Productivity limits and potentials of the principles of conservation agriculture

One of the primary challenges of our time is to feed a growing and more demanding world population with reduced external inputs and minimal environmental impacts, all under more variable and extreme climate conditions in the future^{1, 2, 3, 4}. Conservation agriculture represents a set of three crop management principles that has received strong international support to help address this challenge^{5, 6}, with recent conservation agriculture efforts focusing on smallholder farming systems in sub-Saharan Africa and South Asia⁷. However, conservation agriculture is highly debated, with respect to both its effects on crop yields^{8, 9, 10} and its applicability in different farming contexts^{7, 11, 12, 13}. Here we conduct a global meta-analysis using 5,463 paired yield observations from 610 studies to compare no-till, the original and central concept of conservation agriculture, with conventional tillage practices across 48 crops and 63 countries. Overall, our results show that no-till reduces yields, yet this response is variable and under certain conditions no-till can produce equivalent or greater yields than conventional tillage. Importantly, when no-till is combined with the other two conservation agriculture principles of residue retention and crop rotation, its negative impacts are minimized. Moreover, no-till in combination with the other two principles significantly increases rainfed crop productivity in dry climates, suggesting that it may become an important climate-change adaptation strategy for ever-drier regions of the world. However, any expansion of conservation agriculture should be done with caution in these areas, as implementation of the other two principles is often challenging in resource-poor and vulnerable smallholder farming systems, thereby increasing the likelihood of yield losses rather than gains. Although farming systems are multifunctional, and environmental and socio-economic factors need to be considered^{14, 15, 16}, our analysis indicates that the potential contribution of no-till to the sustainable intensification of agriculture is more limited than often assumed.