundance with fruiting peaks in February to April. Mainly trees of the genus Ficus are found be be major fruit resources for frugivores as found for many other tropical forests. When surveying regeneration of tree seedlings as the first immediate result of dispersal processes taking place, our results indicate significantly lower species numbers of primary forest trees in fragments and on more disturbed sites managed by FD. Our results strongly suggest that not only fragmentation affects biodiversity and processes but that differential management regimes of KWS and FD appear to lead to different habitat quality and disturbance levels. To get a better understanding of the disturbance variable, we have conducted a quantitative disturbance survey at Kakamega Forest in cooperation with subprojects E03 and E10 (Bleher et al., subm. manuscript). Furthermore, a study on the history of forest use in Kakamega Forest has been initiated and successfully completed in cooperation of the subprojects E01, E02, E03, E08, E10, E11 and E12 (Mitchell, in press). The consequences of forest fragmentation and disturbance for dispersal and regeneration processes have been studied in Pruns africana, an endangered medicinally-used CITES species, and in Ficus thonningii, an important fruit-producing keystone trees for frugivores.

During the first phase we built up the infrastructure at our study site (establishment of BDO's, of 9 study plots etc.), strengthened our relationships to existing (NMK, ICIPE) and new counterparts (NATURE Kenya) and to other BIOTA subprojects and enhanced capacity building through involving three Kenyan MSc-students and two local field assistents in our research projects.