Last updated: November 29, 2024
Article
Untangling the Threads of North Cascades’ Marmot Decline
Hoary marmots, chubby rodents closely related to groundhogs, are a familiar sight in mountain meadows and talus fields from the Cascades as far north as Alaska. But in North Cascades National Park near southern end of their range, these hardy mammals are being squeezed by shrinking snowpack. In a new study, scientists ask whether climate change might carry other risks for marmots—risks that walk on four legs.
Washington State University / NPS / Cascades Carnivore Project (color corrected)
Between 2007 and 2016, hoary marmot counts in the park fell by a whopping 74%. In a 2020 study, USGS ecologists in partnership with park biologists linked this decline to chronically dry conditions and low snowpack, combined with acute cold.
Snow acts as a protective blanket during the harsh mountain winter, holding subnivean (below-snow) temperatures constant while marmots slumber in a deep torpor. Without this insulation, hibernating marmots can experience physiological stress, hurting their chances of reproducing in the spring—or even freeze to death.
"We don’t know that it’s catastrophic," cautioned park wildlife biologist Jason Ransom, an author of the 2020 study. Marmot populations go through natural boom-and-bust cycles over time. “Even so, 74%—that’s kind of outside the boundaries of a normal bust.”
More than one factor may be at play. Scientists have linked declines in neighboring Olympic marmots and Vancouver Island marmots, both endemic species with small ranges, to changing patterns of predation. Beyond winter's immediate threat, could retreating snow be opening new doors for resourceful predators?
A group of researchers including Washington State University masters student Logan Whiles, NPS wildlife biologists, and scientists with the nonprofit Cascades Carnivore Project came together to find out. They asked three basic questions:
- Who are the biggest predators of hoary marmots at North Cascades?
- Where in the park’s marmot habitat are these predators most common?
- How are marmots changing their behavior based on predator risk?
Washington State University / NPS / Cascades Carnivore Project
Likely Suspects
The first step was to map the whereabouts of possible predators. Research teams set up camera traps near 32 marmot colonies across the park and adjacent national recreation areas. Using hundreds of thousands of photos taken throughout the summers of 2018 and 2019, they counted the frequency of ten species that marmots may perceive as predators: cougars, wolverines, grey wolves, bobcats, Canada lynx, black bears, coyotes, Pacific martens, Cascade red foxes, and humans.
With this wealth of data, they created a spatial model for how likely predator species were to use marmot habitat across the park.
From Whiles et al., 2024
Four maps of North Cascades National Park Service Complex divided into small, color-coded hexagons. Hexagon color corresponds to the model-averaged probability of use for each of four predator species, ranging from purple (0 probability of use) through dark blue, mid blue, green, and yellow (1.0, or 100%, probability of use).
The model-averaged probability of use by black bear is highest throughout the park out of all four species, with most hexagons colored yellow or yellow-green and only a few blue or purple areas.
The model-averaged probability of use by cougar is lower but more variable, with approximately half the hexagons low-probability (purple or dark blue), while others are mid-blue or green. No hexagons indicate a high probability of use.
The model-averaged probability of use by coyote is uniformly low (purple) throughout the park, except for a cluster of mid-blue to green hexagons in the far southeastern corner.
The model-averaged probability of use by Pacific marten is in the mid-range (mid blue to green) throughout the park, with a gradient of slightly higher probability as you move from west to east.
The model-averaged probability of use by black bear is highest throughout the park out of all four species, with most hexagons colored yellow or yellow-green and only a few blue or purple areas.
The model-averaged probability of use by cougar is lower but more variable, with approximately half the hexagons low-probability (purple or dark blue), while others are mid-blue or green. No hexagons indicate a high probability of use.
The model-averaged probability of use by coyote is uniformly low (purple) throughout the park, except for a cluster of mid-blue to green hexagons in the far southeastern corner.
The model-averaged probability of use by Pacific marten is in the mid-range (mid blue to green) throughout the park, with a gradient of slightly higher probability as you move from west to east.
Researchers predicted that typically low-elevation predators—black bears, cougars, wolves, bobcats, and coyotes—would be seen more often at sites with earlier snowmelt. To their surprise, this link was found only in one predator species: coyotes.
The Proof is in the...
While the cameras were standing sentinel at marmot colonies, researchers themselves were busy with something much smellier. To confirm which carnivores make a habit of munching marmots, they collected over 500 pieces of evidence—scat (wildlife poop).
Photo courtesy Logan Whiles
“One person looks at the center of the trail, the other person scans the perimeter of the trail, and we would just do that the whole time we hiked,” said study author Logan Whiles. “You definitely start noticing a lot more stuff on the ground. It was fun to learn all their little quirks—like the martens love to defecate on what looks like a high point to them on the trail, like a tiny little rock.”
Some scat was more of a challenge. “We just picked up so much black bear—it’s so wet and so heavy and so large,” laughed Whiles. “We’re trying to backpack long distances for the entire summer, and the bear poop was really weighing us down.” (The second field season, researchers skipped scooping the mostly-vegetarian bear scat unless prey remains were evident.)
Photo courtesy Logan Whiles
Once the scat was packed out and dried, genetic metabarcoding was used to confirm the species of origin and check for marmot DNA. One big surprise came from those dainty marten scats—4% contained marmot DNA, the first record of this weasel relative preying on the often-larger marmots. Coyotes stood above the competition, with one in five scats showing a recent meal of marmot. Marmot remains rarely appeared in wolf, cougar, or black bear scats, and never in bobcat or lynx scats. No wolverine scats were collected.
Enemies in High Places
Armed with the answers to both who the most dangerous predators are and where they are most likely to be, researchers crunched the numbers to predict the risk to marmots from each predator across the North Cascades landscape.
From Whiles et al., 2024.
Four maps of North Cascades National Park Service Complex divided into small, color-coded hexagons. Hexagon color corresponds to the risk landscape for each of four predator species, ranging from purple (0) through dark blue, mid blue, green, and yellow (0.3 and higher).
The risk landscape for marmot predation by black bear is uniformly low, with all hexagons purple or purple-blue.
The risk landscape for marmot predation by cougar is uniformly low, with all hexagons purple or purple-blue.
The risk landscape for marmot predation by coyote is uniformly low, except for a distinct cluster of blue, green, and yellow hexagons in the far southeastern corner of the park. High risk areas are concentrated at the very southeast tip of the map.
The risk landscape for marmot predation by Pacific marten is in the mid-range (mid blue to green) throughout the park, with a gradient of slightly higher risk as you move from west to east.
The risk landscape for marmot predation by black bear is uniformly low, with all hexagons purple or purple-blue.
The risk landscape for marmot predation by cougar is uniformly low, with all hexagons purple or purple-blue.
The risk landscape for marmot predation by coyote is uniformly low, except for a distinct cluster of blue, green, and yellow hexagons in the far southeastern corner of the park. High risk areas are concentrated at the very southeast tip of the map.
The risk landscape for marmot predation by Pacific marten is in the mid-range (mid blue to green) throughout the park, with a gradient of slightly higher risk as you move from west to east.
Just one question remained: Are marmots changing their behavior based on risk from predators? More time and energy spent on guard could stress colonies, even if individual marmots avoided an untimely end. The short answer, researchers found, is no.
Observing marmots at their colonies, researchers tracked the amount of time each spent watching for predators compared to other behaviors. (As another measure of vigilance, they also tested the distance at which marmots fled from an approaching researcher.) They expected that in riskier areas—where models predict lots of marmot-munching predators—marmots would keep a closer eye on their surroundings. This pattern was not found in the data, but that doesn’t mean marmots are letting down their guard. The observations showed they spent, on average, over 50% of the time aboveground keeping watch!
A Bad Combination
Like ecosystems themselves, the future of marmots in the North Cascades is interdependent with many factors. One heavyweight in the ring, however, is the coyote.
“Out of all the predator scats that we picked up that had marmot in it, 50% of them came from coyotes,” said Whiles. At the same time, coyote presence at a site was correlated with earlier snowmelt dates—dates which are already creeping earlier as the climate changes.
“When there’s no snow, what we see from this and other studies is that it allows an ‘elevational release’ for carnivores,” added Ransom. “So coyotes are much more likely to wander around up there, because it’s easy.” This points to a possible interactive effect: if the physiological stress of low snowpack meets the risks of easier predator access, the combined effect may be more dangerous than the sum of its parts.
“You’ve got marmots struggling maybe to wake up the same time coyotes are up there,” said Ransom. “That’s a bad combination.”
NPS / Jason Ransom
What's Next?
Park managers can use these findings to avoid adding extra human-caused stress to marmots when planning projects like new trails. But their tools are limited. “We can’t build a snow globe over the mountains,” said Ransom. “But it’s still important to do the science and monitor what’s happening.”
There is lots left to learn. Carnivore communities themselves are constantly reshaped by outside forces, including human development beyond park boundaries and a rise in wildfire frequency. The slow return of wolves to the park and the recovery of wolverine populations are likely also sending ripples through the predator-prey system.
Even with these unknowns, the cocktail of threats to a beloved North Cascades resident can feel like another gloomy climate story. It’s something the researchers think about too.
“I went into this project looking at that 74% number and it made me really sad, and it still totally can,” said Whiles. But spending time with the marmots and finding colonies that beat the odds has given him hope. “Sometimes I’d show up to this one site at Rainbow Lake—and I need to check in on how that little family’s doing. They just had so many babies,” he added. “They ate my backpack and hiking poles one year.”