Zai Pits And Half-Moons — African Dryland Farming Techniques That Reverse Desertification

The history of desertification in the Sahel is inseparable from the history of colonial disruption to land tenure and management systems. Pre-colonial Sahelian societies operated complex rotational systems — managed fallow periods, transhumance routes coordinated between agricultural and pastoral communities, tree tenure rules that protected specific species in farmed landscapes. These systems were not uniform or perfect, but they were calibrated to their environment over centuries of feedback.

Colonial administrations disrupted these systems in multiple ways. Sedentarization policies eliminated transhumant grazing routes, concentrating livestock pressure on smaller areas. Taxation in cash rather than kind pushed farmers to maximize annual production at the expense of fallow periods. Tree clearance for export crops was actively promoted. Customary land governance structures were weakened or replaced by statutory systems that neither farmers nor herders trusted or understood. The biophysical degradation that followed was partly a direct consequence of policy choices — not simply an inevitable result of population pressure or rainfall variability.

By the 1970s drought, the underlying management degradation had left soils poorly positioned to absorb reduced rainfall. The crisis appeared to outside observers as a climate crisis because that was the proximate trigger. The structural cause was a century of management system disruption. This distinction matters because it shapes the appropriate response. If the problem is rainfall, you need rainfall. If the problem is management, you need management.

Zai pit technology specifically deserves precise description. The pits are dug during the October to April dry season, when laborers — typically household members including women and children — work the baked laterite and sandy soils with pointed hand hoes called daba. Spacing is typically 0.8 to 1 meter between pits in both directions, producing roughly 10,000 to 15,000 pits per hectare. Each pit receives a handful of compost, manure, or organic material before the rains. During the rainy season, millet or sorghum is seeded directly into each pit.

The multiple mechanisms operating in a zai pit field have been studied by researchers from ICRISAT, Wageningen University, and various West African agricultural research institutes. Runoff capture is the primary function: rainfall that would run off crusted bare soil is intercepted by the pit geometry. Studies on degraded Sahelian soils show infiltration rates on flat crusted surfaces of as little as 5 to 20 millimeters per hour, while the effective infiltration in zai pit fields is several times higher because the water is concentrated in pits rather than distributed across crusted interfluvial areas.

The termite mechanism is less immediately intuitive but documented in the literature. Organic matter in zai pits attracts Macrotermes and other termite genera, which excavate tunnel systems extending well below the root zone. These channels break up laterite hardpans — compacted iron-rich horizons that prevent root penetration and water infiltration. In established zai fields, porosity measurements in soil below pits are significantly higher than in adjacent unmanaged soil. The termites are doing soil work that would otherwise require mechanized subsoiling.

Yacouba Sawadogo's adaptation of traditional zai practice on his farm near Gourcy, Burkina Faso, added a critical innovation: planting trees in a subset of the pits. Rather than treating the field as purely annual cropland, Sawadogo allowed Faidherbia albida, Piliostigma reticulatum, Guiera senegalensis, and other indigenous trees to establish from seed in zai pits, then managed their canopy to complement rather than compete with crops. Faidherbia albida is particularly notable — it drops its leaves during the rainy season (when crops need light) and leafs out during the dry season (when it provides shade that reduces soil moisture evaporation). This phenological inversion makes it almost uniquely suited to agroforestry in seasonal rainfall systems.

Sawadogo's farm expanded from a few degraded hectares in the 1980s to a documented forest of roughly 20 hectares by the 2000s, on land that neighboring farmers had abandoned as unrestorable. His methods spread through farmer-to-farmer exchange across northern Burkina Faso and into Niger and Mali. The film "The Man Who Stopped the Desert" brought his work to international attention in 2010. What is significant about his approach is that it was not a package delivered by an NGO — it was farmer innovation within a traditional knowledge framework, observable by neighbors who could replicate it without external assistance.

Half-moon earthworks — demi-lunes — operate at a somewhat different scale and are more commonly used for pasture restoration and tree establishment on degraded hillslopes, though they are also used for crop production. The crescent mound is constructed on the contour, with the open side facing upslope to catch runoff. On slopes of 1 to 3 percent, the catchment area feeding each half-moon is typically 4 to 8 times the planted area inside the crescent. This multiplication of effective rainfall can transform a site receiving 300 millimeters of annual rainfall into a microenvironment receiving the equivalent of 1,200 to 2,400 millimeters — sufficient to establish trees and shrubs that could not otherwise survive at that rainfall level.

The World Overview of Conservation Approaches and Technologies (WOCAT) database documents half-moon earthwork projects across Burkina Faso, Niger, Mali, Chad, Senegal, and Ethiopia. Long-term monitoring plots show vegetation cover recovering from less than 10 percent to 40 to 70 percent within 5 to 10 years of establishment. Ground-nesting bird populations, soil macro-invertebrate diversity, and soil organic carbon show corresponding increases. Desertification processes, measured by wind erosion rates and dune movement, slow or reverse in project areas.

Niger's farmer-managed natural regeneration (FMNR) program, documented extensively by researcher Chris Reij and agronomist Tony Rinaudo, represents perhaps the largest-scale dryland restoration success on record. Farmers stopped treating trees emerging from root stocks as weeds to be cleared — a practice enforced under colonial-era laws that vested tree ownership in the state regardless of who owned the land — and instead managed them as productive components of their farm systems. As land tenure reform made trees legally the farmer's property, the incentive to manage rather than clear became rational. By the 2000s, aerial survey and satellite analysis suggested that five million or more hectares of southern Niger had increased vegetative cover, with corresponding improvements in crop yields, household food security, and soil organic matter.

The economic analysis of these systems consistently shows high returns on labor investment. A World Bank study of zai pit systems in Burkina Faso estimated labor investment at 100 to 300 person-days per hectare for initial pit construction, with reduced investment in subsequent years as soil structure improves. At documented yield improvements of 40 to 100 percent for millet and sorghum, the return on that labor — measured as additional calories produced per day of labor invested — compares favorably with purchased fertilizer options at any realistic smallholder farm wage rate.

The development-assistance failure mode here is different from waru waru. Zai and half-moons do not require collective action at the same scale. A single household can implement and benefit from them on its own land. The persistent barriers are therefore different: lack of initial organic matter for pit planting, secure land tenure (farmers do not invest labor in improving land they may lose), and — critically — knowledge transmission. Where zai practice was known, it spread. Where it was unknown, farmers on degraded soils with insufficient rainfall sat on a solution they did not know existed.

This is a knowledge distribution failure at civilizational scale. The techniques exist. The evidence base is substantial. The implementation barriers are manageable. The gap is between where the knowledge lives and where it is needed — and that gap is, in principle, closable.

A final observation: both zai and half-moons work by capturing existing rainfall more effectively, not by adding water. In a world where freshwater scarcity is increasing and irrigation infrastructure costs are prohibitive for most smallholder contexts, techniques that maximize the productivity of rainfall-as-received are structurally important. They are not stopgap measures. They are a category of solution that deserves the same infrastructure investment that irrigation receives — which is to say, far more than it currently gets.