Changes and Improvement in Water Management Practices in the Mwea Irrigation Scheme

n experiment to assess the impact of intermittent irrigation on Anopheles larval populations, rice yields and water use was conducted in the Mwea rice irrigation scheme in Kenya. Four water regimes including intermittent irrigation were tested in a complete randomized block experimental design. Intermittent irrigation was carried out on a weekly schedule, with flooded conditions from Saturday through Tuesday morning. Larval sampling at each plot was conducted every Monday and prior to draining of intermittently irrigated subplots on Tuesday. All the adult anopheline mosquitoes emerging from larvae collected in the experimental plots were identified as being An. arabiensis. By far the highest numbers of An. arabiensis 1st instar larvae were found in the intermittently irrigated subplots, indicating that the water regime provided the most attractive environment for egg laying. However, the ratio between the 4th and 1st instar larvae in the subplots was only 0.08, indicating very low survival rates. In contrast, the 4th/1st instar ratio for subplots with other water management regimes ranged between 0.27 and 0.68, suggesting a correspondingly higher survival than observed with intermittent irrigation. The total number of 4th instars was almost the same in the intermittently irrigated subplots and the irrigation system normally practised by the farmers. The failure to eliminate larval development up to the 4th instar in the former method was attributed to residual pools of water. Larval abundance fluctuated throughout the 12-week sampling period. The highest larval densities were recorded in the 3 weeks after transplanting the rice seedlings. Afterwards, larval numbers dropped dramatically as the height of rice plants increased. Rice yields at harvest did not show statistically significant differences among subplots with different water regimes. The average yield per hectare ranged from 4.8–5.3 metric tonnes. The average daily water percolation/seepage rate was 3.6 mm and did not significantly differ among different water management regimes. Further research is necessary to, among other things, determine whether rice yields could be increased by having flooded and drained intervals that were different from those used in this study. It would likewise be important to assess on a wider scale the feasibility of implementing intermittent irrigation with respect to farmer acceptance and required changes in irrigation system design and management.