Evaluating manual conservation agriculture systems in southern Africa

• The paper evaluates productivity and profitability of different CA systems.
• Direct seeding outyielded the conventional control in nine out of fourteen yield comparisons.
• Based on research data, uniform promotion of basin planting is questioned.
• Basins should only be promoted where moisture is limiting.
• A flexible extension approach is required to increase widespread adoption of CA.

Future threats of climate variability and change and accelerated soil degradation in southern Africa have increased the need for more sustainable and “climate-smart” agriculture practices. Manual systems of conservation agriculture (CA) based on seeding into planting basins or direct seeding techniques have received increased attention over the last decade. However, a critical review of the pros and cons of the different manual seeding systems under different agro-ecologies has been lacking. This paper aims at analysing different manual seeding systems in areas extending from central Mozambique to central Malawi. Results show that CA systems perform differently in contrasting agro-ecological environments. Direct seeded treatments had greater maize yields than conventional tillage practices by an average of 12–27% and outperformed the conventional practice in nine out of fourteen yield comparisons. Basin planted treatments performed well only in Sofala and Manica (15%) with yield penalties of −9% in Tete. The strongest factor influencing maize grain yields in the more variable areas of Manica and Sofala was the quality of season and the location, whereas tillage treatment and location were more important in the higher rainfall areas of Tete. Direct seeding systems out-yielded other treatments in areas of higher rainfall and responded better to a favourable environment than conventional tillage practices. CA systems, especially direct seeding in Malawi, Manica and Sofala, showed greater financial returns to investments and labour productivity due to reduced labour costs and higher yields. Labour savings of up to 43 labour days ha−1 could be achieved with direct seeded treatments in Malawi. The results of this research clearly highlight the need for site-specific recommendations and adaptation of CA systems to different agro-ecological environments. Blanket recommendations of one CA system across many agro-ecologies, as has often been done in the past, will only lead to underperformance of CA in some areas and rejection by smallholder farmers if yield benefits are not achieved.