Frog vs. Fish

It used to be hard not to find a Sierra Nevada yellow-legged frog underfoot when hiking around Yosemite National Park’s alpine lakes and meadows; it was once the most abundant amphibian in the region. But today, with less than 5 percent of the frog’s former numbers remaining, they’re one of North America’s most endangered amphibians. The recent global spread of the deadly chytrid fungus is partly to blame, but the tale of the yellow-legged frog’s demise begins more than a century ago. This is when park officials—aiming to please anglers—introduced trout to the park’s naturally fishless mountain lakes and streams. The trout ate the tadpoles, and frog numbers plummeted. Oops.

In response, for the past several years, Yosemite has dispatched “killing crews” to cull trout from the park’s high-altitude lakes each summer. These amphibian avengers snag the fish, deliver a death blow, and then chuck the corpses back into the water, where they can decompose without attracting sharp-nosed bears. Once the lakes are deemed fish-free and frog-friendly, park biologists—who stopped stocking trout in 1991—reintroduce the amphibians. So far, they’ve released about 20 adult frogs into two of the seven restored lakes. With the dastardly fish out of the way, the hoppers and many other native creatures, like finches and damselflies, are expected to rapidly return, as they have in other Yosemite lakes where trout naturally died out.

It’s a great strategy, except for one problem: How can park biologists be certain that sneaky anglers don't restock the voracious fish, or that the swimmers don't find some other way back into the waters? Enter a CSI-like solution: eDNA.

Caren Goldberg, an ecologist at Washington State University, is knee-deep in the cutting-edge field of environmental DNA, or eDNA. This modern approach to surveying species in freshwater environments takes advantage of the fact that animals leave DNA—from sloughed off skin, feces, or urine—behind in the water. Researchers simply take water samples and then analyze them for eDNA to determine which species are present.

“It has a high power to detect which species are there,” says Goldberg. And the benefits over traditional surveying methods, which can involve extensive field analyses and disturbed habitat, are substantial. “With this approach, you aren’t turning over rocks, or kicking up mud, or searching for well-camouflaged or incredibly rare individual animals. You’re simply taking water samples.”

Goldberg has used eDNA to look for chinook salmon and invasive New Zealand mud snails in Washington, amphibians in Idaho, and endangered Colombia spotted frogs in Oregon and Idaho.

Of course, like all emerging technologies, eDNA isn’t perfect. A major limitation is that while it’s pretty accurate in detecting if a species is present, it can’t determine exactly how many individuals there are. Furthermore, it’s still unclear if the technique can differentiate between dead animals and live ones. In the case of Yosemite, all that decaying trout at the bottom of the park's lakes could lead to false positives. Fortunately, Colleen Kamoroff, a grad student advisee of Goldberg’s (and a five-year Yosemite trout-removal veteran), has been on the case.

In January, Kamoroff launched a study to determine how long eDNA from dead fish persists. After sacrificing several goldfish (killing them “kind of choked me up, even though I’ve killed thousands of trout in the Sierras,” she says) and putting them in buckets of water, she tested for eDNA through winter and spring. She found that as the fish degraded, their genetic material became too small to be captured by filters with large pores—filters she could use for testing in the field. Her study also indicated that eDNA sinks over time. “That tells me that if I’m sampling at the surface, as we do, and I find eDNA, then I would think it’s from live fish.”

Armed with this knowledge, Kamoroff spent the summer sampling two dozen lakes in Yosemite and nearby Sequoia Kings Canyon. Using a jury-rigged brake pump for automobiles and coffee mug–size water filters, she took water samples around each lake, one every 130 feet. She tested both trout-free and trout-ridden areas, so she could compare results. By the end of this month, she expects to have analyzed all of her samples and to know if her trout detection tool is one that park biologists can use.

While in the field, Kamoroff got a glimpse of what Yosemite’s alpine lakes probably looked like a century ago. One spot she tested in Sequoia Kings Canyon was “loaded” with yellow-legged frogs. “It was just so great to see what seemed to be a healthy population,” she says. If eDNA helps frogs bounce back, high-altitude hikers may once again have to watch their step.