Abstract:
As the climate of the West African Sahel changes due to global warming, agricultural systems are facing stress due to the excess water induced by extreme rain events. Our first experiment, implemented in the suburbs of Ouagadougou (Burkina Faso), tested, the effects of different water levels above the soil surface (2-3 cm and 7-8 cm above ground) and their stagnation (1 - 3 and 3 - 6 days) on the growth, development and productivity of Obatampa maize cultivar, at three of its growth stages (stage at six leaves (V6), tasseling (VT) and the milky grain stage (R3)) during two years, under ambiant on-farm conditions. Hypoxia (1 to 3 days of soil saturation) and anoxia (4 to 6 days of soil saturation) at the tasseling stage reduced the grain yield by 53% and 54% respectively. At V6 stage, only anoxia caused 31% grain yield loss.Those grain yield losses were exponentially correlated with the stress days index (SDI) (R2 = 0.7 considering all the growth stages). The second experiment was carried out on the floodplain of Aniabisi (northern Ghana). The cumulative effect of the frequent waterlogging caused by precipitation, at different topographic locations (upslope, middle slope and downslope), associated water management techniques (presence or absence of bunds) and two planting dates for the Wang Data maize cultivar was tested. The results showed that the cumulative effect of waterlogging when it occurs from the vegetative stage to the end of the crop cycle can reduce the growth and productivity of maize drastically. Grain yield losses on downslope plots, represented 91% (2017) and 62% (2018). Under these conditions, a strong exponential relationship (R2 = 0.8) was established between the grain yield loss of Wang Data and the excess water stress factor (SEW30) in the vegetative stage. This factor could be useful as tool for assessing losses and damages associated with hazards due to excess rain events and also as tool in crop insurance scheme. By using the experiments for the calibration and validation of the EPIC model, it simulated well, the periods of waterlogging during these experiments. Indices generated from the simulated soil moisture appear to be a good predictor of grain yield decline during the tasselling stage of Obatampa. Nevertheless, the EPIC model was limited to simulating the reductions of yield, due to temporary waterlogging greater than 3 days at the V6 stage, occurring at the tasseling stage, or frequent from the vegetative stage. The improvement of the model should require the incorporation of different sensitivities to waterlogging at phenological stages.
