Conservation agriculture is being used increasingly around the world (H. Buffett Nature 484, 455; 2012). But many of the claimed benefits of minimal or no-tillage farming — such as carbon sequestration and boosting crop yields — are far from proven (K. E. Giller et al. Field Crop Res. 124, 468–472; 2011).

Several of these conservation techniques are good agricultural practice: to sow with the first rains, to rotate crops, to fertilize them appropriately and to return crop residues to the soil. But in the absence of herbicides (a reality for many African smallholders), tillage can save labour, allows farmers to plant early and controls weeds. It helps to prevent runoff and erosion if the soil is not protected by mulch, for which smallholder farmers often lack the organic resources.

My view is therefore that an across-the-board recommendation of conservation agriculture is misplaced (Nature 483, 525–527; 2012). Given the enormous diversity of African smallholder farming systems, locally adapted, best-fit technologies are needed (K.E. Giller et al. Agric. Syst. 104, 191–203; 2011).

Farming-systems analysis will identify for whom and where different intensification strategies are most likely to help.