July heat wave triggers king salmon die-off in front of Petersburg hatchery

The Crystal Lake Hatchery near Petersburg. (Alaska Department of Fish and Games photo)

By Orin Pierson

Petersburg Pilot

Hatchery managers had hoped that June’s steady rainfall would spare them from having to intervene in this year’s king salmon run, allowing fish to reach the Crystal Lake Hatchery in Petersburg naturally, without the stress of human handling.

Those hopes evaporated in late July when a few hot days right at the wrong time caused a significant mortality event for king salmon transiting the shallow waters off Mitkof Island between Blind River Rapids and the hatchery.

On July 20, after observing 100 or more king salmon dead in the slough’s shallow pools, hatchery staff with help from the U.S. Forest Service, Alaska Department of Fish and Game and volunteers launched a beach seine rescue operation — like last summer — to move as many of the surviving kings as possible up to the hatchery before they succumbed to the warm water and low oxygen conditions.

“It was a bit of a tragedy,” said Bill Gass, production manager for the Southern Southeast Regional Aquaculture Association, which operates Crystal Lake Hatchery.

The die-off is a testament to the gauntlet that salmon face during the last few miles of their approach to the hatchery. Fish that had made it up Blind Slough during cooler and rainier conditions found themselves trapped in isolated pools when temperatures rose. As the water heated to about 75 degrees Fahrenheit, dissolved oxygen levels in the water plummeted while the fish’s metabolic demands accelerated.

“They couldn’t go upstream, so they came downstream, and they hit the rapids,” Gass explained. As “the fish that were up there schooled up” in the shallow pool near the rapids, they ran out of oxygen.

Jeff Rice, Alaska Department of Fish and Game sportfish area manager for Petersburg and Wrangell, described the grim discovery: “We flew a drone over. … I think initially we were talking dozens, then we were talking maybe 100, and then they thought, with the drone up there looking, it could have been more.”

The exact toll remains uncertain because within six to eight hours, a strong tide swept through the area, flushing many of the bodies away. But the observed mortality was enough to trigger a swift response.

The regional aquaculture association had first attempted the beach seining effort last summer, successfully moving 146 live king salmon from the rapids to the hatchery over two July operations with just one mortality. The effort last year proved crucial to support the hatchery’s struggling broodstock returns.

This year’s June weather — cool and wet, ideal conditions for migrating salmon — initially suggested a repeat rescue operation might not be needed at all. Steady rainfall in early July maintained favorable conditions for fish moving naturally upstream.

“I was feeling like, well, these fish should be in good shape,” Gass said. “It’s really cool and we’ve had good water flows.”

But the weather changed; after four days without rain, cloud cover failed to materialize, and temperatures exceeded what the fish confined to shallow pools could survive.

When Gass surveyed the rapids at 1 p.m. on July 20, he found relatively few fish. Returning at 6:30 p.m., the scene had transformed dramatically — dozens of highly agitated kings packed into the shallow pool, their frantic behavior signaling distress. By the next morning, many of those fish were dead.

During the emergency beach seining operation on July 21 managers noted the “very warm” water. Despite moving quickly, 10 of the 80 fish died in the transferring process, a mortality rate they considered too high to continue. “We finally said, yeah, we’re doing more harm than good now,” Gass recalled.

Around 10 days later, cooler conditions allowed for a second, much more successful operation that moved 79 females with zero mortality.

Combined, the rescue operations secured approximately 120 female kings for the hatchery’s broodstock program, representing an anticipated 350,000 to 400,000 eggs critical to the facility’s future production — a shortfall, but much more manageable than the alternative.

Crystal Lake Hatchery aims to release 700,000 juvenile kings annually, but last year’s weak returns left the facility with only about 550,000 eggs.

The hatchery operates as part of a network of three Southeast Alaska facilities that all raise their king salmon from Andrews Creek ancestral broodstock, allowing them to share eggs when one facility has surplus.

In 2024, all three hatcheries fell short of their goals, leaving no excess for sharing. This year appears more promising, with the other two facilities potentially on track to have surplus broodstock.

“I’m hopeful that there will be maybe some help to get us over the hump,” Gass said, though he won’t know for certain until later this month.

• This story was originally published by the Petersburg Pilot.