Freiburger Geographische Hefte, Vol. 70

Summary

In the time of rapid urban transformation, urban and periurban agriculture has evolved as an important livelihood strategy for millions of urban dwellers in sub-Saharan Africa. Especially in small and medium-sized cities, cultivated areas can be seen as an integral part of the urban patchwork. Yet, little is known about the spatial extent of agricultural production in and around these cities and how the spatial parameters of this complex patchwork change along the urban-rural continuum. The purpose of this study therefore was to enhance the understanding of spatio-temporal dynamics of urban and periurban agriculture along the urban-rural continuum with a special focus on small and medium-sized cities in Africa.

Two cities were selected for this study: 1) Moshi, located at the southern foothills of Mt. Kilimanjaro in northern Tanzania; and 2) Bamenda, the capital and largest city of the Northwest Region of Cameroon. In both cities, four transect polygons – 100 metres wide and up to 15 kilometres long – were laid out radially from the city centre, building the spatial framework for the data collection and analysis process. Within these transects, all agricultural land use was mapped and a representative number of households was interviewed (404 in Moshi and 480 in Bamenda). All data was digitised and geocoded, allowing for the spatial analysis of the two datasets. For data analysis, an inductive approach was chosen, meaning that all spatial classifications were derived from the raw data. An Urban-Rural Index (URI) was calculated based on building density and travel isochrones as the foundation of spatial analysis, ensuring that the process was not biased by the selection of conventional categories, such as urban, periurban or rural.

The results of this study revealed that almost all agricultural parameters showed significant correlations with the respective URI score. Even though correlations in land use data were usually larger than in household data, four characteristic patterns of spatial changes along the continuum could be identified. While building density, the availability of infrastructure, such as electricity and water, or formal flat/house ownership steadily decreased with decreasing URI values (Type A), the proportion of area under cultivation, mean patch sizes and the share of agriculturally active households increased (Type B). Even though these changes have hardly been quantified in previous studies, these patterns were expected, as most of these parameters are more or less evident parts of the definition of urban. However, it was found that spatial changes in other parameters were less predictable. While construction activity, crop diversity, and patch size variation were highest in periurban areas (Type C), the numbers on the duration of residence or land ownership were lowest in these areas of transition between the urban core and the rural hinterland (Type D).

The principal conclusion of this thesis was that agricultural parameters usually rather change gradually – forming an urban-rural continuum – than reflecting the artificial administrative boundaries that dichotomously differentiate between urban and rural areas. It could be shown that the periurban areas in between these extremes have distinct characteristics. Consequently, they need to be recognised as distinct parts of the urban-rural continuum, keeping in mind that these areas are likely to bear a significant potential for conflicts, but also great opportunities for sustainable urban development in the future.