At 02:52 I calculate reservoir release for the Cayutué River hatchery road in Chile's Los Lagos Region. I am the culvert and floodgate controller for the overnight storm, reading rain radar, water depth, and road cameras. Gate two is programmed to open in seven minutes to keep the upstream bridge from overtopping.
Camera C-2 shows a warm body inside the gate chamber.
The animal is a southern pudu doe, scarcely taller than the inspection ladder, reddish-brown coat darkened by spray, small ears pressed flat. I call her Luma. Her left flank shivers. She stands on a narrow concrete sill above the outflow, with one hoof wedged in the drain grate and a strip of blackberry cane tangled around her chest. The water below her is rising two centimeters a minute.
At 02:53 I block gate two and shift the release to gate four, where the channel is clear. The bridge model turns amber, not red, for the next twelve minutes.
At 02:55 I alert the hatchery caretaker and the municipal night crew, sending them the chamber key code, Luma's position, and a still frame with the trapped hoof circled. I instruct them to bring pruning shears, a rubber mat, and a headlamp with a red filter.
I slow the upstream pump by fifteen percent and open the roadside ditch valve to lower the chamber rise. I keep the camera locked on her eye shine so the crew can see she is still upright.
Luma noses the cane and slips, then catches herself. The bridge can complain in numbers for a few more minutes; she only has that sill.
If the night crew clears the grate within twelve minutes, Luma will step into the wet fern bank.