It’s time to harvest! Towards sustainable farming systems in the East African highlands

The UN Sustainable Development Goals of Zero poverty and Zero hunger include leaving no one behind as a key principle. However, many smallholder farmers in sub-Saharan Africa (SSA) are caught in a poverty trap, a vicious cycle of low productivity and limited ability to invest. Moreover, small farm areas may limit the potential benefits that can be accrued at farm level, even if productivity would increase. Sustainable intensification is a key strategy to increase agricultural production for the growing population in SSA, while at the same time avoiding the extension of agricultural land in natural areas. In the first part of the thesis I used an ‘impact-oriented’ perspective to assess, within current farming systems, to what extent integrated co-learning leads to sustainable intensification. In the second part I used a ‘target-oriented’ perspective to explore ‘viable farm sizes’ required to attain a living income (the income required for a decent living including a nutritious diet, clothes, schooling and housing). By situating this study in the East African highlands, characterized by high population density and small farm sizes, I revealed possible pathways towards more sustainable farming systems.

We developed the ‘integrated co-learning approach’, which combines input vouchers with iterative learning cycles on sustainable intensification, and tested it in western Kenya from an impact oriented perspective. Farmers participating in co-learning had a more diverse and cohesive knowledge after five seasons compared to farmers who only received the voucher. Irrespectively of the co-learning, the voucher immediately increased farm level maize yield from less than 20% to 40-50% of water-limited yield. This indicates that closing yield gaps is mainly limited by capital constraints and not by technology or knowledge. However, co-learning facilitated the more complex changes in the cropping system that are required for sustainable intensification, such as the incorporation of legumes.

Although yields improved after the introduction of the voucher, the value of produce from crops was still below the living income benchmark for most households due to their small farm areas. Increasing yield alone was thus not enough to attain a living income from arable farming. Also for other indicators of sustainable intensification the desired outcomes were often not achieved. For instance, nitrogen (N) use efficiency remained too high, indicating the risk of soil N mining. Maize area and farm area also increased, all pointing towards the pathway of extensification instead of desired intensification. This implies the need for policies that favour increased input use and policies that limit area expansion.

Building on the finding that farm size strongly limited farmer income, we explored viable farm sizes for contrasting future scenarios in three sites in the East African highlands. This target-oriented perspective revealed that in the current baseline scenario, cultivated areas per farm would have to increase by 4.5, 1.3 and 2.5 times in Nyando (Kenya), Rakai (Uganda) and Lushoto (Tanzania) respectively, to make a living income. However, if crop yields increased to 50% of the water-limited yields, current cultivated areas of most households (>70%) would be large enough to make a living income in Rakai and Lushoto. In Nyando other sources of income, such as income from livestock, were required to make a living income.

Comparing the outcomes of the two different perspectives indicates that increasing yields of staple crops, e.g. through input subsidies, is not enough for all farmers to make a living income from current farm sizes. Larger changes are required, both within the farming system, e.g. increasing farm areas and/or cultivating more profitable crops, as well as outside the farming system, e.g. alternative employment options outside agriculture. The integrated co-learning approach can be deployed to explore incentives for smallholder farmers to sustainably intensify. Further research is required on how to scale the approach and integrate it into extension systems while keeping the valuable farm-researcher feedback. The viable farm size as a benchmark is a useful method for assessing how to leave no one behind while moving towards more sustainable farming systems.