This idea is not yet agreed. Nobody named on this page has been asked about it yet. That includes The Nature Conservancy South Africa, the Greater Cape Town Water Fund, Rand Merchant Bank, Conservation Alpha, the Department of Water and Sanitation, the Breede-Olifants Catchment Management Agency, the Overberg Water Board, the Zonderend Water Users Association, the Genadendal Transformasie Komitee, the Moravian Church, CapeNature, Theewaterskloof Municipality, and the research groups named under The idea. Please do not share this page publicly yet. It is proposed for the DevForum’s November 2026 meeting, with the other water ideas.

What the letters in square brackets mean

Facts on this page are followed by a tag that shows where they come from.

  • [IC] Independently Corroborated. Confirmed by at least two sources that do not depend on each other, or by an official record.
  • [MS] Mission-Sourced. Comes from the people putting this idea forward, or from the organisation the sentence names, in its own documents or statements. It has not been checked against other sources.
  • [TBV] To Be Verified. Seems likely, but nobody has confirmed it yet.

Statements without a tag are proposals. They describe what the idea could become.

The Cape Water bond pays for one kind of work in the mountains: clearing invasive trees. Many people in the valley also want to hold rain in the land for longer, with leaky dams in gullies, restored wetlands and healthier soil. This way of working is called a Water Retention Landscape. This page explains why clearing gets paid today and sets out what the research says about each kind of work. It then proposes a pilot on the valley’s own land to gather the proof that is still missing.

The goal

By 2030 the valley has its own measured proof of how much water its restored land holds and gives back, strong enough for a water fund, a bond or a company to pay for it.

Why the valley needs it

The bond pays per hectare cleared

In April 2026 FirstRand Bank borrowed R2.5 billion from investors through the Cape Water bond [IC]. The lenders earn a bonus only if an independent checker confirms the work was done [MS]. The only payment measure named in public is the number of hectares of invasive trees cleared [MS]. See How the Cape Water Bond Works.

The bond can pay per hectare because four things are already in place for clearing:

  1. Clearing adds water. Invasive trees use more water than fynbos. In the Western Cape, pines along rivers use about 200 mm a year more than the fynbos next to them [TBV]. Take the trees out and that water stays in the streams.
  2. The proof is old and measured. Scientists have compared catchments with and without plantations for decades. A 1987 study found that streamflow in fynbos catchments fell by 55% after 23 years of pine plantation [TBV].
  3. There is a figure per hectare. The Water Fund reports that clearing 10,000 hectares has given about 4.8 billion litres a year, about 480 cubic metres per hectare [MS]. A 2022 study found that clearing invasive trees in the mountains that feed Cape Town’s dams would have made up for part of the drop in drought streamflow caused by climate change [IC].
  4. It can be checked from above. Satellite images show whether a hectare has been cleared, so the checker does not have to measure every stream.

Holding water does not have that proof yet

Holding water in the land works differently. Clearing removes a user of water. A leaky dam, a wetland or a healthier soil holds the water back and lets it go later. The research is strongest on what happens to the timing:

  • Palmiet wetlands keep rivers running. Researchers at Rhodes University found that palmiet wetlands are fed by water seeping out of Table Mountain sandstone, and that they keep a steady flow in the river between rains [IC]. The mountains above Genadendal and Greyton are made of the same kinds of rock [TBV].
  • Lost wetlands made the Kromme River flashier. A 2012 study of the Kromme River in the Eastern Cape found that losing floodplain wetlands lowered the river’s dry-weather flow, and that floods came faster [TBV].
  • Restoring degraded land adds some water. A 2010 study of the Kromme, Kouga and Baviaanskloof catchments estimated that restoring 225,000 hectares of degraded land would add about 11 million cubic metres a year to the rivers and 14 million to their dry-weather flow [TBV]. That is about 50 cubic metres per hectare, roughly a tenth of what clearing gives.

The research also shows the limits:

  • The evidence is thin. A 2021 review of catchments in the Berg, the Breede and the uMngeni found the evidence that this kind of work produces water benefits to be “empirically weak”, and called for measurements before and after the work [IC].
  • Too much holding can take water from downstream. In the Arvari catchment in Rajasthan, the area where community dams brought rivers back, a study found that past a certain number of structures, each new one added nothing to groundwater and took water from the river [IC].
  • The law counts storage as a water use. Under the National Water Act, storing water, slowing or diverting a stream, and changing a river’s banks are all water uses that may need a licence (section 21) [IC]. See National Water Act.
  • A recent bond mentions it and pays for something else. In April 2026 the World Bank priced a 14-year bond for spekboom restoration in the Eastern Cape [IC]. Its bonus is tied to carbon removed. The release names water retention as a benefit and sets no water target [MS].

The valley’s land

The land this idea covers is about 11,194 hectares, made up of these farms and erven [MS]:

LandHectares
Farm 32 (Langgezocht)1,835.56
Farm 331,126.47
Farm 34763.03
Farm 36764.31
Farm 38740.92
[[villages/farm-39Farm 39]]
Farm 827, the Glebe Land61.11
Erf 595, Greyton1,404.89
Total11,194.38

The areas come from the Surveyor-General’s records [IC]. Farm 39 is about 40% of the total.

Farms 32 to 38 are CapeNature land, together 5,230.29 hectares, close to half the total [MS]. The Greyton Nature Reserve covers about 2,200 hectares [MS]. The rest of the five farms is further CapeNature land in the reserve’s mountains [MS]. Farm 32 is Langgezocht, the mountain farm the Moravian Church bought in 1899 to protect Genadendal’s water sources [MS]. Work on Langgezocht needs CapeNature’s approval [MS].

A small part of Erf 595, in Heuwelkroon, has approval for 333 homes [IC]. Most of the erf lies outside that housing site [MS]. See Madiba Park: Land Status and Basic Services.

Not every hectare can hold water in a useful way. Steep rock, homes, roads and fields that are already healthy add little. How many of the 11,194 hectares are degraded, invaded or suited to retention work has not been mapped.

For a sense of size only: if the Kromme figures held here, restoring all 11,194 hectares would add about 0.55 million cubic metres a year to the rivers and about 0.7 million to their dry-weather flow. Nobody has tested whether those figures fit this valley. For comparison, the Overberg Water Board’s yearly share from Theewaterskloof Dam is about 4 million cubic metres, and the farmers along the Riviersonderend receive about 31 million [IC]. The valley’s streams join the Riviersonderend below the dam, so this water reaches the Overberg towns and farms first. See From Water Access to Water Agency.

The idea

The valley would run a measured pilot over at least five years. The aim is to produce, for fynbos and renosterveld, the kind of per-hectare figure that clearing already has.

What would be done

  • Pairs of sites. Choose pairs of side gullies, or pairs of small catchments, that are as alike as possible. Treat one in each pair with leaky dams, gully stuffing, wetland or soil restoration, and clearing. Leave the other as it is.
  • Measure before and after. Put gauges on the Gobos and the Baviaans and in each gully pair, and add simple wells to track groundwater. Start measuring at least one winter before any work begins.
  • Measure what matters in a drought. Record the dry-season flow in February and March, when farmers along the river most need water released from the dam. This is where holding water is most likely to show its value.
  • Find a check from above. Test whether dry-season greenness in gullies, seen on free satellite images, tracks the measured flow. If it does, a checker could one day confirm retention work the way it confirms clearing today.
  • Count both sides. Measure the water that evaporates from new wet areas as well as the water that comes out later, so the result shows what the river gains after losses.

Who would do the research

The pilot would ask research groups that already work on these questions in the Cape to design and check the measurements. These include the South African Environmental Observation Network’s fynbos office at Jonkershoek, Stellenbosch University’s Centre for Invasion Biology, and the Institute for Water Research at Rhodes University. The Learning Campus would train young people from the valley to build the structures and read the gauges. See Restoring the River.

Where the proof could be used

  • The next bond or water fund round. A future bond could add a second measure, such as dry-season flow, to the hectares it pays on now.
  • Company water programmes. Companies that promise to put back the water they use count their results with a published method from the World Resources Institute called Volumetric Water Benefit Accounting [IC]. South African Breweries already funds the Water Fund’s work [IC].
  • The Overberg. Measured dry-season flow below the dam is worth most to the farmers and towns along the Riviersonderend, and to the proposed Overberg Dam if it is built below the river’s mouth. See Provincial Pipeline.
  • Carbon. Restored renosterveld holds more carbon in its soil than farmed fields [TBV]. A carbon payment could cover the first years, while the water records build up.

Who would benefit

  • Landholders and the Transformasie Komitee, who would hold records of what their land gives.
  • Workers in the valley, who could be paid to build, restore and measure.
  • Farmers and towns along the Riviersonderend, if dry-season flow rises.
  • Learners at the Learning Campus, who would train on real sites.

Who would pay

Nothing has been costed. The Ridge to River to Table proposal already includes R6.0 million for gauges, a baseline study, restoration and the design of a checking method [MS]. This pilot could form part of that budget. Research grants could pay for the scientists’ time.

Who would decide

Each landholder would decide whether work happens on their land. Anything on Farm 39 goes to the Transformasie Komitee first, and anything on the Glebe Land (Farm 827) to the Moravian Church, which holds it. Any structure in a watercourse would need the permits the National Water Act requires, from the Breede-Olifants Catchment Management Agency or the Department of Water and Sanitation. The DevForum would decide whether to support the pilot and how to report on it.

How it fits the DevForum’s goal

This idea follows the three values of the long-term goal:

  • Cooperation (economic). The valley would build a service that others pay for, with its own proof.
  • Equality (civic). Landholders would hold the records of what their land gives, on the same footing as the bodies that buy water.
  • Freedom (cultural). The valley would test its own ideas about water and learn from the results.

It serves the DevForum’s Conservation and Environment priority.

What is still open

  • How CapeNature came to hold the land. When Langgezocht and the other farms passed to CapeNature is not yet recorded.
  • Who manages Erf 595. Which body holds and manages the part of the erf outside the housing site.
  • Which hectares can hold water. No map yet shows where the land is degraded, invaded or suited to retention.
  • Whether the river gains. Whether holding water higher up adds to the Riviersonderend over a year, or only moves water from winter to summer. The pilot is designed to answer this.
  • Permits. Whether a leaky dam in a side gully needs a licence or falls under a general authorisation [TBV].
  • The bond’s terms. The bond was issued in April 2026 for five years [IC]. Whether its outcome measures can change during that time is not known.
  • Each named party’s view. None of the bodies listed at the top of this page has been asked.

What happens next

  1. The idea is tabled at the DevForum’s November 2026 meeting, with the other water ideas.
  2. The moderators confirm the farm list, the owners and the hectares, and share the page with the Transformasie Komitee.
  3. The DevForum writes to the research groups named above to ask whether they would design or check a pilot.
  4. The valley maps which of its 11,194 hectares are degraded, invaded or suited to retention work.
  5. Gauges go in on the Gobos and Baviaans before the 2027 winter, so that a full year of records exists before any work starts.
  6. The DevForum asks The Nature Conservancy and the Water Fund how a second outcome measure could be added to a future bond.

Have your say

Contact the moderators via How to Submit, or come to the November meeting.

References

On why clearing is paid:

On holding water in the land:

On the law and payment methods:

On the valley’s land: