In the Okavango Delta of Botswana, flood recession farming (molapo farming) contributes substantially to rural livelihoods by providing better yields than rain-fed dry land farming. However, molapo farming communities experience challenges associated with the unpredictability and unreliability of flooding that are paradoxically part of the dynamism essential for the farming practice. This makes it difficult for farmers to decide on when to plough and plant. These challenges might have contributed to  a reduction in the number of farmers practicing molapo farming. It is not clear if there is corresponding change in the spatial extent of the practice and if such a change may be related to variation in annual flood extent. We investigated the association between maximum inundated area and the size of molapo fields in three rural communities – Tubu, Shorobe and Xobe. Three time-steps representing different hydro-climatic phases (medium phase 1989-1991, dry phase 2000-2003, wet phase 2008-2011) constituted the study period. We used a geographic information system (GIS) and remote sensing approach to show flooding variability and changes in area under molapo farming over time under different flooding conditions in the Okavango Delta. Molapo fields were mapped using GIS and a global positioning system (GPS). The area under molapo farming for the time-steps 1989-1991 and 2000-2003 were calculated from aerial photographs. For the base time-step 2010, corners of the fenced fields were marked with a GPS receiver, from which polygons were generated. Using Landsat imagery we determined maximum flood extent by working with a combination of Tasseled Cap Analyses and spectral band thresholds to create wet-dry classifications. In each of the study sites, there was considerable variation over time in the location, size and distribution of molapo fields, as well as in the total area occupied. Importantly, the response was different for each of the three study sites. Shorobe showed the most dramatic decrease in the 2010/11 time-step, which represents a high water phase, while Tubu had a decrease in 2002/3 (low water phase) with strong resurgence in the area farmed in the subsequent high water phase (2010/11). Because Xobe’s fields are restricted by steep banks there was less variation in area farmed, but otherwise this community’s farming patterns appear to reflect those of Shorobe. The main reason for the decrease in hectarage in Shorobe and Xobe during the high flood phase is that much of the potential farming areas remained inundated for the entire farming season due to heavy and extended floods. Thus, the decline in molapo farming in the short term may in fact be attributed to the changing hydrology of the Delta.

 

Keywords: flood recession farming, geographic information system, hydro-climatic factors, Okavango Delta, remote sensing