Sockeye return to the upper river for the first time in seventy years

The first fish came through on a Tuesday. Nobody was expecting it that early, and the two counters on shift spent a minute arguing about what they had seen before either of them wrote anything down.

By Friday the tally was past four hundred. By the following Wednesday it had cleared four thousand, and the fisheries office had stopped calling it an anomaly.

The fence goes in the second week of June and comes out when the water drops. In between, two people stand in it for two hours at a time, in waders, counting fish by eye and marking a tally sheet that has not changed format since 2017.

What the agreement actually changed

Three things, in the order they were argued over: the summer release schedule, the temperature trigger, and who holds the gauge data. The last one took longer than the other two combined.

The gauge had been read by the utility since it went in. The agreement moved the reading, the publishing and the archive to the fisheries office — the part nobody expected to win, and the part that has changed the most.

Before the agreement the official number came from a gauge the utility read, published quarterly, and reconciled to nobody. The fisheries office kept its own count the whole time. For six years the two numbers sat side by side and only one of them was allowed to matter.

My grandmother described this river to me and I did not believe her. I thought she was describing a different place.

Marcel Sam, fisheries technician

The counts are still a fraction of what the river carried before the dam. Nobody at the fence pretends otherwise. What has changed is that the number is no longer zero, and it is no longer somebody else's number to publish.

Counting continues until the water drops below the trigger, which on current forecasts means the second week of September.

What happens after that is the question nobody at the fence wants to answer out loud. One good year is a good year. The flow agreement runs to 2031, and the temperature trigger — the clause that took the second-longest to write — has not yet been tested by a genuinely hot August.

The office will publish the final count in October, the same week the panel sits. That timing is not an accident.

There is a version of this that is only about fish, and it is not the version anyone at the fence tells. The dam went in without consent and the run went out within a decade of it. Everything since — the studies, the panels, the twenty-year fight over whether the loss was even attributable — has been about who gets to describe what happened to a river.

That is why the gauge clause mattered more than the release schedule. A number nobody here produced can be revised by people nobody here elected. A number produced at this fence, by these staff, on a form that has not changed since 2017, is harder to argue away and easier to build on.

The elders who walked the bank on Sunday were not there for the count. Several of them had been at the hearings in the eighties, and two had testified. One of them asked the counters to say the number out loud rather than show her the sheet.

She said she wanted to hear somebody say it.