Understanding the productivity of cassava in West Africa

Drought stress and sub-optimal soil fertility management are major constraints to crop production in general and to cassava (<em>Manihot esculenta</em> Crantz) in particular in the rain-fed cropping systems in West Africa. Cassava is an important source of calories for millions of smallholder households in sub-Sahara Africa. The prime aim of this research was to understand cassava productivity in order to contribute to improving yields, food security and farm incomes in rain-fed cassava production systems in West Africa. A long-term goal was to contribute to a decision support tool for site-specific crop and nutrient management recommendations. Firstly, we studied farmers’ perception of cassava production constraints, assessed drivers of diversity among households and analysed the suitability of farmers’ resource endowment groups to the intensification of cassava production. The results indicate that farmers perceived erratic rainfall and poor soil fertility to be prime constraints to cassava production. The agricultural potential of the area and the proximity to regional markets were major drivers for the adoption of crop intensification options including the use of mineral and organic fertilizers. While the use of mineral and organic fertilizers was common in the Maritime zone that had a low agricultural potential, storage roots yields were below the national average of 2.2 Mg dry matter per hectare, and average incomes of 0.62, 0.46 and 0.46 US$ per capita per day for the high, medium and low farmer resource groups (REGs – HRE, MRE and LRE, respectively) were below the poverty line requirement of 1.25 US$. In the high agricultural potential Plateaux zone, HRE and MRE households passed this poverty line by earning 2.58 and 2.59 US$ per capita per day, respectively, unlike the LRE households with 0.89 US$ per capita per day. Secondly, we investigated the effects of mineral fertilizer on nutrient uptake, nutrient physiological use efficiency and storage roots yields of cassava since soil fertility was a major issue across the zones. We used an approach based on the model for the Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS). This model was successfully adapted for cassava and it appropriately assessed the response of cassava to N, P and K applications, especially in years with good rainfall. Under high drought stress, the model overestimated cassava yields. Thirdly, we investigated the impact of balanced nutrition on nutrient use efficiency, yield and return on investment compared to blanket fertilizer use as commonly practiced in cassava production systems in Southern Togo, and in Southern and Northern Ghana. The balanced nutrition approach of the QUEFTS model aimed to maximize simultaneously nutrient use efficiency of N, P and K in accordance with the plant’s needs. Larger nutrient use efficiencies of 20.5 to 23.9 kg storage root dry matter (DM) per kilo crop nutrient equivalent (1kCNE of a nutrient is the quantity of that nutrient that has the same effect on yield as 1 kg of N under balanced nutrition conditions) were achieved at balanced nutrition at harvest index (<em>HI</em>) of 0.50 compared to 20.0 to 20.5 kg storage root DM per kilo CNE for the blanket rates recommended by national research services for cassava production. Lower benefit:cost ratios of 2.4±0.9 were obtained for the blanket fertilizer rates versus 3.8±1.1 for the balanced fertilizer rates. Our study revealed that potassium (K) was a major yield limiting factor for cassava production, especially on the Ferralsols in Southern Togo. Hence, we fourthly studied the effect of K and its interaction with nitrogen (N), phosphorus (P), and the timing of harvest on the productivity of cassava in relation to the effects of K on radiation use efficiency (RUE), light interception, water use efficiency (WUE) and water transpiration. The results suggest that K plays a leading role in RUE and WUE, while N is the leading nutrient for light interception and water transpiration. Potassium effects on RUE and WUE depended on the availability of N and harvest time. Values of RUE and WUE declined with harvest at 4, 8 and 11 months after planting. Thus, enhanced K management with sufficient supply of N during the early stage of development of cassava is needed to maximize RUE and WUE, and consequently attain larger storage root yields. Given that erratic rainfall was another major constraint to cassava production according to the results of the farm survey, and due to the inability of QUEFTS modelling to assess drought effects on cassava yield successfully, another modelling approach based on light interception and utilization (LINTUL) was used. We quantified drought impacts on yields and explored strategies to improve yields through evaluation of planting dates in Southern Togo. The evaluation of the model indicated good agreement between simulated and observed leaf area index (Normalised Root Mean Square Error - NRMSE - 17% of the average observed LAI), storage roots yields (NRMSE 5.8% of the average observed yield) and total biomass yield (NRMSE 5.8% of the average observed). Simulated yield losses due to drought ranged from 9-60% of the water-limited yields. The evaluation of planting dates from mid-January to mid-July indicated that the best planting window is around mid-February. Higher amount of cropping season rainfall was also achieved with early planting. These results contradict current practices of starting planting around mid-March to mid-April. However, the results indicate the possibility to increase cassava yields with early planting, which led to less yield losses due to drought. By contrast, late planting around June-July gave larger potential yields, and suggested these periods to be the best planting window for cassava under irrigated conditions in Southern Togo. This shows that appropriate water control and planting periods can contribute to attaining larger yields in Southern Togo. Further improvement of the LINTUL model is required towards using it to assess water-limited yield, which can be used as boundary constraint in QUEFTS to derive site-specific fertilizer requirements for enhanced cassava yield and returns on investments in West Africa.