Deep erosion of a palmiet peat wetland

Site description and problems

The Elandskloof River flows north from the mountains above Hermanus through an ancient palmiet peat wetland which has a fairly steep slope. The floodplain of the lower reaches of the river, just above the confluence with the Hartbees River, a tributary of the Klein River and Hermanus Lagoon, is divided into smallholdings with small scale farming operations. Especially in the lower regions of the river, the banks have become infested with alien vegetation (mainly Acacia mearnsii (black wattles)) which had shaded out and displaced the erosion resistant indigenous wetland vegetation.

In 2003 a small dam for watering horses, constructed in the watercourse on one of the properties, was washed away during a flood. The washing away of the dam initiated a knick-point in the river which progressed upstream. The Department of Agriculture was requested to assist in arresting the then 2m deep gulley which had formed in the wetland. The landowner was not however prepared to assist in removing alien vegetation from the river banks, meaning that indigenous wetland vegetation necessary to stabilize the watercourse would not be able to re-establish. The Department of Agriculture decided to utilize the little available funding for this kind of work elsewhere on farms where landowners were prepared to make a contribution to the project and by so doing, accept some responsibility for the project.

Several large floods were experienced after 2003.  These rapidly extended and deepened the gulley, which by 2015 was close on a kilometre long and up to 13 m deep over a considerable length. At least a 600 m length of pristine peat wetland up to 100 m wide had been destroyed in this process. The damage was not limited to the gulley itself, but downstream another section of the same wetland has been smothered in sediment,  a road bridge was blocked with sediment, and sediment islands moving down the watercourse have diverted the flow of water into the river bank, initiating fresh erosion and de-stabilizing the river further. The impact of the small gulley initiated by the failure of the dam has now affected several kilometres of the wetland.

Objectives for undertaking rehabilitation

Major rehabilitation works have been constructed to arrest further erosion.  The primary objective of the intervention on the site is to prevent the further upstream degradation of the wetland, and stabilize the remaining soil in the gulley, so as to help stabilize the downstream section of river. The intended spin-offs from achieving the primary objective would include the partial restoration of the damaged wetland’s biodiversity and habitat, removal of alien vegetation, and most importantly from an agricultural perspective, reducing the sediment load emerging un-naturally from the catchment.

Intervention that was undertaken

There were several attempts to plan the project, first with 3 weirs, then 4 and eventually 5. Unfortunately with the first attempts to plan the project, delays during the sourcing funding and statutory authorisations compelled the redesign of the project because the erosion had progressed further up the slope. In 2015 construction of a series of gabion weirs each around 4 m high, which would enable the safe dropping of flood-water through the gulley (a total 23 m fall over 600m) began. The project cost was in the order of R20 m, and apart from a size-able earthworks and landscaping operation, it involved the packing of 7000 cubic meters packed in wire gabions.

Specific design issues

The following were design issues specific to this site: –

  • The weirs were all founded on peat of an undetermined depth, and the potential consolidation of the peat was a concern. Gabion weirs were an obvious choice over concrete weirs because of the material’s ability to flex and accommodate foundation movement. To minimize the risk associated with foundation consolidation, the structures were designed with broad foundations which spread the load on the foundation (this was made more economical by creating a large void in the body of the weir which was filled with compacted soil and sealed in a geotextile envelope). 
  • As with all weirs, protection against piping failure was very important. A commonly used practice is to drape a hyper-liner over the upstream edge of the structure and extend it upstream for a few meters before burying its tip in a trench. Although this method is very successful with this kind of weir because it cuts off water flowing under the structure. In a peat soil the concern is that it may work too well, and cause the peat to dry out and crack. Should cracks in the peat develop and oxygen reach the soil, the thousands of years old peat resumes decomposing and undergoes a volume change, leading to cavities under the structure. With the potential drying out of the peat in mind, the hyper-liner was rejected as an option, and two gabion cut-off walls were added below the structure instead to extend the flow path under the structure, but at the same time to allow the peat to remain saturated.
  • There is no alternative watercourse or floodplain for surplus water to be diverted to during high flood events, so the weirs had to be designed to safely accommodate a large flood, and so the spillways were designed for the expected 1:100-year flood. 
  • The primary objective of the weirs is to prevent the un-natural migration of sediment downstream. To achieve this, the channel between the weirs was designed to have a very low flow velocity, not only to keep flow velocities to a level where most sediment won’t move, but also to create an environment where the wetland vegetation may be re-established, and this would further hinder the movement of sediment. The channel is around 16 m wide and has a longitudinal slope of 0.5 %.  The foundation and crest levels of the weirs were arranged to achieve this.  

Outcomes of the rehabilitation action

An important lesson has been learned from this case study. The vulnerability of palmiet peat wetlands to catastrophic erosion once small knick-points have initiated must be recognized. Had there been funding to address the erosion in 2003, the work would have cost less than 10% of the final project cost. Apart from the cost of offsite damage resulting from the abnormal movement of sediment down the river, peat wetlands themselves have a very special value and must be protected (see other blogs on this site). Once knick-points develop in the peat (which has formed over thousands  of years), the peat soils which support highly endangered ecosystems can be destroyed by just a few relatively insignificant floods in a few short years. 

If the few remaining peat wetlands are to be protected, a high priority must be given to rehabilitation works as soon as the initial damage is recognized.  Early repairs are far cheaper and protect a greater area of upstream wetland and at the same time minimize downstream sedimentation impacts.


Survey and design: Western Cape Department of Agriculture

Funding: National Department of Agriculture – Disaster Relief

Main contractor: B-Waitabasa

Project management: CASIDRA