Adaptive management of irrigated rice in the changing environments of the Sahel

Key words: Alternate wetting and drying, Climate change adaptation, Crop growth simulation models, Genotype × environment interaction, N use efficiency, Oryza sativa L., Phenology, Sahelian irrigation schemes, Sowing date, Spikelet sterility, Temperature increase, Water productivity, Weed control. In the vulnerable environment of the Sahel with its erratic rainfall pattern, irrigated rice production is of major importance. To aid Sahelian rice farmers to sustain irrigated rice production, this study explores management options. It includes field experiments performed at two typical Sahelian sites and simulation studies using crop growth simulation models. This thesis provides evidence that it is possible to use less irrigation water while maintaining rice production, thus increasing water productivity. The effects of a temperature increase on the growing cycle and spikelet sterility of new rice varieties in interaction with different sowing dates is quantified. The simulation results show that the sowing window will be restricted and that the cultivar choice may alter; together they will remain the most important determinants of rice production in the coming decades. In Chapter 2, field experiments involving three water saving regimes using combinations of alternate wetting and drying (AWD) and flooding and a fully flooded control show that between 480 and 1060 mm of irrigation water was used in the water saving treatments compared with 800 to 1490 mm in the flooded rice treatment. Water productivity of the water-saving treatments was higher than of the flooded control, and yields ranged between 141 and 56% of the control. When weeds were controlled, crop yields obtained with a combination of AWD and flooding were comparable with those obtained in fully flooded plots receiving the same weed management. In Ndiaye, agronomic N use efficiency was smaller in the AWD treatments compared with fully flooded conditions. An irrigation regime for rice that starts as conventional (flooded), and then changes to AWD can save water with little or no yield loss, while maintaining low weed pressure and efficient use of N. To assess genotype adaptability, in Chapter 3 the results of experiments involving five genotypes, sown on 15 consecutive dates are presented. Yield (0-12t ha–1) and crop cycle duration (117-190 days) varied with sowing date, genotype and site. Rice yield was very sensitive to sowing date and the associated temperature regimes. Spikelet sterility due to cold stress (T oC) was observed when the crops were sown betweenAugust and October,and heat stress (T > 35oC) resulted in spikelet sterility for sowingin April and May. For the simulation studies of Chapter 4, experimental data were used to calibrate both the DSSAT and ORYZA2000 models. Original genetic coefficients of DSSAT did not simulate phenology well, while genetic coefficients that did, resulted in lower than observed yields. Simulations by ORYZA_S and ORYZA2000 resulted in an increase in simulation error at sowing dates in the last three months of the year. The results show that local calibration at the same sowing date is needed. In the African Sahel, a temperature increase of between 1.8 and 4.7oC is predicted by 2080. Simulations by an improved and validated version of ORYZA2000 presented in Chapter 5 show that rice crop cycle length will decrease by 10‒30 days. The results suggest that with projected temperature changes, timing of sowing and consequently of the risk for crop loss due to sterility will remain the major determinant of rice yield. There is an urgent need for heat tolerant rice varieties. Without adaptation, cropping calendars will change, in the worst case scenario only a single crop will be possible. I conclude by suggesting viable options for adaptive management of irrigated rice in the changing environments of the Sahel to sustain production in the 21stcentury.