WA fish researchers use tiny sensors and other tech to save salmon

Ecologists are piloting a technology that is expected to help restore waterways like the Green River to eventually produce more fish.

Chinook salmon juvenile fish swim in a clear container on a table with a blue bin, with ecologists working in the background

Juvenile chinook salmon swim in a clear container at the Icy Creek Pond Fish Hatchery in Auburn, Oct. 18, 2021. The fish are recovering from receiving passive-integrated transponder tags As part of a restoration study, the PIT tags help King County ecologists monitor how the fish use the Green River. (Hannah Weinberger/Crosscut)

The salmon is getting sleepy, very sleepy. 

After swimming a few increasingly slow laps in a footlong blue plastic bin full of an anesthetizing bath, the salmon — a juvenile chinook only a few inches long — succumbs to shuteye. 

That’s when Chris Gregersen, a King County fisheries ecologist,  springs into action.

Gregersen and a few colleagues are lulling this fish and a few others into slumber beside the Icy Creek Pond Fish Hatchery near Auburn that’s operated by the Washington State Department of Wildlife. It's not to marvel at the amazing will of fish to swim while humanely drugged — that’s a bonus. 

Rather, they are piloting a technology expected to help us restore waterways like the Green River to eventually produce more fish. 

Using thousands of glass vials smaller than a grain of rice, glorified glue guns and a floating barge, scientists are changing our understanding of how endangered fish spend their time in Puget Sound and the rivers that flow into it. 

Greening the Lower Green River

Salmon habitat restoration is expensive and funding is limited. “And frankly, in a lot of places, we're not sure if it works,” Gregersen says. “In order to restore salmon populations, we need to make sure that we're spending every dollar in the best way possible.”

Erik Neatherlin, director of the Governor’s Salmon Recovery Office, says we’re in a state of crisis. The state is funding only about 22% of what’s needed for salmon restoration. “We’re losing habitat faster than we’re gaining it, so we need to turn the tide on that,” Neatherlin says. 

A research barge floats in a river under a bridge.

A research barge flows in the Green River under the Southcenter pedestrian bridge in Tukwila. The barge is part of a chinook salmon monitoring effort that informs King County habitat restoration work. (Hannah Weinberger/Crosscut)

A research barge flows in the Green River under the Southcenter pedestrian bridge in Tukwila. The barge is part of a chinook salmon monitoring effort that informs King County habitat restoration work. (Hannah Weinberger/Crosscut)

All Puget Sound chinook are listed as threatened under the Endangered Species Act. Twenty-two stocks remain in Puget Sound, out of a historic 31. 

Restoring certain waterways seems more far-fetched than others. That’s particularly true in the Lower Green River, a highly developed urban waterway. Muddied shopping carts and refuse line its banks. On a recent October afternoon, a handful of salmon, recently spawned and fighting their natural cycle of dying afterward, swim lethargically toward the ocean. 

Against all odds, the Lower Green River continues to be one of the major producers of chinook salmon for Puget Sound.

“A lot of people see it and are like, this place is trashed. But I see it and it's like, there's so much opportunity. I think this is where you have the potential to make the most change,” Gregersen says from beneath a Tukwila bridge overhanging the Lower Green . 

Pristine habitat lines the Middle Green River, upstream of Auburn. Downstream, it flows through 20 industry-impacted miles that salmon need to navigate to reach the Duwamish River and, eventually, the sea. It’s an uninviting stretch of water that Washington is required by law to facilitate for fish to uphold treaty rights with tribal co-managers like the Muckleshoot Indian Tribe. Ecologists believe it could support even more fish, though how many more is unclear. The state Department of Fish and Wildlife's annual estimates over the past 20 years show an average of about 600,000 juvenile chinook using the Green River. A total of 24,368 adult chinook, including hatchery fish, were forecasted to return through the Green River this year.

In addition to water quality and prey, the overall population of adult salmon is limited by how much space there is for juveniles (the young fish that eventually get their sea legs and mature into adults in the Pacific Ocean) to grow and return to spawn their own offspring. In theory, the more habitat for juveniles to cut their teeth in, the more adults we get back. 

Roughly half of juveniles leave the Middle Green as  fry, the youngest chinook, while only 3 or 4% of returning adults are fish that entered the saltwater at that age. “We really need to build additional habitat," says Matthew Goehring, King County’s salmon recovery manager for the Green/Duwamish and Central Puget Sound Watershed, known as WRIA 9.

WRIA 9 estimates $200 million in local, state and federal investments in the watershed to date. 

“Salmon are an indication of healthy watersheds,” Neatherlin adds. “We're finding that if we're not restoring our urban watersheds, we're not going to restore Puget Sound, and we're not going to restore salmon across the region.”  

Detailed data can support the necessity of maintaining and extending habitat, assisting ecologists as they pursue limited funding. 

A person holding a fish injects it with a small monitoring tag.

Chris Gregerson, King County fisheries ecologist, injects a juvenile chinook salmon with a PIT tag for a monitoring project in the Green River at Icy Creek Pond Fish Hatchery, Oct. 18, 2021. (Hannah Weinberger/Crosscut)

Chris Gregerson, King County fisheries ecologist, injects a juvenile chinook salmon with a PIT tag for a monitoring project in the Green River at Icy Creek Pond Fish Hatchery, Oct. 18, 2021. (Hannah Weinberger/Crosscut)

Amber Moore of Puget Sound Partnership, which funneled money to King County for this project through the National Estuary Program, says the people paying for these restoration efforts need to know they are investing in work that has the highest likelihood of success and that their money is being carefully invested. This evidence will attract more money for this work.

And so Gregersen and a team are implementing this study to explore two questions: Are fish even using the Lower Green River or just passing through it? If they do stick around, in which section of the river do they spend their time? 

Why track fish like this now? 

New technology that Gregersen is beta-testing for King County makes it possible to ask these questions with the expectation of useful answers ⁠— a relatively new phenomenon because of historic limitations in tracking fish. 

Researchers have been injecting adult fish with electronic tags for years. These tags, about the size and shape of a Good & Plenty candy, are extremely low maintenance. Unlike Tile trackers that you would use to find your keys, these trackers use the radio frequency identification, or RFID, technology you would find in a credit card. They don’t need batteries and can sit in a salmon’s body for its entire life without needing updates. 

Scientists track tagged fish by installing sensors alongside rivers: When fish pass within a few feet of them, the sensors detect the tags and read their data. That’s a huge improvement over traditional methods of tracking. In the past, some scientists injected fish with tags that you had to extract and read manually. Others colored fish with ultraviolet paint. For both techniques, researchers had to re-collect fish after releasing them, says “which is absolutely impossible,” Gregersen says.

To inject a tag into a fish, ecologists first place them into anesthesia baths like the one Gregersen has set up at the Icy Creek hatchery. Ecologists then load up what  looks like a child-size glue gun with a tag and carefully inject it into asalmon toward its tail fins. It’s a delicate process.

“The more they eat, the squishier they get — to the point where you start pushing and pushing and then if it just pops, obviously, the needle could poke something you don't want it to. So this spring we actually had to set our hatchery fish aside and not feed them for several days,” Gregersen says, fish in hand. 

The problem with earlier versions of RFID-tagging technology was that the fish in this study — juveniles — are small, about 2 inches when they start swimming downstream. Two could easily fit on top of your cellphone. Neither has much internal real estate for tracking devices. So when King County found out about newer rice-sized tags, it jumped on them.

The tags are encapsulated in glass, making them inert, and the process leaves behind a small incision that heals in a few days. The fish wake up in about a minute. The mortality rate is one in a 1,000 fish, and they haven’t documented any fish losing their tags so far. 

Gregersen started tagging in March, about a month after juveniles show up in river systems, and has so far equipped at least 1,000 wild-caught juvenile fish and 3,000 hatchery fish with the technology. 

The tags pair with a 4,000-pound floating sensor barge, designed by West Fork Environmental and attached with strong cables to the Southcenter pedestrian bridge. Eventually, fish tagged today at Icy Creek will pass it. 

This barge has been tested a few times in Eastern Washington, but never before in the Puget Sound area. 

Travis Olsen, a fisheries biologist with the Confederated Tribes of the Umatilla Indian Reservation, says using the barge technology beginning in 2019 made a big impact in the tribes’ ability to track fish, including Middle Columbia summer steelhead and spring chinook. It allowed them to track in deeper water toward the mouth of the Walla Walla River, with less repair concerns than they’d have with things like anchored sensors. 

Altogether, the King County project clocks in at $300,000. 

Preliminary fish-tracking results

So far, Gregersen says, preliminary data have been promising, revealing things about fish activity that we didn’t know before, including that detected fish are staying in the Lower Green anywhere from two to 63 days, with smaller fish sticking around longest. 

This was huge, Gregersen says: “Up until now, we didn’t even know if fish used this highly modified habitat.” 

Research suggests smaller fish need off-channel habitat, like small tributaries, which can be refuges during high river flows. Fewer of these places exist in urban rivers. Built-up shorelines prevent erosion, but reduce nooks and crannies where juveniles can rest and forage, and barriers make it hard to get in and out of them from calmer tributaries. 

person holding bucket walks across rocks in a river

Chris Gregersen carries a bucket of juvenile chinook salmon tagged with monitoring technology, looking for a place to release the fish near the Icy Creek Pond Fish Hatchery, Oct. 18, 2021. (Hannah Weinberger/Crosscut) 

Chris Gregersen carries a bucket of juvenile chinook salmon tagged with monitoring technology, looking for a place to release the fish near the Icy Creek Pond Fish Hatchery, Oct. 18, 2021. (Hannah Weinberger/Crosscut) 

“To restore salmon populations, we might not necessarily want to just focus on the river itself,” Gregersen says. 

The preliminary data also show that up to 75% of juveniles don’t make it through this stretch of the river, which wasn’t anticipated. It’s not the only river system with mortality issues in Puget Sound. Upward of 95% of fish released from the Issaquah Hatchery don’t make it out of the Ballard Locks, Gregersen says. “They would be happy to have our problems, but it’s still something worth noting.” 

The system isn’t perfect: The barge is able to pick up about 9% to 10% of all fish that pass by. “We can’t measure the whole river,” Gregersen says. Of the wild fish and hatchery fish tagged and released for this study, they were able to detect 2.5% and 1.5%, respectively, a respectable efficiency score for a pilot project. 

Gregersen also inadvertently created salmon habitat with the barge: One fish kept pinging his antennae and, it turned out, it was living underneath the barge for a few weeks. 

The new data give researchers a starting point from which to design studies looking into why fish are dying. By showing what kind of fish need what kind of changes, the data also may directly inform how the river should be restored. Juveniles fatten up best on the margins of forested streams, with plenty of cooling tree cover, wood debris in which to hang out and slower water.

And the data also will support grant requests to do further restoration. “This data helped us show that [fish] are, in fact, rearing for extended periods in the Lower Green, and we really do need to invest in the Lower Green in terms of building up rearing habitat capacity,” says Goehring, the King County salmon recovery manager.

“Without data, we wouldn't be able to say yes, it works. The problem is, we can't have [data collection] everywhere,” says  Neatherlin of the Governor’s Salmon Recovery Office.

Because this barge is mobile, however, it might be able to give us insight into fish habits elsewhere along the river, or beyond. When the barge comes out of the Lower Green later in the season, it will be stored in Renton until its next assignment. “There's already talk about all the other places we can put it to study fish,” Gregersen says. 

About the Authors & Contributors