Nothing here happened suddenly. That is the difficult part — a slow change is easy to live alongside and hard to notice until a threshold is crossed.
Six decades of change
- 1960s
A forested catchment
Buhweju's headwaters carried substantially more forest. Flow was steadier, the dry season less severe, and the basin population a fraction of today's.
- 1970s
Pressure begins
Political instability and displacement moved people onto marginal land. Cultivation started climbing hillsides that had held forest.
- 1980s
Wetland cultivation spreads
Wetland margins are drained for crops. The practice is locally rational and cumulatively serious.
- 1990s
Bank vegetation goes
Riverside strips are cleared systematically rather than in patches. Sediment loads rise. Sand extraction becomes commercial.
- 2000s
Urban demand climbs
Mbarara grows fast. Abstraction increases, and so does the volume of untreated discharge entering the channel.
- 2020
City status
Mbarara becomes a city. Investment, population and demand all step up together, and the river's condition becomes a national news story.
- 2021
The evidence firms up
Peer-reviewed work documents heavy metals in the Rwizi ecosystem and assesses the risk to human health. The problem is no longer anecdotal.
- 2022–2024
Coordinated restoration
Government, districts and NGOs map catchment areas, demarcate buffer zones, plant bamboo and restore wetland at sites including Kakigaani in Rwampara. Kilometres of bank are restored around Mbarara.
- Now
Partial, real progress
Restoration works where it is done and follow-up happens. It is not yet happening at the scale of the pressure. That gap is the reason this organisation exists.
On dates in this timeline
Decade attributions are indicative rather than precise. They are drawn from published catchment assessments and Ugandan press reporting, and they describe trends rather than events with fixed start dates.
Why a slow change is the hard kind
Every line in that timeline describes something that was reasonable for the person doing it. Cultivating a wetland margin is the best land available to a household that has run out of hillside. Cutting bank vegetation clears ground that grows well. Selling sand pays school fees. None of it is vandalism, and treating it as vandalism is why a good deal of restoration work in this catchment has failed.
What makes it hard is that no single decision registers. A metre of bank cleared is invisible; six decades of metres is a different river. By the time a threshold is crossed the practice is normal, widespread, and load-bearing for thousands of household budgets — which is why our work spends more time on livelihoods and negotiation than on planting, and why it moves slowly.
The corollary is more hopeful and equally slow. Recovery accumulates the same way. A stretch that holds for three years stops moving; a bank that stops moving stops adding to what the next stretch downstream receives. Nothing about that is dramatic enough to make a timeline entry until enough of it has happened.
What each decade cost the river
| Change | How it reaches the water |
|---|---|
| Headwater forest cleared | Rain runs off instead of soaking in, so the wet season peaks higher and the dry season falls further |
| Wetland margins drained for crops | The papyrus that filtered and stored water is gone, and with it the buffer between land and channel |
| Bank vegetation removed | Nothing holds the bank, so soil enters the channel and the channel widens and shallows |
| Commercial sand extraction | The bed is altered directly, and the disturbance travels downstream of the site |
| Urban growth at Mbarara | More abstracted from the river, and more discharged into it, in the same reach |
| Eucalyptus on wetland margins | A fast-growing plantation species drawing heavily from exactly the ground that used to hold water |
What would have to change for the next entry to read as recovery
Not predictions. These are the conditions the people working on this river generally agree are necessary, and none of them is within any single organisation's gift.
- Buffer zones enforced consistently rather than demarcated and then left, which requires district capacity rather than another mapping exercise.
- Headwater forest in Buhweju protected, since everything downstream inherits what the source does.
- Livelihood alternatives reaching the households that would otherwise have to keep cultivating the margin.
- Discharge from the premises that hold permits actually meeting the terms of those permits.
- Monitoring that continues between funding cycles, so a claim of recovery can be tested rather than asserted.
The next entry is not written yet
Whether the 2030s line reads as recovery or continued decline is decided by what gets funded and enforced in the next few years.