Environmental DNA reveals the secrets of the creek and beyond

Brittany J. Miller, by Monterey Herald


A drop of rain came from a California plain and fell into a creek. There it scraped against the fish and slipped through their gills, stealing the traces of each encounter. The droplet carried genetic mementos down the road until it reached an innovative device that helped unlock the secrets of the creek’s fauna.

“We call it the Microbiology Lab in Cannes,” said Jim Birch, director of the SURF Center at the Monterey Bay Aquarium Research Institute, or MBARI, located in Moss Landing.


That “can” is actually the MBARI’s Environmental Sample Processor, a $200,000 robotic laboratory the size of a 50-gallon drum. It collects genetic clues – cells, mucus, feces – from ecosystems that are collectively called environmental DNA or eDNA.

MBARI staff installs an environmental sample processor to collect water from Scott Creek for coho salmon, steelhead trout and subsequent genomic analysis to detect invasive species. (Kim Fulton-Bennett/MBARI)

In a project on Scott Creek, north of Davenport in Santa Cruz County, the device produced one of the largest single-site eDNA data collections in the country. From April 2019 to April 2020, scientists uncovered details about endangered and invasive species in freshwater ecosystems. Now a scientific paper in the works, the study reinforces a growing interest in the detection and improved protection of hard-to-find species using eDNA surveillance instead of more invasive techniques such as fish counting.


“It has the potential to do this without putting a lot of traps in the water,” said MBARI’s Kevan Yamahara, an expert on the device and one of the paper’s authors.

Worldwide interest in eDNA’s ability to detect rare organisms has expanded over the past few decades. New technology rediscovered a rare aquatic insect population in the United Kingdom. It detected more mammals in the Canadian wilderness than a traditional camera trap. It helped track the spread of coronavirus.

SJM L FISHDNA 0106 90 01At Scott Creek, MBARI’s equipment pumped water from the creek’s flow and pushed it through a filter several times a day. Once the filter has collected enough material, the machine applied a preservative. According to Yamahara, each filter was placed in a carousel similar to a bullet-loaded chamber of a gun. Once the carousel was filled with 132 samples, the researchers collected the data and took it to their Moss Landing Lab.


SJM L FISHDNA 0106 90 01The year-long monitoring revealed about 700 samples. The researchers focused on the creek’s endangered coho salmon and threatened steelhead trout populations, both commercially important fish. Because Scott Creek is one of the southernmost points where coho salmon come to lay eggs, it’s important to know how the species is progressing, said Birch, who is also the author of a soon-to-be-presented research manuscript.

In the creek, the device sat next to a more established monitoring device: a weir operated by the National Oceanic and Atmospheric Administration. Birch said the dam, a perforated “flow-through dam,” has allowed NOAA staff to match, observe and release fish on a seasonal basis for two decades.

Fish counting continued throughout the project and allowed for comparison of traditional and emerging monitoring techniques. From the samples, the team found that the amount of steelhead trout DNA is generally higher than that of coho salmon. According to Ryan Cersei, an environmental engineering doctoral student at Stanford and lead author of the research paper, this complements the numbers seen in NOAA fish nets.


The collected eDNA also provided seasonal data that showed the species’ dubious life history through winter rains, summer dry spells and most of the days in between. The information revealed the best times to do eDNA sampling for some species, Yamahara said.

The highest concentration of coho salmon eDNA, for example, appeared during the winter, when the fish were thought to migrate and lay eggs. During the fall, when the flow of the creek decreased, the amount of salmon eDNA also decreased. Sarsi said the findings give researchers confidence in the data and suggest that new monitoring methods may be suitable for documenting the behavior of migratory fish.

SJM FISHDNA 1225 01 91560502
Alex Hay, left, a researcher at the NOAA Southwest Fisheries Science Center, and Kevin Yamahara, an ESP specialist at MBARI, observe a rainbow trout retrieved from a downstream weir in Scott Creek. (Todd Walsh/Mbari)

The data uncovered other mysteries of the creek: The team found that less than 1% of eDNA came from invasive species. This low number offers hope that species such as New Zealand mudsnail and striped bass are not yet present in the creek, Sarsi said. Yamahara said such monitoring could give scientists early warning signs for the invasive species.

“You don’t really have to physically look for and physically look for those specimens,” Yamahara said. “You can just take a water sample and process it.”

When it was unveiled in the late 2000s, MBARI’s environmental sampling processor traveled around the world, Birch said. The version of the Scott Creek device has since changed to a new model the size of two basketballs.

The researchers incorporated the advanced technology into underwater autonomous vehicles, which are now used to explore marine habitats in Monterey Bay, Birch said. They are also used in the Great Lakes to track harmful algae blooms, Yamahara said.

As eDNA monitoring evolves from stationary machinery to travel efforts in ecosystems, Yamaha expects the technology to go even further for use in freshwater habitats such as Scott Creek. But while this tool could revolutionize ecosystem monitoring, prototyping has its limits.

SJM FISHDNA 1225 04 91560504
Research specialist Kevin Yamahara, left, and mechanical engineer Doug Purget secure the hoses that will carry water samples from the fish net to the ESP. (Kim Fulton-Bennett / MBARI)

For example, the sheer amount of genetic information that new tools can provide can overwhelm laboratories, Birch said. To fix this problem, he wants to see future versions of the technology being analyzed on-site.

He said, “It really is the Holy Grail — the brass ring — that we here at MBARI are trying to push…

There were also discrepancies between the old and new monitoring strategies. The team found fish eDNA more frequently in NOAA’s fish nets than in fish. This is unexpected, but not unheard of in the field, according to Sarsi, especially since the finding that the eDNA sampling site may be related to upstream fish.

Since each technique reveals different details about the species, they should be viewed as complementary at this point, the researchers say. This combination is valuable and rare in the greater Bay Area, said Brian Alley, lead fisheries biologist for the South Bay Clean Creeks Coalition, who was not involved in the study.

While the device’s $200,000 price tag may limit its use, Alley would like to see its eDNA monitoring applied to local urban streams to check for endangered species.

“We really want wild populations to arise on their own on a sustainable basis,” he said. “It’s been a difficult process – one of which technology is important because we can’t turn the clock back to the Lewis and Clark era.”