Analysing and exploring land use decisions by smallholder agrowetland households in rural areas of East Africa

Small wetlands become increasingly important for agricultural production of rural households in sub-Saharan Africa. Changes in wetland systems are event driven and a cumulative result of individual farmer’s decisions of land uses. The overall objective of this dissertation was to develop a method that takes account of individual decision-making to study the current uses of wetlands by smallholder rural farmers and how wetlands may develop in the future. Diverse methods that include rapid rural and participative approaches, wetland mapping and classification, farm typologies, identification of drivers of land use change and farmers’ decision-making were used to develop a decision tree. This model was then used for scenario analyses. Case study wetlands were surveyed within a total area of 484 km2. The wetlands were located in contrasting landscape units (lowland, midland, and highland) in Central Kenya and Laikipia plateau in Kenya and Usambara mountains and Pangani basin in northeastern Tanzania. Fifty-one (51) wetlands were characterised to identify and understand the drivers of diversity of wetlands and uses. Based on wetland type, shape, size, hydrological regime, soil fertility indicators, drainage patterns, use intensity, fertiliser, and market opportunity, wetland-units were categorised into five wetland cluster groups (WCGs). These groups were: 1. Largely unused narrow permanently flooded inland valleys; 2. Extensively used wide permanently flooded inland valleys and highland floodplains; 3. Seasonally flooded and moderately used wide inland valleys and lowland floodplains; 4. Completely drained and intensively used wide inland valleys and highland floodplains; and 5. narrow valleys drained for continuous high-value crops production. Case study farms from four WGCs (2-5) were characterised to identify the diversity of production systems and drivers of wetland use decisions by smallholder farmers in contrasting rural areas. Using a combination of production systems (livestock ownership, type, and its integration with crops), land resource (upland, wetland, and their combination), and production objectives (subsistence, cash, and cultural), households were grouped into 12 Farm Types. Based on these two typologies and their relationships with environmental and socio-economic drivers of wetland use, a decision tree model framework was developed to represent the diversity of farmers’ decision-making and analyse the effects of such diversity on current land uses in the wetlands. The framework was then used to explore changes in land use in scenario-driven analysis. Increasing land scarcity coupled with improved markets could increase the dependency of household’s livelihood on cropland in the wetland up to 100% (e.g. for FT2). This increase would also decrease pastoralism in semi-arid areas (87%) as well as stimulate livestock integration by crop-based farms. Land use intensification across wetlands and specification in midland valleys could accompany such changes. Furthermore, land use displacement from traditional floodplain to rangeland grazing is an unavoidable consequence of land use intensification. Agricultural use of small wetlands offers opportunities to diversify rural livelihood systems. However, wetland farming is challenged by various hazards and shocks such as conflicts, abiotic, biotic, and socio-economic factors that constrain crop production. Small wetland agricultural systems are complex and characterised by interactions between heterogeneous human decision-makers (i.e. farmers) and their biophysical environment (wetland systems). Changes in these systems are event driven and cumulatively results from individual farmer’s decisions of land use in response to endogenous and exogenous drivers. Reconciling livelihood benefits with sustainable land use and natural resource conservation in rural areas is thus a complex and challenging social task that requires the development of an adaptive co-management process with the active participation of all stakeholders. Key words: Farm(er) typologies, farmers’ decision-making, floodplain, households’ production systems, human-environment interactions, inland valley, Kenya, land use decisions, land use model, rural livelihood, scenario, simulation, Tanzania, uplands, wetland, wetland typology.