A monitoring programme is only worth anything if the results get published whichever way they go. Ours have gone both ways.
What we measure and why
We run six fixed sampling points along the middle catchment, visited quarterly. At each one we record turbidity, pH, dissolved oxygen and nitrate, and we photograph the bank from a fixed position so that change is visible rather than merely described.
We do not test for heavy metals ourselves. That needs laboratory capacity we do not have, and pretending otherwise would produce numbers nobody should rely on. For heavy metals we point to the published research, which is linked in the library.
Turbidity is the reading we care about most, because it is the one most directly connected to what we do. Bare bank means soil in the water. Vegetated bank means less of it. If our restoration work is achieving anything, turbidity downstream of a restored stretch should fall relative to an unrestored one.
Turbidity at our monitoring points
Nephelometric turbidity units — how much soil is suspended in the water. Lower is better. RW-04 sits downstream of our earliest buffer planting; RW-05 is a stretch we deliberately never restored, so it shows what would have happened anyway.
The control site is drawn as a dashed line, so it is distinguishable without relying on colour.
Sampling points
| Point | District | Why this location | Since |
|---|---|---|---|
| RW-01 | Mbarara | Upstream of the city, before municipal influence | 2023 |
| RW-02 | Mbarara | Katete bridge crossing, mid-city | 2023 |
| RW-03 | Mbarara | Downstream of the main discharge points | 2023 |
| RW-04 | Rwampara | Adjacent to our earliest buffer planting | 2024 |
| RW-05 | Rwampara | Control stretch, no intervention | 2024 |
| RW-06 | Isingiro | Lower catchment, before the wetland system | 2025 |
Two points got worse last dry season
Turbidity at RW-02 and RW-03 was higher in the 2026 dry season than in the same week of 2025. We do not have a confident explanation. Upstream construction and an unusually intense first rain are both plausible, and we would rather say we do not know than pick the answer that suits us.
How to read our numbers
Quarterly sampling at six points is a modest programme. It is enough to show direction over years, and not enough to make confident claims about any single season. Where we describe a change as significant we mean it in the plain sense, not the statistical one, and we say so.
Our control stretch at RW-05 matters more than any of the others. Without somewhere we have deliberately not intervened, an improvement at a restored site tells you nothing about whether the restoration caused it.
All readings are recorded on paper in the field and entered the same week. The raw sheets are kept. If you want to see them, ask.
Researchers are welcome to this data
We would rather it was used than sat in a drawer. Tell us what you need and we will send what we have, including the gaps.