The focal question of this work is how the population dynamics of local dragonfly populations and metapopulations, and consequently the biodiversity in arid regions may be altered by a change of the distribution and quality of their habitats. For that purpose I developed a spatially explicit model for the distribution of Odonata in the landscape. The components of this model are habitat suitability models, local population models, rules for the dispersal, and a dynamic landscape model. After a test of the model with certain theoretical scenarios, this model will be applied to different scenarios describing the dynamic change of the landscape under current and future climatic conditions. Using this model will make it possible to simulate certain trends and scenarios in the water balance and the reaction of the Odonata fauna to these changes. This work is a part of the international project BIOTA (BIOdiversity Transect Analysis in Africa), which is funded by the German Federal Ministry of Education and Research. BIOTA has been initiated in 1999. Its aim is "a holistic scientific contribution towards sustainable use and conservation of the biodiversity of the African continent". My own research has been part of the subproject, which is entitled "Effects of anthropogenic changes on the diversity of Namibian Odonata: Modelling on different geographical scales". One of the aims of this subproject has been to develop the spatially explicit model mentioned above.