Subproject E03

Dr. Helmut Dalitz, Bielefeld University, Faculty of Biology, Dept. of Ecology, Universitätsstr.25, 33615 Bielefeld, Germany

Regeneration of tropical-montane tree species: dynamics of feedback processes in time and space and their modelling

Summary
In this section we want to examine the spatial distribution of tree species along a gradient of fragmentation. Light availability under the canopy, input of nutrients from the atmosphere and input of nutrients on the forest floor will be investigated. The idea is that variable conditions of water and nutrient supply as well as variations in light conditions are responsible for the regeneration of tree species. Another aspect is the preparation a plant inventory of the Kakamega forest in cooperation with other groups, presenning all plants in a database with digital pictures. The self-generated software product "Visual Plants" will be used. It has already been used in similar projects in Costa Rica and Ecuador.

Methods
Spatial distribution of plants: all trees with more than 5 cm BH-diameter will be recorder on 400 m2 plots with reference to their spatial arangement. Crown architecture will be investigated with hemispherical photographs (HemiView, DeltaT), and light conditions will be monitored through measurements of photosynthetic active radiation (SunScan-system, DeltaT).
Nutrient input: Bulk precipitation of rain and of canopy throughfall will be collected and analysed for all relevant cations. Results are to be related to canopy structure to find possible correlations between nutrient fluxes and forest structure. The collections of precipitation and canopy throughfall should be carried out for two years.
Soil chemistry: Water extractable cations and total content of soil horizons will be investigated with referecne to cation content and temporal variation at different sites within the plots.
Inventory of plants: Flatbed scans, digitized scans, and digitzed herbarium specimen will be collected in the software "Visual Plants". This will allow the presentation of plants pictures on a computer monitor and searching and identifying plants with only a few known vegetative characters.