The positive impact of better agronomic practices on maize yield in Africa has long been known. Reports back from the 1960s show results that are not too different from the ones shown here14–16. However, there has been an incessant call in more recent decades for SSA cropping systems to embrace a range of approaches of diverse ideology, focus, and labeling, including conservation agriculture17, agroecology18–20, climate-smart practices21,22, regenerative agriculture23,24, nature-based solutions25,26, and digital agriculture27–29. These approaches are promoted based on the assumption that they will contribute to a wide range of societal goals associated with food security, climate change, and conservation of natural resources. A concern is that their labeling, broad goals, and a strong focus on practices rather than outcomes can shift the focus away from the urgent need to increase crop yields in SSA and the most obvious agronomic solutions to achieve this goal. Here we showed that farmers in SSA could benefit from an explicit effort from national and international programs to facilitate adoption of better agronomic practices in relation to soil fertility, cultivar selection, sowing date, plant density, and pest control. We show that maize yields in the region more than double when farmers adopt better agronomic practices. In turn, increasing average yield can help increase local and regional food security and improve smallholders’ income30,31. Indeed, when our results were scaled out over the entire maize area in SSA, we found that better agronomy can drastically reduce the need for cropland expansion and/or imports (Table 1). Failure to narrow the yield gap substantially can have severe negative socio-economic and environmental consequences. Therefore, the region should not be used as the testing ground to promote approaches that have not been empirically validated in their capacity to increase on-farm yields.
Our findings are supported by thousands of field observations across major agroecological zones where maize is grown in East Africa, a well-validated spatial framework, and robust statistical methods. Our work overcomes major limitations from previous studies assessing yield constraints in SSA. First, many previous analyses have relied on farmer surveys, which is problematic because farmers frequently overestimate yields, particularly in small plots32–37. In contrast, our study is based on direct yield measurements derived from crop samples collected following robust protocols. Second, previous yield gap analyses in SSA are narrow in geographic scope and range of agronomic practices38–41. In contrast, our study includes thousands of farmers from four countries, across 21 climate zones, including farmers with access to inputs which, in turn, extend the range of agronomic practices in our database, increasing the odds to identify yield-improving practices42. Finally, the combination of a robust climate spatial framework and statistical methods, together with soil and terrain databases, allowed us to control for the confounding effects of varying climate, soil, and terrain factors in relation to the impact of agronomic practices on on-farm yields7–9, 43,44. Application of the same approach over other geographies and crop systems could help identify, with a modest investment and in a relatively short timeframe, key yield constraints for all major cropping systems in SSA, providing a basis for researchers, extension workers, and policy makers to prioritize their efforts.
Reaching maize self-sufficiency in Africa is not ‘mission impossible’ but rather requires investment in AR&D and proper policy to focus on those technologies with greatest probability to deliver yield gains within a short time and at relevant spatial scale. Our study shows that those technologies already exist, for example, use of hybrids combined with proper nutrient, pest, and crop establishment practices. While it is unrealistic to imagine all SSA farmers adopting these practices immediately, the period between now and year 2050, which coincides with the sharp increase in maize demand in SSA, offers a reasonable window of time to promote better agronomic practices and the means for their adoption. In turn, reaching an average yield of 4.4 t ha− 1 by year 2050, equivalent to a yield gap closure of 30%, would require annual rates of yield gains for maize that are ca. three times compared to current ones and comparable to those observed in other tropical and subtropical rainfed maize producing regions in South America and Southeast Asia (Fig. 1). Maize yields nearly doubled in those two regions over the past two decades, and the same scenario is possible for SSA through adoption of better agronomic management. But we note that while adopting yield-improving practices has the potential to narrow ca. one-third of the yield gap, there would still be another two-thirds that remains, suggesting that further yield improvements are possible but would require more inputs and fine-tuning other management practices and interactions. For instance, combining twice the fertilizer rate of higher technology farmers, proper hybrids, good crop stands, and full weed control resulted in yields that reached 75% of the yield potential (Fig. 4).
Our study is subjected to several limitations. First, we note that our study is limited to East Africa. However, similarity in climate and yield between our study region and other maize producing areas in SSA (Supplementary Figs. S1-S2) and previous evidence showing the widespread positive impact of our identified yield-improving management practices41,45,46 further supports extrapolations of our results to the entire region. Second, our pledge to promote good agronomic practices in SSA should not be taken as a prescription. We acknowledge that local tuning in each environment will be needed to account for gradients in soil fertility resulting from differences in field history47,48. While tailoring local recommendations is relevant, we also note that the message of better agronomy across SSA needs to be clear. Finally, our study underscores the importance of re-orienting AR&D programs and policy to those key factors likely to benefit farmers and facilitate farmers’ access to inputs and technical information. For example, low fertilizer application in SSA is a consequence of high fertilizer prices (relative to the grain price) and low economic incentives to intensify crop production, which, in turn, results from a large uncertainty on the return of investment from agriculture and high rural population dependency on off-farm income49–51. Hence, yield improvement will surely require more than good agronomy but also proper markets, institutions, and policy to facilitate their adoption by farmers. Ultimately, an area greater than the US corn belt is waiting for yield intensification in SSA, but no intensification can be achieved without good agronomy.