Adapting to climate change: agricultural system and household impacts in East Africa

Philip K. Thornton and Mario Herrero are ILRI authorsThe East African region exhibits considerable climatic and topographic variability. Much spatial and temporal variation in the response of different crops to climate change can thus be anticipated. In previous work we showed that a large part of this variation can be explained in terms of temperature and, to a lesser extent, water effects. Here, we summarise simulated yield response in two crops that are widely grown in the region, maize and beans, and investigate how the impacts of climate change might be addressed at two levels: the agricultural system and the household. Regionally, there are substantial between-country and within-system differences in maize and bean production responses projected to 2050. The arid-semiarid mixed crop-livestock systems are projected to see reductions in maize and bean production throughout most of the region to 2050. Yields of these crops in the tropical highland mixed systems are projected to increase, sometimes substantially. The humid–subhumid mixed systems show more varied yield responses through time and across space. Some within-country shifts in cropping away from the arid–semiarid systems to cooler, higher-elevation locations may be possible, but increased regional trade should be able to overcome the country-level production deficits in maize and beans caused by climate change to 2050, all other things being equal. For some places in the tropical highlands, maize and bean yield increases could have beneficial effects on household food security and income levels. In the other mixed systems, moderate yield losses can be expected to be offset by crop breeding and agronomic approaches in the coming decades, while more severe yield losses may necessitate changes in crop types, movement to more livestock-orientated production, or abandonment of cropping altogether. These production responses are indicative only, and their effects will be under-estimated because the methods used here have not accounted for increasing weather variability in the future or changes in the distribution and impacts of biotic and other abiotic stresses. These system-level shifts will take place in a context characterised by high population growth rates; the demand for food is projected to nearly triple by the middle of this century. Systems will have to intensify substantially in response, particularly in the better-endowed mixed systems in the region. For the more marginal areas, the variability in yield response, and the variability in households’ ability to adapt, suggest that, even given the limitations of this analysis, adaptation options need to be assessed at the level of the household and the local community, if research for development is to meet its poverty alleviation and food security targets in the face of global change