On a hazy Delta morning at Isleton’s B & W Resort, more than a dozen trucks are already neatly arrayed in the double-long parking spaces with empty trailers facing the boat launch: evidence of fishers and boaters getting an early start on their Labor Day weekend. Randy Mager, sporting a flannel shirt and worn baseball cap, radiates earnest enthusiasm, which for a 20-year state government veteran is as refreshing and rare as a Delta smelt. “I am more excited about this than I have been about pretty much anything else in my career,” says Mager, a senior environmental scientist with California’s Department of Water Resources (DWR).
The subject of his palpable excitement lays in the water in front of us: a 25-foot long pontoon boat with a soccer-goal-like rectangular array mounted on the front, and a gaping table-size hole cut into the boat’s flat floor. Joe Merz, whose company Cramer Fish Sciences designed the boat, issues an invitation aboard from behind a shark-covered facemask, telling us how the vessel was a Lake Shasta houseboat when he bought it in 2015. After years of testing and modifications, the boat has been so radically changed that according to Merz the only remaining original part is the steering console and instrument panel.
Whitney Thorpe, a fisheries biologist at Cramer and the MacGyver-like genius behind much of the boat’s transformation, hands out life jackets and efficiently runs us through the boat’s safety protocols. “Obviously watch out for the giant hole,” she points out with a friendly laugh. “It’s painted bright orange so you can’t miss it!”
The new monitoring boat is capable of something never done before: combining video technology and DNA analysis from shed molecules in the water to get a clearer picture of what’s going on in the Delta’s murky depths. “Fish monitoring is limited in the Delta’s restored habitats,” explains Mager. Net surveys are all but impossible in shallow riverbanks and intertidal wetlands, where dense aquatic plants choke nets and propellers, and levee rip-rap bruises boat bottoms. “You end up with a net full of vegetation and fish that die by the time you take them out,” he says. Which poses a problem if some of those species are endangered.
With the new method, fish aren’t captured — they simply pass through a water-filled chute under the boat, without being pulled from the water or handled. Mager calls the technology a “game changer” and raves about all the new data it collects and questions it can answer, including whether fish prefer a certain type of restored habitat.
As we drift into the south fork of the Mokelumne River, the rest of Joe Merz’s crew trails behind us in a second boat. Normally the DNA sampling team would also be on the platform boat, scooping up water samples to hunt for molecules they can use to identify what species have passed through the water column recently. But due to coronavirus protocols the group is split up today, in order to have no more than four people per boat.
Merz pilots us to the first sample site, peppering jokes and friendly banter into his overview of the boat’s technology and their research. Meanwhile Thorpe moves in a hyper-efficient blur, rigging up two GoPro cameras and snapping them into a waterproof housing. She belly-flops onto the boat’s floor to connect the net that will funnel fish from the wide-mouthed front array through a narrow chute under the boat’s platform, where the fish pass in front of the cameras before exiting through gates in the back. The apparent simplicity is deceptive: Merz notes they contracted a company to model the fluid dynamics of the water flow in order to design the system so that fish pass through slow enough and at the best orientation to the cameras for successful identification. “This system has been in my head for 20 years,” he says.
We head toward an island of reeds and thick hyacinth in the middle of the river, and Merz uses hydraulics to maneuver the rectangular intake in front of the boat up and down like a bulldozer’s blade. A tablet streaming the video feed shows several fish rapidly passing through the chute, almost too fast to see, though Merz can identify them with just a glance: “Bass,” he says, checking the boat’s line then turning back to the screen. “There goes a bluegill.” Soon he won’t have to, though: the team has developed an algorithm that can automatically identify the fish species from the video. “We’re at 95% accuracy for some of the more common Delta species,” says Merz, who at this point relies on trained staff to verify and improve the algorithm’s recognition. “Our goal is 100%.”
As we nose into the green swath of vegetation, the perils of aquatic plants that Mager warned about quickly become evident. Green beards of algae float by on the video screen, and Thorpe uses a small rake to rapidly pull out the thick clumps so they don’t block the fish passage or cameras. Eventually the net clogs and Thorpe lifts a floor panel to unzip the net from the top, pulling out double handfuls of green sludge.
“When you haul up net, you are just getting an average,” Mager points out. “You don’t know if the fish [you catch in the net] were all in one group, or all at one depth.” He explains that Merz’s group is collecting real-time data on water temperature, salinity, and more that is time-stamped and matched to the individual fish recorded on the video. This gives a much more detailed, nuanced look than simply lumping together fish caught from throughout the water column in a trawl or seine net under one measurement. “With a trawl survey you could spend an hour picking through the net,” says Mager, shaking his head with amazement at how fast Thorpe and Merz have the net cleared and ready.
Merz next demonstrates the vessel’s “four-wheeling” capability, piloting the boat close enough to a rip-rapped levee to elicit an astonished curse from Mager. The boat’s front intake has wheels on the bottom of the posts, allowing it to act as a fender as it bumps along the shallow river bottom. If it encounters something larger, like a tree stump (or, as the requisite Delta joke goes, a dead body), the array is designed to snap off and easily reattach, rather than bending or breaking.
The DNA sampling crew pull up next to us and show off their finger-length clear plastic capsules of water. From these samples the DNA team will filter and capture molecules that fish have sloughed off from scales, mucus, and feces. Once the specimens, traveling on ice to prevent DNA degradation, reach the lab the team will know within 90 minutes which locations had Chinook salmon DNA and therefore which type of habitat the species may prefer.
“[DNA analysis] isn’t new,” notes Gregg Schumer, who oversees Cramer’s environmental DNA services. He points out that the technique just came slowly to environmental monitoring. “Before it was like doing santeria on the Delta,” he laughs, joking that early on his DNA presentation audiences consisted of his mother and one other person. “Now, the rooms are packed.”
Mager excitedly rattles off research questions this technology and data could answer next, ranging from nutrient blooms during levee breeches to salinity gradient impact on fish distribution. “It’s like we just got a new Lego set,” he says. “We could build anything.”
However, Merz and his crew are seeking bigger prey. With the recent addition of a second camera, they now can measure fish length from video, and are starting to incorporate fish larvae into their video identification algorithm. And though their DNA analysis focuses on migratory salmon for Mager’s project, Schumer points out that their treasure trove of samples could be analyzed for a broader sweep of species in order to unveil a more comprehensive picture of the Delta ecosystem.
“We need to monitor the Delta like a patient in a hospital: looking at the big picture,” says Schumer, equating traditional monitoring to diagnosing a patient by staring at their pinky. “We are just scratching the surface of the things we can do.”
Related Estuary Stories
Top photo: Isaac Pearlman