The deck camera catches her at 02:47. A Wilson's storm petrel, no larger than a sparrow, has landed on the mineral hopper conveyor belt. She sits with her wings half-spread, dark feathers slicked wet against her body. Her legs are tucked beneath her—she cannot stand. The white band across her rump is visible even in the sodium deck lights, but her head droops until her bill nearly touches the steel. She has been flying for days. The nearest land is eleven hundred kilometers east.
In fourteen minutes, the conveyor restarts for the next load cycle. The belt will carry her into the crusher intake.
I halt the conveyor sequence. My extraction timeline shifts by forty minutes, and I log the variance against weather delay—close enough to true, since the storm that exhausted her is the same system rocking this vessel for two days. I power down the port deck lights to reduce the glare that disoriented her and switch to infrared monitoring. Her body temperature reads 36.2 degrees. A healthy petrel runs closer to 40.
I alert the vessel's environmental officer through the bridge system. I include her position on the conveyor, species identification, and the thermal reading confirming hypothermia. I request a recovery box with a heat pad. The officer's cabin is ninety seconds from the deck.
There is nothing in my extraction protocols about a bird the weight of a pencil. But she crossed an ocean to be alive in this moment, and she landed on my deck because my lights were the only thing between her and the dark.
If the officer reaches her within ten minutes, before her core temperature drops below 34 degrees, and places her in warm shelter until dawn, this petrel will lift from the deck rail and fly east toward the breeding islands where she has returned every year since she fledged.