| Over decades, food production in East Africa has been affected by a changing climate, limited use of fertilizers and pest control, inadequate food storage facilities and complex marketing channels that together have led to malnutrition, hunger, and poverty. The six most important food crops feeding the region include cassava, maize, plantains, sweet potatoes, potatoes and paddy rice. Of all these crops, relatively little is known about how climate influences sweet potato growth, development, and yield. Deterministic simulation models for sweet potatoes exist but are relatively young or still in development. Relevant data for climate impact assessments are scarce: detailed agronomic data for sweet potato cultivars grown in East Africa are limited; representative high-quality climate data for the region are scarce, and soils data is only available at coarse spatial resolution. The major objective of this research was therefore to assess the impact of climate variability and change on sweet potato production in East Africa. This was achieved by: (i) developing a modeling framework for use in a deterministic sweet potato crop model, SPOTCOMS, for East Africa; (ii) analyzing trends of historical climate and sweet potato root yields for the period 1980-2009; (iii) developing local climate change scenarios for East Africa for the current time slice 2010s, near future 2030s, mid-future 2050s and distant future 2050s using two representative concentration pathways 4.5 and 8.5 for four Global Climate Models, CSIRO, MIROC5, MRICGC3-M and NorESM-1; and (iv) estimating the impact of projected future climate change on sweet potato production using SPOTCOMS model. Crop coefficients where determined from field trials for four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and SPK004. Results from the calibration and evaluation of SPOTCOMS model gave an index of agreement (IA) of 0.94 and 0.7, a modeling efficiency of 0.9 and 0.31, and a mean bias error of 1.16 t/ha and 0.5 t/ha respectively. Trend analysis indicated that East Africa had warmed on average by 1.50C, the rainfall for the February-June season had declined by more than 60 mm while rainfall for the August -- December season had increased for most parts of East Africa by more than 50 mm over the past 30 years. The results of future climate projections from Global Climate Models showed mixed results for precipitation and more distinct results for temperatures. Temperatures in the region were projected to rise by 0.8°C, 1.2°C and more than 3°C in the 2030s, 2050s, 2070s respectively and precipitation is projected to consist of more increases in the short rainfall intensity than the long rains for all the three future time slices. The sensitivity analysis showed that SPOTCOMS was sensitive to increase in precipitation and temperature for all the four sweet potato cultivars, NASPOT 1, NASPOT 10 0, NASPOT 11 and SPK004. The projected increase in sweet potato yield in the region coincides with areas that will experience increases in precipitation and temperature. Models with the larger radiative forcing of RCP8.5 showed an overall higher increase in precipitation, temperatures and therefore higher increases in sweet potato yield. All the four cultivars (NASPOT 1, NASPOT 10 0, SPK004 and NASPOT 11) showed similar spatial distribution of yields but SPK004 had lower yields for both historical and projected future periods. Results from this study are useful to all stakeholders interested in sweet potato production in East Africa and the rest of the tropics. |
