Closing the cassava yield gap: An analysis from smallholder farms in East Africa

Cassava yields in Africa are small and it remains unclear which factors most limit yields. Using a series of farm surveys and on-farm and on-station trials in Uganda and western Kenya, we evaluated the importance of abiotic, biotic and associated crop management constraints for cassava production in a range of socio-economic settings as found in smallholder farms in the region. Average yields under farmer management were 8.6 t ha_1, but these were more than doubled to 20.8 t ha_1 by using improved crop establishment, improved genotypes and 100-22-83 kg ha_1 of single-nutrient N-P-K fertilizers. A farm survey revealed large yield differences between farms. Less endowed farmers harvested less cassava per unit area than better endowed farmers (difference of 5.9 and 9.7 t ha_1 in Kenya and Uganda, respectively); differences were associated with less access to labour, poorer soils, and premature harvesting by less endowed farmers. Analysis of 99 on-farm and 6 on-station trials showed that constraints for cassava production varied strongly between sites and years. Poor soil fertility, early water stress and sub-optimal weed management limited cassava production by 6.7, 5.4 and 5.0 t ha_1, respectively, when improved crop establishment and genotypes were used. Pests and diseases were relatively unimportant, while weed management was particularly important in farmer fields during a dry year in Kenya (yield gap of 11.6 t ha_1). The use of complementary analytical tools such as multiple regression and boundary line analysis revealed that many fields were affected by multiple and interacting production constraints. These should be addressed simultaneously if significant productivity improvements are to be achieved. This will be more difficult for less endowed than for better endowed farm households, since the former lack social and financial capital to improve management. (Résumé d'auteur