Date Published: March 7, 2018
Publisher: Public Library of Science
Author(s): Confidence Duku, Sander J. Zwart, Lars Hein, Prasanta K. Subudhi.
In recent decades, there have been substantial increases in crop production in sub-Saharan Africa (SSA) as a result of higher yields, increased cropping intensity, expansion of irrigated cropping systems, and rainfed cropland expansion. Yet, to date much of the research focus of the impact of climate change on crop production in the coming decades has been on crop yield responses. In this study, we analyse the impact of climate change on the potential for increasing rainfed cropping intensity through sequential cropping and irrigation expansion in central Benin. Our approach combines hydrological modelling and scenario analysis involving two Representative Concentration Pathways (RCPs), two water-use scenarios for the watershed based on the Shared Socioeconomic Pathways (SSPs), and environmental water requirements leading to sustained streamflow. Our analyses show that in Benin, warmer temperatures will severely limit crop production increases achieved through the expansion of sequential cropping. Depending on the climate change scenario, between 50% and 95% of cultivated areas that can currently support sequential cropping or will need to revert to single cropping. The results also show that the irrigation potential of the watershed will be at least halved by mid-century in all scenario combinations. Given the urgent need to increase crop production to meet the demands of a growing population in SSA, our study outlines challenges and the need for planned development that need to be overcome to improve food security in the coming decades.
Increasing crop production in sub-Saharan Africa (SSA) is urgently needed. The population of the region is projected to double by 2050 compared to 2015 . About 97% of current cropland area is under rainfed cultivation  and current productivity levels for major food crops, which are the lowest in the world, are inadequate to meet the projected food demand . To meet the food demand of a growing population several options for increasing crop production must be harnessed. These include amongst others crop intensification in rainfed systems to produce higher yields and/or increased cropping frequency, expansion of irrigated area and rainfed cropland expansion. Over the past decades, higher yields, increased cropping frequency (i.e. sequential cropping and intercropping) and cropland expansion have accounted for an estimated 38%, 31% and 31% respectively of the recorded increases in crop production in SSA .
In this study, we used multi-model ensemble climate data projected by 17 GCMs for simulation. In using multi-model ensemble data, extreme climatic values projected by individual GCMs are smoothed. To maintain the climate extremes, climate data from individual GCMs with divergent climatic trajectories can be used. However, in West Africa, where these projections are highly uncertain especially for precipitation, and there is a large spread across various GCMs [8, 68], using climate data projected by individual GCMs may not address the structural uncertainties and independent simulation errors in these models. In these situations, multi-model ensemble data are highly recommended and have been widely used . The multi-model ensemble approach have been reported to outperform individual ensemble members in hindcasting studies and thus provide an improved ‘best estimate’ forecast . Lambert et al.  showed that simulations of multi-model ensemble data of precipitation, temperature, and pressure of current climate are generally closer to observed distributions, as measured by mean squared differences, correlations, and variance ratios, than the results of any particular model. Yira et al.  show that as a result of the high uncertainties in precipitation projections in West Africa, a larger ensemble of climate projections is required to estimate the impacts of climate change on water resources accurately. Giorig et al.  suggest that as a result of these uncertainties, a minimum of four to five multi-model ensemble is needed to obtain robust regional precipitation change estimates.
In this study, we have shown that, in addition to crop yield responses, climate change will affect other options that have been used to increase crop production in recent decades in SSA i.e. rainfed sequential cropping, rainfed cropland expansion and irrigation expansion. Currently, about 41% of cultivated areas in the Upper Ouémé watershed are either used for rainfed sequential cropping or can support it. However, by 2050 this will decrease to between 2% and 16% depending on the climate change scenario. Farmers will therefore have to shift to single cropping systems or adopt improved agronomic practices including drought-resistant and short-cycle cultivars. Farmers may also be driven to expand to hitherto uncultivated areas to make up for lost opportunities to increase crop production. In the Upper Ouémé watershed, over 90% of the land area that can support rainfed cropping is not currently under cultivation and largely consists of forest and woodland savannah. This situation is unlike other parts of Benin where the availability of currently uncultivated land is much lower. A large part of these potential arable lands will still be able to support rainfed sequential cropping in the coming decades despite the loss of between 15 and 30 growing days due to their relatively higher soil moisture storage. If these areas are to be used for rainfed sequential cropping, then fallow periods will have to be shortened or lost completely and improved soil and nutrient management will be needed to increase productivity. However, in a previous paper, Duku et al. , we showed that the conversion of forested and woodland savannah areas to cropland will have negative feedbacks on water availability for irrigation. In the present study, we have shown that even if there is no change in forest cover, at least 50% of irrigation potential will be lost in the coming decades due to climate change. Forest and woodland areas, therefore, will be needed to regulate water flows and increase dry season streamflow in addition to the provision of other ecosystem services. Our paper shows the importance of using an integrated approach to rural development planning, where climate change can be expected to have multiple, major implications on cropping systems and resilience for climate change depends upon maintaining overall landscape integrity including areas that regulate water flows.