Speaker
Description
InnoFoodAfrica was a Horizon 2020 project running from 2020 to 2024, with the aim of improving productivity and sustainability of food systems in Ethiopia, Kenya, South Africa and Uganda. The work-package on crop production practices used Farmer Participatory Research (FPR). Experiments in each country tested the impact of certain inputs on modern cultivars and local landraces, with three iterations. Several (usually pre-existing) teams of 10 – 20 farmers were engaged for work on each crop. The crops were faba bean, teff, maize and orange-fleshed sweet potato (OFSP) in Ethiopia; cowpea, finger millet, amaranth and OFSP in Kenya; cowpea, Bambara groundnut and OFSP in South Africa; and cowpea, finger millet, sorghum and matooke banana in Uganda. Seeds and inputs were provided by the project. The farmers did the work under the guidance of local experts and, importantly, made the decisions on how to proceed based on the outputs of the research. Since communities were involved, eating quality of the harvest was investigated and formed an important part of the decision-making process. In the first iteration, the full factorial of cultivars and inputs, with replicates, was established on 2 – 4 farms. In the second iteration, the number of treatments was reduced, the size of the plots was increased, and several “baby trials” demonstrating favoured combinations of cultivar and input were established on participating farms. The third iteration was a scaling-up phase, with fewer but the best treatments in the replicated experiments and more baby trials.
Both genetics and management contributed to improved yields. The best modern cultivars outyielded landraces 1.2-fold to 5-fold, even with low inputs, and the most effective input packages increased yields by similar margins.
One of the key management factors was sowing in rows instead of broadcasting. This enabled weed control by hoeing instead of hand-weeding, and the application of inputs, such as lime in Ethiopia, to only 25% of the plot. To ease the consistency of row spacing, a farmer in Uganda invented a simple row-marking tool consisting of nails driven into a plank towed by ropes.
Farmers chose which treatments to combine. Manure-fertilized amaranth in Kenya had strong stems, while synthetically fertilized amaranth had large seed heads, so the farmers chose a 50 : 50 blend for the next round, producing plants that both stood well and yielded well. Adding a foliar fertilizer, containing micronutrients, boosted yield further. Wood ash, a traditional seed protectant for finger millet in Kenya, boosted yields in combination with synthetic fertilizer (Figure 1A).
Traditional and modern knowledge conflicted when it came to sowing of Bambara groundnut in South Africa. Culturally, this may not happen until after 15 December, but the climate has changed enough that this is several weeks too late for maximum yield.
Dual-purpose crops were not successful. Amaranth and cowpea leaves are traditional green vegetables, but the best cultivars for vegetable use were not the best for seed production, and leaf harvest set back potential seed yield.
Rhizobium inoculation of faba bean was clearly effective in Ethiopia; that of cowpea in Kenya was less so, suggesting that the soils were deficient in more than just N and that the condition of the inoculum was not optimal.
Eating quality was an important determinant of choices. Nitrogen-fertilized OFSP was almost universally watery and of poor texture, unfertilized being much preferred; and in several cases, the second-highest yielding cultivar of OFSP was preferred over the highest-yielding. One new cowpea cultivar, NaroCowpea-2, was praised by the Ugandan farmers for its excellent flavour, although another often yielded more (Figure 1B). When the faba bean crop was harvested, some of the Ethiopian farmers preferred their landrace, but 6 months later said that the new, 2.5x higher yielding cultivar, was just as good.
The project demonstrated several ways in which yields, profits and nutrition could be raised on farms. Continued access to inputs, such as fertilizers, outside the project is a challenge. It is up to each farmer to decide how to proceed.
Figure 1. Grain yields of (A) two cultivars and a local landrace of finger millet in Kenya, given no fertilizer, wood ash only, synthetic fertilizer only, and the combination; (right), five cultivars of cowpea in Uganda at three sites.
| Keywords | farmer participatory research; cultivar; management; input; crop quality |
|---|
