Spatial and temporal variation in yield of rainfed lowland rice in inland valley as affected by fertilizer application and bunding in North-West Benin

Rainfed lowlands in inland valleys present a high potential for rice (Oryza spp.) production in West Africa. However, rice yield in the lowlands is, in general, low due to various constraints such as poor soil fertility, drought, iron (Fe) toxicity, and poor crop management practices. The objective of this study was to evaluate the efficiency of bunding and fertilizer application for improving rice productivity in the two toposequential positions (upslope and downslope positions) in an inland valley. The experiment was conducted in a researcher-managed on-farm trial located in the North West of Benin Republic over four wet seasons (2007–2010). In addition to the toposequential positions, the experiment included two treatments: (1) fertilizer inputs: no fertilizer and fertilizer (60kgN and 40kgPha−1) and (2) water control: with and without bunds (without drainage). Effect of bunding on rice yield was consistent across four seasons, two fertilizer application treatments, and two toposequential positions, and bunding increased rice yield by 29%. Rice yield was higher in upslope than in downslope, except for 2010. In upslope, high ponded water level in 2010, due to higher rainfall than in other years, increased Fe toxicity, resulting in lower yield in 2010. Year-to-year variation in yield response to fertilizer application was related to differences in N uptake at 38 days after sowing. When Fe concentration in leaves was lower, N uptake was higher. Thus, Fe toxicity at early stage could contribute to yield response to fertilizer application. These results indicate that while bunding is essential for improving rice productivity, improving yield response to fertilizer application requires drainage systems in the areas where risk for Fe toxicity is high.