Check Dams and Gabions: Slowing Erosion in Drylands

The Drainage Channel Problem

Swales address diffuse runoff across a slope. Check dams address concentrated flow in drainage channels and gullies: the parts of a catchment where water has already aggregated into a single high-velocity stream. These are different problems requiring different tools. A swale placed in an active drainage channel will fail; the concentrated flow overtops the berm before it can infiltrate. A check dam placed in that same drainage channel works because it is designed for concentrated flow, not diffuse surface runoff.

Gully erosion is the most visible and the most destructive form of land degradation in drylands. A single storm event in an unprotected catchment can advance a gully headcut by 5-20 metres and export hundreds of tonnes of sediment from a 10-hectare property. Once established, a gully is self-reinforcing: each erosion event deepens the channel, increases flow velocity, and enables larger erosion events at the next rain. Without intervention, a gully starting at 0.5 metres depth can reach 3-5 metres in under a decade in highly erodible soils.

The intervention logic is simple: slow the water, drop the sediment, raise the base level. Check dams are the tool that executes that logic in drainage channels. This page covers the engineering decisions that determine whether a check dam succeeds or fails, the materials available from free rock to engineered gabions, the spacing formula for a series of dams that fills a gully profile with trapped sediment, and the documented outcomes from large-scale check dam programmes including the Loess Plateau rehabilitation (World Bank P056216, 1999-2005) and Yacouba Sawadogo's Burkina Faso work.

The full water-harvesting context connecting check dams to swales and farm ponds is covered in the water harvesting pillar essay. This page focuses on drainage channel interventions specifically.

How Check Dams Work: Flow Velocity, Sediment, and Groundwater

Erosion rate in a drainage channel scales with flow velocity to approximately the third power in mild conditions and higher powers in turbulent flow: doubling flow velocity increases erosive energy by a factor of eight or more. This non-linearity is the central physics of gully management: the smallest reductions in peak flow velocity produce disproportionately large reductions in erosion. A check dam that reduces peak flow velocity by 30 percent reduces erosive energy by roughly 65 percent in that reach.

The mechanism operates through base level adjustment. Before a check dam: water falls freely from the lip of each natural drop in the channel, creating a plunge pool that erodes the channel bed and undercuts the downstream bank. After a check dam: the pool behind the dam raises the effective channel floor to the dam crest. Water entering the pool decelerates immediately due to the increased cross-sectional area and depth. Sediment suspended in the flow drops out of suspension as velocity decreases. Over successive rain events, the pool behind the dam fills with sediment until the effective channel grade across the ponded zone approaches zero. At that point, the formerly eroding section is stable, has a raised water table, and supports vegetation that locks the sediment in place permanently.

The groundwater effect is often overlooked but is economically significant. Check dams that create ponding zones, even briefly during and after rain events, allow infiltration into the streambank and adjacent soils. In arid and semi-arid catchments, this is often the primary mechanism by which stream-adjacent soils maintain enough moisture to support woody vegetation year-round. Brad Lancaster's rainwater harvesting work in Tucson, Arizona, documents how a series of check dams in an ephemeral channel converted dry bare banks into productive riparian corridor with native tree cover within 3-5 years, requiring zero irrigation after establishment (Lancaster 2019 Rainwater Harvesting for Drylands Vol. 2).

The Numbers: Cost, Sediment Load, and the Loess Plateau

Yacouba Sawadogo restored approximately 40 hectares of severely degraded Sahel land in Burkina Faso using traditional Zai pits combined with stone bunds (the Sahel equivalent of check dams across micro-catchments). Zai pits are small depressions dug on degraded hardpan, filled with compost and organic matter, that concentrate rainfall and organic inputs around planted trees and crops. Stone bunds placed along contours slow surface flow and trap sediment. The model spread to over 200,000 hectares across West Africa, making it one of the largest grassroots land restoration programmes ever documented (Reij et al. 2009 IFPRI Discussion Paper; World Future Council 2018 Future Policy Award documentation).

The Loess Plateau Watershed Rehabilitation Project (World Bank P056216) installed check dams on 3,700 gullies as part of its 1999-2005 intervention across 35,000 square kilometres. Biomass cover increased 126 percent across the project area. The Yellow River sediment load fell by approximately 100 million tonnes per year by 2005, reversing a 2,000-year trend of accelerating erosion (Wang et al. 2016 Nature Geoscience 9:38-41). The check dams were a significant element of that sediment reduction: gully systems that had been exporting millions of tonnes of sediment annually were converted to stable depositional features within the project period.

Individual check dam construction costs: rock check dams in accessible terrain cost 300 to 1,500 EUR each depending on size, including rock sourcing and manual labour. Gabion check dams of comparable size cost 800 to 3,000 EUR each depending on gabion basket cost (15-60 EUR per metre of basket), labour, and concrete for the plunge pool. A series of 10-15 dams addressing a 100-metre active gully costs 3,000 to 15,000 EUR. Compare that to the cost of topsoil loss: at 50 tonnes of sediment per event from an active gully, with 5-10 events per year at 10-20 EUR per tonne of agricultural soil value destroyed (conservative), the ongoing annual loss is 2,500 to 10,000 EUR. Payback on a well-installed check dam series is typically 1-5 years (EU rural infrastructure cost data; FAO watershed management guidelines, sources pending vault retrofit).

Building a Check Dam Series: Materials, Spacing, and Sequence

The first installation decision is material selection: loose rock, engineered gabions, or a hybrid approach with rock at low-velocity locations and gabions at headcuts. Loose rock check dams built to 0.3-0.6 metres height are appropriate for minor rills and small ephemeral channels with peak flows below 1 cubic metre per second. They require no tools beyond shovels and wheelbarrows, cost 300-800 EUR per dam in labour and materials, and can be installed in a day by a two-person team. Their failure mode is displacement of individual rocks during high-flow events; inspection after each major storm and replacement of displaced rock is the ongoing maintenance task.

Gabion check dams are the preferred option for active gullies with headcuts deeper than 1 metre or channels with estimated peak flows above 2 cubic metres per second. A gabion basket (typical dimensions 1m x 1m x 2m) is prefabricated galvanised wire mesh assembled on-site, filled with angular rock 100-200mm in diameter, and tied closed with binding wire. The completed basket has sufficient self-weight and structural integrity to resist displacement by flow velocities up to 3-5 metres per second. Adjacent baskets are tied together to form a continuous dam face. The wire mesh must be galvanised or PVC-coated for longevity; ungalvanised wire rusts through in 5-10 years in wet climates.

The construction sequence for a check dam series runs from bottom to top: install the lowest dam first, then work upstream. This allows the pool behind the lower dam to begin filling with sediment while upstream dams are under construction, and gives the operator access to the gully base without disturbing the newly compacted sections above. The key trench at each dam site anchors the structure into stable ground on both banks and the gully base, preventing flankingflow from bypassing the dam around the edges. Flanking is the primary failure mode for check dams without adequate key trenches.

The plunge pool below each overflow notch requires armouring with large angular rock (riprap) or concrete. The overflow drop creates a jet of water with significant kinetic energy at the impact point. An unarmoured plunge pool scours rapidly: the scour hole undercuts the downstream toe of the dam structure and causes failure within 3-10 storm events. Riprap sized at 1.5-2x the median rock size in the flow provides adequate armour if keyed 0.3-0.5 metres into the plunge pool floor.

The maintenance calendar after installation: inspect after every storm event in the first season. Replace displaced rocks, clear debris blocking overflow notches, and monitor the sediment fill level behind each dam. Once dams are 80 percent full of sediment and vegetation is established on the depositional surfaces, maintenance frequency drops to annual inspection. A well-built check dam series in stable geology with established vegetation requires minimal attention after year five.

Where Check Dams Fit in the Catchment System

Check dams address the drainage channel network: the 10-20 percent of a catchment's area that drains via concentrated linear flow paths. The other 80-90 percent of the catchment, the slope surface draining as diffuse sheet flow, is addressed by swales and keyline cultivation. These two intervention types are complementary rather than alternative: a catchment managed with swales on the slopes and check dams in the gullies has addressed both flow regimes and converted both from erosion sources into productive storage systems.

The interaction between check dams and farm ponds is important to sequence correctly. Check dams in the drainage lines upstream of a farm pond reduce the sediment load reaching the pond, extending the pond's operational lifespan before dredging is required. A pond without upstream check dams in actively eroding catchments may silt up within 5-15 years, requiring costly dredging or complete reconstruction. The same 3,000-10,000 EUR spent on check dams upstream typically extends the pond's productive life by 20-40 years.

The Zai pit and stone bund method documented by Yacouba Sawadogo in Burkina Faso demonstrates what check dam principles achieve at the smallest scale: individual Zai pits function as micro-check-dams, each intercepting the flow from a few square metres of degraded hardpan and concentrating it around a planted seed. Stone bunds along contours function as shallow check dams across broad shallow drainage lines. The combination of micro-scale water concentration and meso-scale flow control restored severely degraded land to productive agricultural use with no mechanical equipment and minimal capital cost. The model that spread to over 200,000 hectares across West Africa validates the fundamental principle: slow the water, drop the sediment, give biology a chance to work (Reij et al. 2009 IFPRI; World Future Council 2018 documentation).

For regenerative agriculture systems transitioning degraded conventional land, check dams in drainage lines are often the first earthworks intervention. They address the most acute ongoing capital loss (sediment export) before any other investment is made. Stopping the bleeding precedes any other farm improvement. A property losing 100 tonnes of sediment per year from active gullies cannot build soil fertility with cover crops and compost while that loss continues. Check dams close that hole first. Then swales, keyline cultivation, and pond storage can begin accumulating the water and soil capital that make the whole system productive.