It’s a cold morning in early February, and Chris Vallee of the U.S. Geological Survey is motoring upriver along Steamboat Slough. His two-man crew is hunched in the bow with backs to the wind, wrapped to the ears in water-resistant jackets above warm layers. Vallee pilots the vessel in relative comfort behind the shelter of the windscreen.

The boat passes the usual Delta sights: greenish-brown water drifting past road-crowned levees, the occasional dock and house, a greater egret the color of alabaster lifting off from a eucalyptus branch. After glancing periodically at the open laptop perched on his console, which displays a GPS map of our destination waypoints, Vallee turns the bow toward shore.

The bank is armored with riprap and sheltered by a few trees. As we nose up to dry land, hydrologic technicians Norbert “Nubs” VandenBranden and Ryan Johnson scramble ashore, feeling with bare hands for the steel cable they know is tucked beneath the rocks. After detaching the cable from a nearby tree with leg-length bolt cutters, both hop back into the boat.

Vallee and his team are here to maintain an array of hydrophones used to track migrating native fish. The work is part of a multi-agency effort to provide more timely and detailed information about the movements of salmon, steelhead, and sturgeon in the Central Valley. Deploying hundreds of listening stations across the watershed, the program lets scientists follow thousands of tagged fish as they navigate from hatcheries and headwater streams toward the Pacific Ocean.

The goal of the program is to keep migrating fish safer while continuing to supply water to San Joaquin farmers and Southern California cities. Whether this relatively new technology ultimately helps conserve native fish populations remains to be seen.

As Vallee maneuvers the boat back toward the center of the channel, VandenBranden and Johnson hook the cable to a winch. Soon, a triangular metal frame rises from the depths. They ease the dripping hulk into the bow, and set to work swapping out the white plastic cylinder bolted to its center: a hydrophone.

Vallee types the ID number of the receiver, now stowed carefully on deck, into the computer, and a fresh instrument gets lowered into the water. A few checks to make sure the hydrophone is located and oriented correctly, and they move on to the second receiver at this site. The four receivers here form two acoustic “gates” able to hear the high-frequency sound emitted by a tagged fish no matter where it’s located along this cross section of the channel.

The team moves with the smooth efficiency of long practice, communicating with few glances and fewer words. It’s no wonder: they revisit these autonomous receivers every three months to install units with fresh batteries and download their fish detection data.

The USGS is responsible for 80 to 100 receivers positioned at key junctions along Delta waterways. But the agency was visiting this site and others long before the fish telemetry program existed. The USGS has mounted its receivers adjacent to equipment that tracks water quality characteristics such as water level, flow, salinity, temperature, and turbidity. That’s because layering velocity, discharge, and water quality data atop tagging information adds tremendous scientific value.

“It gives us a lot better information to pass on to water managers for what conditions are ideal for fish, and on the flip side what conditions are lethal,” says UC Santa Cruz fisheries biologist Cyril Michel, part of the large team collaborating on salmon tracking. For example, several studies have shown that higher flows in the river and Delta seem to be the strongest predictor as to whether fish survive that first migration to sea.

That’s good news for fish, according to Michel. “We have the ability to increase survival of fish if the right decisions are made. We have the levers to control the water that comes down the river.”

“Everyone wants more water, but fish are the regulatory constraint on the water flows we have,” Vallee says.

From fin snips to fish blips

The ability to follow fish on their watery travels is the realization of a long-held dream for biologists. The paths fish hatchlings take to reach the Golden Gate, the length of their journey to the ocean, and even the likelihood of a new fry surviving the journey have all long been a mystery.

Over the decades, methods of marking fish for tracking have undergone a sea change. Fin snips were the first method used to distinguish hatchery-born from wild fish. Then, in the 1960s, scientists came up with the idea of also injecting smolts with a sliver of wire etched with a code revealing where and when the fish was born. Traps and trawls were set up at spots downstream to recapture tagged fish. Any fish with snipped fins had its head chopped off and taken back to the lab, where scientists could prise out and read the tag.

But coded wire tags gave limited information. “The recapture rate was only a tenth of a percent. At this rate it took about 20 years to start to get meaningful results,” says Russ Perry, a USGS research fish biologist.

Nor did wire tags reveal the routes fish took to get to a capture site. “We might see high mortality between the Red Bluff Diversion Dam and the Chipps Island trawl [opposite Pittsburg], but we didn’t necessarily know where that occurred,” says Michel.

In the mid-2000s, technology that made WiFi and cell phones possible began to transform fish tracking. Tags got small enough to fit inside the belly of a juvenile salmon. When a tagged fish swims past a listening station, the receiver automatically logs its tag number.

Over the past 15 years, the practice of acoustic telemetry has changed from small-scale use for basic science research to large-scale use by professional agency scientists to guide management decisions. On the order of 20,000 fish have been released since the technology first became available.

Real-time receivers are the latest improvement in the world of fish tracking. Connected to a modem, these devices upload detections as they are received to an online database. The data is instantly available and free for anyone to download from the online portal Cal Fish Track.

“The fish swims by, and through the beauty of cell technology you can see on the website the fish survived to be detected there,” says Rachel Johnson, a research fish biologist with NOAA.

Real-time data promises to enable water managers to adapt their operations based on the season’s fish detections. “This is one of the most exciting parts of the program — the ability for us to collect information in real time and translate the information [into something that helps us] understand if there’s an effect from management choices,” says Josh Israel, a fish biologist with the U.S. Bureau of Reclamation.

Many scientists view real-time as a major advance over autonomous receivers, which must be hauled out of the water to have their data downloaded every few weeks. “A lot of that information is generated on timelines not necessarily reflecting agency needs,” Israel says.

The era of cooperation

Scientists from all the major fish tracking and water agencies are now working together to build an acoustic telemetry network offering broader and more detailed coverage of the Central Valley watershed. The Interagency Telemetry Advisory Group, or ITAG, met for the first time in August of 2018. The member list reads like a who’s who of fish research and water management in the region: California Department of Fish and Wildlife, National Oceanic and Atmospheric Administration, California Department of Water Resources, UC Santa Cruz, UC Davis, U.S. Army Corps of Engineers, U.S. Bureau of Reclamation, U.S. Fish and Wildlife Service, U.S. Geological Survey.

“It’s been easy for agencies to support this because ITAG is focused on creating the kind of information that can be used for fisheries and water management,” says Israel, who co-chairs the group. “ITAG is a big backbone that can support both basic research and investigations into a variety of questions.” 

Much of the funding for ITAG coordination has come through Israel’s employer, the U.S. Bureau of Reclamation. The bureau plays a major role in dam operations and water deliveries in California. Other state and regional investors include the Delta Stewardship Council, whose managers want to see continued advances in science coordination, synthesis and communication.

“We have repurposed the telemetry data to describe broad patterns in outmigration and life history diversity for salmon,” says Pascale Goertler, a Delta Stewardship Council scientist who has synthesized 10 years of telemetry data. Her project helped illuminate ITAG’s data integration needs. “We want to understand how juvenile salmon navigate the risks and rewards of freshwater residency under changing conditions.”

ITAG has made major progress in its year and a half of existence. Participants are now using compatible equipment. Duplicate tag numbers, where two different fish had the same number because different agencies were tracking them, are a thing of the past. All receivers in the network recognize all participants’ tags, producing more detections across the watershed. Care and maintenance of the receivers is divvied up among the participants. All agencies are now using the same river mile designations, making it easier to find and service receivers.

“It’s a great idea, and it’s working very well. I can see it in the way people help each other,” says Flora Cordoleani, a fish biologist for NOAA as well as an ITAG facilitator. For example, she says, participants are now willing to lend gear or replacement parts.

Everyone wants to contribute as best they can because people want to ensure the agencies have the best information,” Israel says.

ITAG also organizes workshops to teach protocols for inserting tags as well as estimating survival and movement rates from the detection data. Unified by ITAG, scientists are now implanting acoustic tags by the thousands in fish runs of interest in the Central Valley, more specifically 7,200 green sturgeon, Chinook, and steelhead in 2020.

Advancing the science

There’s no denying that native fish, particularly salmon, need all the help they can get. A juvenile salmon born in a headwaters stream has just a 3 percent chance of making it through the Central Valley and out to the Golden Gate. In many years, a larger proportion die within the weeks they spend outmigrating than during the whole one to three years they spend in the ocean.

“Survival for Central Valley salmon is terrible,” Michel says.

Acoustic telemetry data has the potential to improve conditions for fish by giving scientists new insights into their habits. Tagged fish are released in groups of hundreds of animals apiece to answer specific questions. The routes fish take can reveal everything from the progress of a run of interest to how fish react to infrastructure like gates, screens, and pumps, or certain water conditions.

These insights could be used to nudge fish away from danger zones and toward better habitat. For example, researchers have found salmon grow larger and faster in the Yolo Bypass. Learning how salmon ply the waters near the bypass’ Fremont Weir entrance will help engineers design a gate to pull fish into this fishy Eden.

Simulated salmon runs

Right now, scientists’ best guess about how a given fish group’s migration will play out comes from the Delta STARS model. This computer simulation predicts where fish will go, and how fast, under given water operation conditions. The model lets scientists and managers release virtual fish at the top of the Delta. River flow and channel information helps the model fill in the details through eight segments of the watershed.

USGS biologist Russ Perry based the model on five years of acoustic tag data from late-fall-run Chinook. “That predictive model basically would be impossible if it weren’t for this amazing network of telemetry receivers and acoustic tag releases over a number of years,” he says.

The model lets managers break down how components such as a given route or water conditions affect chances of survival. “You can look at the differences between scenarios and understand whether one is better or worse,” Perry says.

NOAA Fisheries has already used Delta STARS to evaluate the effects of the California WaterFix plan on fish, while the USBR and DWR used it to study potential fish impacts from proposed operations changes to the Central Valley Project and State Water Project.

Perry is now using telemetry information from winter and spring runs to build more flavors of the model for those imperiled stocks.

“It’s really exciting to have the acoustic telemetry data get to a state where it’s informing models that affect the management of natural and water resources,” Perry says.

The costs of tracking

New knowledge from acoustic tagging comes at high cost. Each autonomous acoustic receiver currently costs up to $6,500. Real-time receivers cost even more due to the need for solar panels for power and communications links. The fish tags can run a few hundred dollars apiece. Multiply this by dozens of receivers and thousands of tags, and the bill for the equipment runs into the millions of dollars.

Then there’s the need to refresh the equipment as improvements come along, the same way businesses must keep their computers and other equipment current. Right now, most of the network’s receivers are two or three years old. Plans are to upgrade the system every five years or so.

As advanced as it seems, the technology severely limits which fish can be tracked. Current tag designs are too large for many juvenile fish. Smolts must be at least 80 millimeters long to accommodate a tag.

“We’re looking via telemetry at only the largest size class of juvenile salmon emigrating through the Delta,” Perry says. “The information is amazing, but we have to keep in mind the other populations and life stages out there when these different environmental conditions and water operations are occurring.”

Questionable stand-ins

To simulate these other types of runs, hatchery smolts could be tagged and released when the wild fish begin their migrations. “But then the question is whether these fish are true surrogates for the population of interest,” says Michel.

The lion’s share of tagged fish have always been from hatcheries. Managers want to know how well these stalwarts of the commercial and recreational salmon fisheries fare. In the case of the winter run, they want to see whether wild fish are protected (the winter run is the most endangered of the salmon runs). Tagged animals are released together with the larger group of fish, and all mingle on their journey to the ocean. Yet many scientists question how well hatchery-raised animals surgically altered to carry a tag truly represent the behavior of untouched, wild-born salmon.

Hatchery fish are believed to have lower migration survival rates than their wild brethren. After all, they’ve grown up in an environment without predators, and no natural culling has occurred before release.

However, both hatchery and wild fish should respond similarly to environmental conditions like flows, temperatures, or routes. So using hatchery fish to make inferences about relative changes in survival due to environmental conditions tends to be more sound than using them to infer absolute values of survival, says Perry.

It could also be argued that tagging a fish as small as a salmon smolt may be invasive enough to alter the animal’s responses. Fish are poured into water laced with sedative, have their belly cut open with a scalpel, and get a tag poked into their abdominal cavity before being surtured back up again.

If a major reason for the telemetry effort is to help guide water-delivery decisions, data from tagged hatchery fish might carry more weight than it can scientifically bear.

Seeing is believing

Around USGS they have a name for the instant eye candy that can be seen so quickly onscreen when the telemetry yields detections: “one fish, two fish, red fish, blue fish.”

Perry says managers must remain alive to the shortcomings of the data, however. “It’s tempting to use the raw detections coming out of the telemetry system,” he says. “But it has to be statistically analyzed first.” These manipulations involve quality filtering of the data, as well as calculations that can add in measures such as of the number of tagged survivors that evaded detection.

In fact, scientists say acoustic tagging should be considered just one of many sources of information used to inform water management, along with trawl and screw trap data, fish surveys, and historical trends.

“There’s a worry that when people see pretty dots on a screen, they interpret that as understanding, but it’s only raw material,” says Steve Culberson, lead scientist for the Interagency Ecological Program. “What I’m more concerned about is that we will collect all this data but not invest in the intellect— the seats we need to fill—to ingest, analyze, understand and communicate what it all means.”

Nor do pretty dots always have statistical significance in terms of the number of detections obtained from each release group within certain regions of the Delta. Only a tiny percentage of any fish cohort gets tagged due to costs. Vanishingly few of those are detected as far as the Delta or pumps. Vallee worries that a lack of detection data will be used by water managers to approve exports. “It’s the fox guarding the henhouse,” he says.

In times of crisis, like extreme droughts, any insights into where endangered salmon are located can be valuable. Consider what happened during the 2015 drought. Declining water quality in the interior Delta threatened the water supply pumped to the San Joaquin Valley. Water managers knew they could boost freshwater levels by opening the Delta Cross Channel gates in Walnut Grove to allow Sacramento River water to improve conditions. Yet opening the channel can pull endangered outmigrating winter-run salmon on an often fatal detour through the interior Delta.

Throughout that drought year, administrators kept asking about the location of the troubled runs. Data from tagged fish and real-time receivers kept them informed. “Having the data and models to see what’s going on now is helpful when decision makers are confronted with balancing fish and water,” Johnson says.

This scenario would only happen in an emergency. “Acoustic tagging data isn’t used regularly to operate the Cross Channel because tagged fish are released after we’ve closed the channel to protect passing salmon,” Israel says. “The only time we may deviate is when the state and federal water projects are at risk of not meeting Bay-Delta water quality standards.”

Vallee feels several additional technologies could help inform water and fisheries management in the Delta. These include split-beam sonar, which detects the air in fish bladders to count animals but doesn’t identify species, and acoustic cameras, which are towed behind a boat and automatically image any fish that pass through their sound beams. A video version of the latter, dubbed SmeltCam, is already being used to study the fragile Delta smelt. 

Overall, this ambitious analytical endeavor is still evolving in terms of being able to answer management questions. The number of fish tags and receivers being deployed is steadily increasing, making sample sizes more representative. And new data filters are able to discern when tags have been eaten by predators. 

 “We’ve also made a lot of progress educating managers and exporters about how to use our statistical models to confirm the error range of the real time results,” says USGS Project Chief Jon Burau. “So if we can ramp all this telemetry analysis up, and combine it with really large scale restoration of dendritic tidal channels, the fish will have half a chance.”

Real-time data: an emerging experiment

Current wildlife and water policies explicitly embrace the use of real-time telemetry data, even as many scientists retain a healthy skepticism about outcomes.

“The policy landscape is allowing more flexibility in decision-making based on current conditions versus calendar-based ones,” Perry says. “People see value in trying to manage for the complexity around water reliability and fish based on what is happening.”

“Having the fish movement data integrated with other information about the environment, and maintaining this array over years and years, gives scientists the chance to get a full picture, over time, of how fish use the system,” says Louise Conrad, Deputy for Science for the Delta Stewardship Council.

“Sociologically, it improves our ability to communicate among our different tribes, and that’s a good thing,” says IEP’s Culberson. “It has moved the needle on what we know and given a sense of pride to the people doing the work, because they are solving problems.”

 “We’re Californians. We’re Delta folk. We’re fishermen. We’re water users,” says Vallee. “We want decisions to be made with the most accurate data, as precisely and truthfully as possible.”

CalFish Track – STARS

CalFish Track – Current Migrations

USGS Innovations Flow Stations Network Delta

UC Davis Biotelemetry Data Base

Related Prior ESTUARY stories

Back to the Bones of the Delta – Restoration is Key to Fish Health

Options for Orphan Species – When it Gets Really Bad What Can We Do?

Federal Research Crew Bucks Headwinds – More on the Best of Federal Science

Back to rest of issue

Receive ESTUARY News for FREE

Top fish video: Salmon smolts in a tank after tagging. Film: Arnold Ammann