From Olympic National Park, August 2024.
A hard question
"So, how many elk are there in the park, anyway?" Visitors ask it often at Olympic National Park, and it's surprisingly hard to answer.
Managers want to track the size and makeup of the park's Roosevelt elk herds for warning signs. Elk can't be hunted in the park, but diseases like treponeme-associated hoof disease and chronic wasting disease in Washington have biologists on alert, and some herds move in and out of the park with the seasons — so the Park Service needs data to work with outside agencies on sustainable herds.
Obstacle course: in summer 2024, staff retrieved the 30 game cameras spread through the Hoh watershed, often off-trail through dense rainforest. Right: wildlife technician Em Wymore packs up an SD card of images. NPS / Em Wymore; NPS / Hazel Galloway.
Two chapters, both paused
| Years | Method | Why it stopped |
|---|---|---|
| 1984–2010 | NPS and USGS helicopter surveys of winter range on the Hoh, South Fork Hoh and Queets floodplains | logistics and survey design; the data showed declines in all three over 27 years |
| 2008–2015 | a new program with Mount Rainier surveying subalpine meadows used in summer | elk shifted when they use the meadows — likely linked to climate change — and aerial surveys grew more complex and costly; suspended after 2015 |
Chapter three: game cameras
Starting in 2019, crews set 30 cameras in remote spots in the Quinault and later the Hoh drainages, taking snapshots at set intervals for a pilot study.
Camera's-eye view: after collecting the cards, technicians recorded each scene. Right: Scientists in Parks interns Jessica Fox and Eliana Hernandez use a compass and a reference photo to measure how far the camera sees. NPS / Hazel Galloway; NPS / Dan Johnson.
Two infrared images from a Hoh Rainforest camera just after sunset, 30 minutes apart. Can you spot both elk in the first? Could a computer program? That's one of the pilot's open questions: how well machine learning can find elk in different light, weather and distances.
Selfies for science: technicians stood in for elk at five-meter steps from each camera, so analysts can estimate the odds of seeing an elk at five, ten, fifteen or twenty meters. NPS / Dan Johnson; NPS / Hazel Galloway.
What cameras can and can't tell
Fixed cameras give different information from aerial counts. "We probably wouldn't get a total count," says Olympic wildlife biologist Miranda Terwilliger, "but we could get occupancy—are elk using this area, or are they not? The question is, can we also get things like cow-calf ratio and sex ratio? And if we can get occupancy, can we tell if the population is increasing or declining?"
- This summer, staff backpacked and bushwhacked by GPS and handwritten notes to bring in the last cameras.
- The catch: cameras produce huge amounts of data. Machine learning may help, but much is uncertain; the images are still being sorted into usable data before analysis can show whether cameras answer the old questions.
Why elk matter
Tracking elk again is a priority for Olympic and the North Coast and Cascades Inventory and Monitoring Network: herd health is both a sign of, and a contributor to, the park's health.
"They're big ecosystem drivers," Terwilliger says. "They're prey to a lot of predators; they drive a lot of vegetation changes therefore they have cascading impacts on smaller mammals and things like that. And clearly they are changing how they use the habitat based on climate change, which maybe has some much more significant implications for parks in the future."
More: the 2008–2017 monitoring report for Olympic and Mount Rainier, and science at Olympic.
Sources
Based on "Snapshots: A Third Chapter in Monitoring Olympic's Elk?," Olympic National Park, National Park Service; a work of the United States government in the public domain.
In these publicationsOlympic National Park
Licence: CC0 1.0 (public domain) · Adapted from www.nps.gov
1
0
0
0
Comments
