Zai Pits and Bunds: African Dryland Regeneration Techniques

The Question This Page Answers

The land manager or development practitioner arriving at this page is typically dealing with severely degraded dryland soils where conventional earthworks are not practical. The ground sheds water as a sheet. There are no excavators. There is no budget for backhoe hire. The question is: what can you do with hand tools, local materials, and a working knowledge of how water concentrates on flat degraded land?

The answer from the West African Sahel is zai pits combined with stone bunds. These are not primitive substitutes for proper earthworks. They are a sophisticated two-layer intervention that addresses both the field-scale hydrology (stone bunds slow runoff across the slope) and the point-scale soil biology (zai pits break up hardpan at each planting point using termite activity). The combination restores crop production on soils that were producing nothing and does it in a single growing season, with ongoing improvement over three to five years.

The technique is applicable well outside the Sahel context. Any degraded soil with a compacted surface layer, anywhere rainfall intensity routinely exceeds soil infiltration capacity, is a candidate for zai-style planting pockets. That includes degraded pastures in sub-Saharan Africa, overgrazed rangelands in the Middle East and Central Asia, and compacted orchard soils in Mediterranean climates where mechanical access is limited by slope or tree density.

The broader context for dryland water harvesting, including how zai pits relate to the full system of swales, terraces, and check dams, is in the water harvesting pillar essay. This page goes deep on zai pits and stone bunds specifically: the mechanics, the Sawadogo case record, and the conditions under which this method outperforms mechanised alternatives.

How Zai Pits and Bunds Work: Soil Physics and Termite Biology

Degraded Sahel soils develop a surface condition locally called zipele in Mooré: a smooth, sealed crust of fine particles cemented by alternating wet-dry cycles and the complete removal of surface vegetation. This crust has a hydraulic conductivity near zero. Rainfall during the monsoon season, even at low intensities, runs off as a sheet because the crust absorbs nothing. The result is no infiltration, no groundwater recharge, no crop production, and accelerating erosion in the concentrated flow paths the runoff creates.

Zai pits interrupt this process at two levels simultaneously. At the surface level, the pit creates a depression that collects and holds runoff from the surrounding sealed crust. Rainfall that falls within the pit's catchment radius (roughly 30 to 50 centimetres around each pit) pools in the pit rather than running off. At the subsurface level, the handful of organic matter placed in each pit triggers the mechanism that makes zai pits far more effective than a simple planting hole: termite foraging. Termites are chemically attracted to decomposing organic material. Within weeks of pit digging and organic matter placement, Macrotermes and Trinervitermes species begin excavating channels downward from the pit floor, breaking up the hardpan layer at 20 to 40 centimetres depth. Each termite channel becomes a preferential infiltration pathway. A single pit with active termite activity can have hydraulic conductivity improvements of 10 to 50 times the surrounding soil within one season (Reij et al. 2009 IFPRI Discussion Paper; Fatondji et al. 2006 Plant and Soil).

Stone bunds operate at the field scale. A stone bund is a line of rocks placed loosely on contour across the slope, typically 20 to 40 centimetres high. It is permeable: water seeps through the gaps between stones rather than overtopping in a concentrated flow. The effect is to slow and spread runoff velocity across the field, increasing the residence time of surface water between bunds and giving the soil beneath more time to absorb it. The sediment trapping function compounds over years: as fine soil particles settle uphill of each bund, a soil terrace forms gradually without mechanical earthmoving. In the Sahel, stone bunds have been documented raising organic matter and fine soil content on the uphill face by measurable amounts within five to ten years of installation.

The combination of bunds and pits addresses both the field and the point. Bunds reduce the velocity and volume of water arriving at each pit from the wider field; pits capture what arrives and direct it into the soil profile through termite-opened channels. Neither method alone is as effective as the two together. Bunds without pits slow runoff but cannot overcome the hardpan infiltration failure on severely degraded soils. Pits without bunds capture water at each plant but do nothing to reduce the erosive force of runoff in the inter-pit zones.

The Numbers: Scale, Yield Data, and Cost

Yacouba Sawadogo began his experimental work with zai pits on his degraded farm near Gourga in the Yatenga Province of Burkina Faso in the early 1980s, during the severe Sahelian drought period of 1983 to 1985. He restored approximately 40 hectares of land that his neighbours had abandoned as permanently unproductive. More significantly, he observed and documented the tree recovery that resulted from the improved infiltration: by the late 1990s his 40-hectare site supported a growing agroforestry canopy of native tree species that had not been present for decades. The World Future Council awarded him the Future Policy Award in 2018, by which point the technique had spread to over 200,000 hectares across Burkina Faso, Niger, Mali, and Senegal (Reij et al. 2009 IFPRI Discussion Paper; World Future Council 2018 documentation).

The yield data from zai pit trials in Burkina Faso documents average sorghum yield increases of 300 to 500 kilograms per hectare above non-zai controls on the same degraded soils, in seasons where the control plots produced near zero. On better soils, the absolute gain is smaller but still significant: 100 to 200 kg/ha advantage over conventional planting. Given that the input cost of zai pit preparation is essentially the labour of digging (approximately 60 to 80 person-days per hectare to dig 25,000 pits, including organic matter collection and transport), the return on labour investment is extremely high in contexts where labour is available and alternative employment is limited (Fatondji et al. 2006 Plant and Soil 282:1-2; source for labour figures: vault_atom_TBD).

The scaling numbers from the broader Farmer-Managed Natural Regeneration (FMNR) movement, of which Sawadogo's work is a founding example, document what happens at regional scale. Across southern Niger, FMNR practitioners using zai-adjacent techniques on approximately 5 million hectares of cropland have measurably increased tree cover, reduced wind erosion, and improved crop yields over a 30-year period, with the additional economic benefit of harvestable tree products from the recovered agroforestry canopy. The regreening of southern Niger is one of the largest verified land rehabilitation outcomes in the world and was largely accomplished with no external capital inputs: just farmers applying their own labour and knowledge (Reij and Winterbottom 2015; vault_atom_TBD).

Yacouba Sawadogo and the Sahel Rehabilitation Record

Sawadogo's contribution was not the invention of zai pits, which have antecedents in traditional Sahelian farming practice, but the systematic intensification and documentation of the method. The traditional zai was a small planting depression used opportunistically. Sawadogo's innovation was to dig pits during the dry season, add organic material rather than planting directly, and manage the spacing and bund combination as a deliberate system rather than an ad-hoc intervention. He also documented his results over decades and shared the method with other farmers in the region, creating the knowledge diffusion pathway that eventually reached hundreds of thousands of practitioners.

The scale of the recovery at Gourga is documented in satellite imagery as well as ground truth data. Land that appeared as bare soil or sparse degraded cover in 1983 showed measurable tree cover increase by 1999 and a substantially recovered agroforestry system by 2010. The variety of tree species that returned includes species not planted by Sawadogo: the improved infiltration and organic matter cycling created conditions in which wind-dispersed and bird-dispersed seeds could establish, and termite-improved soil structure allowed root penetration where it had been impossible for decades. The rehabilitation was not just of the soil surface; it was of the entire below-ground biological system that determines whether any plant can establish.

The broader FMNR regreening across Niger, documented by Reij, Tappan, and Smale (2009) at IFPRI, applies similar principles at a scale that makes it one of the most significant land rehabilitation stories of the 20th century. Approximately 200 million trees were added to southern Niger's croplands over the period 1985 to 2004, increasing household food security and income for millions of farmers. The capital cost was near zero: the technique is labour-intensive in the dry season pit-digging period and requires no purchased inputs beyond the organic matter farmers generate from their own livestock and crop residues.

For the regenerative agriculture practitioner in West Africa or any comparable dryland context, the zai pit system represents the lowest-capital-cost entry point into water harvesting. It requires no machinery, no imported materials, and no specialist knowledge beyond what can be learned from a single season of observation. The limitation is labour: digging 25,000 pits per hectare in the dry season heat is physically demanding, and operations above ten hectares typically require community or household labour pools rather than individual capacity. That is not a technical constraint; it is a social organisation question that traditional community land management systems in the Sahel have solved across generations.

Where Zai Pits Fit in a Water-Harvesting System

Zai pits occupy the lowest-capital, highest-labour end of the earthworks spectrum. On the cost and equipment axis, they sit below on-contour swales (which require a backhoe or at minimum a tractor) and far below bench terracing (which requires significant mechanised earthmoving). On the effectiveness axis for severely degraded hardpan soils where machinery cannot operate or is unavailable, they are the only method that works. A backhoe cannot fix a soil with zero infiltration capacity; only the biological soil-breaking mechanism of termites operating inside organic matter-filled pits can restore infiltration from zero.

The decision framework for choosing zai pits over other earthworks methods is:

• If soil surface is sealed hardpan with near-zero infiltration and you have labour but no machinery: zai pits and stone bunds are the primary method.
• If soil is degraded but not hardpan and you have machinery access: on-contour swales will be faster and more durable per hectare.
• If the site combines hardpan in lower areas and compacted soil on steeper sections: combine zai pits in the flat degraded zones with stone bunds on the slopes and check dams in drainage channels.

The connection to agroforestry is direct and historically documented. Sawadogo's observation that trees returned to his restored site without planting established what practitioners now call Farmer-Managed Natural Regeneration: the principle that if infiltration and soil organic matter reach a threshold level, the seed bank and existing root stock in the soil will regenerate woody cover on its own. This is the foundation of agroforestry rehabilitation in dryland systems: the earthworks create the conditions; biology does the recovery.

In the longer run, successful zai pit rehabilitation creates the soil conditions that allow more capital-intensive earthworks to be added profitably. Once a previously degraded hectare is producing crops and beginning to accumulate soil organic matter, the economic argument for swale installation or small pond construction becomes viable. The zai pit is the first-year intervention that makes everything else possible. This sequencing logic, from low-capital biological soil-building to higher-capital hydrological infrastructure, is the economic structure underlying the earthworks economics analysis in this pillar's dedicated comparison page.

The lesson from the Sahel at scale is that there is no capital floor for water harvesting. The method at the bottom of the capital ladder, executed consistently across millions of hectares by farmers acting individually without external coordination, produced one of the largest land rehabilitation outcomes ever documented. The economic mechanism is not charity or aid; it is that each farmer who restores a hectare of degraded land captures the full value of that restoration in improved yields, reduced hunger risk, and harvestable tree products. The incentive and the technique aligned, and 200 million trees resulted.