Spatially explicit models for interacting populations in a changing landscape

Abstract

In arid regions like the ephemeral river catchments of western Namibia the population dynamics of dragonflies is strongly influenced by the dynamic of drying and re-flooding of the habitats. Global climate change and shortage of water as a resource will deteriorate the life conditions for dragonflies. To draw conclusions about the influences of the landscape on the biodiversity of dragonflies, the present thesis shows the development and application of a spatially explicit population dynamic model for Odonata in changing landscapes. The model consists of four parts. In the first submodel habitat suitability models are introduced, which describe the presence of the species in dependency of habitat parameters. The second model is a population dynamic model at the base of Leslie matrices. The matrix parameters are mainly determined by food supply in the habitat.

This enables the model to depict inter- and intraspecific competition effects. With this model the importance of colonisation sequences at the habitat is being analysed. The third submodel describes the implementation of theoretical and real landscapes by means of a grid based approach (= cellular automaton). Effects of landscape dynamics on the development of the species' populations are shown by the use of landscape metrics. The fourth submodel deals with the rules for the dispersal of different ecological types of Namibian dragonflies. All four submodels are integrated into the spatially explicit dispersal model. The effects of the scenarios on the modelled species communities are discussed. The model has shown amongst others that permanent habitats with a well-developed vegetation structure are of great importance for the less mobile resident species as a kind of refuge. The results of the model showed a good accordance with actual monitoring data for the modelled species